Sage Journals: Discover world-class research

Abstract

Drawing on panel data collected from 281 Chinese cities at the prefecture level between 2012 and 2020, this study develops a multidimensional food security index that accounts for four key aspects: usability, availability, sustainability, and stability. Initially, the research investigates spatial distribution, evolutionary patterns, and β-convergence dynamics of this index. Subsequently, it empirically evaluates how Digital Inclusive Finance (henceforth named DIF) influences food security, delving into underlying mechanisms, regional disparities, and spatial spillover effects. Results demonstrate that: (1) DIF significantly strengthens food security, a finding supported by comprehensive robustness checks and endogeneity tests. (2) The beneficial influence of DIF operates through advancing industrial structures and narrowing income disparities. (3) The impact is particularly strong in major grain-producing regions, areas with substantial agricultural insurance investment, low fluctuation, and humid or semi-humid climates. Conversely, impacts are minimal or insignificant in non-major grain-producing regions, regions with limited insurance inputs, higher fluctuation, and arid or semi-arid climates. (4) Spatial spillover assessments, employing a rural revitalization-weighted matrix, reveal adverse spillover effects in central and western China, while the eastern region shows no meaningful spatial transmission.

Keywords

digital inclusive finance food security industrial structure upgrading income gap spatial spillover effect

Introduction

Global food security remains challenged by multiple persistent threats, such as the COVID-19 pandemic, extreme climatic events, and geopolitical conflicts. These developments have intensified food insecurity both within China and globally (Fan et al., 2024). Food security refers to a state where every person has consistent physical, social, and economic means to obtain sufficient, safe, and nourishing food that meets both their nutritional needs and personal tastes, enabling them to maintain a vibrant and healthy life. According to the Food and Agriculture Organization (FAO), about 205 million individuals across 45 nations and regions faced severe food insecurity at “crisis” levels or worse in 2022. In China, food security faces four major challenges: usability, availability, stability, and sustainability (Lee et al., 2024a). Regarding usability constraints, profit incentives increasingly threaten grain production systems, as exemplified by the 2022 wheat-to-silage diversion in Shandong and other regions. Such incidents demonstrate how economic pressures may redirect critical ration grain toward non-food uses, directly undermining production stability and self-sufficiency (Jin et al., 2023). In terms of availability, persistent weaknesses in the grain distribution system, such as insufficient storage facilities and underdeveloped logistics, have increased the risk of spoilage and food loss. Regarding stability, China’s growing reliance on grain imports has made it increasingly vulnerable to international market fluctuations (Y. Liu & Zhou, 2021). Furthermore, the national grain early warning system remains outdated, lacking a unified, standardized monitoring platform and a science-based alert mechanism. With respect to sustainability, the overall quality of China’s arable land has declined, with black soil degradation in the northeast, acidification in the south, and salinization in the north affecting nearly one-third of total cultivated land (Norse & Ju, 2015; Ye & Van Ranst, 2009). Food security is closely tied to national welfare and livelihoods, requiring continuous improvements in these four dimensions to mitigate domestic and international risks (Griggs et al., 2013). Therefore, safeguarding food security has become an urgent priority in China’s high-quality development agenda.

A key underlying factor contributing to China’s food security challenges is the systemic financial gap. Finance has been an essential wellspring for agricultural growth and food security measures, yet the present level of financial backing falls short in meeting the growing needs of the agri-industry sector. Despite this, farmers are still grappling with challenges such as insufficient safeguarding, restricted coverage, and steep obstacles to accessing affordable credit. Given the pressing need for sustainable solutions to enhance finance’s role in ensuring food security, Digital Inclusive Finance (henceforth named DIF) emerges as a promising innovation. By leveraging cloud computing, big data analytics, and internet platforms, this fintech innovation reconfigures traditional financial inclusion paradigms to address systemic gaps in agricultural financing. By significantly expanding both the reach and depth of conventional inclusive banking solutions, this innovative model has emerged as the prevailing paradigm in modern finance (Huang & Wang, 2022). Its positive influence on agricultural advancement is well documented (Zhao et al., 2022). Specifically, DIF addresses farmers’ financial constraints in upgrading equipment, as well as enhancing storage, processing, and transportation infrastructure, thus improving food usability, availability, and stability. Furthermore, it fosters technological progress in agriculture, curtails dependence on chemical fertilizers, and alleviates adverse environmental effects, thereby reinforcing sustainability aspects of food security (J. Ma & Li, 2021). Consequently, the growth of DIF offers a viable and effective strategy for tackling contemporary food security challenges.

Literature Review

The relevant literature for this study can be categorized into two primary streams. The first is about food security, this study categorizes the existing literature into three main facets: definition, measurement, and influencing factors. In defining food security, scholars have expanded upon the FAO’s interpretation by emphasizing three critical components: availability, access, and utilization (Ingram, 2011). An alternative perspective posits that food security comprises ensuring the availability of safe, nutritionally adequate food at all times and the ability to access required sustenance through socially acceptable means (Coleman-Jensen et al., 2016). Food security measurement methodologies vary widely. Most studies use singular metrics like yield per hectare (Bouteska et al., 2024), while others employ adapted instruments such as the USDA food security module (H. Lu & Carter, 2024). Cross-national research frequently relies on the World Development Indicators’ food production index (Ashraf & Javed, 2023; Subramaniam & Masron, 2021). More comprehensive approaches include J. T. Gao et al. (2020)'s DEA model with first-level indicators covering import dependence, production control, living environment, and international competitiveness. While existing research has yielded substantial progress in developing food security evaluation systems, several issues remain, including indicator redundancy, high subjectivity in weight assignment, and insufficient data dimension. Concerning the factors influencing food security, existing studies primarily examine the roles of government policy, urbanization, and environmental conditions. First of all, in the policy domain, China has recently implemented a range of measures, including agricultural subsidies and land transfer initiatives. Agricultural subsidies have been shown to significantly enhance food security in China (M. Gao & Wang, 2021). However, research findings on the impact of land transfer remain inconclusive (Z. Li et al., 2021; Jin et al., 2023). Internationally, Ethiopia’s Productive Safety Net Program (PSNP) effectively mitigated the adverse effects of the COVID-19 pandemic on food security, providing substantial support to impoverished households and those in remote regions (Abay et al., 2023). Secondly, regarding urbanization, research presents divergent perspectives. In the case of Ethiopia, Abebe (2024) reported that rapid urban expansion leads to reduced food production and higher food prices, thereby undermining regional food security. Conversely, some scholars argue that although urbanization encroaches on arable land, advances in agricultural technology can significantly enhance productivity and the quality of agricultural products, thus mitigating negative impacts on food security (S. Li et al., 2023). Third, in terms of ecological factors, existing research predominantly focuses on how changing climate patterns impact the stability of food supply systems. Scholars broadly recognize that erratic weather conditions pose significant challenges to global food networks (Cui & Zhong, 2024; Hadley et al., 2023; Lee et al., 2024b). Although earlier studies have identified the potential drawbacks of climate change mitigation strategies, they often fall short in distinguishing the distinct effects of individual policies under various scenarios. Addressing this gap, Fujimori et al. (2022) found that afforestation has a more pronounced impact on food security compared to strategies aimed at reducing non-CO₂ emissions.

Concerning DIF, it is formally defined by the World Bank’s Consultative Group to Assist the Poor (CGAP) as digital delivery of formal financial services to populations excluded or underserved by traditional financial systems. Chinese scholarship conceptualizes DIF as an innovative delivery model leveraging digital technologies to democratize financial access. Through mobile payments, virtual banking, and similar innovations, it fundamentally addresses financial inclusion barriers, particularly universality and affordability constraints (H. M. Tang & Zhao, 2022). In terms of measurement, most existing studies utilize the Digital Inclusive Finance Index developed jointly by Peking University and Ant Financial (J. Li et al., 2020; Jiao et al., 2024). Some researchers have further refined this approach by constructing county-level DIF indices to capture localized development trends (Feng et al., 2021). The literature broadly concurs on the positive influence of DIF across multiple sectors. In the Chinese context, DIF contributes significantly to macro-level outcomes such as economic growth (E. Li et al., 2024), green inclusive development (E. Li et al., 2024), and the pursuit of common prosperity (D. Guo et al., 2024). At the micro level, it also facilitates corporate digital transformation (B. Guo et al., 2023), encourages green innovation (K. Ma, 2023), and improves ESG performance (H. Lu & Cheng, 2024). Cross-country evidence confirms digital finance’s growth-enabling role. Daud and Ahmad (2023), analyzing 84 post-financial crisis economies, identify digital-inclusive financial integration as a significant economic catalyst. Complementing this, Frimpong et al. (2022) demonstrate that digital finance solutions substantially enhance SME viability in Ghana’s Central Region. Their findings further establish financial literacy as a key moderating factor in entrepreneurs’ effective leveraging of these tools. Focusing on agricultural development, existing research indicates that DIF alleviates resource mismatches in agriculture (Hong et al., 2024), supports the sector’s digital transformation (X. Liu et al., 2023), and enhances overall agricultural productivity (Cao & Wang, 2024). In terms of promoting green agricultural development, DIF contributes to lower agricultural carbon emissions by fostering rural entrepreneurship, green agribusiness activities, and agricultural innovation (X. Zhang & Li, 2025), thereby supporting sustainable development goals (J. Guo et al., 2024). Furthermore, Jiang et al. (2025) revealed that DIF and its specific dimensions significantly enhance the development of agricultural producer services, underscoring its broad influence across the agricultural value chain.

