Research on the Mechanism of Digital–Real Economic Integration Enhancing Industrial Structure Upgrading

Abstract

The integration of the digital and real economies (DRI) is an inevitable trend in future economic growth. This study measures DRI levels across 30 Chinese provinces from 2012 to 2022 using a coupling coordination model with panel data and empirically examines DRI’s impact on industrial structure upgrading (ISU) through fixed-effects models, mediation effect models, and panel threshold models. The findings reveal that (1) DRI promotes industrial structure upgrading, a conclusion that remains valid under robustness tests and endogeneity tests; (2) DRI can facilitate ISU by enhancing consumption levels, correcting factor distortions, and accelerating the marketization process; (3) there exists a threshold effect, with a positive effect of DRI on ISU based on the level of digital economy and the scale of the real economy as threshold variables; (4) the impact of DRI on ISU differs across different regions due to differences in policy support and resource allocation; (5) ISU has a significant spatial spillover effect, as shown by spatial econometric analysis. These conclusions offer a new perspective, practical policy implications for China’s high-quality economic development, and strategic insights to enhance industrial competitiveness in the global value chain.

1. Introduction

To achieve high-quality development, it is key to optimize the industrial structure and replace old growth drivers with new ones. An up-to-date industrial framework forms the material and technical bedrock of a contemporary nation, making it critical to center economic development on the real economy. Grounded in this real economy foundation, efforts should prioritize bolstering cutting-edge manufacturing, advancing new-style industrialization, upgrading conventional sectors, expanding burgeoning industries, and proactively laying the groundwork for future industries. These steps will hasten the development of an advanced industrial system with cutting-edge manufacturing at its core.

However, there are still some obstacles to the current transformation and upgrading of China’s industrial structure, such as the urgent need to strengthen innovation transformation, insufficient digital penetration in manufacturing leading to an inverted relationship between structural progression and productivity growth (H. Zhang, 2023), and regional disparities and coordination deficiencies in digital development that exacerbate Baumol’s disease and premature deindustrialization (X. Chen & Xu, 2024). The bottleneck constraints are caused by the lack of “patient” capital and international technology blockades, which severely hinder technological innovation and further exacerbate the imbalance of the industrial structure (G. Zhang, 2023). It is key to supporting high-quality economic development that promotes ISU and advances the construction of a modern industrial system.

Against the backdrop of rapid digitalization, the innovation, application, and deep integration of digital technology with the real economy provide a new engine for high-quality economic development in the new era. Data from the National Bureau of Statistics indicate that between 2020 and 2024, the share of value-added of China’s core digital economy industries in GDP increased from 7.8% to approximately 10%. However, industrial digitalization is characterized by unbalanced development. The eastern regions dominate in terms of scale, innovation capacity, and industrial transformation outcomes. In 2024, their digital industry revenue grew by 6.5% year-over-year, accounting for 73.6% of the national total. In contrast, the central, western, and northeastern regions recorded growth rates of 4.2%, 0.8%, and 2.5%, respectively. This highlights the pivotal role of the digital economy as a catalyst for ISU. However, it is crucial to recognize that the foundational status of the real economy necessitates integration as the locus of digital economic value. Without integration, the digital economy risks becoming an “air castle”. Meanwhile, the depth and breadth of such integration, as well as the degree of regional digital synergy, equally shape its enabling effects. Therefore, effectively unlocking digital dividends, promoting digital transformation in traditional real sectors, and constructing a mutually reinforcing and sustainable economic model are critical agendas for overcoming bottlenecks in industrial upgrading.

In recent years, studies on the impact of the DRI on ISU have focused on theoretical and empirical research on the role of the digital economy in ISU, as well as the theoretical elaboration of the impact of DRI on ISU. In terms of the theoretical and empirical research on the impact of the digital economy on ISU, the digital economy is considered to be new kinetic energy leading the upgrading of the industrial structure in the future (X. Chen & Yang, 2021), which can promote the deep integration of manufacturing industry with the Internet, research and development, service industry, and new technologies, providing strong support for industrial transformation (Jiao, 2020). Its organic integration with the service industry can continuously deepen the industrial structure of China’s economy (X. Dong et al., 2022). Y. Liu and Chen (2021) and T. Zhang and Jiang (2022) used provincial panel data from China to investigate the promotional effect of the digital economy on ISU, as well as the mechanisms of human capital, technological innovation, factor structure, and industrial structure. Deevi et al. (2024) used data from 25 regions in China from 2014 to 2022 to confirm the driving effect of the digital economy on ISU, considering sustainable entrepreneurial growth as an important path to realization. However, the realization of the efficacy of the digital economy is fundamentally dependent on the supporting infrastructure provided by the real economy. The aforementioned studies are confined to the single-dimensional enabling effects of the digital economy and fail to delve into the synergistic forces generated by the dynamic process of DRI on ISU.

In terms of the theoretical impact of DRI on ISU, Y. Liu and Xiu (2022) systematically elaborated on the promotional effect of DRI on ISU, the number of domestic patent applications and acceptances, and investments in research and experimental development. The DRI aims to fully drive the transformation and upgrading of the real economy through the dual engines of digital technology and data elements (Hong & Ren, 2023), which greatly encourages the leading role of enterprises in innovation-driven, employment promotion, and international competition, and is a new kinetic energy for the high-quality development of the modern industrial system (Xia & Su, 2024). The promotion of new industrialization by DRI has the stage characteristics of innovation-driven, efficient and intensive, and internal and external linkage (Ren & Li, 2024), and its empowerment in the digital transformation and upgrading of the manufacturing industry is reflected in flexible production and smart manufacturing (X. Xu & Du, 2024).

The above studies have explored the impact of the digital economy and DRI on ISU, focusing on the empirical testing of the promotional effect of the digital economy on ISU at the empirical analysis level, laying the research foundation for this study. The theoretical discussion of DRI and ISU has laid a solid theoretical foundation for this study. However, few studies have extensively explored the intrinsic mechanism and realization path of DRI’s effect on ISU; additionally, whether DRI has a nonlinear impact and spatial spillover effects on ISU remains to be tested.

Therefore, this paper uses panel data from 30 Chinese provinces from 2011 to 2022, focusing on “the specific impact mechanism and path exploration of DRI on ISU”, constructing a comprehensive analysis framework to explore how the digital economy interacts with and penetrates the real economy during the DRI process, and uses fixed-effect models, panel threshold models, and spatial econometric models to explore its driving effect on the evolution of the industrial structure towards higher efficiency and optimization.

