Empowering Agricultural Economic Resilience with Smart Supply Chain: Theoretical Mechanism and Action Path

Abstract

As agriculture faces increasingly complex risk shocks, improving the resilience of the agricultural sector is imperative. Therefore, as a result of their geographical penetration, smart supply chains in agriculture have attracted attention. This study examines the relationship between agricultural economic resilience and smart supply chains using panel data from 30 provinces in China from 2008 to 2021. Benchmark regression results indicate that smart supply chains promote agricultural economic resilience, depending on geographical location. The threshold model test results indicate that the integration level of rural industries and increases in farmers’ disposable income result in smart supply chains having nonlinear effects on agricultural economic resilience. Moreover, the spatial spillover effect test indicates that the development of regional smart supply chains will also promote resilience in surrounding areas. The findings provide helpful insights for sustainable agricultural development.

1. Introduction

In the past decade, frequent extreme weather, trade protectionism, etc., have had a significant impact on agricultural product prices, affecting the stabilization of the agricultural economy. As a key to China’s poverty alleviation and sustainable development, the safety and stability of agriculture cannot be ignored [1,2]. Therefore, the resilience of the agricultural economy has received much attention. “Implementation Opinions of the Ministry of Agriculture and Rural Affairs on Effectively Advancing Key Tasks in Rural Revitalization in 2023 as Deployed by the State Council” pointed out that China must continue to make agricultural supply stability and security a top priority [3]. The government must provide strong supply guarantees, construct a strong management system, and guarantee strong industrial resilience.

Agricultural economic resilience focuses on the ability of the system to resist external risk shocks and, on this basis, adapt to market environment fluctuations to adjust the industrial structure and achieve future innovation and upgrading, thereby forming an agricultural economic system with a high degree of self-repair and self-optimization capability [4]. Enhancing resilience has become an important part of agricultural modernization, especially in a modern socialist country. However, agricultural development in China is still constrained by the small-scale and scattered agricultural management structure, institutional barriers that restrict the free flow of factors, and relatively low agricultural productivity [5]. The regional collaborative risk resistance capacity and agricultural management efficiency need to be improved urgently. This study argues that scientifically evaluating agricultural economic resilience and deeply exploring its temporal and spatial changes and influencing factors are of great significance for resolving various risks in agricultural modernization [6].

The smart supply chain is based on the traditional supply chain and deeply integrates new digital technologies to form an integrated management system [7]. The concept of a smart supply chain has been highly valued by the country since it was proposed. In May 2021, the “Notice on Further Strengthening the Construction of the Agricultural Products Supply Chain System” proposed that the agricultural products supply chain system should be further strengthened and supply chain information should be carried out to accelerate the formation of an agricultural products supply chain system with closer ties between production and marketing. In August 2023, the document “Continuously promote the downward movement of supply chain, logistics distribution, goods and services, and the upward movement of agricultural products and establish a county-level commercial system within three years” pointed out that it is important to encourage rural areas with the conditions to explore and develop new models such as smart logistics.

In recent decades, the agricultural economy has developed rapidly. Many large-scale farmers represented by family farms and professional large-scale farmers have emerged under the moderate-scale operation model. Hence, there is an urgent need for a large-scale and intelligent supply chain system to follow up in coordination to assist leading enterprises, distributors, and other operators to complete resource integration and ensure its security. Meanwhile, the development of a smart supply chain relies on emerging in-formation technologies to effectively connect the business flow, logistics flow, and capital flow of all links in agricultural production and sales, realize the intelligence of all links in agriculture, and reduce agricultural operating costs and risks [8]. In view of this, building a smart supply chain to improve the agricultural economy is self-evident.

Therefore, China’s agriculture is facing increasingly complex risk shocks, and it is urgent to improve its resilience [9,10,11,12]. The smart supply chain, with its strong geographical penetration, can achieve coordinated development and optimize the rural industrial structure. Therefore, can the smart supply chain improve the agricultural economy’s resilience? What is the mechanism? In response to these issues, this study intends to examine the impact, mechanism, threshold effect, and spatial spillover characteristics between smart supply chain and agricultural economic resilience. Research on such issues will not only help to fully tap the dividend effects brought about by the upgrade to a smart supply chain but also help enhance the resilience of the agricultural economy and accelerate the realization of the strategic goal of building a “strong agricultural country”.

This paper is composed of six parts. The first part is the introduction. The second part is a literature review. The third to sixth parts are the theoretical mechanisms and hypotheses, models, the details of variables and data sources, and the results of empirical analysis, respectively. At the end, the seventh part gives the conclusion and recommendations.

2. Literature Review

Current studies on agricultural economic resilience can be divided into several categories. The first category focuses on defining agricultural economic resilience. This type of literature usually studies resilience in other fields based on the concept in traditional physics. The concept of resilience was proposed by Holling (1973) [13] to measure the ability of an ecosystem to resist and recover from uncertain shocks. Subsequently, the concept was extended to economics. Based on the perspective of evolutionary resilience, Martin and Sunley (2015) [14] defined economic resilience as the ability of the economy to resist, recover, adapt, and renew. Among the studies on agricultural economic resilience, most of them are purely theoretical. For example, Foster (2007) [15] defined agricultural economic resilience as the ability to maintain its original characteristics and not lose key system functions under external shocks.