Despite extensive prior research, several gaps persist in the existing literature. First, many studies continue to rely on a singular index to assess food security, often characterized by subjective weighting and limited detail. Second, there is insufficient analysis of the spatial distribution and evolutionary patterns of food security in China. Third, as the literature review reveals, current research on food security determinants predominantly examines policy, urbanization, and ecological factors, with limited attention to DIF. Crucially, few studies analyze the relationship between multidimensional food security and DIF. While existing DIF research focuses on economic growth and poverty alleviation, its implications for agriculture, particularly its causal mechanisms affecting food security, remain underexplored. There are still research gaps in the intersection of these two major themes that need to be filled. Thus, given the rapid development of DIF and the urgency of food security issues, several critical questions arise: How can food security be more accurately measured? What are its spatial characteristics, temporal dynamics, and convergence trends? Can DIF effectively enhance food security, and through what mechanisms? Does its impact vary across regions? Furthermore, considering the spatially expansive nature of DIF, do spatial spillover effects influence food security outcomes? This study analyzes 2012 to 2020 panel data from 281 Chinese prefecture-level cities. Against mounting food security pressures and rapid DIF expansion, we demonstrate DIF’s critical role in enhancing food security—primarily through industrial upgrading and bridging urban-rural income gap. These findings offer new pathways for optimizing China’s food policies and contribute actionable insights for global food security strategies aligned with sustainable development goals.

This paper makes several key contributions to the field. First, existing research on food security assessment is beset with notable limitations, including an overemphasis on singular metrics, inadequately detailed indicators, and substantial bias in assessment methodologies. To address these gaps, our study proposes a comprehensive food security framework that assesses four key dimensions, usability, availability, sustainability, and stability, thereby offering a more nuanced and comprehensive analytical approach. Secondly, while considerable research has investigated DIF and food security as distinct fields, few studies have probed the nexus between these two critical domains. This study addresses this research gap by examining the impact of digital inclusive financial on food security, disentangling the mediating mechanisms and heterogeneities underlying this relationship. Thirdly, the spatial interrelationship between DIF and food security remains underexplored in existing literature. This paper tackles this research gap by employing the Spatial Durbin Model to examine how digital financial inclusion generates spillover effects that impact food security in adjacent regions. By adopting this spatial econometric approach, the study not only enriches the methodological framework but also affords deeper insights into the regional propagation of these economic benefits.

The remainder of this paper is structured as follows. Section 2 provides a literature review of existing relevant studies. Section 3 presents the theoretical framework and formulates the research hypotheses. Section 4 details the methodology and experimental design. Section 5 offers an in-depth empirical analysis of the findings. Section 6 investigates heterogeneity effects across different subgroups. Section 7 conducts additional analyses to further validate the results. The paper concludes with a summary of key findings and proposes actionable policy recommendations.

Theoretical Analysis and Research Hypothesis

DIF impacts food security through its unique advantages and characteristics. DIF distinguishes itself from conventional banking by leveraging advanced technologies such as cloud computing and big data analytics. This innovative model substantially reduces operational costs, enhances financial accessibility, and streamlines the delivery of financial services to end-users—all while upholding strict regulatory standards. In response to challenges such as increasing capital demands and rising input costs in agriculture, DIF can alleviate farmers’ financial constraints, support productive investment, and enhance their motivation to produce (Osabohien et al., 2020). Moreover, it helps farmers manage market risks and mitigate potential losses arising from price fluctuations, thereby acting as a stabilizing mechanism and reinforcing food security (Chen, 2022). Furthermore, by harnessing information technology, DIF improves access to traditional grain market services and real-time information, thereby alleviating usability issues and reducing post-harvest losses associated with information asymmetry—ultimately contributing to improved food security. In addition, it leverages the Internet’s “network effect” to reinforce agricultural digital platforms, integrating online and offline channels to deliver advanced green production technologies and more efficient input procurement options. This dual-channel approach facilitates the diffusion of sustainable agricultural practices (S. Xu & Wang, 2023), promoting the long-term development of agriculture and supporting overall food security. Accordingly, Hypothesis 1 was proposed:

H1: DIF can significantly improve the level of food security.

The rise of DIF represents more than mere expansion of financial access—it denotes a quantum leap in operational efficiency. Characterized by its broad accessibility and deep market penetration, this digital transformation optimizes resource allocation, drives regional innovation, and enhances consumer purchasing power. Collectively, these effects generate robust momentum for industrial restructuring (Jiao et al., 2024). The influence of upgrading industrial structure on food security primarily manifests in the following aspects: Firstly, according to the Hicks-Hayami-Ruttan-Binswager Hypothesis, the increased cost of agricultural labor resulting from industrial structure upgrading will drive progresses in agricultural technology toward factors beyond labor and land, foster the augmentation of agricultural capital, and subsequently elevate food security levels (J. J. Tang et al., 2022). Secondly, the upgrading of the industrial structure has attracted increased capital and resources toward the agricultural sector, resulting in improved infrastructure conditions. By concurrently enhancing levels of production technology and product quality, this upgrading has mitigated issues related to grain spoilage and corruption stemming from inadequate storage facilities and transportation shortcomings, thereby advancing food security. Thirdly, the upgrading of industrial structure signifies the modernization, organic cultivation, and environmentally friendly practices in agriculture, reducing harm to natural resources like cultivated land, establishing a solid foundation for sustainable agricultural development (H. Zhang et al., 2022), and ensuring a consistent enhancement in food security. Moreover, industrial upgrading can increase urban residents’ demand for high-quality agricultural products, thereby encouraging the optimization of agricultural industry and product structures, which in turn contributes to the enhancement of food security. Accordingly, Hypothesis 2 was proposed:

H2: DIF can improve the level of food security by promoting the upgrading of industrial structure.

Leveraging the Internet and e-commerce platforms, DIF can assist farmers in promptly understanding beneficial farmer policies, seizing opportunities, accessing resources, fostering rural industries’ growth, enhancing rural entrepreneurship, and enabling the creation of online financial products like stocks and bonds in rural regions. These programs establish innovative pathways and structural frameworks designed to enhance income for rural populations while narrowing the urban-rural income gap (Barbu et al., 2021; Jackson, 2022). As farm households experience income growth, the financial disparity between urban and rural communities diminishes, credit constraints for agricultural workers are alleviated, and investments in farm productivity increase substantially (Ito & Kurosaki, 2009). These developments not only facilitate the enhancement of rural infrastructure and expedite agricultural modernization (Cen et al., 2022), but also provide efficient support for grain production and storage, but also foster the creation of new rural entities, encouraging the moderate-scale operation of agriculture. This, in turn, enhances agricultural labor productivity, land output rates, and subsequently bolsters levels of food security. Furthermore, the reduction of income disparity between urban and rural areas can effectively deter young and middle-aged rural laborers from seeking employment far from their home, beckon back rural laborers working outside the locality, mitigate the issue of excessive rural labor depletion stemmed from unbalanced regional development and significant income gaps, leading to favorable effects on the enhancement of food security (J. L. Xu, 2013). Accordingly, Hypothesis 3 was proposed:

H3: DIF can improve the level of food security by narrowing the urban-rural income disparities.

Considering the significant regional socio-economic disparities in China, the influence of DIF on food security is likely to exhibit substantial heterogeneity. Policies in major grain-producing regions encompass a range of initiatives, such as price support mechanisms, technical subsidies aimed at boosting productivity, and the development of national modern agricultural demonstration zones. These measures are designed to reinforce food security in key production areas through targeted policy interventions and concentrated investment, thereby exerting a positive influence on national food security outcomes (S. S. Lu et al., 2013). Compared to non-major grain-producing regions, areas benefiting from such policy support are more likely to leverage DIF effectively to enhance food security. As a result, the impact of DIF on food security is expected to exhibit regional heterogeneity between major and non-major grain-producing areas. Furthermore, investments in agricultural insurance emerge as a key determinant of food security performance (Mârza et al., 2015). Agricultural insurance supports producers in improving technological adoption and production efficiency, while also serving as a buffer against risks from natural disasters, market volatility, and other emergencies (Xie et al., 2024). By mitigating these risks, it increases farmers’ willingness to invest in production. Consequently, regions with higher levels of agricultural insurance investment may be more inclined to utilize financial instruments, including DIF, to safeguard food security. Therefore, regional differences in agricultural insurance investment contribute further to the heterogeneity in the impact of DIF on food security. Accordingly, Hypothesis 4a & 4b was proposed:

H4a: The impact of DIF on food security level is heterogeneous in major grain producing areas and non-major grain producing areas.

H4b: The impact of DIF on food security level will be heterogeneous due to the level of agricultural insurance premium input.

Food security is closely influenced by natural environmental factors (Mok et al., 2020). Given China’s vast geographic expanse and diverse ecological conditions, the effectiveness of DIF in promoting food security is subject to regional variation. One key determinant is terrain. In areas with significant topographic variation, agricultural production and infrastructure development incur substantially higher costs than in flat regions. These regions often face challenges such as land fragmentation and abandonment, which hinder efforts to improve food security (Xiao et al., 2021). Thus, the impact of DIF on food security is less pronounced in areas with highly uneven topography compared to flatter regions. Precipitation is another critical factor influencing agricultural productivity and regional food security disparities (Holtermann, 2020). In regions with abundant rainfall and favorable climatic conditions, DIF can more effectively enhance food security by supporting agricultural development and risk management. Conversely, in areas with low precipitation, even well-developed DIF systems may fall short in improving food security due to the environmental constraints that limit agricultural output (Figure 1). Accordingly, Hypothesis 5a & 5b was proposed:

H5a: The impact of DIF on food security level will be heterogeneous in different topographic relief areas.