The marginal contributions of this paper include three aspects. First, from the unique perspective of “DRI”, systematically analyzing and testing the promotional effect of the DRI on ISU and the specific path mechanism. Compared with the literature focusing on the impact of the digital economy or the real economy on industrial structure, this perspective can more comprehensively reveal the synergistic effects and deep-seated reasons for the industrial structure transformation brought about by the “dual-wheel” driving effect of the integration of the two. Moreover, this paper conducts empirical tests using panel data from 30 provinces and cities across the country, providing empirical evidence for the promotion of ISU by DRI and enriching the research on the factors affecting ISU. Second, this paper explores the nonlinear relationship between DRI and ISU using the panel threshold method, finding that the impact of DRI on ISU varies with the level of development of the digital economy and the real economy, which provides a thought for synergizing the development of the digital economy and the real economy to promote ISU and provides a theoretical basis for formulating more appropriate supportive policies. Third, this paper employs a spatial panel model to examine the spatial spillover effects of DRI and ISU, confirming the spatial spillover effects of DRI and ISU, exploring the spatial transmission effects of DRI, providing a new perspective for related research, and offering policy insights for the coordination of DRI and ISU among regions.

2. Literature Review and Theoretical Foundation

2.1. Literature Review

Scholars have widely focused on measuring ISU and identifying its influencing factors. As the digital economy evolves, research has turned to examining how it affects ISU. The profound integration of the digital economy with the real economy has injected new momentum into ISU, sparking academic debates around DRI and theoretical inquiries into its impact on ISU.

2.1.1. Research on the Influencing Factors of ISU

Many scholars have conducted extensive and in-depth discussions on the factors that may affect ISU. They believe that technological innovation, economic system reform, factor cost and factor price, financial agglomeration, trade opening, financial reform, digital economy innovation, human capital, environmental regulation, and population aging may all have an impact on ISU (Peneder, 2003; Du et al., 2021; Yin et al., 2022; J. Yu, 2024).

2.1.2. Connotation of DRI and Theoretical Exploration of Its Impact on ISU

Connotation and interaction mechanism of DRI: DRI refers to the real economy sector transforming itself by achieving factor innovation and efficiency upgrades through the purchase and utilization of data elements from the digital economy sector, supported by integration infrastructure, development environment, and digital economy supply (D. Wang et al., 2023). It focuses on technology, industry, and enterprise levels (Hong & Ren, 2023), with approaches involving the digital economy at enterprise, industry, and society levels (Y. Chen, 2023), and features industrial digital upgrading, enterprise digital transformation, and labor skill changes (Ding et al., 2024). Relevant studies have explored its economic benefits from multiple perspectives (Llopis-Albert et al., 2021), defined it as the systematic embedding of digital technologies in industrial chains and product lifecycles to form a new techno-economic paradigm (Meng et al., 2023), and showed that deep integration of digital twin systems in production enhances efficiency (Alsakka et al., 2024). He et al. (2024) enriched its connotation with digital technology and practice, while Xia and Li (2024) noted that it empowers high-quality development through ecosystem construction, real economy foundation solidification, and environmental optimization.

Empirical test of the impact of the Digital Economy on ISU: The digital economy utilizes the technical–economic characteristics of data elements to promote the formation of scale effects and drive technological progress (Guo et al., 2024), overcoming the “Baumol disease” and improving the efficiency of the service industry sector (G. Liu et al., 2023). The realization of the value of data elements promotes price discovery and reduces the cost of information search and principal–agent delegation (Luo et al., 2024), effectively promoting corporate innovation. Highly qualified and skilled labor is the power source of ISU (W. Wang et al., 2015), and the digital economy promotes the industrialization, high-technology, and high-skilled transformation of the employment structure (Ye et al., 2021), effectively alleviating the constraint of high-quality human capital bottlenecks, promoting innovation-driven and ISU, and facilitating supply-side structural reform and sustained economic growth (Z. Liu et al., 2018).

Theoretical exploration of the impact of DRI on ISU: The deep integration and coordinated promotion of the digital economy and the real economy are key to industrial transformation and upgrading, an important driving force for high-quality economic development, and a key factor for economic modernization (Han et al., 2023). At the macro level, the deep DRI has promoted the transformation of the economic form (W. Liu et al., 2024), playing a core driving role in fostering a new economic ecosystem characterized by new industries, new formats, and new models. The dual-wheel driving mechanism of DRI promotes green innovation and green development (Cui & Feng, 2024), accelerates the free flow of factors among industries, realizes the cooperation of the whole industry chain, and promotes the construction of a unified national market (B. Shi & Hu, 2024). At the micro level, based on external perspectives such as cooperative innovation networks (L. Wang et al., 2024), supply chain transformation (Ren & Miao, 2024), and internal perspectives such as promoting production and innovation paradigm shifts (Li & Dong, 2024), reducing corporate R&D manipulation behavior (S. Dong et al., 2023), and improving corporate innovation quality and efficiency, total factor productivity (Y. Xu & Shi, 2024; X. Huang & Gao, 2023), the positive role of DRI in corporate innovation is explored, providing a micro perspective and path for ISU.

Compared with existing studies, this study makes several crucial contributions by examining the influence of DRI on ISU from a novel perspective. The existing studies primarily focus on theoretical research concerning the impact of DR on ISU. In contrast, this paper adopts an empirical approach as its research foundation. Furthermore, the existing literature on the relationship between digitalization and ISU has predominantly explored the linear correlation between DRI and ISU. This study employs the panel threshold method to investigate the nonlinear relationship between DRI and ISU, demonstrating that the impact of DRI on ISU depends on the development levels of the digital and real economies. Finally, by applying a spatial panel model, the study confirms the spatial spillover effects of DRI and ISU, providing valuable insights for regional coordination and policymaking.

2.2. Theoretical Foundation

2.2.1. Direct Effect of DRI on ISU

DRI driven by the digital economy transforms the real economy through digital means, realizing a dynamic model of mutual promotion and circular development between the digital economy and the real economy (M. Wang et al., 2024). Specifically, it is the application of data, processed by digital technology using modern information networks, in various fields of society, which is reflected in enterprises. By using high technology to upgrade old equipment and production modes, it improves productivity and rates of factor utilization, solving the problems of resource waste and environmental pollution in traditional industries; by using modern information networks to optimize and reconstruct old management methods, it improves corporate operation efficiency and reduces labor costs, enhances market information grasp, and reduces production risks and overcapacity caused by information asymmetry and information gaps. More importantly, the application of digital elements in the real economy to address the sustainable development of traditional industries has found a new economic force, leading to high-quality economic development, thereby promoting the intelligent development of industries and the rationalization of industrial structure. Therefore, hypothesis H1 is proposed:

H1. 

DRI can promote ISU.

2.2.2. Indirect Effect of DRI on ISU

Expanding domestic demand is the strategic basis for building a new development pattern of “dual circulation”. Q. Shi et al. (2009) evaluated the relationship between consumption upgrading and industrial structure using input–output tables from 2000 and 2005 and found that consumption upgrading can explain 29.4% of the changes in industrial structure. K. Liu and Feng (2024) found that the strategy of expanding imports can significantly promote ISU, and optimizing consumption structure and expanding consumption scale are important ways of promoting ISU. From this, we can infer that increased consumption demand has a positive effect on ISU (refer to Figure 1). Based on this, hypothesis H2 is proposed:

H2. 