The second body of literature selects indicators from different dimensions to establish a measurement system based on the concept of agricultural economic resilience. For example, Yao et al. (2024) [16] established three first-level indicators and fourteen second-level indicators based on risk resistance ability and used the entropy method to measure it. Zhong and Wang et al. (2024) [17] constructed an indicator system for agricultural economic resilience from three dimensions: resistance, regulation, and innovation. In terms of regional characteristics, Ye et al. (2022) [18] measured the differences in agricultural economic resilience levels in the north, central, northeast, and south of Jiangxi based on the spatial–temporal perspective, which investigated the spatial differences.

The third body of literature focuses on the impact of other agricultural factors on agricultural economic resilience. For example, many scholars found that factors such as industrial integration [15,18,19], agricultural infrastructure construction [20], and rural economic growth [21,22] can play a positive role under the condition of controlling relevant variables. Research on the smart supply chain has also gradually attracted attention from the academic community. The existing literature mainly focuses on the construction of indicator systems, the upgrading of advantages compared with traditional supply chains, and its application in different industries. In terms of indicator system construction, Wang et al. (2023) [23] integrated the four aspects of supply chain coordination to evaluate supply chain performance. Lee et al. (2024) [24] established a smart supply chain evaluation system based on the manufacturing industry and studied the role of smart technologies in the operational performance of the smart supply chain. In terms of upgrading advantages over traditional supply chains, Ning and Yao (2023) [25] believe that the application of digital technology enables the smart supply chain to achieve updates in sustainable competitive performance in the supply chain context. Moreover, Sharma et al. (2025) [26] explore the role of the smart supply chain on sustainable business performance by improving their agility and resilience from the view of natural resources.

Although the existing literature has conducted comprehensive research on the relative advantages of smart supply chains in reducing agricultural resource mismatch, expanding trade entities, and increasing farmers’ income [26,27], almost none of it has directly explained the logic and influencing path of the relationship between the smart supply chain and agricultural economic resilience. Only some scholars have paid attention to the relationship between supply chain finance and the agricultural economy [28,29] and between the digital technology economy and agricultural economic resilience [30,31]. Therefore, this study constructs a model to explore the impact and path of smart supply chains in enabling agricultural economic resilience.

Therefore, the contributions of this study are as follows: (1) Incorporating the smart supply chain and agricultural economic resilience into the same framework, analyzing their mechanism, and providing a theoretical basis for ensuring agricultural economic stability; (2) Exploring, using benchmark regression, moderation test, and threshold test, the detailed influencing mechanism and providing a new idea for promoting the construction of the smart supply chain and improving agricultural economic resilience; (3) Using regional heterogeneity and spatial spillover effect tests to analyze the impact of location factors on the correlation.

3. Research Hypotheses

3.1. The Direct Impact of Smart Supply Chains on Agricultural Economic Resilience

Agricultural economic resilience emphasizes both the resistance and resilience of agricultural sectors facing risk and its adaptability and innovation in long-term adjustment and upgrading after risks. The multi-dimensional application of the smart supply chain can achieve coordinated development from agricultural production to sales, curb a series of problems, and stimulate the vitality of agricultural product market circulation. At the same time, it increases the supply of high-quality talent, enables technological innovation in agriculture, improves resource utilization efficiency, and forms sustainable development [32,33,34]. Based on this, this study analyzes the direct impact of the smart supply chain on agricultural economic resilience from the following perspectives.

3.1.1. Strengthen the Basic Support from Data Elements

Smart supply chains are based on traditional supply chains and rely on new digital technologies to explore modes of operation. They optimize the security and stability of traditional agricultural support and enhance the ability to resist external uncertainty shocks [35]. The digital transformation of the agricultural sector has made the application of data elements more frequent in the supply chain. Data elements are less sensitive to external economic fluctuations and have strong stability. Therefore, when the agricultural sector is hit by external risks, smart supply chains can rely on its “moat effect” to strengthen stability, curb the negative impact caused by risk shocks [36], and ensure the stability of the supply and marketing level under extreme conditions. Meanwhile, the application of smart supply chains divides the agricultural production process. The status of demand entities reasonably connects the agricultural industry chain supply chain process, supply, and marketing of agricultural products, improves the connection between the main entities to ensure the stability of the supply chain, reduces the risk of chain breakage caused by shocks, and ensures resilience in a risky environment [37,38].

3.1.2. Optimize the Production Inputs by Process Upgrade

The smart supply chain plays a positive role in optimizing seedling selection at the front end of the agricultural industry, production process management in the middle, and market circulation at the back end [39], realizing the optimization of the entire process and improving efficiency. As for the front end of the sector, the embedding of smart supply chain technology enables agricultural producers to use digital channels to query the quality of seedlings, guiding the transformation of agricultural production from “experience-oriented” to “data-oriented”. As for the middle section of the industry, smart supply chain technology provides environmental information for agricultural entities and provides agricultural producers with suggestions based on land conditions and crop types, realizing the “number-based production” of the supply chain [38,40] and improving production efficiency. As for the back end of the industry, the smart supply chain activates information flow, logistics flow, and capital flow through data information elements and explores market demand through big data analysis to achieve destination-oriented transportation, effectively improving efficiency.