H5b: The impact of DIF on food security level will be heterogeneous due to the difference of precipitation.

Figure 1.

Theoretical analysis diagram.

Research Design

Food Security Index Calculation and Information Mining

This study utilizes agricultural data from prefecture-level cities to construct a comprehensive food security index (FS). Informed by existing scholarship in the field (Lee et al., 2023; Qi et al., 2023), the index is structured around four core dimensions: usability, availability, sustainability, and stability. These constructs are operationalized through seven meticulously selected tertiary indicators, as detailed in Table 1. The usability dimension comprises three indicators: production capacity, government support, and production efficiency. Production capacity is measured by unit grain output and the proportion of cropland sown area. Unit output directly captures technical efficiency in converting land into food, while the sown area proportion indicates the intensity of cropland use—both critical for assessing regional food production potential. These data are readily available from national statistical yearbooks, ensuring data availability and cross-regional comparability. Government support is captured through expenditures on agriculture, forestry, and water affairs. Government expenditures for irrigation, rural infrastructure, and agricultural R&D directly enhance production usability by mitigating natural risks and improving technical adoption. Production efficiency is reflected in agricultural mechanization levels. Higher mechanization reduces dependency on manual labor, enhances production consistency, and is a key driver of modern food systems. The availability dimension encompasses degree of satisfaction and supply infrastructure, represented by per capita grain output and per capita road area, respectively. The former is a core metric for supply sufficiency, directly comparing regional production to consumption needs. The latter captures logistical capacity for food distribution. Road networks reduce transportation costs, shorten supply chains, and improve market connectivity—critical for ensuring timely food availability, especially in remote regions. Sustainability dimension and stability dimension are respectively represented by two negative indicators: Fertilizer application intensity and grain yield fluence index. Excessive fertilizer use degrades soil quality, pollutes water resources, and undermines ecosystem resilience. By incorporating this variable as a negative indicator, the framework highlights the balance between short-term productivity and lasting ecological health. Grain yield fluence index measures year-to-year production volatility caused by climate events, pests, or policy changes. A lower fluence index indicates more stable production, reducing risks of supply disruptions. This methodology not only mitigates subjective bias associated with methods such as expert scoring but also prevents information gaps that may arise from principal component analysis.

Table 1.

Evaluation System of Food Security Index.

Primary index	Secondary index	Three-level index	Measurement method	Sign	Weight
Food security	Usability	Production capacity	Unit grain output	+	0.061
		Production capacity	Proportion of sown area of crops	+	0.254
		Government support	Proportion of expenditure on agriculture, forestry and water affairs	+	0.087
		Production efficiency	level of mechanization	+	0.165
	Availability	Degree of satisfaction	Per capita grain output	+	0.333
	Availability	Supply level	Per capita road area	+	0.081
	Sustainability	Sustainability level	Fertilizer application intensity	−	0.009
	Stability	Fluctuation degree	Grain yield fluctuation index	−	0.01

After the index is determined, the main calculation steps of weight and comprehensive score are as follows:

Suppose there are $n$ samples and $m$ indicators, which form the original data matrix (X_ij)_nm, and X_ij represents the j-th indicator of the i-th sample. First of all, the original data is normalized by the method of range standardization, and the specific formula is:

$\begin{matrix} Positive index : Y_{ij} = \frac{X_{ij} - min X_{ij}}{max X_{ij} - min X_{ij}}; \\ Negative index : Y_{ij} = \frac{min X_{ij} - X_{ij}}{max X_{ij} - min X_{ij}} \end{matrix}$

Step 2, calculate the corresponding index proportion:

$P_{ij} = \frac{Y_{ij}}{\sum_{i = 1}^{n} Y_{ij}}$ (1)

When P_ij is 0, replace it with 0.000001.

Step 3, calculate the information entropy e_j of the j-th indicator:

$e_{j} = - k \sum_{i = 1}^{n} P_{ij} \ln P_{ij}, k > 0, 0 \leq e_{j} \leq 1$ (2)

Step 4, calculate the difference coefficient g_j of indicators:

$g_{j} = 1 - e_{j}$ (3)

Step 5, normalize the difference coefficient of the j-th index obtained from the above formula, and calculate the corresponding weight w_j:

$w_{j} = \frac{g_{j}}{\sum_{i = 1}^{m} g_{j}}$ (4)

Finally, calculate the comprehensive score, that is, the food security index:

$F S_{i} = \sum_{i = 1}^{m} (w_{ij} Y_{ij}), i = 1, 2, 3, \dots, n$ (5)

After the construction of measurement index, in order to further mine effective information from it, this paper analyzes its convergence and dynamic evolution process respectively.

First of all, this paper uses the method of absolute β convergence to examine the convergence of China’s food security level, and explores whether low-level areas can approach low-level areas with high growth rate over time, and finally converge to the same level. The formula is as follows:

$\ln \frac{F S_{it + 1}}{F S_{it}} = α + β F S_{it} + ε_{it}$ (6)

Where FS_it represents the food security level of the city i in the t year; $\ln \frac{F S_{it + 1}}{F S_{it}}$ indicates the growth rate of food security level in city i from t to t + 1; α is a constant term; β is the convergence coefficient, and when β is significantly less than 0, it means that there is absolute convergence; On the contrary, there is divergence.

Secondly, by using kernel function to estimate probability density function, this paper explores the dynamic evolution process of China’s food security level in order to better reflect the distribution characteristics of food security level. Kernel density estimation is nonparametric estimation, independent of the selected interval length, and has good continuity. Given a group of independent and identically distributed samples x₁, x₂, x₃,…, x_n, let its probability density function be f, and the kernel density function is as follows:

$f_{h}^{\land} (x) = \frac{1}{n} \sum_{i = 1}^{n} K_{h} (x - x_{i}) = \frac{1}{n h} \sum_{i = 1}^{n} K (\frac{x - x_{i}}{h})$ (7)

K(x) is a kernel function; n is the total number of samples; h is the estimated bandwidth for density.

Model Construction

In order to explore the effect and mechanism of DIF on food security level, this paper adopts the time individual double fixed effect model according to the results of Hausman test and joint significance test of time dummy variables. The specific model is set as follows:

$F S_{it} = α_{0} + α_{1} Inde x_{it} + α_{2} \sum Control s_{it} + μ_{i} + θ_{t} + ε_{it}$ (8)

$M_{it} = β_{0} + β_{1} Inde x_{it} + β_{2} \sum Control s_{it} + μ_{i} + θ_{t} + ε_{it}$ (9)

$\begin{matrix} F S_{it} = γ_{0} + γ_{1} Inde x_{it} + γ_{2} M_{it} + γ_{3} \sum Control s_{it} + μ_{i} \\ + θ_{t} + ε_{it} \end{matrix}$ (10)

FS_it represents the food security index of the city i in the year t. Index_it represents the logarithm of the DIF index of the city i in the year t. In addition, the DIF index is refined, and its three sub-dimensions: coverage, depth of use and digitalization are used to make regression respectively, so as to verify its influence on the food security level. Controls_it represents the control variable. M_it represents the intermediate variable; μ_i indicates the individual-fixed effect; θ_t indicates the time-fixed effect. ε_it represents a random perturbation term.

Furthermore, this paper adopts the Spatial Dobbin Model (SDM) with double fixed effects, and studies the spatial correlation and spillover effect of the explained variables and explanatory variables by introducing the spatial lag term. The model settings are as follows:

$F S_{i t} = φ_{1} I n d e x_{i t} + φ_{2} W I n d e x_{i t} + φ_{3} \sum C o n t r o l s_{i t} + μ_{i} + θ_{t} + ε_{i t}$ (11)

Windex_it indicates the spatial lag term of the core explanatory variables, and φ₁ and φ₂ indicate the core coefficients to be estimated.

Data Source and Variable Setting

This research employs panel data from 281 prefecture-level cities in China spanning 2012 to 2020 to investigate the impact of DIF on food security. The primary measure of DIF is derived from the PKU Digital Inclusive Finance Index. Key agricultural indicators are primarily sourced from the Yearbook of Regional Economic Statistics of China. For periods where data are unavailable, supplementary information is obtained from statistical yearbooks, municipal bulletins, local government websites, and the EPS database. Data on control variables are sourced from the Statistical Yearbook of China. Some missing values are supplemented using interpolation.

This paper designates food security (FS) as the dependent variable, measured by an entropy-based composite index. The methodology for constructing this index is detailed above. The independent variable is the development of prefecture-level DIF (Index), gauged by the Digital Inclusive Finance Development Index co-compiled by Peking University’s Digital Finance Research Center and Ant Financial Group. This index comprises several sub-dimensions, including coverage breadth (Cov), depth of use (Use), and degree of digitization (Dig) (F. Guo et al., 2020). To control for potential confounding factors, this paper incorporates several control variables, including: (1) the urbanization rate (Town), measured as the proportion of urban population in the total population; (2) foreign direct investment (FDI), expressed as the FDI-to-GDP ratio; (3) economic development level (GDP), gauged by the natural logarithm of per capita GDP; (4) the material input level (FIX), proxied by the natural logarithm of fixed asset investment; (5) education investment (EDU) is defined as the percentage of educational spending in total local fiscal expenditure; (6) science and technology investment (TECH), quantified as the proportion of sci-tech expenditures in total local fiscal expenditure.

