Research on the impact of smart logistics on the the manufacturing industry chain resilience

Abstract

The development history of smart logistics and smart manufacturing in China demonstrates a high degree of synergy. This study aims to analyze the impact, mechanisms, and heterogeneous performance of smart logistics on the manufacturing industry chain resilience. The analysis is based on panel data collected from 30 provinces in China, covering the period from 2012 to 2023. The empirical findings suggest a significant positive impact of smart logistics on the manufacturing industry chain resilience, and the research findings are relatively robust. This impact can be primarily attributed to the reduction of transaction costs and improvements in logistics efficiency. In a comparison of the impact relationships between the three different economic regions of the East, Center and West, it was found that the Eastern region had a stronger facilitating effect, the Central region had a less pronounced facilitating effect, and the Western region had a weaker facilitating effect relative to the Eastern region. Moreover, Intelligent logistics has a greater role in promoting the resistance and renewal of the manufacturing industry chain, and a relatively weaker role in promoting the recovery of the manufacturing industry chain. Futhermore, threshold test shows that, in terms of long-term dynamics, there is a threshold effect of smart logistics development on the enhancement of manufacturing industry chain toughness; at the same time, higher level of advanced industrial structure and innovation environment, smart logistics is conducive to the enhancement of manufacturing industry chain toughness level.

Introduction

Within the context of ongoing globalization, characterized by the deepening international division of labor and the global allocation of resource elements, the modern logistics system has emerged as a critical component deeply integrated within the supply chain. It plays a pivotal role in expediting innovation, industry upgrading, production, and circulation¹. As a result of the continuous development and extensive application of information technology, the logistics and supply chain industry is undergoing an intelligent transformation akin to an industrial revolution. Smart logistics encompasses a contemporary logistics model that intelligently manages and optimizes the entire logistics process by leveraging information technology, communication technology, Internet of Things (IoT) technology, data analytics, artificial intelligence, and other advanced tools. This approach enables real-time collection, transmission, processing, and feedback of logistics information, ultimately enhancing logistics efficiency, reducing costs, and improving service quality and users experience². Key features of smart logistics include informatization (real-time information sharing and data-driven decision-making), intelligence (automated operation and intelligent scheduling), integration (system integration and multimodal transportation), visualization (comprehensive visual monitoring and data visualization), greening (environmentally friendly logistics and intelligent recycling), and customer orientation (personalized service and omni-channel service)³.

China has a rich history of smart logistics development, with the government specifically emphasizing the application of smart logistics technology as a crucial driver for the transformation and upgrading of the modern manufacturing industry chain. Tables 1 and 2 highlight notable events and government policies in this regard.

Table 1.

Representative events in the development history of smart logistics in China.

Timing	Representative events
2003	CR Vanguard commenced the implementation of automatic sorting equipment.
2012	Consumer Testing Technology Co., Ltd. (CTT) provided a logistics and warehousing system for LG Chem’s Nanjing base.
2012	The footwear industry initiated the transformation of smart logistics equipment.
2016	Solution providers for express smart logistics equipment began large-scale shipments.
2016	JD Logistics, Inc. initiated the establishment of a smart logistics network.
2017	BlueSword successfully deployed its self-developed sorting system in VIPSHOP’s southwest warehouse.
2019	SF launched a pilot program for unmanned delivery vehicles in Shenzhen.
2020	JD Logistics, Inc. achieved complete unmanned delivery for the first time during the “Double 11” period.
2020	Cainiao Network implemented nationwide deployment of smart warehousing.
2021	JD Logistics, Inc. collaborated with China Mobile to achieve 5G network coverage in several warehouses.

Table 2.

Representative policies for empowering manufacturing development through smart logistics in China.

Timing	Representative policies
2017	The Guiding Opinions on Deepening “Internet + Advanced Manufacturing” and Developing Industrial Internet, issued by the State Council of China, aims to expedite the construction and development of the industrial internet, facilitate the integration of the internet, big data, artificial intelligence, and the real economy, foster advanced manufacturing, and support the optimization and upgrading of traditional industries.
2019	The Opinions on Promoting High-Quality Development of Logistics and Facilitating the Formation of a Strong Domestic Market, issued by China’s National Development and Reform Commission, proposes the establishment of a national logistics hub network, enhancement of the national public information platform for logistics, and implementation of initiatives for intelligent transformation of the logistics sector. These measures aim to facilitate the creation of a unified national market.
2020	China’s National Development and Reform Commission issued the “Implementation Plan for Promoting the Deep Integration and Innovative Development of the Logistics Industry in the Manufacturing Sector”, which encourages manufacturing enterprises to undertake intelligent transformations in logistics, promote the adoption of logistics robots, intelligent warehousing, and other new logistics technologies and equipment. The plan also strives to advance the deep integration and innovative development of the logistics industry within the manufacturing sector, reduce logistics costs, enhance efficiency, and promote the transformation and upgrading of the manufacturing industry.
2022	The Opinions of the Central Committee of the Communist Party of China and the State Council on Accelerating the Construction of a National Unified Market, issued by the State Council of China, emphasizes the need to construct a modern circulation network, optimize the infrastructure layout for commerce and trade circulation, expedite digitalization efforts, promote the convergence of online and offline development, and foster new platforms for commerce and trade circulation in diverse forms and modes.

Manufacturing industry chain resilience refers to the capacity of the manufacturing industry to rapidly respond, adapt to changes, maintain operational stability and continuity, and quickly recover to a normal state when faced with external shocks such as natural disasters, market fluctuations, policy changes, and supply chain disruptions. This is achieved through improving adaptability, flexibility, synergy, redundancy design, innovation, and risk management⁴. It represents an extension of the concept of “resilience” from physics to the research on manufacturing industry chains. The report of the Twentieth National Congress of the Communist Party of China (CPC) emphasizes the need to accelerate the construction of a modernized economic system, with a particular focus on enhancing total factor productivity and ensuring the resilience and security of industrial and supply chains. The outbreak of the novel coronavirus pneumonia (COVID-19) a few years ago severely disrupted global supply chains and industrial production in several countries, leaving a lasting impact⁵. Intelligent manufacturing heavily relies on the robust support of smart logistics, as smart logistics serves as the guarantee system for intelligent manufacturing⁶. In the context of China and the world, there is a growing urgency to accelerate the integration of the digital economy with the real economy and construct a modern industrial and supply chain system. Exploring the interactive relationship between smart logistics and manufacturing industry chain resilience holds significant theoretical and practical value. It is also crucial to examine the heterogeneous impact of the potential enabling effect of digital infrastructure.

The current difficulty in providing satisfactory answers to the above question can be attributed to several reasons. Firstly, existing research on smart logistics primarily focuses on concept definition, interpretation of its connotation⁷, and the impact of smart logistics on enhancing supply chain intelligence⁸. However, there is a limited amount of cross-industry engineering research specifically studying the direct influence of smart logistics on the development of the manufacturing industry. Secondly, there is a scarcity of studies examining the manufacturing industry chain resilience and its ability to withstand external shocks. The concept of manufacturing industry chain resilience has received less attention in research endeavors. Thirdly, although the development of China’s manufacturing industry and the adoption of smart logistics technology have shown a high degree of synergy in their evolution, it is possible that various favorable policies such as the intelligent manufacturing project, industrial Internet demonstration zones, and the establishment of manufacturing innovation centers have contributed to the growth of the manufacturing industry, rather than being solely driven by the empowerment of smart logistics. Moreover, there is a lack of clear empirical evidence to substantiate the facilitating effect of smart logistics on improving the manufacturing industry chain resilience.

Through the combing of related research, this paper clarifies the connotation of smart logistics, that is, smart logistics is a kind of intermediate input industry that connects production, sub-circulation, and consumption together, and is associated with industries such as agriculture, manufacturing, mining, and trade circulation, and has the characteristics of numerical intelligence, network, mobility, diversity, and evolution⁹. At the same time, this paper is based on the economic resilience of the evolution of resilience theory and related research results^10,11, the industry chain resilience is divided into industry chain resistance, recovery and renewal of three dimensions, specifically expressed as follows: resistance dimension mainly reflects the manufacturing industry chain in response to the pressure or impact to maintain their own stability, to prevent the industry chain from breaking the “stable chain” ability; recovery dimension mainly reflects the manufacturing industry chain in response to the pressure or impact to maintain their own stability, to prevent the industry chain breakage. The recovery dimension mainly reflects the ability of the manufacturing industry chain to adapt to pressure or shock and adjust and restore to the original state as soon as possible, so the size of the recovery depends on the speed and degree of the manufacturing industry chain’s adjustment and restoration to a stable state in the face of pressure or shock; the updating power dimension mainly reflects the ability of the manufacturing industry chain to realize the transformation and upgrading of the industry chain through updating the operation mode and opening up the industry chain to a new state. The dimension of renewal force mainly reflects the ability of the manufacturing industry chain to realize the transformation and upgrading of the industry chain through updating the operation mode and opening up new development paths, and whether it can realize the “path breakthrough” is the key to the improvement of the renewal force of the manufacturing industry chain.