3.1.3. Matching Supply and Demand to Ensure Green Ecology

On the one hand, the traditional supply chain pays more attention to the quantity and price level of industrial products. As green agriculture is gradually becoming the focus of agricultural product production and consumption, consumers are shifting from pursuing “eating enough and wearing warm clothes” to “eating well and wearing fashionable clothes”. The smart supply chain drives the production and consumption of agricultural products to gradually shift to being green, healthy, and low-carbon [27,41,42]. It conforms to the emission reduction policy orientation and realizes the green development of agricultural production through the transmission of the supply chain system. On the other hand, the combination of smart supply chains and agricultural fields realizes the diversification of agricultural planting. Through the establishment of regional industrial logistics information platforms, it breaks through the regional restrictions on agricultural crop sowing selection and product sales. At the same time, combined with the green agricultural product promotion policy, it enriches agricultural resilience.

3.1.4. Stabilizes Agricultural Output Through Resource Synergy

Digital agricultural productivity can improve the level of inter-departmental and urban–rural factor resource synergy and drive steady growth in agricultural economic output [43]. The emergence of new service industries, such as rural e-commerce, rural digital logistics, and rural big data platforms, has realized the connection between the production and deep processing of primary agricultural products, promoted production conditions, and ensured the steady increase of agricultural output [44]. On the other hand, the development of the smart supply chain makes it possible for agricultural products that had high requirements for logistics speed and storage to enter the city and for new large-scale agricultural machinery and medical supplies to go to the countryside, ensuring steady progress in agricultural economic output.

3.1.5. Digital Technology Empowers Scientific and Technological Innovation Upgrades

The smart supply chain provides agricultural technology innovation in many fields. In addition to the above-mentioned technology upgrades, it also makes technological changes to the traditional supply chain in the field of commodity circulation [38]. The cloud-based “industrial ecological chain” combines social media, big data, and Internet e-commerce to create a new platform that combines multiple factors and spans multiple subjects. The combination of agricultural product information transmission, commodity trading, and logistics transmission with a C2B business model realizes the supply chain technology upgrade and innovation of the entire chain. Hence, it proposes hypothesis H1.

H1. 

The smart supply chain can promote the improvement of agricultural economic resilience.

3.2. The Moderating Effect Analysis

Rural industrial integration is based on traditional agriculture and relies on emerging technologies. Through agricultural management organizations such as small farmers and family farms, it builds a diversified new model of industrial chain extension and internal integration and promotes rural resource and factor reorganization [45]. Rural industrial integration can promote integration and penetration within and between industries and is important in stabilizing the agricultural economy. Therefore, this study intends to further examine its moderating effect on the impact of smart supply chains on agricultural economic resilience. On the one hand, it can guide the optimization of the operating structure and the coordinated upgrading of smart technology in the supply chain [46] and stimulate market competition. According to the cost-effectiveness theory, the enhancement of agricultural industry competitiveness forces various departments and entities in the agricultural industry to cooperate, boosting agricultural economic resilience. On the other hand, it also helps to match the geographical advantages of different locations with the needs of market consumption upgrades [47] and improve resilience. Therefore, this paper proposes hypothesis H2.

H2. 

Rural industrial integration has a significant moderating effect between the smart supply chain and agricultural economic resilience.

3.3. Threshold Effect Analysis

In the early stage of rural industrial integration, digital technology has infiltrated various entities of the agricultural sector, promoting agricultural production from large-scale to flexible manufacturing [48] and promoting the transformation of agricultural production entities from “single-handed” to contractual cooperation; hence, the smart supply chain greatly enhances the overall impact resistance of the agricultural economy. With the deep integration of various agricultural departments and the maturity of smart technology in the agricultural field, the marginal space for smart supply chains to exert momentum is reduced, and the role of improving agricultural economic resilience may be flat. However, the impact might also increase as the per capita income increases. Hence, this study proposes H3.

H3. 

There is a threshold effect for the impact of the smart supply chain on agricultural economic resilience.

4. Research Design

4.1. Benchmark Regression Model

Drawing on the literature [49,50], this study constructs model (1) to examine how smart supply chain construction impacts agricultural economic resilience.

$$Re{s}_{it}={\alpha }_0+{\alpha }_{1}Su{p}_{it}+\lambda Control{s}_{it}+{\tau }_i+{\mu }_t+{\epsilon }_{it}$$

(1)

The subscripts i and t represent time and province, respectively. The explained variable Res_it represents agricultural economic resilience, and the explanatory variable Sup_it represents the level of the smart supply chain; Controls_it represents control variables, which include per capita disposable income, agricultural innovation level, agricultural mechanization level, agricultural power facilities, agricultural planting structure, and ecological environment. In addition, this study also controlled a series of fixed effects, $\tau$ and $\mu$, representing individual and time fixed effects. $\epsilon$ is a random error term. This paper focuses on the significance of the coefficient in model (1). According to the previous hypothesis, if the ${\alpha }_1$ of is significantly positive, it means that the smart supply chain improves agricultural economic resilience.

4.2. Moderating Effect Model

Rural industrial integration can smooth the transmission channels and then deepen the role of the smart supply chain in promoting agricultural economic resilience. To verify hypothesis H2, this study selects rural industrial integration as a moderating variable and establishes the following model:

$$Re{s}_{it}={\alpha }_0+{\alpha }_{1}Su{p}_{it}+{\alpha }_{2}Merg{e}_{it}+{\alpha }_{3}Su{p}_{it}\times Merg{e}_{it}+\lambda Contro{l}_{it}+{\tau }_i+{\mu }_t+{\epsilon }_{it}$$

(2)

Merge_it represents the level of rural industrial integration. Sup_it × Merge_it is the cross-term of rural industrial integration and the smart supply chain, which measures the moderating effect. When ${\alpha }_3$ is positive, it indicates that rural industrial integration has a positive moderating effect on impact.