This study investigates how digital financial inclusion impacts food security by examining two key mediating variables: industrial structure (Ind) and the urban-rural income gap (Gini). Drawing on established methodologies from Sun et al. (2022), the analysis explores the mechanisms through which DIF initiatives influence food security. The industrial structure index gives higher weight to the proportion of the added value of the primary, secondary and tertiary industries in GDP in turn:

$Ind = \sum_{n = 1}^{3} S_{n} n$ (12)

S_n indicates the proportion of the added value of the n-th industry in GDP, and the range of this index is 1 to 3. The greater the index, the higher the level of industrial structure.

Gini coefficient of income gap between urban and rural areas According to DSMP/OLS night lighting data, the average night lighting brightness in the region is used as a proxy indicator of per capita income level to measure the income gap between urban and rural areas. The Gini coefficient ranges from 0 to 1, and its specific calculation formula is:

$Gini = \frac{\sum_{i = 1}^{n} \sum_{j = 1}^{n} | y_{j} - y_{i} | / n (n - 1)}{2 η}$ (13)

y_i is the average value of city i lighting data brightness, and n is the number of prefecture-level cities in a certain province. η is the average value of light brightness data in this province.

Heterogeneous grouping variables are divided into social factors and natural factors. In terms of social factors, the main grain-producing areas and non-grain-producing areas designated by the state in 2001 are denoted by 0 and 1, correspondingly. The agricultural premium income in China Insurance Yearbook will be used to measure the agricultural premium. Natural factors are considered, with topographic relief data at the prefecture-level city level and annual average precipitation data of prefecture-level cities being taken into account. Descriptive statistics for all essential variables are presented in Table 2.

Table 2.

Descriptive Statistics.

Variable type	Variable	Symbol	Mean	SD	Min	Max
Explained variable	Food security index	FS	0.202	0.082	0.034	0.587
Explanatory variables	DIF index	Index	5.185	0.347	4.017	5.813
	Coverage breadth	Cov	5.123	0.373	3.587	5.788
	Depth of use	Use	5.158	0.368	3.932	5.857
	Degree of digitization	Dig	5.378	0.355	3.942	6.365
Mechanism variables	Industrial structure index	Ind	2.307	0.141	1.831	2.836
	Gini index	Gini	0.497	0.215	0.001	0.918
Control variables	Urbanization rate	Town	0.559	0.148	0.223	1.101
	Foreign direct investment	FDI	0.017	0.018	0.000	0.210
	Per capita GDP	GDP	10.762	0.560	9.007	13.056
	Fixed-asset investment	FIX	11.765	0.915	8.691	14.538
	Education expenditure	EDU	0.176	0.039	0.036	0.356
	Science and technology expenditure	TECH	0.017	0.017	0.001	0.207

Empirical Analyses

Index Analysis of Food Security Level

Time Evolution Analysis

This study employs kernel density plots (Figure 2) using data samples from 2012, 2014, 2016, 2018, and 2020 to examine the temporal dynamics of food security across 281 prefecture-level cities. By analyzing the shifting center of gravity in these density curves, we observe an overall upward trend in food security levels. However, this positive trajectory experienced a mild decline in 2020, likely attributed to the COVID-19 pandemic. Secondly, from the perspective of peak height, it declined in 2012, 2014, 2016, and 2018, and rose in 2020, indicating that the widening trend of food security level differences among cities has eased, but it is still lower than that in 2012. Thirdly, from the situation of the left and right tails, the right tail is larger than the left tail, and it tends to thicken and lengthen in 2012, 2014, 2016, 2018, and 2020, indicating that the proportion of cities in high-value areas with food security level has increased. Finally, from the number of peaks, the figure also shows the coexistence of a main peak and multiple peaks, which shows that the multipolar differentiation of food security level is obvious during this period. To sum up, the food security level of 281 cities studied in this paper has the characteristics of different development levels and multi-polarization.

Figure 2.

Kernel density estimation.

Spatial Distribution

Within this study, the food security levels of the 281 cities examined during the study period were averaged and categorized into four groups utilizing the natural breakpoint classification method. The colors from deep to light were high-level areas, high-level areas, low-level areas, and low-level areas respectively. The spatial distribution of food security levels was mapped using ArcGIS 10, as depicted in Figure 3. The visualization indicates that areas with high and moderately high food security are primarily concentrated in Northeast China, the North China Plain, and the middle-lower Yangtze River basins. These areas are the main areas of grain production in China with superior natural conditions such as topography and soil. The low-level areas and sub-low level areas are mainly distributed in the southeast hills, parts of Yunnan-Guizhou Plateau in the southwest, central mountains, and parts of Loess Plateau. These regions face numerous constraints on food security attributed to challenges such as significant relief variations, poor soil quality, and water resource scarcity. Overall, food security in China demonstrates a spatial pattern of higher levels in eastern and northern regions and lower levels in western and southern areas, underscoring significant regional disparities.

Figure 3.

Spatial distribution of food security level.

Convergence Analysis

This study employs the absolute β-convergence framework to examine whether food security levels are converging across urban areas. Following a Hausman test, the analysis employs a fixed-effects model, with detailed results presented in Table 3. As indicated, the absolute β-convergence coefficient is −3.703 and statistically significant at the 1% level. This finding confirms the presence of absolute β-convergence in the food security levels of the 281 prefecture-level cities under analysis. In other words, regions with initially lower food security levels exhibit a significant “catch-up effect” over time.

Table 3.

Convergence Analysis Result.

Coefficient	Absolute β convergence
β	−3.703*** (.118)
α	.744*** (.024)
N	2,248
R ²	.332
Regression model	Fixed effect
F test	.000
Hausman test	.000

Note. Standard errors in parentheses.

p < .1. **p < .05. ***p < .01.

Analysis on the Effect of DIF on Food Security

Benchmark

This study investigates the impact of DIF and its subdimensions on food security using a fixed-effects regression model. As presented in Table 4, baseline regression results demonstrate a robust positive association—the coefficient of the DIF index is statistically significant at the 1% level (Column 1). These findings confirm that promoting digital financial inclusion plays a critical role in enhancing food security. This finding aligns closely with China’s strategic policy direction of integrating digital technologies into national food security planning and highlights the pivotal role of DIF in advancing rural revitalization under the framework of the 14th Five-Year Plan. H1 is verified. The results presented in columns (2) through (4) indicate that both the coverage breadth and utilization depth of DIF have a significant positive effect on food security, underscoring the critical role of these subdimensions in enhancing food security level. By contrast, the digitization degree does not exhibit a statistically significant impact. Two potential explanations may account for this. First, in several regions of China, DIF remains in a nascent stage, with relatively low levels of digitalization, thereby limiting its effectiveness. Second, the limited dissemination and adoption of digital technologies in rural areas have resulted in low levels of digital literacy and limited understanding among rural residents. This lack of proficiency constrains their ability to effectively engage with digital financial tools, thereby diminishing the observable impact of digitalization on food security outcomes.

Table 4.

Benchmark Results.

Variable name	(1)	(2)	(3)	(4)
	FS	FS	FS	FS
Index	0.104*** (0.017)
Cov		0.082*** (0.013)
Use			0.041*** (0.011)
Dig				−0.008 (0.006)
Town	−0.007 (0.027)	−0.014 (0.027)	0.005 (0.027)	0.007 (0.026)
FDI	0.095* (0.049)	0.087* (0.049)	0.109** (0.048)	0.121** (0.050)
GDP	−0.001 (0.008)	−0.004 (0.008)	0.004 (0.007)	0.005 (0.007)
FIX	−0.006* (0.003)	−0.007** (0.003)	−0.005 (0.003)	−0.005 (0.003)
EDU	−0.053 (0.038)	−0.046 (0.039)	−0.058 (0.039)	−0.068* (0.040)
TECH	−0.109* (0.062)	−0.084 (0.057)	−0.147** (0.069)	−0.164** (0.072)
_cons	−0.165* (0.088)	−0.024 (0.071)	0.039 (0.079)	0.262*** (0.077)
N	2,373	2,373	2,373	2,373
R ²	.488	.493	.475	.469
Individual-fixed effect	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES

p < .1. **p < .05. ***p < .01.

Robustness Test

The baseline regression analysis shows that digital financial inclusion has a statistically significant positive effect on food security enhancement. To verify the reliability of these results, this study conducts several robustness checks using the following methods: (1) Replace the explained variables. The food security index is recalculated by factor analysis (FS2). (2) Excluding the four first-tier cities of Beijing, Shanghai, Guangzhou, and Shenzhen from the sample. (3) Avoid the impact of the epidemic in 2020. Due to the outbreak of COVID-19 epidemic in 2020, considering that agricultural input and output were affected by the epidemic (Fan et al., 2021), this paper excluded the data samples in 2020. (4) Adding new control variables. Climate change and related agricultural policies may also influence food security; therefore, this paper will incorporate the extreme weather ratio (Pan et al., 2022) and the “zero growth” policy for fertilizers (Lin et al., 2024), as new control variables in the analysis. Table 5 reports the results of the four robustness test methodologies.

Table 5.

Robustness Test Results.