Based on the above background analysis, this study focuses on the following core issues: first, how the intelligent development of the logistics industry works with the manufacturing industry chain; second, under the conditions of economic downturn or potential downturn, how to promote the smart logistics to effectively serve the manufacturing industry chain, in order to promote the manufacturing industry chain and product life cycle to achieve green development, enhance the industry’s basic innovation capacity, enhance the autonomy of the industry chain, promote the manufacturing industry chain resilience and enhance international competitiveness. Manufacturing industry chain resilience and enhance international competitiveness to accumulate power? Given these circumstances, this study aims to analyze the impact, mechanism, and heterogeneous performance of smart logistics on manufacturing industry chain resilience. The analysis is conducted using provincial panel data from China covering the period from 2012 to 2023.

Compared to existing studies, this paper aims to make three distinct contributions. Firstly, from the research perspective, the existing research mainly focuses on the impact of digital economy, industrial structure and technological innovation on the manufacturing industry chain resilience, and the research on the causal effect between smart logistics and manufacturing industry chain resilience is still in the early stage. Therefore, this paper takes China’s inter-provincial region as the research object, explores the role mechanism and path of smart logistics on the manufacturing industry chain resilience, analyzes the blank spot of the research on the industrial relationship between the smart logistics industry and the manufacturing industry, opens up the research ideas and analytical paradigm about the causal relationship between smart logistics and the manufacturing industry chain resilience, and provides a theoretical support for the further investigation of the impact of smart logistics development on the manufacturing industry chain resilience. Theoretical support.

Secondly, this paper introduces the mutation level model to measure the development level of the development level of manufacturing industry chain resilience, which expands the application of the mutation level model in the empirical research.

Thirdly, on the index measurement, this paper adopts 19 secondary indicators to characterize the manufacturing industry chain resilience in three dimensions of resistance, renewal and recovery, which overcomes the inadequacy of the existing literature on the manufacturing industry chain resilience.

Fourthly, on the mechanism path, this paper analyzes and verifies the mechanism and conduction path of smart logistics affecting the manufacturing industry chain resilience from two perspectives of transaction cost and logistics operation efficiency, deepening the understanding of the theoretical black box of both.

Fifthly, the third is that the policy recommendations closely match the theoretical analysis, empirical test results and the characteristics of the research object, and thus have more guiding significance and practical value.

The above possible innovations may help to fill the research gaps in the existing literature and provide new theoretical guidance for solving the impact of smart logistics on industry chain resilience in the manufacturing industry, and provide scientific decision-making basis for related policy formulation.

The remainder of the paper is structured as follows: The second part presents a theoretical analysis and research hypotheses. The third part outlines the research design. The fourth part presents the results of the empirical analysis. Finally, the last part concludes and discusses the findings.

Theoretical analysis and research hypothesis

Direct effect of smart logistics on manufacturing industry chain resilience

Smart logistics, as a strategic and fundamental industry and an innovative logistics management model, plays a crucial role in China’s supply-side structural reform and demand-side management optimization. It has the potential to enhance the stability and interconnectedness of each link within the manufacturing industry chain. The direct impact of smart logistics on the manufacturing industry chain resilience is primarily achieved through the knowledge spillover effect¹².

The concept of knowledge spillover refers to the phenomenon of knowledge transfer between complementary industries within the upstream and downstream segments of the industrial chain¹³. The application of smart logistics technology facilitates the sharing and dissemination of new theories and technical knowledge, resulting in a positive knowledge spillover effect that enhances the manufacturing industry chain resilience. This effect can be attributed to several factors.

Firstly, the knowledge spillover effect generates a chain reaction, driving force, exchange effect, and incentive effect that reduce the technological gaps between different segments of the manufacturing industry chain¹⁴. Consequently, it strengthens technical exchanges and collaborations among various subsectors.

Secondly, the advancement and adoption of smart logistics promote the development of digital capabilities within the manufacturing industry chain and supply chain ecosystem⁹. This, in turn, assists in establishing robust partnerships and building a more mature manufacturing industry ecosystem. As a result, the overall service functions and capabilities of the manufacturing industry are enhanced, enabling it to better meet the needs of high-end manufacturing, intelligent manufacturing, and other sophisticated and demanding customers in a timely and flexible manner.

Based on the aforementioned analysis, the following hypothesis is proposed in this paper:

Hypothesis 1

Smart logistics plays a positive role in promoting the manufacturing industry chain resilience.

Indirect effects of smart logistics on manufacturing industry chain resilience

(i) Smart logistics improves the manufacturing industry chain resilience by reducing transaction costs.

Reducing transaction costs is a key factor in enhancing the risk resistance of the manufacturing industry chain. Drawing on transaction cost theory¹⁵, this paper provides three explanations on how manufacturing enterprises can reduce transaction costs with the assistance of smart logistics.

Firstly, transaction frequency. Traditional logistics and manufacturing industry collaborations often involve outsourcing logistics services for the transportation of raw materials and finished goods¹⁶. Manufacturing enterprises typically invest significant time and effort in locating and establishing partnerships with logistics companies to ensure smooth operations. However, smart logistics enterprises can establish long-term strategic collaborations with manufacturing enterprises through network-based platforms. This significantly reduces the time and transaction costs associated with seeking logistics support and building stable supply relationships.

Secondly, uncertainty. In traditional business collaborations, unresolved matters that are not explicitly specified in contracts often require negotiation, primarily relying on trust¹⁷. With the advancements in digital technology and the emergence of smart logistics, the manufacturing industry can analyze changes in supply and demand and customer preferences more accurately. This reduction in information barriers enables manufacturing enterprises and industrial investors to improve their judgment accuracy and psychological expectations¹⁸. As the collaboration between the smart logistics-enabled manufacturing industry chain deepens, trust between the two industries strengthens, which further inhibits or reduces opportunistic behavior in transactions.

Thirdly, asset specialization. Long-term collaborations between manufacturing enterprises and smart logistics firms facilitate the specialization of each other’s assets¹⁹. Logistics companies can invest in specialized equipment such as transportation tools, warehousing facilities, and information systems based on the specific needs of manufacturing enterprises. Simultaneously, manufacturing enterprises can outsource product transportation, allowing them to allocate more funds to their core business operations, such as process improvements or advanced product development. This allocation of specialized assets aligns with the production characteristics of manufacturing enterprises and reduces the occurrence of idle or wasted assets, thus significantly lowering transaction costs²⁰.

Based on the aforementioned analysis, the following hypothesis is proposed in this paper:

Hypothesis 2

Smart logistics improves the manufacturing industry chain resilience by reducing transaction costs.

(ii) Smart logistics improves manufacturing industry chain resilience by enhancing logistics efficiency.

Smart logistics, through the promotion of logistics efficiency optimization, plays a significant role in enhancing the manufacturing industry chain resilience. The specific pathways through which this is achieved are as follows:

Firstly, decision support. Smart logistics systems, enabled by advanced information technology, can collect, analyze, and share real-time information on the location, transportation status, and estimated arrival time of goods²¹. This real-time data provides manufacturing companies with accurate decision support, enabling them to quickly respond to changes in market demand and supply chain disruptions. By accessing real-time information, manufacturing companies can adjust production plans or seek alternative suppliers in the event of delays in key raw material transportation, thereby avoiding production line stoppages²².

Secondly, resource allocation. Smart logistics systems utilize intelligent scheduling and optimization algorithms to effectively allocate transportation resources and improve transportation efficiency. This optimization encompasses various aspects, such as route planning, cargo loading, and transportation mode selection²³. In large-scale manufacturing production, any delays in the delivery of raw materials can lead to disruptions in the entire production process, affecting delivery times and customer satisfaction²⁴. By facilitating efficient transportation arrangements, smart logistics ensures that manufacturing companies can maintain stable production capacity, thereby enhancing the resilience of the industry chain.

Thirdly, risk management. Smart logistics systems enable predictive analysis, allowing enterprises to identify potential problems in advance and take timely countermeasures. Enhanced transparency enables manufacturing companies to quickly adjust transportation routes or select alternative transportation modes in the face of natural disasters or traffic congestion, reducing disruptions to production²⁵. This flexible risk management capability significantly enhances the resilience of the industry chain, enabling enterprises to maintain stability and development in an uncertain environment²⁶.