4.3. Threshold Effect Model

To measure the possible nonlinear characteristics of the impact, this study sets the following threshold model.

$$Re{s}_{it}={\alpha }_0+{\alpha }_{1}Su{p}_{it}\times I(Ad{j}_{it}\le \rho )+{\alpha }_{2}Su{p}_{it}\times I(Ad{j}_{it}>\rho )+\lambda Contro{l}_{it}+{\tau }_i+{\mu }_t+{\epsilon }_{it}$$

(3)

Here, Adj_it is the threshold variable. I (.) is the indicator function, and it takes the value 1 if Adj_it is less than the threshold ρ. Formula (3) is used to measure the single threshold, which can expand to a double or multi-threshold model if two or more thresholds exist after the tests. Other variables are the same as in Formula (1).

5. Variables and Data Source

5.1. Explained Variables

5.1.1. Variable Selection and Measurement

This study selects agricultural economic resilience as the explanatory variable. Drawing on the studies by Zhong and Wang (2024) [17] and Li et al. (2024) [51], from the three dimensions of risk resistance, adaptive adjustment, and innovative transformation, five secondary indicators and 23 tertiary indicators are selected to measure it (

Table 1. Index system of agricultural economic resilience.

First Level	Second Level	Third Level	Impact
Risk resistance	Basic support	Per capita contribution of farmers to the total agricultural, forestry, wood, and fishery production value	+
		Contribution of agricultural, forestry, fishery, and agricultural production per mu of cultivated land	+
		Number of people employed in the primary industry in rural areas	+
		The original value of productive fixed assets in agriculture, forestry, animal husbandry, and fishery owned by rural households	+
		Level of public financial support for agriculture	+
		Level of financial support for agriculture
	Production input	Crop Diversification Index	+
		Effective irrigation area	+
		Average total mechanical power per unit area	+
		Electricity consumption of rural residents in production and life	+
		The proportion of disaster-affected area to non-disaster-affected area	−
	Green ecology	Amount of agricultural fertilizer (pure) per unit sowing area	−
		Pesticide application per unit sowing area	−
		Amount of agricultural plastic film applied per unit sowing area	−
Adaptive adjustment	Agricultural output	Total output value of agriculture, forestry, animal husbandry, and fishery	+
		Total grain production	+
		GDP growth rate of primary industry	+
		Average annual wage of rural residents	+
		The proportion of agricultural product processing industry in the total agricultural output value	+
innovative transformation	Scientific & technological innovation upgrade	Agricultural R&D expenditure	+
		Rural broadband network penetration rate	+
		Number of Taobao Villages	+
		Recreational agriculture business income	+

).

5.1.2. Calculation of Agricultural Economic Resilience

Using the index system in Table 1 and employing the entropy method, the agricultural economic resilience in 30 provinces over the country across the years is calculated. Considering the length of the article,

Table 2. Provincial agricultural economic resilience across the years.

Province	2009	2011	2013	2015	2017	2019	2021	Average	Rank
Beijing	0.07	0.07	0.09	0.10	0.11	0.14	0.15	0.10	24
Tianjin	0.17	0.18	0.08	0.09	0.10	0.11	0.14	0.12	17
Hebei	0.14	0.16	0.18	0.20	0.21	0.24	0.29	0.20	7
Shanghai	0.08	0.09	0.14	0.15	0.16	0.18	0.17	0.13	13
Jiangsu	0.18	0.22	0.26	0.30	0.33	0.40	0.38	0.29	1
Zhejiang	0.13	0.17	0.18	0.23	0.30	0.41	0.46	0.26	3
Fujian	0.09	0.11	0.13	0.15	0.16	0.19	0.23	0.14	12
Shandong	0.18	0.21	0.24	0.27	0.31	0.35	0.41	0.27	2
Guangdong	0.14	0.17	0.18	0.21	0.26	0.33	0.38	0.23	4
Hainan	0.05	0.08	0.09	0.13	0.16	0.19	0.12	0.11	20
Shanxi	0.06	0.07	0.08	0.10	0.10	0.12	0.10	0.09	28
Anhui	0.10	0.12	0.16	0.18	0.20	0.22	0.23	0.17	8
Jiangxi	0.08	0.09	0.10	0.11	0.12	0.13	0.16	0.11	22
Henan	0.16	0.17	0.19	0.20	0.21	0.24	0.28	0.20	6
Hubei	0.11	0.13	0.14	0.16	0.17	0.19	0.21	0.15	11
Hunan	0.11	0.12	0.14	0.16	0.18	0.20	0.23	0.16	10
Neimenggu	0.08	0.10	0.11	0.13	0.14	0.16	0.18	0.12	15
Guangxi	0.08	0.09	0.10	0.11	0.13	0.14	0.16	0.11	18
Chongqing	0.06	0.07	0.10	0.12	0.14	0.13	0.16	0.11	23
Sichuan	0.12	0.15	0.18	0.21	0.24	0.27	0.32	0.20	5
Guizhou	0.06	0.07	0.08	0.09	0.10	0.12	0.13	0.09	26
Yunnan	0.07	0.09	0.10	0.11	0.12	0.14	0.16	0.11	19
Shaanxi	0.06	0.08	0.09	0.10	0.11	0.12	0.14	0.10	25
Gansu	0.06	0.07	0.08	0.09	0.09	0.11	0.12	0.08	29
Qinghai	0.07	0.05	0.06	0.07	0.08	0.11	0.13	0.08	30
Ningxia	0.14	0.17	0.06	0.06	0.06	0.08	0.10	0.09	27
Xinjiang	0.08	0.09	0.11	0.13	0.14	0.16	0.18	0.12	16
Heilongjiang	0.10	0.13	0.17	0.17	0.18	0.20	0.22	0.16	9
Jilin	0.07	0.08	0.10	0.11	0.12	0.14	0.16	0.11	21
Liaoning	0.09	0.12	0.13	0.14	0.14	0.14	0.15	0.13	14
Average	0.10	0.12	0.13	0.15	0.16	0.19	0.21	0.14	/