Consumption upgrading plays a mediating role in the relationship between DRI and ISU.

The core feature of factor market distortion is that its price has not reached the market-clearing level under competitive equilibrium, mainly manifested as a mismatch between total factor supply and demand, and an unreasonable allocation of structure (C. Yu & Shen, 2024). The development of the digital economy is the premise for DRI to improve factor market distortion (W. Liu et al., 2024). Under the promotion of the digital economy, data elements, as knowledge carriers, enhance marginal output through non-rivalry and spillover effects, synergizing with capital and labor, consistent with Romer’s knowledge spillover theory. When integrated with physical capital as complementary inputs (Acemoglu et al., 2018), they bypass price rigidity in traditional factor markets, optimizing resource allocation to elevate industrial productivity and economic gains. Digital technology achieves real-time sharing of information by digitizing and dynamizing factor information, breaks the temporal and spatial constraints between supply and demand (Y. Zhang & Wang, 2020), and helps to achieve more accurate matching between factor supply and demand, promoting ISU (refer to Figure 1). Based on the above analysis, hypothesis H3 is proposed:

H3. 

The degree of factor distortion plays a mediating role in the relationship between DRI and ISU.

DRI relies on digital platforms, utilizes digital technology, fully demonstrates the characteristics of data resource sharing and virtual substitution, and promotes the digital transformation of industrial structure. With the widespread application of digital technology, the loss of market players caused by information asymmetry is gradually reduced, which can effectively prevent the occurrence of market monopoly and make the market more inclined towards a state of full competition. The acceleration of the marketization process can help stimulate market vitality, extend the length of the industrial chain, increase the division of labor, and increase market demand, thereby expanding the profit space of enterprises and enhancing the power of enterprise transformation and upgrading (G. Liu & Li, 2024). The marketization process plays an indirect driving role in the process of DRI to promote ISU (refer to Figure 1); thus, hypothesis H4 is proposed:

H4. 

The marketization process plays a mediating role in DRI and ISU.

2.2.3. Nonlinear Effect of DRI on ISU

Analysis of the hypothesis shows that DRI can promote ISU, which assumes regional homogeneity and coordinated development between DRI and ISU. Recent studies have confirmed the heterogeneity in the development of the digital economy, the scale of the real economy, and the industrial structure among the eastern, central, and western regions of China. In terms of the time dimension, the digital economy, the real economy, and ISU may not achieve coordinated development; therefore, the level of development of the digital economy and real economy may cause differences in the impact of DRI on ISU. First, the digital economy has a nonlinear impact on ISU. Specifically, the marginal effect on the advancement of industrial structure decreases, and the marginal effect on the rationalization of industrial structure increases (Y. Liu & Chen, 2021). The empowering effect of the digital economy also shows a nonlinear characteristic of “diminishing marginal utility” due to different levels of digitalization (W. Zhang & Zhou, 2023). Second, the impact of the digital economy on ISU in China has a phased feature, with a stronger promoting effect after 2013 (X. Chen & Yang, 2021). Finally, there are differences in the application foundation, revenue, and digital personnel foundation of digital industries among different regions or cities, and the impact of the digital economy may be influenced by city size (Lu, 2024). Based on the above analysis, hypothesis H5 is proposed: DRI has a threshold effect on ISU due to the level of digital economic development and the level of real economic development.

2.2.4. Spatial Effects of DRI on IUS

The digital economy’s openness, sharing, and networking characteristics not only foster high-quality local economic development but also create significant spatial spillover effects in neighboring regions through mechanisms such as technology diffusion, factor mobility, and industrial synergy (X. Zhang, 2019). The non-exclusive and non-rivalrous nature of data elements enables the free flow and efficient allocation of factors such as technology, capital, and talent across regions. This, in turn, expands the scope and depth of economic cooperation (Ming & Peng, 2024). Such spatial spillover effects provide new impetus for IUS. On the one hand, regions with more advanced digital economies can leverage the spillover effects of technological innovation to drive industrial structure upgrading in neighboring regions; on the other hand, inter-regional collaborative cooperation can mitigate the limitations of industrial structure layout in a single region. Therefore, the DRI not only promotes IUS within a province but also leverages spatial spillover effects to drive industrial structure transformation in neighboring provinces. This enhances the coordination and interdependence of IUS across neighboring provinces. Based on the above analysis, Hypothesis H6 is proposed: The impact of DRI on IUS has a positive spatial spillover effect.

3. Research Design

3.1. Variable Description

This study utilizes panel data from 30 Chinese provinces from 2012 to 2022. The original data sources include the China Statistical Yearbook, statistical yearbooks of various provinces, and the websites of provincial and municipal statistical bureaus.

3.1.1. The Dependent Variable

ISU includes two aspects: industrial structure advancement and industrial structure rationalization. Industrial advancement reflects the orderly evolution of the industrial structure from low to high levels. This paper uses the industrial hierarchy coefficient as a measure, with the following formula:

$$RIS={\sum }_{i=1}^3{y}_i\times i$$

(1)

where $y_i=Y_i/Y$ represents the proportion of the value of the ith industry’s output to the total output.

Industrial structure rationalization reflects the degree of coordination between industries and the effective utilization of resources, providing an effective measure of the level of the industrial structure. This paper uses the Thiele Index to measure industrial structure rationalization based on the following formula:

$$Theil={\sum }_{i=1}^3{\displaystyle \frac{Y_i}{Y}} \ln [({\displaystyle \frac{Y_i}{Y}})/({\displaystyle \frac{L_i}{L}})]$$

(2)

where $Y_i$ and $L_i$ represent the added value and the number of employees in the ith industry, respectively.

3.1.2. The Core Independent Variables

Regarding the measurement of the integration level between the digital economy and the real economy, this research follows the common practice in existing studies and uses the coupling coordination degree model for measurement. It refers to Zhao et al. (2020) for the construction of a digital economy subsystem, which includes digital infrastructure, digital applications, and digital innovation as secondary indicators. It refers to Q. Huang (2017) for structuring a real economy subsystem constructed with economic vitality, trade activity level, and economic scale as indicators, as shown in

Table 1. Comprehensive evaluation of the digital economy and the real economy.

First-Level Indicator	Second-Level Indicator	Third-Level Indicator
Digital Economy	Digital Infrastructure	Length of long-distance optical cables per unit area
		Number of Internet access ports
		Mobile phone penetration rate (per 100 people)
	Digital Applications	Total volume of postal services (billion yuan)
		Total volume of telecommunications services (billion yuan)
		Number of websites per 100 enterprises
		E-commerce transaction volume (billion yuan)
		Number of information technology enterprises
		The proportion of enterprises engaged in e-commerce transactions
		Software business revenue (billion yuan)
	Digital Innovation	R&D expenditure of industrial enterprises above the designated size (ten thousand yuan)
		Number of IPv4 addresses (million)
		Number of patent applications granted
Real Economy	Economic Vitality	Total retail sales of consumer goods (billion yuan)
	Trade Activity Level	Total import and export volume of goods (billion yuan)
	Economic Scale	Gross production value (excluding finance and real estate)

.