Based on the aforementioned analysis, the following hypothesis is proposed in this paper:

Hypothesis 3

Smart logistics improves manufacturing industry chain resilience by enhancing logistics efficiency.

Threshold effect of smart logistics on manufacturing industry chain resilience

Firstly, in terms of long-term dynamic perspective, in the initial stage of the development of wisdom, logistics enterprises master intelligent equipment and technology scale is small, intelligent infrastructure and human resources equipped with insufficient reasons, so that intelligent logistics has not reached the scale of economies of scale and network effect of the efficiency of the scale of the threshold, intelligent drive industry knowledge overflow and the formation of technological innovation of the potential to promote the role of the qualitative change has not yet been formed, that is, knowledge of the intelligent logistics Knowledge overflow and cost savings and a series of promotional effects there is a lag effect. With the accelerated development of the level of intelligence, intelligent logistics enterprises occupy the level of intelligence to reach the efficiency release of the order of magnitude, at the same time, knowledge overflow, cost savings also make intelligent logistics enterprises innovation ability gradually improve, these prompt intelligent logistics enterprises to the depth of the application of intelligent technology, at this time, the value of intelligence has been fully released, the manufacturing industry produces a strong knowledge overflow and promote the enhancement of its industrial innovation ability, to Promote the steady improvement of the manufacturing industry chain resilience. Therefore, the impact of the development of intelligent logistics on the manufacturing industry chain resilience has non-linear characteristics.

Secondly, based on the framework of “technology-economy paradigm”, it has been found that the digital economy is an effective path to promote the efficiency of the real economy and economic growth, and the comprehensive and spatially dispersed characteristics of the logistics industry also determine that the logistics industry is a data-intensive and digital economy-dependent industry, which can inject impetus into the development of smart logistics²⁷. Through innovative R&D and application of digital technology, a favorable regional innovation environment can effectively enhance the input level of green technology of logistics enterprises, promote green innovation, and then promote the logistics industry to serve related industries. According to the theory of “technology-economy” paradigm²⁸, this paper argues that, under the technological change promoted by the digital economy, when the level of regional innovation environment is low, the synergistic effect with the development of smart logistics is reduced, which is mainly due to the fact that, when the level of innovation environment is low, the integration depth of smart logistics and technological innovation is insufficient, and logistics enterprises are still difficult to fully absorb the technological innovation. Logistics enterprises are still difficult to fully absorb the transformational achievements of technological innovation, and its role in serving the manufacturing industry chain is quite limited. When the level of regional innovation environment is high, with the increase of regional R&D investment, the kinetic energy of digital transformation has been fully released, and the quality and efficiency of labor, knowledge, management, capital and technology factors have been improved, which promotes the logistics enterprises to more efficiently serve the change and upgrading of related industries. Therefore, the regional innovation environment has a positive impact on smart logistics enterprises to reduce enterprise search costs, stimulate enterprise innovation initiative and drive enterprise management change, but there is a threshold effect in the process of its role in the resilience of smart logistics to the manufacturing industry chain.

Thirdly, the role of smart logistics on the manufacturing industry chain resilience will be gradually strengthened with the improvement of the level of advanced industrial structure. Industrial structure upgrading is the key driver of regional economic growth, and is also a key factor affecting the resilience of the smart logistics and manufacturing industry chain. On the one hand, smart logistics can use digital technology to promote industrial organizational change, realize the rational distribution of factors and promote inter-industry synergistic development, and then help industrial structure upgrading, and at the same time, the industrial structure to the advanced, rationalized transformation process of the “economies of scale, the competition effect” to form a global driving effect, and promote the rational use of labor resources²⁹. Rational use of labor resources. On the other hand, smart logistics broadens the industrial production space with the characteristics of networking and intelligence, drives the manufacturing industry to carry out independent innovation and subversive technological innovation, breaks through the bottlenecks in the basic fields, and forms the development trend of deep cross-fertilization and multi-point breakthrough, and promotes the upgrading of industrial structure. At the same time, the upgrading of industrial structure strengthens the cooperation and communication between upstream and downstream industries, improves the synergistic effect of the industrial chain while opening up the circulation channels of factors, realizing the sharing of resources and complementing each other’s strengths, and enhances the industry chain’s risk perception ability, resistance ability and governance ability, which empowers the industry chain to enhance the resilience level³⁰. According to the theory of technology-economic paradigm, this paper argues that when the level of advanced industrial structure is low, the promotion effect of smart logistics on the manufacturing industry chain resilience is relatively weak; when the level of advanced industrial structure breaks through a certain threshold, the promotion effect of smart logistics on the manufacturing industry chain resilience is enhanced. Therefore, the role of smart logistics on the manufacturing industry chain resilience will be gradually strengthened with the improvement of the advanced level of industrial structure.

Based on the aforementioned analysis, the following hypothesis is proposed in this paper:

H4: There is a threshold effect in the influence of smart logistics on the manufacturing industry chain resilience. At the same time, the innovation environment and the level of advanced industrial structure have a threshold effect in the process of the role of smart logistics on the manufacturing industry chain resilience.

Research design

Modeling

(i) Benchmark regression model.

To examine whether smart logistics can effectively contribute to the manufacturing industry chain resilience, this study formulates the following econometric model:

1

Where:andrepresent the level of the manufacturing industry chain resilience and of smart logistics development in region i and year t, respectively.represents a series of control variables.and represent the fixed effects of province and year, respectively. is a random error term.is the constant term.and are the estimated coefficients of the relevant variables. This study focuses on estimating the coefficient , which measures the impact of smart logistics on the manufacturing industry chain resilience. A positive coefficient would suggest that smart logistics indeed enhances the resilience level of the manufacturing industry chain.

(ii) Mechanism test model.

To further investigate the indirect impact of smart logistics on the manufacturing industry chain resilience, this paper adopts a two-step approach. In the first step, the following model is constructed to examine the effect of smart logistics on the mediating variables:

2

Where: represent the mediating variable of transaction cost and logistics efficiency. The meanings of other variables are consistent with Model (1). In the second step, the following model is constructed to verify the effect of the mediating variables on the explanatory variables, aiming to enhance the design of the mechanism test:

3

(iii) Threshold test model.

To further examine the non-linear spillover effects of smart logistics on the manufacturing industry chain resilience, this study constructs a spatial threshold model using innovation environment (IE) and industrial structure (IS) as threshold variables. Among them, the innovation environment is characterized by the number of invention patent applications per 1,000 people, and the industrial structure is accounted for by the ratio of the value added of the secondary industry to the value added of the tertiary industry. The model is designed as follow:

4

Where:are the threshold variables.denote the coefficients of their influence on the manufacturing industry chain resilience under different level intervals of the threshold variables.denotes the schematic function, and are the threshold values to be estimated.

Description of variables

(i) Explained variable.

Manufacturing Industry Chain Resilience (MICR): Based on comprehensive reference to academic research on supply chain resilience measurement in other industries³¹, considering the characteristics of the manufacturing industry, the defined connotation of industry chain resilience (1), and the availability and reliability of relevant data, this study constructs an evaluation index system for the manufacturing industry chain resilience. The evaluation system includes three dimensions: resistance, recovery, and renewa1 (as presented in Table 3). The comprehensive level of MICR is measured using the entropy weight method³².

Table 3.

Evaluation index system of manufacturing industry chain resilience.

Primary indicators	Secondary indicators	Tertiary indicators	Methods of measurement	Unit	Causality
Industry chain resistance	Scale	Gross output	Sum of output value by sub-sectors	Million RMB	+
		Manufacturing value added/ person	Manufacturing value added / Average annual employment in manufacturing	Million RMB / Person	+
	Structure	Rationalization of industrial structure	Divide manufacturing into labor-intensive, capital-intensive, and technology-intensive categories, then calculate using the Theil entropy index	/	-
		Advanced industrial structure	Total output value of high-tech manufacturing / Total output value of mid-tech manufacturing	%	+
		Open-door structure	Export sophistication of manufacturing	%	+
Industry chain recovery	Development effectiveness	Labor productivity	Manufacturing output value / Average annual employment in manufacturing	Million RMB / Person	+
		Profitability	Total profit of manufacturing / Main business revenue of manufacturing	%	+
		Deficit	Sum of losses by sub-sectors	Million RMB	-
	Green development	Composite index of environmental regulation	Completed investment in industrial pollution control / Proportion of secondary industry	/	-
		Energy consumption per unit of value added	Total energy consumption / Manufacturing value added (Electricity consumption, wastewater discharge, and exhaust gas emissions)	Tons of standard coal / RMB	-
	Adjustment capacity	Diversification index	Reciprocal of the Hirschman-Herfindahl Index (HHI)	/	+
		Specialization index	Location entropy	/	+
		Manufacturing employment	Employment in private enterprises and self-employed individuals in manufacturing (10,000 persons) + Employment in urban units in manufacturing (10,000 persons)	Million people	+
Industry chain renewal	Innovation capacity	Intensity of R&D investment	Internal expenditure on R&D in manufacturing	Million RMB	+
		R&D expenditures Patent applications	Number of patent applications in manufacturing	Pcs	+
		Profitability of developing new products	Sales revenue of new products in manufacturing / Expenditure on new product development in manufacturing	%	+
	Innovation potential	Percentage of main business income from high-tech	Main business revenue of high-tech manufacturing / Main business revenue of manufacturing	%	+
		Industrial Robot Penetration	/	Pcs / Person	+
		Percentage of technical staff	R&D personnel in manufacturing / Average annual employment in manufacturing	%	+