only lists the results every two years from 2009 to 2021. As shown in Table 2, national agricultural economic resilience shows an overall upward trend. The average increase every two years is 18.33%, which shows that agricultural economic resilience has steadily increased, and the trend of continuous improvement is obvious. In terms of provinces, the average value of agricultural economic resilience in Jiangsu is highest (0.29). In recent years, China has been committed to solving the “three rural” problems and has implemented policies such as the construction of modern rural infrastructure, which has significantly improved agricultural economic resilience. In general, resilience in eastern China has improved rapidly, mainly concentrated in Jiangsu, Shandong, and other provinces. Improvement in central China has been slow, and, as of 2021, some provinces still have poor resilience. Overall, since 2009, the speed of improvement in agricultural economic resilience has shown a distribution of high around the four sides and low in the middle. The eastern provinces generally reached a high level around 2017, and the northern, northeastern, southwestern, and northwestern regions showed a relatively high level of agricultural economic resilience around 2021, while, as of 2021, some central provinces still have low levels.

5.2. Core Explanatory Variables

This study selects the smart supply chain as the explanatory variable, which is the Sup_it in equation (1). Referring to the measurement of the level of smart supply chain by Li et al. (2022) [52], an index system based on three dimensions and twenty secondary indicators is constructed. The specific indicator selection and measurement are in

Table 3. Indicator system of smart supply chain.

First Level	Second Level	Impact
Supply chain foundation	Road density	+
	Railway density	+
	Density of inland waterways	+
	Investment in transportation, warehousing, and postal services	+
	Transport, warehousing, and postal workers	+
	Logistics and warehousing land area	+
	Traffic congestion	−
Supply chain collaboration	Agricultural value added	+
	Manufacturing value added	+
	Cargo transportation volume	+
	Cargo turnover	+
	E-commerce sales	+
	E-commerce purchases	+
	Proportion of e-commerce transaction activities	+
Supply chain innovation	Number of transportation science and technology institutions	+
	Number of transportation research laboratories and research centers	+
	Whether a national pilot city for supply chain innovation and application	+
	Mobile device penetration	+
	Number of Internet users	+
	Total number of IT professionals	+

. Then, by using the index system in Table 3 and employing the entropy method, the level of the smart supply chain in 30 provinces across the years is calculated.

5.3. Moderating Variables

This study selects rural industrial integration as a moderating variable, which is an important path to achieve the steady improvement of the rural economy. Building a rural primary, secondary, and tertiary industry integrated development system will help enhance the cohesion of the agricultural and rural economy and is crucial to improving the overall ability of agriculture to resist risk shocks. Considering the possible moderating effect of an industry integrated development process, this paper draws on Wang and Li (2019) [53] and uses the quotient of the proportion of regional agricultural output value in the national agricultural output value and the proportion of the province’s total output value in the national total output value for measurement.

5.4. Control Variables

According to the literature [5,54,55], this study selects the agricultural innovation level (Inn), agricultural mechanization level (Mech), rural power facilities (Elec), agricultural planting structure (Stru), and ecological environment (Eco) as control variables. Among them, agricultural innovation level (Inn) is measured by the number of agricultural utility model patent applications; agricultural mechanization level (Mech) is measured by the ratio of total power of agricultural machinery to the number of people employed in the primary industry; agricultural power facilities (Elec) are measured by rural per capita electricity consumption; agricultural planting structure (Stru) is measured by the ratio of grain planting area to total planting area; ecological environment (Eco) is measured by the ratio of regional soil erosion control area to the total area of urban area.

5.5. Data Sources and Descriptive Statistics

5.5.1. Data Sources

Considering data availability, this study selects data from 30 provinces from 2008 to 2021 as samples. The data come from “China Statistical Yearbook”, “China Rural Statistical Yearbook”, “Digital Inclusive Finance Index”, “China Agricultural Yearbook”, “China Electronic Information Industry Statistical Yearbook”, and “China Transportation Statistical Yearbook”, among others. Individual missing values are filled by interpolation.

5.5.2. Descriptive Statistics

The descriptive statistics are in

Table 4. Descriptive statistics.

Type	Name	Variables	Obs.	Mean	St.d.	Min.	Max.
Explained	Res	Agricultural economic resilience	420	0.233	0.098	0.060	0.082
Explanatory	Sup	Smart supply chain	420	0.134	0.114	0.010	0.840
Moderating	Merge	Rural industrial integration	420	0.144	0.111	0.010	0.560
Control	LnPgdp	Per capita disposable income	420	9.662	0.942	6.799	11.731
	Inn	Agricultural innovation level	420	17.700	23.573	0.000	159.000
	Mech	Agricultural mechanization level	420	4.352	2.168	0.943	13.396
	Elec	Agricultural power facilities	420	19.984	51.509	1.130	412.75
	Stru	Agricultural planting structure	420	0.660	0.142	0.355	0.971
	Eco	Ecological environment	420	1.015	1.143	0.001	10.182

.