	(1)	(2)	(3)	(4)
	Replace Explained variable	Eliminate first-tier cities	Eliminate 2020	Adding control variables
Variable name	FS2	FS	FS	FS
Index	0.590^*** (0.223)	0.101^*** (0.018)	0.097^*** (0.017)	0.104^*** (0.017)
_cons	−1.643^* (0.841)	−0.246^*** (0.094)	−0.090 (0.087)	−0.157^* (0.089)
N	2,372	2,337	2,149	2,373
R ²	.022	.565	.582	.013
Control variable	YES	YES	YES	YES
Individual-fixed effect	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES

p < .1. ***p < .01.

The results reported in Table 5 confirm the robustness of the baseline regression. First of all, when the dependent variable is replaced, the impact of DIF on food security is still significantly positive, which confirms the robustness of the core relationship. Second, excluding the four first-tier cities does not alter the results, as the coefficient for independent variable remains positive and significant, further reinforcing the robustness of the baseline estimates. Third, even after excluding 2020 data to mitigate potential COVID-19 pandemic disruptions, the DIF coefficient retains both statistical significance and directional consistency. Lastly, including additional control variables does not alter the index’s significance, which continues to show a positive effect at the 1% level. These results collectively affirm the stability and reliability of the benchmark regression outcomes.

Endogenous Test

Because the model may have two-way causality, missing variables bias and other endogenous problems, the tool variable method is used to deal with the possible endogenous problems. In this paper, two instrumental variables of DIF are used: (1) In this paper, the Bartik instrumental variable (IV1) is constructed with reference to related research. Specifically, the instrumental variable consists of the interactive term of the difference between the DIF index and the national DIF index. On the one hand, because the national DIF index comes from 281 prefecture-level cities, it will not be affected by a single city, so it meets the exogenous requirements. On the other hand, demand shocks, except inclusive finance, may also lead to estimation bias, but as long as a single city is not so important that its internal demand shocks is significantly related to inclusive finance of the whole country, this instrumental variable is valid; (2) In addition, we select the shortest distance from the administrative center of each prefecture—level city to the “Eight Vertical and Eight Horizontal” optical cable transmission backbone network construction node city as the instrumental variable (IV2). As the geographical distance remains constant over time, we endow this variable with a time trend by multiplying it by the international internet user count from the previous period. In terms of relevance, the “Eight Vertical and Eight Horizontal” optical cable trunk network forms the fundamental framework of China’s communication network, and proximity to the node cities brings greater convenience for the development of DIF. Regarding exogeneity, the distance from the administrative center of each prefecture—level city to the “Eight Vertical and Eight Horizontal” optical cable transmission backbone network construction node city established in 1998 is a historical geographical variable. It is difficult for this distance to have an impact on the current development of DIF. Both instrumental variables have passed the weak tool variable test, over-recognition test and unrecognizable test, and the results are shown in Table 6.

Table 6.

Endogenous Test Results.

	(1)	(2)	(3)	(4)
Variable name	Index	FS	Index	FS
IV1	0.009^*** (0.0004)
Index		0.487^*** (0.062)
IV2			0.00002^*** (0.00004)
Index				0.271^*** (0.095)
N	2,098	2,098	2,373	2,373
KP Wald test		502.368		19.098
LM test		0.000		0.000
Control variable	YES	YES	YES	YES
Individual-fixed effect	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES

***

p < .01.

The estimation results presented in Table 6 indicate that the selected instrumental variables are valid and appropriate. Specifically, the KP Wald statistic exceeds the 10% critical value for the weak identification test, suggesting that weak instrumental variable is not a concern. Moreover, the KP LM statistic significantly rejects the null hypothesis of under-identification at the 1% level, confirming the adequacy of instrumental variables. After addressing potential endogeneity using instrumental variable estimation, the effect of DIF on food security remains significantly positive. These findings are consistent with the benchmark regression results, further affirming the robustness and reliability of the core conclusions.

Mechanism Analysis

This study investigates whether DIF impacts food security by promoting industrial structure upgrading and reducing urban-rural income disparities, as discussed in the preceding section. To test these hypotheses, the research analyzes the mediating roles of these two key factors in influencing food security. The regression results supporting this analysis are presented in Table 7.

Table 7.

Mechanism Analysis Results.

Variable name	(1)	(2)	(3)	(4)	(5)
	FS	Ind	Gini	FS	FS
Index	0.104*** (0.017)	0.152*** (0.024)	−0.280*** (0.051)	0.088*** (0.017)	0.094*** (0.017)
Ind				0.104*** (0.033)
Gini					−0.040*** (0.012)
_cons	−0.165* (0.088)	1.428*** (0.152)	2.624*** (0.266)	−0.313*** (0.106)	−0.070 (0.089)
N	2,373	2,373	2,344	2,373	2,344
R ²	.488	.778	.832	.500	.494
Sobel test				0.015	0.011
Bootstrap test				0.000 [0.006, 0.025]	0.000 [0.006, 0.025]
Control variable	YES	YES	YES	YES	YES
Individual-fixed effect	YES	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES	YES

p < .1. ***p < .01.

The results in column (2) of Table 7 indicate a statistically significant positive association between DIF and industrial upgrading, highlighting its critical role in promoting industrial upgrading. Meanwhile, column (4) shows that both the independent variable and the mediating variable exhibit robust positive effects, reinforcing the conclusion that DIF enhances food security primarily by driving industrial upgrading. The results in column (3) indicate a clear negative correlation between DIF and the urban-rural income gap, as proxied by the Gini index. The substantially negative coefficient of the independent variable suggests that DIF plays a significant role in narrowing the economic gap between urban and rural residents. Additionally, column (5) shows that DIF retains its positive impact, while the Gini index exhibits a persistent negative association. These findings highlight that reducing income inequality represents an additional channel through which digital financial inclusion enhances food security. On this basis, the results of Sobel test and Bootstrap test also confirm the above results. These findings offer a nuanced perspective on how common prosperity policies manifest in agriculture. Fundamentally, industrial upgrading serves as the economic backbone for shared wealth creation. When digital financial tools drive agricultural modernization, food security evolves beyond mere production quotas to emphasize quality—a strategic shift that aligns with national self-sufficiency priorities. Meanwhile, narrowing income disparities remains central to the common prosperity agenda. By enabling rural communities to access digital economic opportunities, these initiatives embody the ethos that progress should benefit all segments of society. This approach not only distributes prosperity more equitably but also fosters the social cohesion essential for long-term food system sustainability. H2 and H3 are verified.

Heterogeneity Analysis

Social Factors

Based on the two social factors, grain production and marketing areas and agricultural insurance premium input, which were divided in China in 2001, this paper makes a heterogeneity test. Grain production and marketing are divided into main producing areas and non-main producing areas; The agricultural premium input is divided into two groups according to the average value. The cities whose average agricultural premium input is higher than the overall average value in these 9 years are classified into the high premium input group, and the cities whose average agricultural premium input is lower than the overall average value are classified into the low premium input group. The regression results are shown in Table 8.

Table 8.

Heterogeneity Analysis Results Based on Social Factors.

	(1)	(2)	(3)	(4)
	Main producing areas	Non-main producing areas	High premium input	Low premium input
Variable name	FS	FS	FS	FS
Index	0.139*** (0.028)	0.044*** (0.014)	0.150*** (0.024)	0.072*** (0.019)
_cons	−0.178 (0.142)	−0.235*** (0.088)	−0.187 (0.125)	−0.102 (0.105)
N	1,474	899	899	1,474
R ²	.563	.508	.631	.417
Inter-group difference test	0.000		0.001
Control variable	YES	YES	YES	YES
Individual-fixed effect	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES

***

p < .01.

Table 8’s regression results reveal a clear pattern: DIF exerts a significant impact on food security in major grain-producing regions, but its effect diminishes in non-major grain-growing areas. This disparity is further validated by the group difference test. Concerning agricultural insurance, the data indicates that DIF significantly enhances food security in regions with high insurance premiums. By contrast, in areas with minimal premium levels, the impact is statistically insignificant. The group difference test, with a p-value below .01, provides strong evidence that the effect of digital finance on food security varies significantly with the level of agricultural insurance premiums. Assume that H4a and H4b are verified.

Natural Factors

Agriculture is highly sensitive to natural environmental conditions, which vary significantly across regions. These differences influence both agricultural inputs and outputs, shaping distinct modes of agricultural production. As a result, the effect of DIF on food security is also likely to vary across different environmental contexts. To explore this heterogeneity, the sample cities are grouped based on topographic relief and annual precipitation. For topographic relief, the study calculates the average value across all sample cities. Cities with relief above the average are categorized as high-relief zones, while those below are defined as low-relief zones. Regarding precipitation, the average annual precipitation for each city from 2012 to 2020 is computed. Based on China’s official classification of climate zones, cities with average annual precipitation exceeding 400 mm are categorized as humid and semi-humid areas, while those with 400 mm or less are classified as arid and semi-arid areas. Table 9 reports the results of the subgroup regression analysis.

Table 9.

Heterogeneity Analysis Results Based on Natural Factors.

	(1)	(2)	(3)	(4)
	High topographic relief	Low topographic relief	Humid and semi-humid areas	Arid and semi-arid areas
Variable name	FS	FS	FS	FS
Index	0.013 (0.015)	0.165*** (0.022)	0.108*** (0.018)	0.006 (0.042)
_cons	−0.011 (0.116)	−0.365*** (0.103)	−0.174* (0.093)	0.243 (0.211)
N	763	1,610	2,257	116
R ²	.608	.496	.490	.586
Inter-group difference test	0.000		0.007
Control variable	YES	YES	YES	YES
Individual-fixed effect	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES

p < .1. ***p < .01.