First, following X. Wang et al. (2024), this paper uses the entropy weight method to calculate the comprehensive scores of the digital and real economies. Then, the coupling coordination degree model is used to calculate the integration level of the digital economy and the real economy, with the calculation process as follows:

$$C_{it}={\displaystyle \frac{2\sqrt{R{E}_{it}\times D{E}_{it}}}{R{E}_{it}+D{E}_{it}}}$$

(3)

where $R{E}_{it}$ and $D{E}_{it}$ represent the comprehensive scores of the digital economy and the real economy calculated by the entropy weight method for the t^th period, and $C_{it}$ represents the degree of coordination between the digital economy and the real economy for the t^th period. Considering the possibility that both $R{E}_{it}$ and $D{E}_{it}$ may be low while the value of C is large, to truly reflect the coupling level between the two, a systemic coupling coordination degree model for the digital economy and the real economy is constructed using the following formula:

$$D{R}_{it}=D_{it}=\sqrt{C_{it}\times T_{it}}$$

(4)

$$T_{it}=\alpha R{E}_{it}+\beta D{E}_{it}$$

(5)

where $D{R}_{it}$ represents the coupling coordination level of the digital economy and the real economy for the t^th period, which is used as a proxy variable for DRI. $T_{it}$ represents the comprehensive evaluation index of the digital economy system and the real economy system, $\alpha +\beta =1$. Based on the equal importance of DRI and the method proposed by Cui and Feng (2024), both α and β are set to 0.5. The level of DRI ($D{R}_{it}$) measured in this paper ranges from 0 to 1.

3.1.3. The Mechanism Variables

(1)

Social Consumption Demand (CON): Total retail sales of consumer goods refer to the sum of all consumer goods sold to end consumers through diverse channels, with the supply side covering wholesale and retail trades, catering services, and other relevant industries. This indicator reflects the market scale of the retail sector and the growth in social consumption demand at the current stage of development. The level of social consumption demand is measured by the ratio of total retail sales of consumer goods to regional GDP.

(2)

Degree of Factor Distortion (FAC): Following the methodology proposed by Bai and Liu (2018), this study estimates the factor misallocation indices for both capital and labor across different regions. The overall degree of factor distortion in each region is then measured by taking the average of these two indices.

(3)

Degree of Marketization (MAR): This study uses the Marketization Index developed by Fan et al. (2011) as a proxy for the degree of marketization, where higher index values signify more robust market-driven mechanisms in resource allocation.

3.1.4. The Threshold Variables

The development of the digital economy and the expansion of the real economy both influence industrial structure. Moreover, given regional disparities in the digital and real economies, the impact of DRI on ISU is likely to be nonlinear. For this reason, this paper uses the digital economy and the real economy as threshold variables to empirically test the effect of DRI.

3.1.5. The Control Variables

To reduce the estimation bias caused by the omission of variables and improve the robustness and reliability of the model estimation results, considering that factors such as government, technology market, and international market may also affect ISU, this paper selects openness level (Open), technology market development level (Tech), tax burden level (Tax), informatization level (Infm), labor level (Labor), and government intervention level (Gov) as control variables.

3.2. Model Construction

3.2.1. Basic Regression Model

To examine the impact of DRI on ISU, this paper constructs a basic regression equation to reveal the causal relationship, as follows:

$$RO{I}_{it}={\alpha }_0+{\alpha }_{1}D{R}_{it}+{\alpha }_j{contols}_{it}+u_i+{\lambda }_t+{\epsilon }_{it}$$

(6)

where $RO{I}_{it}$ represents the situation of ISU in the i^th province during the t^th period; $D{R}_{it}$ represents the situation of DRI in the i^th province during the t^th period; ${contols}_{it}$ represents a series of control variables; $u_i$ represents individual fixed effects of the corresponding provinces; $u_i$ represents a random disturbance term; and the coefficient value $\alpha$ represents the degree to which DRI improves or worsens the industrial structure.

3.2.2. Mediating Effect Model

Drawing on the research of T. Jiang (2022), a mediating effect model is constructed using the following equation:

$$Me{d}_{it}={\beta }_0+{\beta }_{1}D{R}_{it}+{\beta }_j{controls}_{it}+u_i+{\lambda }_t+{\epsilon }_{it}$$

(7)

where $Me{d}_{it}$ represents the mediating variable, and the specific measurement of each mediating variable is described in the previous text; the coefficient value represents the degree to which DRI improves or worsens the social consumption level.

3.2.3. Panel Threshold Model

Based on Equation (5), this study employs a panel threshold model to further explore the nonlinear effects of the integration of the digital economy and the real economy on ISU (as shown in Equation (8)) and to determine the threshold values of the nonlinear relationship. The threshold values are set in the variables for threshold effect testing.

$$RO{I}_{it}={\delta }_0+{\delta }_{1}D{R}_{it}\times I\left(T{h}_{it}\le \gamma \right)+{\delta }_{2}D{R}_{it}\times I\left(T{h}_{it}>\gamma \right)+{\delta }_j{contols}_{it}+u_i+{\lambda }_t+{\epsilon }_{it}$$

(8)

Here, $T{h}_{it}$ represents the estimated value of the panel threshold variables, such as the development of the digital economy and the expansion of the real economy, that pass the test. $I$ is the indicator function, which takes the value of 1 if the threshold variable in the parentheses meets the standard, and 0 otherwise.

3.2.4. Spatial Econometrics Model

ISU can reflect a region’s industrial layout, talent, and factor allocation. With the elimination of inter-regional barriers, the degree of inter-provincial and inter-city economic linkage gradually increases, and there is a significant spatial spillover effect of ISU between provinces. The degree of numerical and real integration and control variables affecting ISU in this region may have a positive impact on ISU in neighboring regions. Therefore, this paper introduces the spatial lag terms of ISU, DRI, and control variables based on Equation (5), and sets up the model as follows:

$$RO{I}_{it}=\rho {\sum }_{j=1}^n{w}_{ij}RO{I}_{it}+{\alpha }_{1}D{R}_{it}+{\alpha }_j\sum {contols}_{it}+\phi {\sum }_{j=1}^n{w}_{ij}D{R}_{it}+\varphi {\sum }_{j=1}^n{w}_{ij}{contols}_{it}+u_i+{\gamma }_t+{\epsilon }_{it}$$

(9)

where $\rho$ is the spatial autoregressive coefficient, $w_{ij}$ represents the spatial weight matrix, and $\phi$ and $\varphi$ are the coefficients of the spatial lag terms of the numerical real fusion and control variables. The paper employs an economic distance matrix as the spatial weight matrix, which is calculated using the reciprocal of the absolute difference between the average per capita GDP of two provinces during the sample investigation period.