6. Results of Empirical Analysis

6.1. Benchmark Regression

According to the test for autocorrelation and the Hausman test, there was no significant correlation between the variables, and the fixed effect model could be employed.

Table 5. Benchmark regression.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
Sup	0.508 *** (0.046)	0.368 *** (0.053)	0.418 *** (0.050)	0.406 *** (0.050)	0.398 *** (0.052)	0.421 *** (0.053)
Inn		0.001 *** (0.000)	0.001 *** (0.000)	0.001 *** (0.000)	0.001 *** (0.000)	0.001 *** (0.000)
Mech			0.010 *** (0.001)	0.010 *** (0.001)	0.011 *** (0.002)	0.010 *** (0.002)
Elec				0.000 ** (0.000)	0.000 ** (0.000)	0.000 * (0.000)
Stru					−0.026 (0.043)	−0.061 (0.047)
Eco						−0.004 * (0.002)
Constant	0.036 *** (0.007)	0.050 *** (0.007)	0.014 * (0.009)	0.014 * (0.009)	0.032 (0.030)	0.057 * (0.033)
Fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Obs.	420	420	420	420	420	420
R2	0.714	0.733	0.763	0.765	0.766	0.768

Note: The value in brackets are standard errors. *, ** and *** indicate significance at the 10%, 5% and 1% levels, respectively.

presents the results of the benchmark regression, with columns (1) to (6) gradually adding control variables. The results show that the coefficient of the smart supply chain has no significant change before and after the addition of control variables. That is, under the premise of controlling related interference items, the smart supply chain plays a stable and positive role in promoting agricultural economic resilience. From an economic point of view, every 1% increase in the level of the smart supply chain increases the resilience of the agricultural economy by 0.421%. The smart supply chain can drive the steady improvement of agricultural economic resilience. Hence, H1 is established. Besides, in terms of control variables, taking column (6) as an example, the coefficients of agricultural innovation, agricultural mechanization, and agricultural power facilities are significantly positive, which indicates that they can improve agricultural economic resilience. The ecological environment is negative, but the significance level is only at the 10% level.

6.2. Robustness Test

6.2.1. Considering Endogeneity

Considering potential endogeneity, this study selects instrumental variables (IV) and uses the original least squares (OLS) method to perform robustness tests. This study selects a one-period lagged smart supply chain as IV (Instru) and performs robustness tests through OLS. The results of IV are shown in columns (1) and (2) of

Table 6. Robustness test.

	(1) Sup	(2) Res	(3) Res	(4) Res
Instru	1.067 *** (0.000)
Sup		0.326 *** (0.028)	0.575 *** (0.148)	0.397 *** (0.052)
Controls	Yes	Yes	Yes	Yes
Fixed effects	Yes	Yes	Yes	Yes
Obs.	390	390	120	420
R2	0.995	0.776	0.590	0.779

Note: The value in brackets are standard errors. *** indicate significance at the 1% levels.

. The coefficients of the smart supply chain and IV are significantly positive, and the results of other control variables are consistent with previous results. Furthermore, the Kleibergen-Paap rk Wald F value of the test is 13,354.9 > 10, and the unidentifiable test Anderson LM statistic p value is less than 0.01. It shows that there is no weak IV problem, the null hypothesis of “insufficient identification of instrumental variables” is rejected, and the IV method test is passed.

6.2.2. Narrowing the Sample Interval

Referring to Tan et al. (2023) [56], the interval was narrowed to 2017–2020 to examine whether the role of the smart supply chain in promoting agricultural economic resilience during the construction period is still significant. The test results are shown in column (3) of Table 6. The coefficient of smart supply chain is consistent with the previous text.

6.2.3. Eliminating Outliers

This study performs 1% and 99% tailing processing on the samples to eliminate the possible influence of outliers; the results are in column (4) of Table 6. It shows that the results after tailing processing are consistent with the previous text. Comparing the above results, it shows that the main conclusions are still valid.

6.3. Moderating Effect Test

To further explore how rural industrial integration affects the impact of the smart supply chain on agricultural economic resilience, this study uses rural industrial integration as a moderating variable; results are in

Table 7. Moderating Effect Test.

Variables	Res
Sup	0.287 *** (0.071)
Merge	−0.117 *** (0.044)
Sup × Merge	0.396 *** (0.136)
Controls	Yes
Constant	0.080 ** (0.036)
Fixed effects	Yes
Obs.	420
R2	0.773

Note: The value in brackets are standard errors. ** and *** indicate significance at the 5% and 1% levels, respectively.

. The coefficient of the interaction term between the smart supply chain and rural industrial integration (Sup × Merge) is significantly positive, which shows that rural industrial integration significantly promotes the impact of the smart supply chain on agricultural economic resilience, and H2 is established.

6.4. Threshold Effect Test

According to the theoretical analysis, the smart supply chain may have nonlinear effects on agricultural economic resilience. Therefore, this study selects relevant variables for the threshold effect test, and the results are in

Table 8. Results of threshold effect test.

Variables	Threshold	F Value	p Value	Threshold Value	Critical Values			BS Times
					10%	5%	1%
Merge	Single	48.06	0.063	0.240	37.996	51.913	72.395	300
	Double	30.88	0.057	0.270	25.108	31.439	46.362	300
LnGdp	Single	51.86	0.077	10.724	47.629	55.478	77.820	300

. It found that rural industrial integration passed the double threshold test and per capita disposable income passed the single threshold test. Therefore, the corresponding threshold model was applied for regression, and the results are in

Table 9. Threshold regression results.