The first two columns of Table 9 display the regression outcomes based on topographic relief. In high-relief areas, the coefficient of the independent variable is not statistically significant, whereas in low-relief areas, the coefficient is significantly positive. The results in columns (3) and (4) classify outcomes according to annual average rainfall. DIF exhibits a significantly positive association in regions with greater precipitation, such as humid and semi-humid zones, whereas its impact is statistically insignificant in arid and semi-arid regions. Additionally, group difference tests between these two climatic groups produce p-values below .01, confirming the pronounced regional disparities observed. In summary, the findings suggest that DIF has a more pronounced positive impact on food security in regions with natural environmental conditions conducive to agricultural development. Assume that H5a and H5b are verified.

Further Analyses

Currently, as the market becomes increasingly open and agricultural production factors flow more freely across regions, DIF transcends time and space barriers. Therefore, when investigating the correlation between DIF and food security, it is crucial to assess not only the direct impact but also consider the presence of spatial autocorrelation. Food security is intricately linked to the rural revitalization strategy - ensuring food security effectively serves as a key element of the rural revitalization strategy (Guan et al., 2022). Moreover, with the extensive implementation of the rural revitalization strategy, significant advancements have been witnessed in China’s agricultural and rural development, leading to a notable enhancement in comprehensive agricultural productivity and subsequently influencing China’s food security level. Additionally, the rural revitalization strategy fosters the development of rural infrastructure, facilitates information exchange between rural areas and the external environment, thereby influencing the extent and direction of spatial spillover effects on food security levels (Engås et al., 2023). Drawing from prior research (Y. Liu et al., 2020), this study establishes the rural revitalization index for each prefecture-level city and utilizes this index to create a spatial weight matrix for examining the spatial spillover effect of DIF on food security. In comparison to geographical distance, the level of rural revitalization development exhibits a stronger correlation with food security. Utilizing the rural revitalization index in constructing a spatial weight matrix can better capture the developmental disparities between agriculture and rural regions, consequently enabling a more precise evaluation of interregional connections and interactions in food security. This study first employs the global Moran’s I spatial autocorrelation method to analyze the relationship between food security and DIF. Using Stata 17 for regression analysis, the results show significant spatial autocorrelation, as reported in Table 10.

Table 10.

Moran’s I Index.

Year	Adjacency matrix		Rural revitalization matrix
Year	Food security	DIF	Food security	DIF
2012	0.634***	0.419***	0.680***	0.562***
2013	0.587***	0.401***	0.615***	0.552***
2014	0.680***	0.387***	0.723***	0.460***
2015	0.639***	0.403***	0.652***	0.548***
2016	0.696***	0.386***	0.756***	0.539***
2017	0.721***	0.421***	0.766***	0.595***
2018	0.701***	0.489***	0.743***	0.640***
2019	0.661***	0.494***	0.779***	0.654***
2020	0.662***	0.510***	0.788***	0.666***

***

p < .01.

Table 10 shows that Moran’s I value consistently lie between −1 and 1, with each value significant at the 99% confidence level. This evidence confirms strong spatial autocorrelation patterns in both the adjacency matrix and the rural revitalization spatial weight matrix. Given these results, spatial interdependence cannot be overlooked.

After confirming the presence of spatial autocorrelation in the variables, it is necessary to verify the suitability of the econometric model. To this end, this study conducts the Hausman test, LM test, and Wald test based on Equation (13). The results of the Hausman test support the use of a fixed effects model, while the LM and Wald tests confirm the suitability of the Spatial Durbin Model (SDM) for capturing spatial dependencies. Furthermore, this study uses two types of spatial weight matrices, the adjacency matrix and rural revitalization matrix, to test the robustness of spatial econometric results. A comparison of regression results derived from these two matrices is reported in Table 11.

Table 11.

Spatial Spillover Effect.

Variables	Adjacency matrix	Rural revitalization matrix
Index	0.060*** (0.010)	0.080*** (0.010)
WIndex	0.018** (0.008)	−0.017 (0.010)
N	2,529	2,529
Control variable	YES	YES
Individual-fixed effect	YES	YES
Time-fixed effect	YES	YES

p < .05. ***p < .01.

The regression results in Table 11 show that the core explanatory variable coefficients are significantly positive under both the adjacency matrix and rural revitalization spatial weight matrix. This confirms that DIF significantly improves local food security, aligning with the benchmark regression findings. The spatial lag term exhibits notable differences between the two models. Under the adjacency matrix, DIF shows a significantly positive spatial lag coefficient. This indicates that its benefits go beyond improving food security within individual cities, generating measurable positive spillovers for neighboring urban areas. In contrast, when using the rural revitalization matrix, the spatial lag coefficient is not statistically significant. To further explore this discrepancy, the study divides the 281 cities into two groups: eastern and central–western regions. A spatial econometric regression is then conducted using the rural revitalization matrix for each subgroup. The results of this regional analysis are presented in the following table.

Table 12 reveals heterogeneous regional effects of digital financial inclusion on urban food security via spatial spillovers. While the eastern region exhibits a significantly positive spillover effect under the adjacency matrix, the central and western regions show no statistically significant impact. When employing the “rural revitalization” spatial weight matrix, the eastern region no longer exhibits statistically significant spatial spillover effects, whereas the central and western regions demonstrate significantly negative impacts. This regional disparity is likely attributable to the uneven development of DIF. In the central and western areas, where digital financial services remain in the early stages of development, the profit-driven nature of digital capital leads to its concentration in cities with robust food security—particularly in regions where rural revitalization initiatives are more advanced. This clustering effect may divert financial resources from less-developed urban areas, ultimately weakening their capacity to enhance food security. Moreover, the negative spatial spillover observed in China’s central and western regions highlights deeper institutional tensions stemming from unbalanced regional development. The eastern region, characterized by a high degree of marketization and a mature digital ecosystem, has cultivated a comparative institutional advantage led by DIF. In contrast, the central and western regions face constraints due to an insufficient supply of rural revitalization mechanisms—for instance, the slow progress of land management rights mortgage loan pilot programs. As a result, these areas are caught in a “resource outflow–security weakening” cycle, where financial and digital resources increasingly concentrate in more developed areas, undermining food security in less-developed regions.

Table 12.

Heterogeneity Test of Spatial Spillover Effect.

Variables	Adjacency matrix		Rural revitalization matrix
Variables	East	Mid & West	East	Mid & West
Index	0.084*** (0.015)	0.058*** (0.013)	0.116*** (0.016)	0.062*** (0.014)
WIndex	0.023** (0.009)	0.004 (0.011)	−0.206 (0.017)	−0.031** (0.014)
N	900	1,629	900	1,629
Control variable	YES	YES	YES	YES
Individual-fixed effect	YES	YES	YES	YES
Time-fixed effect	YES	YES	YES	YES

p < .05. ***p < .01.

Conclusions and Policy Recommendations

This study utilizes panel data from 281 Chinese cities spanning 2012 to 2020 to examine how DIF affects food security. The key findings are as follows: (1) DIF significantly enhances food security, with the breadth of coverage and depth of use identified as the primary contributing factors. (2) In terms of mechanisms of influence, DIF promotes food security by facilitating industrial structure upgrading and narrowing the urban–rural income gap. (3) The impact of DIF on food security exhibits notable heterogeneity across both social and natural dimensions. Regarding social factors, the positive impact of DIF is more pronounced in major grain-producing areas and regions with high agricultural insurance investment. In terms of natural conditions, DIF significantly improves food security in low-relief and humid/semi-humid areas. (4) Based on a spatial weight matrix constructed from the rural revitalization index, the analysis reveals a significantly negative spatial spillover effect of DIF on food security in the central and western regions, whereas the effect is not significant in the eastern region. Recent research quantifies DIF’s dual role in food-climate systems. J. Liu and Ren (2023) empirically analyze how DIF balances agricultural output stability with emissions control, demonstrating its capacity to simultaneously advance food security and carbon reduction. Their framework establishes DIF as a mechanism reconciling economic and environmental objectives. Complementing this, Osabohien et al. (2020) confirms financial access improves food production in Nigeria. While consistent with broader evidence, our study centers on food security as the primary outcome advancing the literature through an innovative causal framework that explicitly models regional heterogeneities and spatial spillovers. This dual methodology addresses persistent limitations in current DIF-food security research.

To effectively enhance national food security, rural areas must accelerate the development of DIF and leverage its potential to support agricultural production. First, traditional financial institutions should deepen digital transformation by expanding DIF coverage through innovative financial instruments tailored to farmers’ needs. DIF can improve usability by expanding access to targeted credit for digital agricultural tools (e.g., precision farming technologies and smart storage systems), and by promoting blockchain-based financing to streamline supply chains and reduce post-harvest losses. To improve availability, DIF should support the development of smart warehousing financial products, enabling farmers to upgrade digital storage infrastructure. A “Blockchain + Logistics” financing platform can optimize rural logistics and reduce grain circulation losses. For sustainability, DIF can enable green financial products, including climate-resilient insurance and carbon finance mechanisms, incentivizing organic and water-efficient farming practices. Real-time environmental monitoring via satellite and IoT data can help channel funding toward regenerative agriculture and mitigate soil degradation. On stability, DIF-powered early warning systems that integrate real-time supply chain and climate data can predict systemic shocks. In parallel, the creation of digital emergency liquidity tools—such as automated disaster-relief loans and fast-claim crop insurance—can ensure production continuity during crises. However, it is essential to recognize that such policies may require significant investment and may take time to implement effectively. Additionally, while promoting DIF, concerns about data privacy and security must be carefully managed to maintain public trust.