Among them, the dimension of resistance mainly revolves around the scale and structure of the manufacturing industry chain. The scale characteristic measures the expansion capacity of the manufacturing industry chain. A larger-scale manufacturing industry chain can provide greater production capacity and supply security, reducing dependence on a single market or resource and dispersing risks. Structural characteristics consider the diversity of the manufacturing industry chain. A resilient manufacturing industry chain should have diverse supply chains and division of labor networks³³. When disruptions occur in one link, other links can complement and support each other, reducing dependence on specific links or enterprises and enhancing the ability to withstand external shocks. The larger the scale and the more balanced the structure of the manufacturing industry chain, the smaller the impact of external pressures on it.

The recovery dimension is primarily measured by the development effectiveness, green development, and adjustment capacity of the manufacturing industry chain. An efficient manufacturing industry chain can quickly adapt to market demand changes, strengthen supply chain integration and synergy, and minimize losses during recovery. Green development focuses on the environmental and ecological effects of the manufacturing industry chain during the recovery process. A resilient manufacturing industry chain should be capable of adjusting its production methods in response to environmental pressure, adopting environmentally friendly technologies and processes, and reducing resource consumption and pollution emissions. Adjustment capacity is achieved by adjusting manufacturing layout, optimizing industrial structure, and innovating technology to adapt to new environments and demands, transforming or optimizing production methods, and maintaining competitiveness.

The renewal dimension is primarily reflected in the innovation capacity and potential of the manufacturing industry chain³⁴. The innovation ability feature characterizes the manufacturing industry chain’s ability to innovate in the face of market changes and intensified external competition. A resilient manufacturing industry chain actively promotes technological innovation, product innovation, and business model innovation. The innovation capability considers factors such as R&D investment, technological innovation capabilities, and innovation management. The innovation potential feature takes into account the manufacturing industry chain’s potential and activity for future development. A resilient manufacturing industry chain should have a robust innovation ecosystem, including a concentration of R&D institutions, innovative enterprises, and talented individuals, as well as favorable policy support and access to innovation resources to facilitate enterprise innovation.

(ii) Core explanatory variable.

Smart logistics (SL). The development level of smart logistics is evaluated using an index system constructed from three dimensions: development elements, development environment, and development effectiveness (as shown in Table 4). The comprehensive level of smart logistics is measured using the entropy weight method. Due to challenges in directly obtaining relevant data for the logistics industry, the transportation, warehousing, and postal industry is used as a substitute. The indicators of smart logistics in the industry are proportionally converted. For instance, the degree of logistics industry agglomeration is measured by improving the location entropy. The energy consumption intensity of the logistics industry is measured by converting various types of energy consumption into standard coal consumption and calculating the ratio of the converted result to the value added of the logistics industry.

Table 4.

Evaluation index system of smart logistics.

Primary indicators	Secondary indicators	Tertiary indicators	Unit	Causality
Development elements	Technological innovation	Number of logistics industry invention patents	Pcs	+
	Financial support	Ratio of transportation and fiscal expenditure / Fiscal expenditure	%	+
	Talent scale	Number of logistics industry employees / Number of higher education personnel	/	+
		Proportion of employees in information transmission, computer services, and software industry	%	+
	Education investment	Ratio of fiscal education expenditure / Total expenditure	%	+
Development environment	Industrial structure	Logistics industry agglomeration	%	+
		Logistics industry output / Total output	%	+
	Digital finance	Digital Inclusive Finance Index	/	+
	Hardware sharing	Per capita internet broadband access ports	Pcs / Person	+
		Number of postal outlets per unit area	Pcs / Square meter	+
		Per capita fiber optic cable length	Meters / Person	+
	Software sharing	Mobile phone penetration rate	Set / 100 persons	+
		Number of national intelligent warehousing and logistics demonstration bases	Pcs	+
	Road density	(Road + railway mileage) / area	/ km	+
	Investment intensity	Logistics industry investment / Fixed asset investment	%	+
Development effectiveness	Intelligent Applications	Per capita telecommunications service volume	Pcs / Person	+
		Revenue from data processing, storage services / Logistics industry value added	%	+
		Revenue from platform operation and maintenance services / Logistics industry value added	/	+
	Hub capacity	Freight turnover	Pcs	+
	Economic benefit	Logistics industry revenue / Total population of the region	RMB / Person	+
	Environmental benefit	Logistics industry energy consumption / Logistics industry value added.	Tons of standard coal / RMB	-

(iii) Intermediary variables.

This study examines the mediating variables that explain the mechanism of the impact of smart logistics on manufacturing industry chain resilience. Two mediating variables are considered. Firstly, transaction cost represents the external transaction cost faced by manufacturing enterprises and is measured by the level of marketization (MK). Secondly, logistics operation efficiency represents the productivity of manufacturing industry factors in terms of micro-network chains. It is measured by the ratio of regional freight traffic to labor inputs.

(iv) Control variables.

To comprehensively analyze the impact mechanism of smart logistics on manufacturing industry chain resilience, this study selects control variables based on a systematic methodology categorized as “physical-factual-humanistic”. The control variables include economic growth as physical factor, financial development, opening up, urban and rural structure as factual factors, and investment efficiency and government functions as human factors. These control variables are comprehensively designed (as shown in Table 5). Among them, government debt is expected to have a negative correlation with manufacturing industry chain resilience, while all other control variables are expected to have a positive relationship with manufacturing industry chain resilience.

Table 5.

Control variables and their measurement.

Categorization	Control variables	Measurement
Physical factor	Economic growth (EG)	Regional GDP growth rate
Factual factor	Financial Engineering (FE)	Financial sector value added as a percentage of GRP
	Foreign Direct Investment (FDI)	Total exports and imports as a percentage of regional GDP
	Urban-rural structure (UR)	Urbanization rate
Humanistic factor	Investment efficiency (EI)	Incremental capital output rate (ICOR)
	Government debt (GD)	Government debt balance as a percentage of regional GDP

This paper proposes the following research framework: Firstly, the entropy weight method will be employed to calculate the comprehensive evaluation indices of smart logistics and manufacturing industry chain resilience respectively. A baseline regression model will then be constructed to empirically examine the magnitude and direction of smart logistics’ impact on manufacturing industry chain resilience, with instrumental variable methods adopted to address potential endogeneity issues. Secondly, robustness checks will be conducted through multiple approaches, including replacing core explanatory variables, substituting dependent variables, excluding outlier influences, removing special samples, and omitting data from 2020 to 2021. Thirdly, mechanism tests will be implemented to verify the validity of the proposed transmission channels. Fourthly, a comprehensive heterogeneity analysis will be performed, encompassing dimensional heterogeneity, temporal heterogeneity, and regional heterogeneity, to deeply investigate the variations in how smart logistics affects manufacturing industry chain resilience across different dimensions, time periods, and geographical locations. Fifthly, by introducing threshold variables and constructing a threshold regression model, the threshold effects of smart logistics on manufacturing industry chain resilience will be examined. This multi-dimensional analytical framework aims to provide a rigorous and comprehensive examination of the relationship between smart logistics development and manufacturing industry chain resilience enhancement.

Data sources and descriptive statistics

This study uses a sample of 30 provinces in China (excluding Tibet, Hong Kong, Macao, and Taiwan) from 2012 to 2023. The required index data are primarily sourced from publications such as the China Statistical Yearbook, China Industrial Statistical Yearbook, China Energy Statistical Yearbook, China High-Tech Industry Statistical Yearbook, China Electronic Information Industry Statistical Yearbook, and statistical yearbooks of various provinces and cities. Industrial robot data are obtained from the official website of the International Federation of Robotics (IFR), and marketization index data come from the report by Fan Gang’s marketization index. Missing data are supplemented using the weighted average method. To minimize the impact of absolute differences between the data and ensure data smoothness, logarithmic transformation is applied to the original data of the control variables.