4. Empirical Analysis

4.1. Benchmark Regression

Table 2. Results of the benchmark tests.

	(1)	(2)	(3)	(4)	(5)	(6)
Variable	Theil	Theil	Theil	RIS	RIS	RIS
DR	−0.3989 ***	−0.5637 ***	−0.4872 ***	0.3710 ***	0.2662 ***	0.2579 ***
	(−11.71)	(−8.39)	(−5.33)	(10.23)	(3.81)	(3.75)
Controls	No	No	Yes	No	No	Yes
Fixed Effects	No	Yes	Yes	No	Yes	Yes
N	330	330	330	330	330	330
$R_{adj}^2$	0.2927	0.7055	0.7279	0.2471	0.9582	0.9703

Notes: () values represent the t value; *** indicate significance levels of 1%.

reports the baseline regression results of the impact of digital–real economy integration on ISU. Columns (1)–(3) examine the impact of digital–real economy integration on industrial structure rationalization, presenting the results of mixed OLS regression, fixed effects regression, and regression with control variables, respectively. The coefficients of DRI are −0.3989, −0.5637, and −0.4872, all significant at the 1% level, indicating that DRI can significantly promote industrial structure rationalization. Columns (4)–(6) report the impact of DRI on ISU, with coefficients of 0.3710, 0.2662, and 0.2579, all significant at the 1% level, suggesting that digital–real economy integration drives ISU. This means that, assuming all other factors are held constant, a one-unit increase in digital readiness (DR) will result in a 0.4872-unit improvement in industrial structure rationalization and a 0.2579-unit enhancement in industrial structure advancement. Empirical analysis indicates that DRI has a significant positive effect on ISU, supporting Hypothesis 1.

4.2. Robustness Test

4.2.1. Substitute Dependent Variable

To test the robustness of the above empirical results, this study uses the degree of industrial structure deviation (ER) as a substitute for the rationalization of industrial structure. Following the method of Gan Chunhui et al., the index of industrial advancement (HIS) is calculated, and regression analysis is conducted. As shown in panel A column (1) and panel B column (1) of

Table 3. Results of the robustness tests.

	Replace the Explained Variable.	Outlier Year	Exclude Municipalities	Outliers Eliminator
Panel A: Industrial structure rationalization
	(1)	(2)	(3)	(4)
Variable	ER	Theil	Theil	Theil 95
DR	−2.3798 ***	−0.4646 ***	−0.4800 ***	−0.5193 ***
	(−4.57)	(−4.16)	(−3.67)	(−7.48)
controls	YES	YES	YES	YES
Province FE	YES	YES	YES	YES
Year FE	NO	NO	NO	NO
N	330	300	286	297
R2	0.7839	0.7684	0.7287	0.8204
Panel B: industrial structure advancement
	(1)	(2)	(3)	(4)
Variable	HIS	RIS	RIS	RIS95
DR	11.5908 **	0.2402 ***	0.3158 ***	0.2550 ***
	(2.10)	(3.32)	(3.90)	(3.96)
controls	YES	YES	YES	YES
Province FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
N	330	300	286	297
$R_{adj}^2$	0.9089	0.9692	0.9084	0.9321

Notes: () values represent the t value; **, and *** indicate significance levels of 5%, and 1%, respectively.

, the coefficients of industrial rationalization and industrial advancement are −2.3798 and 11.5908, respectively, and both are significant at the 5% level, further confirming the promoting role of DRI in ISU.

4.2.2. Exclude Abnormal Years

Considering that the outbreak of the pandemic may have affected the normal operation of the economy and thus impacted the credibility of the regression results, this study excludes the 2021 data and conducts the regression analysis again. As shown in panel A column (2) and panel B column (2) of Table 3, the coefficients of DRI on ISU after excluding abnormal years are −0.4646 and 0.2402, respectively, and both are significant at the 1% level, indicating that the results of the baseline regression are robust.

4.2.3. Exclude Directly Administered Cities

Given that the policies, economic positioning, development models, and development levels of directly administered cities differ significantly from those of other provinces and cities, this may lead to biased results; therefore, this study excludes Beijing, Tianjin, Shanghai, and Chongqing, the four directly administered cities, from the sample and conducts the regression analysis again. As shown in panel A column (3) and panel B column (3) of Table 3, the coefficients of DRI are −0.4800 and 0.3158, respectively, and both are significant at the 1% level, indicating that the impact of DRI on ISU remains significant even after excluding a portion of the sample.

4.2.4. Exclude Outliers

To reduce noise and redundant information in the data, this study truncates the industrial structure rationalization and industrial advancement indicators by 5% on both sides and conducts regression analysis again. As shown in panel A column (4) and panel B column (4) of Table 3, DRI has a significant impact on industrial structure rationalization and industrial advancement at the 1% level.

4.3. Mechanism Analysis

To verify hypotheses H2, H3, and H4, regression analysis was performed on Equation (7).

The results of regression analysis in column (1) of

Table 4. Results of the mechanism tests.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
Variable	CON	Theil	RIS	FAC	Theil	RIS	MAR	Theil	RIS
DR	0.9367 ***			−0.3119 **			2.7614 *
	(6.39)			(−2.10)			(1.73)
CON		−0.367 ***	0.106 ***
		(−4.68)	(3.15)
FAC					−0.0905 *	−0.0891 **
					(−1.87)	(−2.12)
MAR								−0.0146 **	0.0229 ***
								(−2.31)	(6.87)
controls	YES	YES	YES	YES	YES	YES	YES	YES	YES
Province FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
N	330	330	330	330	330	330	330	330	330
$R_{adj}^2$	0.6397	0.7924	0.9700	0.8990	0.8657	0.9169	0.9459	0.8280	0.9288

Notes: () values represent the t value; *, **, and *** indicate significance levels of 10%, 5%, and 1%, respectively.

show that the impact of the DRI on the level of social consumption demand is significant at the 1% level, indicating that DRI promotes the growth of consumption demand. The results in columns (2) and (3) clearly show that the coefficient of social consumption demand level is significant at the 1% level, indicating that an increase in social consumption can promote the rationalization and advancement of industrial structure, which suggests that the increase in consumption demand promotes industrial restructuring. The integration of digital and real economies has an indirect impact on industrial restructuring, thus verifying hypothesis H2.

As shown by the results in column (4), the DRI significantly reduces the degree of factor distortion at the 5% level, which indicates that the integration of digital and real economies plays a positive role in the effective allocation of resource factors. The results in column (5) indicate that the coefficient of the factor allocation structure is negative and statistically significant at the 10% level. This suggests that factor misallocation contributes to the rationalization of the industrial structure to some extent. Meanwhile, the coefficient of factor misallocation in column (6) is negative and statistically significant at the 5% level, indicating that factor misallocation hinders the advancement of the industrial structure. These findings support the logical chain of “digital-real integration → factor misallocation → industrial structure upgrading”. This implies that digital–real integration can influence the upgrading of the industrial structure by improving the factor allocation structure; thus, research hypothesis H3 is verified.