Variables	(1)	(2)
	Merge	LnPgdp
Sup·I (Adj ≤ 0.240)	0.437 *** (0.042)
Sup·I (0.240 < Adj ≤ 0.270)	0.618 *** (0.050)
Sup·I (Adj > 0.270)	0.509 *** (0.041)
Sup·I (Adj ≤ 10.724)		0.371 *** (0.042)
Sup·I (Adj > 10.724)		0.505 *** (0.041)
Controls	Yes	Yes
Fixed effects	Yes	Yes
Obs.	420	420
R2	0.790	0.791

Note: The value in brackets are standard errors. *** indicate significance at the 1% levels.

.

Column (1) of Table 9 lists the results with rural industrial integration as the threshold variable. After crossing the left threshold, the coefficient of the smart supply chain increased from 0.437 to 0.618 and slightly decreased after crossing the right threshold, but it always remained significant. Column (2) shows the results with per capita disposable income as the threshold variable. After crossing the threshold, the coefficient of the smart supply chain increased from 0.371 to 0.505 and always remained significant. As rural industrial integration deepens, the impact gradually increases. As industrial integration reaches a certain threshold, the dividend effects gradually weaken. However, as per capita disposable income reaches a certain level, the stimulation effect of the smart supply chain increases. Therefore, hypothesis H3 is established.

6.5. Heterogeneity Test

To test the heterogeneity of the impact of smart supply chains on agricultural economic resilience among regions, according to the division of China’s geographical location by the National Bureau of Statistics, 30 provinces were divided into the eastern, the central, and the northeastern regions for the heterogeneity test. The results are shown in

Table 10. Heterogeneity test.

Variables	(1)	(2)	(3)	(4)
	Eastern	Central	Western	Northeastern
Sup	0.276 *** (0.092)	0.303 *** (0.070)	1.037 *** (0.142)	0.182 (0.137)
Constant	0.068 (0.066)	0.118 *** (0.031)	−0.063 (0.069)	0.039 (0.144)
Controls	Yes	Yes	Yes	Yes
Fixed effects	Yes	Yes	Yes	Yes
Obs.	140	84	154	42
R2	0.852	0.981	0.741	0.993

Note: The value in brackets are standard errors. *** indicate significance at the 1% levels.

. According to columns (1) to (4), the coefficients of smart supply chains in the eastern, the central, and the western regions are all significantly positive, while those in the northeastern region are not significant. Among them, the coefficient in the western region is the largest (1037), significantly higher than that in the central and western regions, while that in the northeastern region is not significant. The reason may be that the smart supply chains in the eastern region started earlier, the infrastructure construction is more complete, and the integration with agriculture is relatively close. The further development of smart supply chains has a limited effect on improving agricultural economic resilience; the western region has a vast territory and has many undeveloped and relatively suitable grain planting environments. Therefore, the upgrade to smart supply chains has a greater impact on the agricultural economic resilience in the western region; while the northeastern region is restricted by production environment and other conditions, the impact of smart supply chains is not significant.

6.6. Spatial Spillover Effect

6.6.1. Spatial Spillover Effect Test

To further explore the spatial correlation between the smart supply chain and agricultural economic resilience, this study uses the economic geography nested matrix to test whether there is spatial autocorrelation between them in each province through Moran’s I, and the results are shown in

Table 11. Moran Index.

Year	Res			Sup
	I	Z	P	I	Z	P
2008	0.009	0.903	0.367	0.122	3.332	0.001
2009	0.008	0.879	0.380	0.122	3.335	0.001
2010	0.030	1.324	0.186	0.121	3.291	0.001
2011	0.037	1.465	0.143	0.118	3.218	0.001
2012	0.043	1.616	0.106	0.112	3.092	0.002
2013	0.050	1.766	0.077	0.101	2.878	0.004
2014	0.050	1.762	0.078	0.107	2.980	0.003
2015	0.062	2.013	0.044	0.114	3.134	0.002
2016	0.053	1.837	0.066	0.097	2.769	0.006
2017	0.066	2.091	0.037	0.094	2.702	0.007
2018	0.070	2.182	0.029	0.091	2.649	0.008
2019	0.073	2.259	0.024	0.098	2.804	0.005
2020	0.072	2.227	0.026	0.102	2.886	0.004
2021	0.054	1.842	0.065	0.108	2.989	0.003

. Except for some years, the Moran indices are significantly positive. This shows that there are positive spatial effects between them, and the spatial econometric model should be further used to examine its spillover effects.

6.6.2. Spatial Spillover Effect Test

To select a suitable model, this paper conducts the LM and LR tests. The results of the LM test indicate that there are spatial lag and spatial error terms. Moreover, the result of the LR test is significant. Hence, the spatial Durbin model is used; results are in

Table 12. Spatial Durbin test results.

Variables	(1)	(2)	(3)	(4)
Sup	0.556 *** (0.047)
Jc		0.709 *** (0.124)
Xt			0.348 *** (0.028)
Cx				0.208 *** (0.037)
W × Sup	1.804 *** (0.226)
W × Jc		0.566 (0.538)
W × Xt			0.836 *** (0.141)
W × Cx				1.372 *** (0.167)
Controls	Yes	Yes	Yes	Yes
Fixed effects	Yes	Yes	Yes	Yes
Log-likelihood	1068.914	1008.715	1063.670	1031.603

Note: The value in brackets are standard errors. *** indicate significance at the 1% levels.