Table 6 presents the descriptive statistics of the main variables. The maximum value of manufacturing industry chain resilience is 0.724, the minimum value is 0.041, and the standard deviation is 0.099. These figures indicate significant variations in the level of development of manufacturing industry chain resilience across different regions of China, highlighting the validity of differentiation. Similarly, there are substantial differences in the level of development of smart logistics among various regions.

Table 6.

Descriptive statistical results.

Variable	N	Mean	SD	Min	p50	Max
MICR	360	0.155	0.099	0.041	0.131	0.724
SL	360	0.196	0.107	0.035	0.179	0.662
FE	360	3.314	1.146	1.568	3.113	8.131
FDI	360	0.259	0.277	0.008	0.141	1.441
EG	360	1.065	0.0372	0.827	1.075	1.135
EI	360	0.798	0.467	−2.669	0.877	1.628
UR	360	0.594	0.120	0.354	0.583	0.950
GD	360	0.151	0.113	0.0130	0.129	0.636
IE	360	0.017	0.012	0.005	0.015	0.065
IS	360	1.553	1.926	0.572	1.176	23.78

Results and discussion

Benchmark regression analysis

The benchmark regression results based on the earlier described benchmark regression model are presented in Table 7. In Column (1), the analysis initially confirms the positive contribution of smart logistics to the manufacturing industry chain resilience. To enhance the robustness of the results, a method of gradually increasing control variables is employed. It is observed that the coefficients of smart logistics are consistently and significantly positive across all estimations, with minimal changes. Taking Column (3) as an example for analysis, the estimated coefficient of smart logistics passes the significance test at the 1% level, indicating that smart logistics indeed improves the manufacturing industry chain resilience. This empirical evidence supports Hypothesis 1.

Table 7.

Benchmark regression results.

Variable	(1)	(2)	(3)
	MICR	MICR	MICR
SL	0.773*** (26.31)	0.737*** (15.17)	0.668*** (10.07)
IE		4.539*** (4.82)	4.704*** (4.96)
IS		0.000 (0.28)	0.000 (0.30)
FE		0.005 (1.17)	0.003 (0.45)
FDI		−0.030 (−1.31)	−0.060** (−2.44)
EG			−0.099 (−1.22)
EI			−0.006* (−1.69)
UR			0.345*** (3.42)
GD			−0.040 (−0.45)
Constant	0.003 (0.46)	0.010 (1.06)	−0.233* (−1.82)
Year FE	No	Yes	Yes
ID FE	No	Yes	Yes
N	360	360	360
Adj R2	0.698	0.958	0.964

Note: * p < 0.1, ** p < 0.05, *** p < 0.01; t-values are clustered by province level and are shown in parentheses. These notation conventions apply to all subsequent tables and will not be repeated.

The regression results for the control variables indicate the following: The coefficient estimate of the innovation environment is 4.701 and statistically significant at the 1% level, suggesting that a region with a robust innovation environment can provide favorable conditions such as innovation resources, research and development institutions, and talent clustering. By meeting the R&D needs for technological and business model innovation in the manufacturing industry, building innovative technology platforms, and promoting the utilization and value creation of R&D results, the resilience and competitiveness of the manufacturing industry chain can be enhanced³⁵. The estimated coefficient of opening up to the outside world is −0.060and passes the significance test at the 10% level. This implies that increased openness to the outside world may bring risks such as intensified domestic market competition and volatility in the international market³⁶. Overreliance on foreign markets restricts the autonomy and control of China’s manufacturing industry chain. When facing intense competition from the international market and lacking the ability to withstand external shocks, the resilience level of the manufacturing industry chain may be reduced. The coefficient estimate of investment efficiency is significantly negative at the 10% level. This suggests that the availability of capital is an important factor affecting the resilience and security of the manufacturing industry chain³⁷. Adequate access to capital is crucial for maintaining and improving the resilience level of the manufacturing industry chain. The coefficient estimate of urban-rural structure is 0.3452 and statistically significant at the 1% level. This indicates that optimizing the urban-rural structure can promote the manufacturing industry chain resilience. The rational arrangement of urban-rural structure provides abundant labor resources and industrial support, enhancing the coordination and interconnection between manufacturing industry chains. Additionally, balanced urban-rural development reduces dependence on a single region or market, thus reducing the risk faced by the manufacturing industry chain. The remaining control variables do not have a significant impact on the manufacturing industry chain resilience.

Overcoming endogeneity: instrumental variable approach

To address the issue of endogeneity arising from potential reverse causality or bidirectional causality, this paper employs an instrumental variable approach. Firstly, the interaction terms between the number of landline telephones per 100 people and the number of post offices per million people in China in 1984 (IV1) and the number of people who accessed the Internet in the previous year (IV2) are used as instrumental variables for smart logistics. IV2 is lagged by one period. Secondly, the instrumental variables are tested using the two-stage least squares method (2SLS). The selection of instrumental variables satisfies the principles of relevance and exogeneity. In terms of relevance, the number of landline telephones, the number of post offices, and the number of people accessing the Internet reflect the level of infrastructure construction and digitalization development in a region, which align with the infrastructure attributes and intelligent features of smart logistics. In terms of exogeneity, these variables are historical and have no direct and significant impact on the manufacturing industry chain resilience.

The results of the two-stage least squares regression for the instrumental variables of smart logistics are presented in Table 8. The Kleibergen-Paap rk LM statistic rejects the hypothesis of under-identification. The Kleibergen-Paap rk Wald F statistic exceeds the critical value of 10% bias in the Stock-Yogo weak instrument test (19.93), suggesting the absence of weak instrumental variable problems. The p-value of the Hansen J statistic is greater than 0.1, indicating that the original hypothesis of exogeneity of the instrumental variables can be accepted. Column (1) shows a strong correlation between the instrumental variables IV1 and IV2 with smart logistics. Column (2) demonstrates that, after overcoming potential endogeneity issues, smart logistics continues to significantly contribute to enhancing the manufacturing industry chain resilience in China.

Table 8.

Two-stage least squares regression results.

Variable	(1)Stage I	(2)Stage II
	MICR	MICR
SL		0.752*** (5.04)
IV1	−0.258** (−2.14)
IV2	0.888*** (15.03)
Control Variables	Yes	Yes
Year FE	Yes	Yes
ID FE	Yes	Yes
N	240	240
Kleibergen-Paap rk LM statistic	0.000***
Kleibergen-Paap rk Wald F statistic	146.589
Hansen J statistic	0.383
Adj R2	0.501

Robustness test

This paper uses the practices of replacing the core explanatory variables, excluding the effects of outliers, excluding the effects of special samples, and excluding the impact of the COVID-19 pandemic to test the robustness of the empirical results in the previous paper, and the results of the robustness test are shown in Table 9.

Table 9.

Robustness test results.

Variable	(1)	(2)	(3)	(4)
SL2	0.616*** (11.99)
SL		0.448*** (7.45)	0.556*** (6.22)	0.789*** (16.56)
Control Variables	Yes	Yes	Yes	Yes
Constant	−0.107 (−0.87)	−0.263** (−2.12)	−0.058 (−0.43)	−0.051 (−0.26)
Year FE	Yes	Yes	Yes	Yes
ID FE	Yes	Yes	Yes	Yes
N	360	360	260	330
Adj R2	0.968	0.970	0.972	0.793

(i) Replacement of core explanatory variables.

The entropy weight method used in the previous accounting of the level of smart logistics development is assigned by comparing the degree of dispersion of the distribution of each indicator’s value, and the difference in the weight of each indicator may be on the large side. The coefficient of variation method is based on the average value of each indicator and the initial value of each indicator to assign weights, and the weight difference of each indicator is smaller. Thus, this paper uses the coefficient of variation method to re-measure the level of smart logistics development of Chinese provinces (SL2), and brings the new measurement results into the model (1) to re-regression analysis. The results show that the coefficient estimates of SL2 after replacing the core explanatory variables are significantly positive, and the research conclusions have not changed as a result.

(ii) Troubleshooting outliers.

In order to eliminate the impact of outliers on the regression results, this paper shrinks the upper and lower 1% of the explanatory variables, the core explanatory variables, and the control variables. The robustness test results in column (2) of Table 7 show that the coefficient estimates of smart logistics are still significantly positive after the shrinking treatment, indicating that the impact of outliers on the empirical results of the previous paper is relatively small, which can be inferred that the conclusions obtained from the study are more robust and credible.

(iii) Excluding the effect of special samples.