The results in column (7) show that the regression coefficient of the marketization process is 2.7614 and significant at the 10% level, which indicates that the integration of digital and real economies has a significant effect on accelerating the process of marketization. As shown in columns (8) and (9), the coefficients of the marketization index are −0.0146 and 0.0229, respectively, and both are statistically significant at least at the 5% level. These results suggest that an increase in the level of marketization not only promotes the rationalization of the industrial structure but also drives its advancement. This confirms the validity of the logic chain: “digital-real integration → factor marketization → industrial structure upgrading”. Based on the above analysis, Hypothesis 4 is validated.

4.4. Endogeneity Treatment

A potential endogeneity problem may exist in the model; thus, instrumental variable methods were employed for two-stage least squares regression analysis.

First, following the approach of Q. Huang et al. (2019), the number of telephone lines and postal services were used as instrumental variables. As shown in columns (1) and (2) of

Table 5. The results of endogeneity treatment.

	IV1: Number of Telephone Lines, Postal Services		IV2: Bartik Instrumental Variable
	(1)	(2)	(3)	(4)
	Theil	RIS	Theil	RIS
DR	−0.6883 ***	0.8801 ***	−0.2604 **	0.5513 ***
	(−3.98)	(11.72)	(−2.44)	(8.05)
Controls	YES	YES	YES	YES
Fixed Effects	YES	YES	YES	YES
Kleibergen–Paap rk LM	89.3270	89.3270	91.065	91.065
	[0.0000]	[0.0000]	[0.0000]	[0.0000]
Kleibergen–Paap rk Wald F	469.834	469.834	2621.856	2621.856
	{19.93}	{19.93}	{16.38}	{16.38}
N	330	330	300	300
$R_{adj}^2$	0.2358	0.6818	0.2055	0.6719

Notes: () values represent the t value; **, and *** indicate significance levels of 5%, and 1%, respectively.

, the p-values of the Kleibergen–Paap rk LM test are both less than 0.01, passing the under-identification test of instrumental variables. The Kleibergen–Paap rk Wald F statistic is greater than the critical value of 19.93, passing the weak instrumental variable test. This suggests that using the number of telephone lines and postal services as instrumental variables is reasonable. The regression coefficients of using the number of telephone lines and postal services as instrumental variables for the integration of digital and real economies are −0.6883 and 0.8801, respectively, and both are significant at the 1% level, indicating that, under the consideration of endogeneity, the DRI has a positive effect on industrial restructuring.

Second, the Bartik instrumental variable was constructed following the approach of Goldsmith-Pinkham et al. (2020). As shown in columns (3) and (4) of Table 5, the Bartik instrumental variable passes the under-identification and weak instrumental variable tests. The coefficients of the instrumental variable regression results are −0.2604 and 0.5513, respectively, and both are significant at the 5% level, further confirming the positive effect of the DRI on industrial restructuring. The results of the endogeneity test further illustrate the robustness and reliability of the conclusion that the integration of digital and real economies promotes industrial restructuring.

These studies provide empirical support for the theoretical frameworks proposed by Han et al. (2023) and W. Liu et al. (2024), while also serving as a form of cross-validation. They reinforce both the methodological rigor of our investigation and the representativeness of our sampled population.

5. Further Analysis

5.1. Test of Regional Heterogeneity

The unbalanced development characteristics between regions in China may influence the effect of the DRI on ISU due to regional distribution characteristics; therefore, this study divides 30 provinces and cities into eastern, central, and western to explore the spatial heterogeneity of the impact of DRI on ISU.

Table 6. Results of the heterogeneity tests.

	Eastern		Central		Western
	(1)	(2)	(3)	(4)	(5)	(6)
Variable	Theil	RIS	Theil	RIS	Theil	RIS
DR	−0.1406	−0.1516 **	−2.2485 ***	0.9831 ***	−1.3886 *	0.4256 ***
	(−0.89)	(−2.06)	(−4.34)	(5.03)	(−1.72)	(3.46)
Controls	YES	YES	YES	YES	YES	YES
Province FE	YES	YES	YES	YES	YES	YES
Year FE	YES	YES	YES	YES	YES	YES
N	121	121	88	88	121	121
$R_{adj}^2$	0.6154	0.9919	0.7358	0.9495	0.7710	0.8980

Notes: () values represent the t value; *, **, and *** indicate significance levels of 10%, 5%, and 1%, respectively.

reports the impact of DRI on ISU in different regions. Columns (1) and (2) show the impact on the eastern region, with the regression coefficient of DRI on industrial structure rationalization being −0.1406, but not significant, which suggests that DRI did not have a significant impact on industrial structure rationalization in the eastern region. The coefficient for the industrial structure advanced is −0.1516, significant at the 5% level, indicating that DRI has reduced the level of industrial structure advanced in the eastern region. This is because in eastern China, the large-scale digital economy, rational industrial structure, and developed tertiary industry have weakened the marginal effect of DRI. Data from the Ministry of Industry and Information Technology and the National Bureau of Statistics show that in 2021, the eastern digital economy development index reached 7818.25, over 2.5 times that of the central and western regions. Most eastern areas have a service sector GDP share above the national average of 53.3%, e.g., Beijing (81.7%), Shanghai (73.3%), and Tianjin (61.3%). Meanwhile, China’s 14th Five-Year Plan reverses the service-sector priority approach through its “stable manufacturing share” objective, redirecting resources to manufacturing and driving sectoral restructuring marked by service contraction and industrial agglomeration.

Columns (3) and (4) of Table 6 show the impact of DRI on ISU in the central region, with coefficients of −2.2485 and 0.9831, both significant at the 1% level. Columns (5) and (6) show the impact on the western region, with regression coefficients of −1.3886 and 0.4256 for industrial structure rationalization and industrial structure advanced, respectively, significant at the 10% level. The results indicate that DRI can promote ISU in the central and western regions, with a stronger effect in the central region. The reason is that the national strategy for the rise of the central region emphasizes the deep integration of scientific and technological innovation with industrial innovation, which helps to cultivate new drivers of economic growth. The traditional manufacturing industry accounts for a large proportion of the central region, with abundant talent resources and a good talent mobility mechanism. Statistics reveal that the number of regional R&D personnel increased from 632,000 to 855,000 person-years between 2015 and 2019, achieving an annual growth rate of 7.8%, which exceeds the national average. Notably, personnel growth rates in basic research, applied research, and experimental development all outpaced those in other regions. This robust talent ecosystem provides fertile ground for the penetration of the digital economy and further promotes the development of DRI, thereby enhancing the effect of ISU.

5.2. Empirical Test of Threshold Effect

To determine whether there is a threshold effect and the number of thresholds in the impact of the level of DRI on ISU, the Bootstrap method was used for repeated sampling (300 times); the test results are shown in

Table 7. Results of the threshold effect test.