. The smart supply chain is likely to affect the agricultural economic resilience of surrounding areas. According to the above theoretical analysis, there are three sub-indices of the smart supply chain, namely supply chain foundation (Jc), supply chain collaboration (Xt), and supply chain innovation (Cx). This paper uses the spatial Durbin model for testing (results in

Table 13. Decomposition of spatial spillover effects.

Variables	Direct	Indirect	Total
Sup	0.490 *** (0.052)	0.714 *** (0.129)	1.204 *** (0.121)
Jc	0.708 *** (0.129)	0.340 (0.465)	1.049 ** (0.463)
Xt	0.325 *** (0.030)	0.352 *** (0.084)	0.677 *** (0.082)
Cx	0.165 *** (0.039)	0.819 *** (0.129)	0.984 *** (0.131)

Note: The value in brackets are standard errors. ** and *** indicate significance at the 5% and 1% levels, respectively.

). The coefficients of the smart supply chain and its sub-indices are significantly positive, which indicates promoting effects. In addition to the supply chain foundation, the spatial lag term coefficients of the smart supply chain and other sub-indices are significantly positive, which indicates that the development of smart supply chains and smart technologies has led to the improvement of resilience in neighboring provinces. The development of smart supply chains will attract high-quality logistics companies and agricultural science & technology companies to gather in the region, thereby improving resilience and promoting surrounding areas. However, the level of supply chain infrastructure development is limited by local transportation and government policies, which cannot be improved in the short term.

The decomposition of spatial spillover effects is in Table 13. The direct effect of the smart supply chain and its sub-indices on resilience is significantly positive. Except for the supply chain foundation, the coefficients of the smart supply chain and other sub-indices are significantly positive, which indicates that the smart supply chain in surrounding areas will drive improvement, while the process of supply chain foundation requires more resource support and coordination with local government policies and cannot effectively drive improvement in the short term. In terms of the total effect, the coefficients of the smart supply chain and its sub-indices are significantly positive, further verifying the positive spatial spillover effect of the smart supply chain on agricultural economic resilience in surrounding areas.

7. Conclusions and Policy Recommendations

7.1. Conclusions

Based on the panel data from 2008 to 2021, this paper analyzes the impact of the smart supply chain on agricultural economic resilience and further examines the specific action path and impact mechanism. The findings are as follows. First, the smart supply chain has a significant effect on agricultural economic resilience, and this conclusion passed several robustness tests. Second, there is a significant moderating effect; that is, rural industrial integration can enhance the role of the smart supply chain in agricultural economic resilience. Third, there is a significant threshold effect in the impact of the smart supply chain on agricultural economic resilience. Fourth, from the perspective of spatial heterogeneity analysis, the western region shows the strongest promoting effect of the smart supply chain. Fifth, the spatial Durbin test found that the impact of the smart supply chain has a spatial spillover effect, and the smart supply chain can simultaneously promote resilience in local and surrounding areas.

7.2. Policy Recommendations

First, the government should increase investment in new rural infrastructure and promote the modernization of agricultural sector governance capabilities. Meanwhile, the local government should accelerate the scientific layout of agricultural and rural smart logistics parks and smart warehousing in rural areas, make up for the shortcomings of hardware facilities in rural supply chains, and enhance agricultural economic resilience with new technologies. Supply chain-related companies should take the initiative to accelerate the integrated application of emerging technologies in the field of agricultural product supply chains to realize the real-time, visible, and perceptible supply chain and drive the steady improvement of rural economic resilience.

Second, the government should expand the scale of agricultural sector integration and promote the deep integration and development of the industrial chain. The government should encourage the deep integration of various entities and departments of the agricultural sector and create an industrialization consortium. It should help leading agricultural industrialization enterprises to establish large agricultural enterprise groups, encourage them to cooperate with different industrial sectors, effectively reduce the time and space distance between different industrial sectors, and achieve the safe and stable development of the agricultural economy. Meanwhile, local governments should make rational use of resource advantages to develop new leisure agricultural formats and enhance economic strength.

Third, a supply chain technology-sharing platform should be created to achieve deep integration among regions. The construction of an enterprise-led, industry-university-research-application cooperation supply chain innovation network should be encouraged, and agricultural product service technology sharing mechanisms provided. Moreover, cross-industry and cross-field supply chain technology collaboration platforms should be established to narrow the differences in resilience among regions. In addition, the government should encourage social capital in the central and western regions to establish supply chain innovation industry investment funds, provide financing support for local enterprises to carry out supply chain innovation, and drive the stable development of China’s agricultural economy.

7.3. Research Limitations

This research investigates the impact of the smart supply chain on agricultural economic resilience and examines the detailed mechanisms. The results are robust, and the mechanisms are clear. However, it also has some limitations. For example, the indicator system of measuring the level of the smart supply chain constructed in this study is probably not that comprehensive and is subject to certain constraints, which may influence agricultural economic resilience. Moreover, there might exist some other indirect influencing paths and threshold effects that are not involved in this study. In addition, this paper mainly focuses on the correlation between smart supply chains and agricultural economic resilience in China, while future research can also pay attention to related studies in other developing countries to explore related issues, especially in large agricultural countries like India, Vietnam, Thailand, and so on.