The four municipalities directly under the central government, Beijing, Tianjin, Shanghai and Chongqing, have a special position in China’s political economy and are relatively well-developed in terms of manufacturing. There are part of the municipality directly under the central government for the jurisdiction of the manufacturing industry development facing the reality of the blockage, the introduction of a highly targeted local manufacturing industry policy situation. For example, the Beijing Municipal Bureau of Economy and Information Technology issued the Beijing “New Smart 100” Project Implementation Plan (2021–2025); the Shanghai Municipal Commission of Economy and Information Technology, together with the Shanghai Municipal Commission of Development and Reform, drafted the Guidelines for Accelerating the Development of the Four New Economies in the City. Thus, there are reasons to suspect that it may be these favorable policies that significantly improve the manufacturing industry chain resilience in municipalities directly under the central government, rather than the empowering effect played by smart logistics that makes the main contribution. As a result, the regression analysis of this paper was re-conducted after excluding the sample of municipalities directly under the central government, and the new empirical results show that the coefficient of smart logistics is still significantly positive, which to a certain extent confirms that the smart logistics does have a strong facilitating effect on the enhancement of the manufacturing industry chain resilience.

(iv) Excluding the impact of the COVID-19 pandemic.

The COVID-19 pandemic has been proven to lower the manufacturing industry chain resilience. Is this hindering effect achieved by constraining the empowering effect of smart logistics on enhancing the manufacturing industry chain resilience? This study excluded the samples from 2021 to 2022 and retested the model, obtaining more significant empirical results compared to the original results. This suggests that the COVID-19 pandemic has played a certain masking effect on the process of smart logistics empowering the manufacturing industry chain resilience. This masking effect may manifest in the following ways: First, the pandemic has disrupted global supply chains, including factory closures, transportation restrictions, and logistics disruptions, severely disrupting the flow of logistics and preventing smart logistics from fully leveraging its advantages³⁸. Second, the pandemic has caused significant fluctuations and changes in market demand across industries, making it difficult for manufacturing companies to accurately forecast and plan logistics needs, thus affecting the effectiveness and efficiency of smart logistics. Third, the shortage of human resources during the pandemic, including logistics personnel and warehouse managers, has led to a decrease in operational efficiency, thereby impacting the normal operation of smart logistics systems.

Mechanism test

Based on the previous theoretical analysis, the mechanism through which smart logistics indirectly impacts the manufacturing industry chain resilience is primarily achieved by reducing transaction costs and improving logistics efficiency. The regression results of the mechanism test are presented in Table 10.

Table 10.

Mechanism test results.

Variable	(1)	(2)	(3)
	MICR	MK	MI
SL	0.668*** (10.07)	0.368** (2.04)	0.485** (2.59)
Control Variables	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
ID FE	Yes	Yes	Yes
N	360	360	360
Adj R2	0.964	0.925	0.615

(i) Pathway of reducing transaction costs.

From column (2) of Table 9, we observe a significantly positive coefficient estimate for the marketization level (MK), indicating that smart logistics has a clear positive effect on marketization. This suggests that smart logistics can indeed reduce transaction costs through various channels, such as reducing transaction frequency, mitigating uncertainty arising from information asymmetry, and enhancing asset specialization. By strengthening the security of manufacturing industry and business operations, smart logistics enables market players to gain incentives for expanding production and reinvestment. This, in turn, provides financial support for technological research and development across the upstream and downstream of the manufacturing industry chain. Therefore, empirical evidence supports Hypothesis 2.

(ii) Pathway of improving logistics efficiency.

As evident from column (3) of Table 9, the estimated coefficient of logistics efficiency is 0.485 and statistically significant at the 5% level. This finding confirms the hypothesis that smart logistics improves the manufacturing industry chain resilience by enhancing logistics efficiency. As discussed in the theoretical analysis section, smart logistics technology enhances asset reorganization and resource integration efficiency in the logistics industry. This, in turn, fosters the development of related industry clusters. The internal economies of scale generated by the logistics industry’s agglomeration transform into external economies of scale for enterprises in need of logistics services. This promotes the rationalization and advancement of the manufacturing industry’s industrial structure, enhancing its ability to swiftly adapt to internal fluctuations and external shocks. Furthermore, the ability of intelligent monitoring throughout the logistics process is greatly improved, reducing the likelihood of resource mismatch and significantly enhancing the stability of logistics operations.

Threshold effect analysis

The promotion effect of smart logistics on the manufacturing industry chain resilience may exhibit non-linear characteristics. Therefore, to further explore the phased characteristics of this impact, this study conducts a threshold test on innovation environment and industrial structure as threshold variables. Assuming single, double, and triple threshold conditions, the significance and quantity of thresholds are analyzed, and the P-value and F-statistic of the threshold effect test are calculated 300 times based on the Bootstrap method, as shown in Tables 11 and 12.

Table 11.

Threshold effect test results.

Variable	Threshold	F-statistic	P-value	10%	5%	1%	Bootstrap
SL	Single	188.19	0.000	22.545	27.341	35.419	300
	Double	9.82	0.620	22.479	27.030	175.065	300
IE	Single	105.92	0.003	36.253	46.050	76.059	300
	Double	80.82	0.000	31.625	41.905	56.986	300
	Triple	84.19	0.610	172.898	199.207	243.794	300
IS	Single	34.23	0.047	23.461	33.603	51.011	300
	Double	43.59	0.020	22.541	27.990	50.526	300
	Triple	13.59	0.203	16.830	21.468	32.465	300

Table 12.

Threshold effect Estimation results.

Variable	Threshold interval	Coefficient	t-value	R 2	Control variable	Time fixed effect	Region fixed effect
SL	SL ≤ 0.449	0.273***	4.56	0.8623	Yes	Yes	Yes
	SL > 0.449	0.528***	6.45
IE	0.025 ≤ IE	0.349***	4.09	0.8160	Yes	Yes	Yes
	0.025 < IE ≤ 0.031	0.617***	7.59
	IE > 0.031	0.097	1.52
IS	IS ≤ 0.456	0.244	1.64	0.8388	Yes	Yes	Yes
	0.456 < IS ≤ 1.945	0.648***	4.27
	IS > 1.945	0.381***	3.78

(i) Smart logistics.

Threshold regression analysis is performed using smart logistics as the threshold variable. The results of the test indicate that only a single threshold is statistically significant, suggesting the existence of a threshold value for smart logistics estimated at 0.429. When the level of smart logistics development is less than 0.429, the coefficient is 0.273 and significant at the 1% level. Conversely, when the level of smart logistics development is greater than 0.449, the coefficient is 0.528 and significant at the 1% level. This implies that in the early stages of smart logistics development, its empowerment capability for enhancing the manufacturing industry chain resilience is relatively weak. As smart logistics development matures and surpasses a specific threshold, its enhancing effect on economic and innovative output becomes increasingly prominent, thereby demonstrating a stronger impact on the resilience enhancement of the local manufacturing industry chain.

(ii) Innovation environment.

To examine whether the marginal effect of smart logistics on the manufacturing industry chain resilience is influenced by regional innovation environment differences, regression analysis is conducted using innovation environment as the threshold variable. The results indicate a significant double threshold in the innovation environment. When the quality of the innovation environment is less than 0.025, the coefficient for smart logistics is 0.349, significantly positive. When the quality of the innovation environment is greater than 0.025 but less than 0.031, the coefficient for smart logistics is 0.617, significantly positive. However, when the quality of the innovation environment exceeds 0.031, the coefficient is 0.097 and not significant. This implies that when the innovation environment is at a moderate level, it positively drives the upgrading of the manufacturing industry chain in conjunction with smart logistics. However, as the second threshold is crossed, the driving effect of the innovation environment weakens. One possible explanation is that in the early stages of low innovation levels, regions have limited research and development funds, and the necessary conditions for promoting innovation in the logistics industry are not yet in place, leading to a smaller coefficient for smart logistics. As digital development gradually gains attention, local governments tend to use various policy tools to allocate financial resources towards smart logistics to achieve a competitive advantage in the digital economy. This results in an increase in research and development funds, thereby highlighting the support capacity of smart logistics in enhancing the manufacturing industry chain resilience. However, in the subsequent development process of smart logistics, the marginal contribution of digital technology diminishes, leading to a decline in the coefficient for smart logistics.

(iii) Industrial structure.