	Variable	Thresh Value	Threshold	F-Statistic	p-Value	Critical Value
						10%	5%	1%
Theil	REAL	0.0298	Single	64.84 ***	0.010	37.710	41.911	56.853
		0.2699	Double	25.64	0.187	34.440	48.217	74.386
	DIGT	0.0179	Single	35.44 **	0.050	31.603	35.435	57.408
		0.0341	Double	27.73 *	0.087	26.925	30.182	39.917
		0.2852	Triple	15.38	0.663	33.271	39.168	50.483
RIS	REAL	0.0481	Single	33.68 *	0.090	31.875	38.455	48.838
		0.0279	Double	12.50	0.583	30.931	35.615	48.602
	DIGT	0.3112	Single	69.95 ***	0.000	34.238	38.417	50.310
		0.0678	Double	24.63	0.127	25.787	28.705	47.368

Notes: () values represent the t value; *, **, and *** indicate significance levels of 10%, 5%, and 1%, respectively.

.

Table 7 shows that when the real economy is used as the threshold variable, there is a single threshold for both industrial structure rationalization and industrial structure advancement, with threshold values of 0.0298 and 0.0481, respectively. When the digital economy is used as the threshold variable, there is a double threshold for the impact on industrial structure rationalization, with threshold values of 0.0179 and 0.0341, and a single threshold for the impact on industrial structure advanced, with a threshold value of 0.3112. To test the consistency between the estimated and actual threshold values, an LR likelihood estimation graph is drawn, which shows that the single threshold of the real economy is below the LR benchmark value (7.35) at the 95% confidence level, confirming the consistency of the estimated and actual threshold values. The estimated values of the single and double thresholds of the real economy are both below the LR benchmark value at the 95% confidence level, confirming the consistency of the threshold values and the actual values.

As shown in

Table 8. Results of the threshold effect tests.

	Theil			RIS
DR × I(ST < 0.0298)	0.6586 ***		DR × I(ST < 0.0481)	0.7608 ***
	(3.84)			(16.45)
DR × I(ST > 0.0298)	−0.4951 ***		DR × I(ST > 0.0481)	0.6658 ***
	(−5.98)			(15.58)
DR × I(DE < 0.0179)		0.6467 ***	DR × I(DE < 0.3112)		0.7839 ***
		(3.77)			(18.05)
DR × I(0.0179 < DE < 0.0341)		−0.0735	DR × I(DE > 0.3112)		0.6802 ***
		(−0.66)			(16.70)
DR × I(DE > 0.0341)		−0.3679 ***
		(−4.35)
Controls	YES	YES		YES	YES
N	330	330		330	330
$R_{adj}^2$	0.5029	0.5460		0.7336	0.7585

Notes: () values represent the t value; *** indicate significance levels of 1%.

, there exists a nonlinear relationship between DRI and industrial structure rationalization. When the level of real economy development is less than 0.0298, the coefficient of DRI is 0.6586, significant at the 1% level, indicating that DRI inhibits the rationalization of industrial structure. When the level of real economy development exceeds 0.0298, the coefficient of DRI is −0.4951, significant at the 1% level, which means that DRI will improve industrial structure rationalization. For example, in Germany’s industrial evolution, early digital–physical integration faced resistance from traditional manufacturing rigidity, hindering structural rationalization. As Industry 4.0 matured, data-driven supply chains and smart factories emerged, reshaping sectors like automotive and machinery into high-value clusters, exemplifying the U-shaped trajectory. When the level of digital economy development is lower than 0.0179, the coefficient of DRI is 0.6467, significant at the 1% level. When the level of digital economy development is between 0.0179 and 0.0341, the coefficient of DRI is −0.0735, but not significant. When the level of digital economy development exceeds 0.0341, the coefficient of DRI is −0.3679, significant at the 1% level. This is because at low levels of digital economy, traditional industries face “digital divide” and “technological lock-in” due to weak digital infrastructure, lagging management, and skill shortages. Compounded by cost–benefit inversion and scenario adaptability gaps, this phenomenon worsens supply–demand imbalances and resource misallocation. Once the digital economy scales past a threshold, network externalities and positive feedback drive industrial rationalization. The above empirical results indicate that when the level of digital and real economy development is low, DRI will inhibit the adjustment of industry to a rational structure; however, with the development of the digital and real economies, DRI will promote the rationalization of industrial structure, showing a “U” shape characteristic. The empirical results indicate that in areas with underdeveloped digital economies or in the early stages of digital economic development, the DRI may hinder industrial structure rationalization. Conversely, in regions with more advanced digital economies or when the digital economy reaches a higher stage of development, the DRI can effectively promote industrial structure rationalization.

There also exists a nonlinear relationship between DRI and industrial structure advancement; when the level of real economy development is less than 0.0481 and the coefficient of the DRI index is 0.7608, both are significant at the 1% level. When the level of real economy development exceeds 0.0481, the coefficient of the DRI index is 0.6658, significant at the 1% level. When the level of digital economy development is less than 0.3112, the coefficient of the DRI index is 0.7839, and when the level of digital economy development is greater than 0.3112, the coefficient of the DRI index is 0.6802; both are significant at the 1% level. The above empirical results show that as the digital and real economies develop, the coefficient of DRI gradually decreases, indicating that the impact of DRI on industrial structure advancement shows a marginal diminishing effect; thus, research hypothesis H5 is verified.

5.3. Empirical Test of Spatial Spillover Effect

To verify the spatial spillover effect of DRI, this paper first uses the global Moran index to test the spatial autocorrelation of industrial structure advancement and industrial structure rationalization in 30 provinces in China. The global Moran index for each year is greater than 0 and significant at the 1% level, indicating that the distribution of industrial structure advanced and industrial structure rationalization has a spatial correlation. To select an appropriate spatial econometric model, the LM test, the Hausman test, the LR test, and the Wald test are conducted sequentially before regression analysis. The results of these tests indicate that the spatial error model and the spatial Dubin model should be used for regression analysis of industrial structure rationalization and industrial structure advancement, respectively.

As shown in

Table 9. The results of spatial spillover tests.

	Theil	RIS
	X	X	W×X	X(SEM)	X(SAR)
DR	−0.3626 ***	0.3710 ***	0.1950 *	0.4698 ***	0.3365 ***
	(−2.90)	(5.89)	(1.84)	(8.20)	(7.19)
$\rho$		0.4871 ***			0.5477 ***
		(7.31)			(10.82)
$\lambda$	0.7202 ***			0.7105 ***
	(18.83)			(13.77)
Log-like	533.6656	818.0483		794.1949	791.2875
controls	YES
Fixed Effects	YES
N	330
$R_{adj}^2$	0.2120	0.7992		0.6636	0.7179

Notes: () values represent the t value; * and *** indicate significance levels of 10% and 1%, respectively.