Secondly, to leverage the role of DIF in industrial structure upgrading, governments can implement industrial policies that guide financial resources toward high-value-added agricultural sectors and agro-processing industries. For example, providing low-interest loans and financial incentives to enterprises that pioneer cutting-edge agricultural innovations, eco-friendly farming practices, and the seamless integration of production, processing, and service sectors in rural areas. These policies can facilitate the transformation of traditional agriculture into modern, efficient, and diversified agricultural systems, which in turn can enhance food security. This could involve offering financial incentives to enterprises adopting advanced agricultural technologies. In terms of narrowing the urban-rural income gap, policymakers should focus on improving rural education and vocational training systems. By investing in digital educational platforms and agricultural vocational schools, rural residents can acquire skills that enable them to access better-paying jobs in both agricultural and non-agricultural sectors. Yet, it is important to acknowledge that these policies may compete for limited government funds with other crucial services. Additionally, promoting the development of rural e-commerce through DIF can help farmers expand their markets, increase the added value of agricultural products, and boost rural incomes. Policies such as establishing rural e-commerce incubation centers and providing financial support for e-commerce entrepreneurs can play a significant role in achieving this goal. However, it is worth noting that market saturation and digital divide may be potential challenges that limit the effectiveness of policies.

Third, while DIF holds transformative potential, its effectiveness varies significantly across regions due to disparities in infrastructure, financial literacy, and risk management capacity (Jack et al., 2023; Zins & Weill, 2016). In underdeveloped rural areas, particularly in central and western China, priority should be given to scaling up the “contract farming” financial model. Through the DIF platform, full-chain financing can be provided—from seed acquisition to post-harvest storage—while establishing credit risk early-warning systems to prevent over-financialization. While cross-regional digital financial alliances and technology-sharing initiatives have potential, their success depends on overcoming significant coordination challenges. Strengthening rural-related policies in these regions must also consider the complex interplay of various rural development factors. Simultaneously, digital financial literacy programs should be implemented regularly, with a focus on unbanked individuals and elderly populations over 60. Face-to-face training, simplified interfaces with voice interaction, and streamlined functions—such as microcredit applications and price inquiries—should be developed to ensure accessibility and adoption across all user groups. For major grain-producing regions such as Northeast China and the North China Plain, which are critical to national food security, policies should prioritize scaling digital financial support for large-scale, modern agriculture. This includes developing targeted DIF products like “precision farming credit lines” to subsidize the adoption of IoT-based irrigation systems, drone monitoring, and smart harvesters, with agtech partnerships integrating real-time yield data into credit assessments to reduce collateral requirements. For non-major grain-producing regions such as the Southeast Hills and Yunnan-Guizhou Plateau, which rely more on cash crops and face structural food security constraints, policies should focus on DIF-driven diversification and supply chain resilience. This involves launching “digital micro-credit for niche crops” with flexible repayment terms aligned to harvest cycles, using big data to align production with market demand, subsidizing “last-mile logistics financing” via DIF platforms to cover 50% of cold-chain vehicle and rural hub costs. Moreover, for arid/semi-arid and high-relief regions where natural constraints weaken DIF effectiveness, policies should combine DIF with targeted environmental adaptations: offering “DIF green loans” for drought-resistant seeds, drip irrigation, and terrace equipment, with disbursement tied to soil conservation participation.

Moreover, the main topic of this essay presents numerous avenues for further exploration. First, some indicators, such as sub-county-level agricultural insurance premiums and micro-level farmer behavior data, were unavailable. This restricted our ability to conduct more granular analyses (e.g., heterogeneous effects at the village level). Secondly, in the aspect of proxy variable constraints, while our multidimensional index incorporates usability, availability, sustainability, and stability, it excludes subjective indicators (e.g., household food access perceptions), which might capture nuanced aspects of food insecurity. Thirdly, as for omitted variable risks, local policies such as grain subsidies, land tenure reforms, and digital financial regulations vary across cities but were not explicitly modeled due to data unavailability. These policies may moderate the relationship between DIF and food security. Fourthly, China’s unique agricultural policies and DIF development model may not replicate in countries with different institutional contexts.

Footnotes

We are grateful to anonymous reviewers and editor for thoughtful comments and suggestions.

Disclaimer

Sage does not take a position with respect to geographic designations or the legal status of any country or region that may be a disputed territory.

ORCID iDs

Shan Haonan

Zhao Kai

Ethical Considerations

This article does not contain any studies with human participants performed by any of the authors.

Consent to Participate

This article does not contain any studies with human participants performed by any of the authors.

Funding

The author(s) disclosed receipt of the following financial support for the research,authorship,and/or publication of this article: This research was funded by National Social Science Fund of China (grant number 24&ZD119) and Shandong Provincial Natural Science Foundation (grant number ZR2023MG075).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Data Availability Statement

The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request.

References

Abay

K. A.

Berhane

Hoddinott

Tafere

(2023). COVID-19 and food security in Ethiopia: Do social protection programs protect? Economic Development and Cultural Change, 71(2), 373–402. https://doi.org/10.1086/715831

Abebe

M. G.

(2024). Impacts of urbanization on food security in Ethiopia. A review with empirical evidence. Journal of Agriculture and Food Research, 15, 100997. https://doi.org/10.1016/j.jafr.2024.100997

Ashraf

Javed

(2023). Food security and environmental degradation: Do institutional quality and human capital make a difference? Journal of Environmental Management, 331, 117330. https://doi.org/10.1016/j.jenvman.2023.117330

Barbu

C. M.

Florea

D. L.

Dabija

D. C.

Barbu

M. C. R.

(2021). Customer experience in fintech. Journal of Theoretical and Applied Electronic Commerce Research, 16(5), 1415–1433. https://doi.org/10.3390/jtaer16050080

Bouteska

Sharif

Bhuiyan

Abedin

M. Z.

(2024). Impacts of the changing climate on agricultural productivity and food security: Evidence from Ethiopia. Journal of Cleaner Production, 449, 141793.

Cao

Wang

(2024). Impact of digital finance on agricultural output: From the perspective of digital development of agriculture. Finance Research Letters, 66, 105698. https://doi.org/10.1016/j.frl.2024.105698

Cen

Lin

(2022). How does digital economy affect rural revitalization? The mediating effect of industrial upgrading. Sustainability, 14, 16987. https://doi.org/10.3390/su142416987

Chen

Y. J.

(2022). Research on the mechanism of digital inclusive finance promoting rural revitalization and Development. Modern Economic Research, 06, 121–132.

Nord

Andrews

, et al. (2016). Household food security in the United States in 2016. Economic Research Report, 2016.

10.

Cui

Zhong

(2024). Climate change, cropland adjustments, and food security: Evidence from China. Journal of Development Economics, 167, 103245. https://doi.org/10.1016/j.jdeveco.2023.103245

11.

Daud

S. N. M.

Ahmad

A. H.

(2023). Financial inclusion, economic growth and the role of digital technology. Finance Research Letters, 53, 103602. https://doi.org/10.1016/j.frl.2022.103602

12.

Engås

K. G.

Raja

J. Z.

Neufang

I. F.

(2023). Decoding technological frames: An exploratory study of access to and meaningful engagement with digital technologies in agriculture. Technological Forecasting and Social Change, 190, 122405. https://doi.org/10.1016/j.techfore.2023.122405

13.

Fan

Teng

Chew

Smith

Copeland

(2021). Food system resilience and COVID-19 - lessons from the Asian experience. Global Food Security, 28, 100501. https://doi.org/10.1016/j.gfs.2021.100501

14.

Fan

Zhao

Wang

(2024). Transforming agri-food systems for multiple wins in nutrition, inclusion and environment. Journal of Integrative Agriculture, 23(2), 355–358. https://doi.org/10.1016/j.jia.2024.01.017

15.

Feng

X. Y.

Sun

T. Q.

Dong

Yan

(2021). The development of digital inclusive finance in Chinese counties: connotations, index construction and measurement results. Chinese Rural Economy, 10, 84–105.

16.

Frimpong

S. E.

Agyapong

(2022). Financial literacy, access to digital finance and performance of SMEs: Evidence from Central region of Ghana. Cogent Economics & Finance, 10(1), 2121356. https://doi.org/10.1080/23322039.2022.2121356

17.

Fujimori

Doelman

Frank

Hristov

Kyle

Sands

van Zeist

W. J.

Havlik

Domínguez

I. P.

Sahoo

Stehfest

Tabeau

Valin

van Meijl

Hasegawa

Takahashi

(2022). Land-based climate change mitigation measures can affect agricultural markets and food security. Nature Food, 3(2), 110–121. https://doi.org/10.1038/s43016-022-00464-4

18.

Gao

J. T.

Shao

J. M.

(2020). Assessment of China grain industry security based on DEA model. Statistics & Decisions, 36(23), 61–65.

19.

Gao

Wang

(2021). Influence of China’s agricultural subsidy policy on food security and reform direction. Hua Nan Nong Ye da Xue Xue Bao, 20(05), 14–26.

20.

Griggs

Stafford-Smith

Gaffney

Rockström

Ohman

M. C.

Shyamsundar

Steffen

Glaser

Kanie

Noble

(2013). Policy: Sustainable development goals for people and planet. Nature, 495(7441), 305–307. https://doi.org/10.1038/495305a

21.