Based on the regression results with industrial structure as the threshold variable, the level of industrial structure rationalization significantly affects the manufacturing industry chain resilience in the context of smart logistics, exhibiting a nonlinear relationship characterized by an initial increase followed by a decrease. Specifically, when the level of industrial structure rationalization is below the threshold value of 0.456, the impact coefficient of smart logistics on the manufacturing industry chain resilience is not significant. When the level of industrial structure rationalization is between 0.456 and 1.945, the impact coefficient increases to 0.648, passing the test at a significance level of 1%. However, when the level of industrial structure rationalization exceeds the threshold value of 1.945, the impact coefficient decreases to 0.381. These findings indicate an overlapping and interactive relationship between industrial structure rationalization and the development of smart logistics. In a specific range of industrial structure rationalization, the positive signals generated by the synergy between smart logistics and industrial structure rationalization act as amplifiers. This stimulates the innovation and proactive engagement of logistics and manufacturing enterprises, while enhancing the ability of different segments within the manufacturing industry chain to maintain production stability and adjust for recovery. In other words, all the independent variables in this scenario tend towards an optimal level of synergy, leading to the generation of the best matching results in the system. However, once the level of industrial structure rationalization surpasses a certain threshold, the marginal benefits of the synergy between industrial structure rationalization and smart logistics on the manufacturing industry chain resilience rapidly decline. This may be attributed to a mismatch between external and internal factors within the influencing factors of the manufacturing industry chain resilience, indicating the potential for further development towards a higher level of matching.

Heterogeneity analysis

Regional heterogeneity

After confirming the enabling effect of smart logistics on the enhancement of the resilience level of the manufacturing industry chain in the overall sample region of China, it is important to analyze whether this effect varies in intensity depending on the geographical location of the provinces. To address this issue, this paper divided the research sample into eastern, central, and western regions, and the new regression results are presented in Table 13.

Table 13.

Regional heterogeneity regression results.

Variable	(1)	(2)	(3)
	Eastern region	Central region	Western region
SL	0.772*** (6.48)	−0.132 (−0.62)	0.241** (2.25)
Control Variables	Yes	Yes	Yes
Constant	−0.105 (−0.23)	0.394** (2.42)	−0.373* (−1.74)
Year FE	Yes	Yes	Yes
ID FE	Yes	Yes	Yes
N	110	80	110
Adj R2	0.960	0.922	0.933

Smart logistics has demonstrated a significant positive contribution to the manufacturing industry chain resilience in eastern China. This can be attributed to several key factors: Firstly, strong manufacturing foundation. The eastern region of China, encompassing provinces such as Jiangsu, Zhejiang, and Shanghai, has long been a hub for manufacturing activities. This region benefits from a robust industrial base, well-developed transportation networks, and access to large consumer markets. The concentration of manufacturing enterprises in this area is driven by the availability of skilled labor, advanced infrastructure, and proximity to international trade routes. Despite the gradual relocation of labor-intensive industries to central and western regions due to rising labor and land costs, as well as environmental constraints, the eastern region has strategically shifted its focus toward high-value-added industries. These include digital manufacturing, advanced materials, and core technology research and development (R&D). Smart logistics has become a critical enabler in this transformation. By integrating technologies such as the Internet of Things (IoT), artificial intelligence (AI), and big data analytics, smart logistics systems optimize supply chain operations, reduce lead times, and minimize costs. Secondly, advanced digital infrastructure. The eastern region’s well-developed digital infrastructure provides a solid foundation for the widespread adoption of smart logistics technologies³⁹. Over the past decade, significant investments have been made in building high-speed internet networks, cloud computing platforms, and data centers. These investments have facilitated the seamless flow of information across the supply chain, enabling real-time tracking, predictive analytics, and automated decision-making. Thirdly, conducive innovation environment. The eastern region’s innovation ecosystem is another critical factor driving the adoption and application of smart logistics technologies. The region is home to a dense network of R&D institutions, technology enterprises, and innovation hubs, which foster collaboration between academia, industry, and government. This ecosystem supports the development of cutting-edge technologies and their integration into industrial practices⁴⁰. For example, Shanghai’s Zhangjiang High-Tech Park, often referred to as China’s “Silicon Valley,” hosts numerous technology companies specializing in AI, robotics, and IoT. These companies collaborate with local manufacturers to develop customized smart logistics solutions that address specific industry challenges. During the COVID-19 pandemic, a Shanghai-based electronics manufacturer utilized a smart logistics platform developed in collaboration with a local tech firm to real-time adjust its supply chain strategy. By shifting raw material procurement from overseas to domestic suppliers, the company avoided production interruptions and maintained its market competitiveness. Forthly, policy support. The Chinese government has played a crucial role in promoting the adoption of smart logistics technologies in the eastern region. Policies such as the “Made in China 2025” initiative and the “Digital China” strategy have provided financial incentives, technical support, and regulatory frameworks to encourage the integration of advanced technologies into manufacturing and logistics operations. These policies have created a favorable environment for innovation and investment in smart logistics.The heterogeneity analysis results reveal that smart logistics does not significantly improve the manufacturing industry chain resilience in the central region of China. This stands in stark contrast to the positive impact observed in the eastern region. The challenges hindering the effective application of smart logistics in the central region are: Firstly, insufficient R&D funding. One of the primary obstacles to the adoption of smart logistics in the central region is the lack of sufficient R&D funding. The development and implementation of smart logistics technologies require substantial financial investment and human resources⁴¹. However, the central region has consistently lagged behind the national average in terms of R&D expenditure as a percentage of GDP. This funding gap limits the ability of local enterprises to invest in advanced technologies and innovate their supply chain processes. For example, in provinces such as Henan and Hubei, many manufacturing enterprises operate in the middle and low ends of the value chain, producing goods with low added value and weak market competitiveness. These enterprises often lack the financial resources to invest in smart logistics technologies, such as IoT-enabled tracking systems or AI-driven demand forecasting tools. As a result, they remain reliant on traditional logistics methods, which are less efficient and more vulnerable to disruptions. The lack of R&D funding also hampers the region’s ability to attract and retain skilled talent, further exacerbating the technological gap between the central and eastern regions. Secondly, institutional mechanism problems. Institutional inefficiencies pose another significant challenge to the adoption of smart logistics in the central region. These inefficiencies manifest in several ways, including poor coordination among government departments, insufficient policy support, and irrational industrial planning. For instance, local governments in the central region often lack a cohesive strategy for promoting smart logistics, resulting in fragmented and inconsistent policies. This lack of coordination creates uncertainty for enterprises and discourages investment in smart logistics technologies. Moreover, the central region’s industrial planning and layout are often suboptimal, with a focus on low-value-added manufacturing activities. This limits the region’s ability to transition to higher-value industries that could benefit more from smart logistics. For example, in Anhui Province, the manufacturing sector is dominated by traditional industries such as textiles and machinery, which have limited demand for advanced logistics solutions. Without a clear policy framework and strategic planning, enterprises in the central region struggle to integrate smart logistics into their operations, further hindering their competitiveness and resilience. Thirdly, discriminatory and hidden regional market barriers. Local protectionism and regional market barriers are significant impediments to the free flow of manufacturing production factors and the adoption of smart logistics technologies in the central region. Local governments often implement discriminatory policies to protect the interests of local enterprises, creating an uneven playing field for external businesses. For example, in procurement processes, local companies are often favored over external ones, even if the latter offer more competitive products or services. This protectionism is not only evident in formal policies but also in covert practices, such as imposing stricter qualification requirements on external companies or creating bureaucratic hurdles in approval processes. These market barriers hinder the ability of central region enterprises to collaborate with external partners and adopt advanced technologies.

In the western region of China, the estimated coefficient of smart logistics is significantly positive at the 5% level, indicating that it plays a crucial role in enhancing the manufacturing industry chain resilience. This positive impact is driven by several factors: Firstly, industrial transfer from the eastern region. The western region has become a key destination for industrial transfer from the more developed eastern region. As labor and land costs rise in the east, many manufacturing enterprises are relocating to the west, where costs are lower and urbanization is accelerating. This industrial transfer has created opportunities for the western region to upgrade its manufacturing capabilities and integrate advanced technologies, including smart logistics. Secondly, policy dividends and strategic initiatives. The western region has benefited significantly from national policies and strategic initiatives aimed at promoting economic development and regional integration. Key policies include the Western Development Strategy, the construction of new land and sea corridors, the Belt and Road Initiative, and the opening of the China-Europe Railway Express. These initiatives have connected the western region with markets in the Middle East, Southeast Asia, and Europe, fostering the development of a modern logistics system⁴². For instance, the China-Europe Railway Express has transformed Chongqing into a major logistics hub, linking the region directly to European markets. This has not only reduced market segmentation but also enhanced the region’s ability to participate in global supply chains. Smart logistics technologies have played a pivotal role in this transformation by enabling efficient coordination and management of cross-border trade. For example, the use of blockchain technology in the China-Europe Railway Express has improved supply chain transparency and security, reducing the risk of delays and disruptions. Thirdly, enhancing resource utilization efficiency. The western region faces challenges related to resource dispersion and an imperfect supply chain. Smart logistics technologies have been instrumental in addressing these challenges by optimizing resource allocation and improving supply chain efficiency. For example, intelligent warehousing systems and inventory management solutions have helped reduce inventory backlogs and minimize resource waste. In the Liangjiang New Area of Chongqing, manufacturing enterprises have adopted smart logistics systems to achieve full-process digital management, from raw material procurement to finished product distribution. These systems use advanced data analytics to optimize warehouse layouts and transportation routes, resulting in significant cost savings and efficiency gains.