, the DRI has an improvement effect on ISU. For industrial structure rationalization, the spatial error term coefficient is 0.7202, significant at the 1% level, indicating significant spatial autocorrelation. This suggests that the rationalization of industrial structure in neighboring cities has a spatial spillover effect on the rationalization of industrial structure in this city. For an advanced industrial structure, the coefficient of DRI is 0.3710, significant at the 1% level. The coefficient of neighboring city DRI is 0.1950, significant at the 10% level, indicating that the development of DRI in neighboring cities can promote the advanced industrialization of this city. The regression coefficient of the spatial Dubin model is 0.4871, significant at the 1% level, indicating that the advanced industrialization of neighboring cities can enhance the advanced industrialization of this city. The spatial lag model and the spatial error model also support this conclusion; thus, research Hypothesis H6 is verified.

5.4. Research Significance

This paper integrates DRI and IUS into a unified analytical framework. It provides insights into the application of DRI and offers new perspectives for exploring the factors influencing IUS. Empirical testing using econometric analysis methods extends the theoretical research on the DRI and IUS to the empirical level, providing empirical evidence for theoretical development. The use of panel threshold models and spatial econometric models to explore the impact of the DRI on IUS addresses the shortcomings of existing research. This research has previously focused solely on linear causal relationships while neglecting nonlinear effects and spatial dependencies. Therefore, this study provides a theoretical basis for policy formulation to promote IUS.

6. Conclusions and Implications

6.1. Conclusions

This paper explores the impact, mechanism, and nonlinear impact of DRI on ISU using panel data collected from 30 Chinese provinces and cities from 2012 to 2022. The research shows that DRI has a significant positive effect on ISU. After robustness tests and endogeneity treatment, DRI still has a direct impact on ISU, further verifying the authenticity of the conclusion. In the heterogeneity analysis, this paper finds that there are differences in the impact of DRI on ISU among provinces, with the effect in the central region being significantly better than that in the eastern and western regions. DRI indirectly affects ISU through the level of consumer demand, the degree of factor distortion, and the process of marketization. In the threshold effect test, the nonlinear impact of DRI on ISU with the level of digital economy development and the scale of the real economy as threshold variables is verified; furthermore, the spatial correlation test shows that there is interaction among provinces, and ISU has spatial spillover.

6.2. Policy Implications

Based on the above findings, DRI is an important driving force for upgrading the industrial structure, and it can work by promoting consumption, correcting factor mismatch, and accelerating marketization. There also exists a threshold effect and a spatial spillover effect on DRI to empower the upgrading of the industrial structure. These results are of importance in accelerating the upgrading of the industrial structure. The main policy implications are as follows:

Firstly, promoting the development of the digital economy in multiple dimensions and accelerating the deep integration of the digital economy with the real economy. Provinces should pay attention to the construction of digital economy infrastructure such as sensor terminals, 5G networks, big data centers, artificial intelligence, etc., promote the innovation and application of digital technologies to enhance the penetration capacity of digital technologies into the real economy, and lay a solid material foundation and technological support for the in-depth fusion of the digital economy and the real economy. It is necessary to accelerate the promotion of the in-depth integration and application of data elements and digital technologies in industries to empower the digital transformation of traditional industries and stimulate their innovative vitality.

Secondly, strengthening the government’s positioning as a leader and supporter, implementing a “policy toolkit” to foster a robust DRI ecosystem. During industrial dataization, the government should provide fiscal incentives such as tax cuts and R&D super-deductions, reduce digital transformation costs, establish an inclusive yet prudent regulatory framework, promote flexible DRI while monitoring risks, build an open and shared data platform, break down the information silos, promote the reasonable flow and efficient allocation of data resources, and provide a broader development space and rich resource support for the formation and development of the modern industrial system.

Thirdly, balancing the development of the digital economy and the real economy, and expanding the path of DRI to empower the upgrading of the industrial structure. Consolidate the mechanism of the digital economy in expanding the level of social consumption demand, improving the distortion of factor allocation, promoting the marketization process and other mechanisms to promote the industrial structure, use digital technology to innovate the consumption mode, broaden the consumption path, build personalized and convenient consumption scenarios for consumers, and satisfy the growing diversified needs of consumers. Make full use of the flexibility and innovativeness of the digital economy and improve the distortion of factor allocation. The flexibility and innovation of the digital economy will be fully utilized to improve the distortion of factor allocation, so that more capital elements will be invested in the real economy and innovation activities; the dividends of the digital economy will be tapped to accelerate the marketization process, and the vitality and advantages of the socialist market economy will be better utilized, so as to promote the evolution of the industrial structure towards rationalization and advanced. Starting from policy guidance and support, industry chain integration, supply chain management, digital transformation of enterprises, and other dimensions, we will further explore the “black box” of DRI to promote the upgrading of industrial structure and inject new vitality into the construction of the modern industrial system.

Fourthly, to strengthen regional coordination, enhance the eastern region’s leading and exemplary role, reinforce the central region’s pivotal role, intensify efforts to foster a favorable economic development environment in the western region, and jointly promote industrial structure upgrading across regions. For the eastern provinces, it is crucial to fully leverage their regional advantages, reduce regional barriers, and enhance joint development and sharing with neighboring provinces in technology, resources, and human capital. For the central provinces, efforts should focus on facilitating the flow of technology, resource exchanges, and talent mobility to promote synchronized industrial upgrading across regions. As for the western provinces, the Party and the state should leverage their role as “Eastern Data, Western Computing” hub nodes to accelerate new infrastructure, such as 5G base stations and data centers, thereby enhancing digital infrastructure capabilities. Utilizing eastern digital technology spillovers, central and western regions could establish cross-regional technology transfer platforms and collaborative industrial parks, forming an innovation ecosystem of “basic R&D—application scenarios—industrial incubation”, to accelerate the construction of a modern industrial system.

In the current context of rapidly evolving digital technologies, the convergence of the digital and real economies is an inevitable trend. This presents developing countries with a new path for development through leapfrogging. As the DR-driven ISU effect becomes increasingly apparent in China, other developing countries can incorporate China’s experience to accelerate the modernization and upgrading of their industrial systems.

6.3. Research Limitations and Further Studies

This paper examines the influence of digital–physical integration on industrial structure upgrading. When developing a digital–physical integration indicator system, it is essential to take into account both the digital economy and the real economy subsystems; however, in constructing the real economy sub-indicator system, this paper mainly focuses on output, imports and exports, and social retail sales, which may not fully capture the development of the real economy. Therefore, the real economy sub-indicator system should be further enhanced, particularly by incorporating industry-related indicators.

In terms of application research on the integration of digital and real economies, this paper analyzes industrial structure upgrading at the macro level. However, whether the integration of digital and real economies affects other macroeconomic variables remains to be explored. Additionally, the mechanisms and practical effects of this integration in promoting the transformation and upgrading of enterprises, especially manufacturing enterprises, require further investigation.

Regarding the measurement of the degree of digital–physical integration, current methods are primarily confined to the macro level. Developing a subsystem for the integration of the digital economy and the real economy at the micro-enterprise level will be an important direction for future research.