Sub-dimensional heterogeneity

The three dimensions within the resilience of the smart logistics and manufacturing industry chain are regressed separately to examine the difference in the degree of influence of smart logistics on different dimensions of the manufacturing industry chain resilience. From the regression results in Table 14, the wisdom of logistics are on the manufacturing industry chain resistance, recovery and renewal of the three dimensions show a significant positive effect, and all through the 1% significance level test. It indicates that the development of smart logistics can effectively enhance the resistance, renewal and recovery of the manufacturing industry chain resilience. However, the promotion effect of smart logistics on the recovery dimension is smaller than that of the resistance dimension and the renewal dimension, which indicates that the recovery dimension of the manufacturing industry chain resilience, i.e., the three dimensions of the industry chain development efficiency, the industry chain green development, and the industry chain adjustment ability, are more likely to highlight the development difficulties.

Table 14.

Sub-dimensional heterogeneity regression results.

Variable	(1)	(2)	(3)
	Industry chain resistance	Industry chain recovery	Industry chain renewal
SL	0.6600*** (9.55)	0.3093*** (5.39)	0.6590*** (11.18)
Control Variables	Yes	Yes	Yes
Constant	−0.4342** (−1.99)	−0.0502 (−0.28)	−0.3574* (−1.92)
Year FE	Yes	Yes	Yes
ID FE	Yes	Yes	Yes
N	360	360	360
Adj R2	0.7182	0.3596	0.6426

Conclusions

Main conclusions

This paper utilizes panel data from 30 provinces in China from 2012 to 2023 to investigate the impact of smart logistics on the manufacturing industry chain resilience. The research findings lead to the following main conclusions:

The main findings of this paper are: (i) the impact of smart logistics on the manufacturing industry chain resilience is significantly positive, and the research conclusions are relatively robust; (ii) the promotion effect of smart logistics on the resistance and renewal of the manufacturing industry chain is greater, and the promotion effect on the manufacturing industry chain resilience is relatively weaker; (iii) in the comparison of the impact relationship between four different economic regions, namely, the east, the central region, the west, and the northeast, it is found that the eastern region’s promotion role is stronger, the central region’s promotion role is not obvious, and the promotion role of the western region and the northeastern region is weaker and weaker relative to the eastern region. (iv) The mechanism test shows that smart logistics further improves the manufacturing industry chain resilience by reducing transaction costs and promoting logistics operation efficiency. (v) Threshold test shows that, in terms of long-term dynamics, there is a threshold effect of the development of smart logistics on the enhancement of the manufacturing industry chain resilience; at the same time, the higher level of advanced industrial structure and innovation environment, smart logistics is conducive to enhancing the level of manufacturing industry chain resilience.

Policy Implications

Based on the existing literature, theoretical analysis, and empirical findings, this paper proposes the following policy recommendations to optimize smart logistics policies and empower the construction of a modernized industrial system while enhancing the manufacturing industry chain resilience:

• (i)The collaborative development of logistics intelligence and manufacturing intelligence should be market-oriented and enterprise-oriented, guided by government policies, and implemented step-by-step in accordance with local conditions⁴³. Considering China’s logistics and manufacturing industry development, the following ideas can assist in enhancing the manufacturing industry chain resilience through smart logistics empowerment. Firstly, the government should formulate policies to incentivize enterprises to invest in and innovate smart logistics technology. This can include providing financial support, tax incentives, and policy subsidies to promote the application of automation, unmanned systems, and intelligent management technologies in logistics. Secondly, fostering partnerships between enterprises should be encouraged to achieve synergistic development between the logistics and manufacturing industries, optimizing the management of supply chains and industrial chains through information, resource, and technology sharing⁴⁴. Thirdly, efforts should be made to cultivate and attract talent in the fields of logistics intelligence and manufacturing intelligence. Encouraging collaboration between businesses and universities/research institutes can help improve the skills and innovation capabilities of professionals in these areas.
• (ii)Reducing transaction costs and improving logistics efficiency should be prioritized to promote the development of smart logistics and smart manufacturing. The mechanism test results indicate that smart logistics can enhance the manufacturing industry chain resilience by reducing transaction costs and improving logistics efficiency. Firstly, optimizing the design and layout of the logistics network is essential. Utilizing intelligent algorithms and big data analysis can determine optimal logistics paths and transportation schemes, improving logistics efficiency and capacity. Secondly, widespread adoption of Internet of Things and sensing technology in logistics can enable accurate positioning and real-time monitoring of goods, enhancing the visualization and controllability of logistics operations. Thirdly, the application of automated warehousing equipment, intelligent robots, and automated sorting systems can reduce labor costs and errors, establish a robust supply chain risk management system⁴⁵, predict and assess potential risks, and formulate response strategies to improve the ability to handle supply chain disruptions and emergencies.
• (iii)Regions with varying levels of manufacturing development should adopt development strategies suited to their respective stages to realize the empowering effect of smart logistics on the manufacturing industry chain resilience. Developed countries can leverage their socioeconomic advantages, strong manufacturing foundations, and innovation environments to transform residents’ consumption capacity and market scale into profitable opportunities for manufacturing enterprises. Under the influence of smart logistics, the development of digital industry clusters and advanced manufacturing industries should be emphasized. Developing countries with better development should focus on ensuring continuous R&D expenditure on supply chain technologies such as smart logistics. This should be coupled with efforts to optimize urban and rural structures, integrate logistics networks, and address institutional mechanism deficiencies and inter-regional market barriers that impede the promotion and application of smart logistics technologies by enterprises. For developing countries in general, undertaking the transfer of manufacturing industries from developed and better-developed developing countries, continuously introducing favorable policy measures, and improving resource utilization efficiency by manufacturing enterprises can serve as reference ideas to enhance the manufacturing industry chain resilience empowered by smart logistics.

Research Shortcomings and Prospects

The motivation behind this paper stems from the synergistic characteristics of smart logistics and smart manufacturing development in China, as well as the limitations of existing studies in explaining the detailed relationship between smart logistics and the manufacturing industry chain resilience. This paper provides empirical evidence supporting the enhanced manufacturing industry chain resilience empowered by smart logistics. It establishes the indirect mechanism of smart logistics on manufacturing industry chain resilience through the paths of opportunity cost and logistics efficiency. Additionally, it provides personalized insights for regions at different stages of manufacturing industry development by analyzing regional heterogeneity. However, there are two main shortcomings that may be present in this paper.

• (i)In the research design, this paper does not consider the exclusion of other favorable policies or elements that promote the manufacturing industry chain resilience. As a result, it is worth considering that the pathway through which smart logistics empowers the manufacturing industry chain resilience may not be limited to the two paths mentioned in the paper. There might be other internal factors or external forces that, together with smart logistics, promote the enhancement of the manufacturing industry chain resilience. Future research should explore and incorporate these factors to provide a more comprehensive understanding of the relationship.
• (ii)The explanation provided in this paper for the regional heterogeneity of the enabling effect of smart logistics on the manufacturing industry chain resilience in eastern, central, and western China is mainly based on empirical and subjective conjecture derived from the macroeconomic development of these regions. Further empirical testing of these explanations was not conducted in this paper. Therefore, the accuracy of these explanations needs to be confirmed or challenged by future researchers who can conduct empirical analysis specific to the regional variations in China.

Moving forward, future research should address these shortcomings and explore additional factors and mechanisms that contribute to the enhancement of the manufacturing industry chain resilience through smart logistics. Further empirical testing and analysis can provide more robust and comprehensive insights into the relationship between smart logistics and the manufacturing industry chain resilience.

Author contributions

Conceptualization, Jianmin Du and Jun Liang; methodology, Jianmin Du; software, Jianmin Du; validation, Jianmin Du and Jingling Wang; formal analysis, Jianmin Du; data curation, Jianmin Du; writing—original draft preparation, Jianmin Du; writing—review and editing, Jianmin Du and Jun Liang; visualization, Jingling Wang and Ruilin Liang; supervision, Jun Liang and Ruilin Liang; project administration, Jianmin Du and Jingling Wang; funding acquisition, and Jianmin Du and Jingling Wang. All authors have read and agreed to the published version of the manuscript.

Data availability

The raw data used in the study are publicly available and their sources are stated in the text. In case of doubt, please consult the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.