Guan

Z. G.

Zhang

S. Y.

Fang

J. B.

(2022). Food security: “Pilot” to boost rural revitalization. Journal of Northwest A&F University (Social Science Edition), 22(06), 29–37.

22.

Guo

Feng

Lin

(2023). Digital inclusive finance and digital transformation of enterprises. Finance Research Letters, 57, 104270. https://doi.org/10.1016/j.frl.2023.104270

23.

Guo

Pang

(2024). How does digital inclusive finance affect county's common prosperity: Theoretical and empirical evidence from China. Economic Analysis and Policy, 82, 340–358. https://doi.org/10.1016/j.eap.2024.03.016

24.

Guo

Wang

J. Y.

Wang

Kong

Zhang

Cheng

Z. Y.

(2020). Measuring China’s financial inclusion: Index compilation and spatial characteristics. China Economic Quarterly, 19(04), 1401–1418.

25.

Guo

Chen

Kang

(2024). Digital inclusive finance and agricultural green development in China: A panel analysis (2013–2022). Finance Research Letters, 69, 106173. https://doi.org/10.1016/j.frl.2024.106173

26.

Hadley

Talbott

Reddy

Wheat

(2023). Impacts of climate change on food security and resulting perinatal health impacts. Seminars in Perinatology, 47(8), 151842. https://doi.org/10.1016/j.semperi.2023.151842

27.

Holtermann

(2020). Precipitation anomalies, economic production, and the role of ‘first-nature’ and ‘second-nature’ Geographies: A disaggregated analysis in high-income countries. Global Environmental Change, 65, 102167.

28.

Hong

Chen

Wang

(2024). The impact of digital inclusive finance on the agricultural factor mismatch of agriculture-related enterprises. Finance Research Letters, 59, 104774. https://doi.org/10.1016/j.frl.2023.104774

29.

Huang

Wang

(2022). The role of digital financial inclusion in the development of digital agriculture. Issues in Agricultural Economy, 5, 27–36.

30.

Ingram

(2011). A food systems approach to researching food security and its interactions with global environmental change. Food Security, 3(4), 417–431. https://doi.org/10.1007/s12571-011-0149-9

31.

Ito

Kurosaki

(2009). Weather Risk, wages in kind, and the Off-Farm labor supply of agricultural households in a developing country. American Journal of Agricultural Economics, 91(3), 697–710.

32.

Jackson

A. B.

(2022). Financial statement analysis: A review and current issues. China Finance Review International, 12, 1–19. https://doi.org/10.1108/cfri-10-2021-0208

33.

Jack

Kremer

de Laat

Suri

(2023). Credit access, selection, and incentives in a market for asset-collateralized loans: Evidence from Kenya. Review of Economic Studies, 90(6), 3153–3185. https://doi.org/10.1093/restud/rdad026

34.

Jiang

Gao

Chen

Zhao

(2025). How does digital inclusive finance affect agricultural productive services?—An economic model in the Chinese context. Sustainability, 17(2), 475. https://doi.org/10.3390/su17020475

35.

Jiao

Wang

G. F.

C. Y.

Pan

(2024). Digital inclusive finance, factor flow and industrial structure upgrading: Evidence from the yellow river basin. Financial Research Letters, 62, 105141. https://doi.org/10.1016/j.frl.2024.105141

36.

Jin

W. D.

Fang

Chen

P. Q.

(2023). Does land circulation help improve food security? — Based on the analysis of change of land management mode. Journal of China University of Geosciences, 23(04), 105–121.

37.

Lee

C. C.

Zeng

Luo

(2024a). The impact of urbanization on food security in China. International Review of Economics & Finance, 93, 1159–1175. https://doi.org/10.1016/j.iref.2024.05.037

38.

Lee

C. C.

Zeng

Luo

(2024b). How does climate change affect food security? Evidence from China. Environmental Impact Assessment Review, 104, 107324. https://doi.org/10.1016/j.eiar.2023.107324

39.

Lee

C. C.

Zeng

Luo

(2023). Food security and digital economy in China: A pathway towards sustainable development. Economic Analysis and Policy, 78, 1106–1125. https://doi.org/10.1016/j.eap.2023.05.003

40.

Tang

Zhang

(2024). Mechanism research on digital inclusive finance promoting high-quality economic development: Evidence from China. Heliyon, 10(3), e25671. https://doi.org/10.1016/j.heliyon.2024.e25671

41.

Xiao

J. J.

(2020). The impact of digital finance on household consumption: Evidence from China. Economic Modelling, 86, 317–326. https://doi.org/10.1016/j.econmod.2019.09.027

42.

Lin

F. W.

Dai

M. H.

(2024). Will fertilizer “zero-growth” policy affect food security: RD test based on quasi-natural experiment. China Soft Science, (1), 12–23.

43.

Liang

Lü

Yan

Jin

Wang

(2023). Urbanization does not endanger food security: Evidence from China’s Loess Plateau. Science of the Total Environment, 871, 162053. https://doi.org/10.1016/j.scitotenv.2023.162053

44.

Liu

Ren

(2023). Can digital inclusive finance ensure food security while achieving low-carbon transformation in agricultural development? Evidence from China. Journal of Cleaner Production, 418, 138106. https://doi.org/10.1016/j.jclepro.2023.138016

45.

Liu

Wang

(2023). Opportunity or challenge? Research on the influence of digital finance on digital transformation of agribusiness. Sustainability, 15(2), 1072. https://doi.org/10.3390/su15021072

46.

Liu

Zang

Yang

(2020). China’s rural revitalization and development: Theory, technology and management. Journal of Geographical Sciences, 30, 1923–1942. https://doi.org/10.1007/s11442-020-1819-3

47.

Liu

Zhou

(2021). Reflections on China's food security and land use policy under rapid urbanization. Land Use Policy, 109, 105699. https://doi.org/10.1016/j.landusepol.2021.105699

48.

Jin

Cui

(2024). Does digital inclusive finance promote regional green inclusive growth? Finance Research Letters, 62, 105163. https://doi.org/10.1016/j.frl.2024.105163

49.

Wang

F. W.

Feng

L. T.

(2021). Research on the impact of land circulation policy on food security. Finance and Economics, (3), 94–105.

50.

Carter

(2024). Social determinants of rural food security: Findings from Michigan’s Upper Peninsula. Journal of Rural Studies, 107, 103256.

51.

Cheng

(2024). Digital inclusive finance and corporate ESG performance: The moderating role of executives with financial backgrounds. Finance Research Letters, 60, 104858. https://doi.org/10.1016/j.frl.2023.104858

52.

S. S.

Liu

Y. S.

Long

H. L.

Guan

X. H.

(2013). Agricultural production structure optimization: A case study of major grain producing areas, China. Journal of Integrative Agriculture, 12(1), 184–197. https://doi.org/10.1016/s2095-3119(13)60218-x

53.

(2021). Does digital financial inclusion affect agricultural eco-efficiency? A case study on China. Agronomy, 11(10), 1949. https://doi.org/10.3390/agronomy11101949

54.

(2023). Digital inclusive finance and corporate green technology innovation. Finance Research Letters, 55, 104015. https://doi.org/10.1016/j.frl.2023.104015

55.

Mârza

Angelescu

Tindeche

(2015). Agricultural insurances and food security. The new climate change challenges. Procedia Economics and Finance, 27, 594–599. https://doi.org/10.1016/s2212-5671(15)01038-2

56.

Mok

W. K.

Tan

Y. X.

Chen

W. N.

(2020). Technology innovations for food security in Singapore: A case study of future food systems for an increasingly natural resource-scarce world. Trends in Food Science & Technology, 102, 155–168. https://doi.org/10.1016/j.tifs.2020.06.013

57.

Norse

(2015). Environmental costs of China’s food security. Agriculture Ecosystems & Environment, 209, 5–14. https://doi.org/10.1016/j.agee.2015.02.014

58.

Osabohien

Adeleye

Alwis

(2020). Agro-financing and food production in Nigeria. Heliyon, 6(5), e04001. https://doi.org/10.1016/j.heliyon.2020.e04001

59.

Pan

Liu

H. Y.

Cheng

Z. S.

(2022). The impact of extreme climate on commercial banks’ risk-taking: Evidence from local commercial banks in China. Journal of Financial Research, (10), 39–57.

60.

Y. B.

Wang

Y. H.

(2023). Temporal and spatial evolution and impact analysis of digital economy development and food security capacity in Yangtze River economic belt. Rural Economy, (2), 48–58.

61.

Subramaniam

Masron

T. A.

(2021). Food security and environmental degradation: Evidence from developing countries. GeoJournal, 86, 1141–1153.

62.

Sun

W. Z.

Niu

X. D.

Wang

G. H.

(2022). Transportation infrastructure and industrial structure upgrading: Evidence from China’s high-speed railway. World Journal of Management, 38(03), 19–34.

63.

Tang

H. M.

Zhao

(2022). Digital financial inclusion, industrial structure, and inclusive growth. Modern Economic Science, 44(06), 71–83.

64.

Tang

J. J.

Gong

J. W.

Song

Q. H.

(2022). Digital financial inclusion and agricultural total factor productivity: The role of factor flow and technology diffusion. Chinese Rural Economy, 7, 81–102.

65.

Xiao

Zhou

Z. Y.

Luo

Q. Y.

(2021). Spatial-temporal differentiation and spatial effects of grain production in the Yangtze River economic belt. Chinese Journal of Agricultural Resources and Regional Planning, 42(12), 89–101.

66.

Xie

Zhang

Xia

Chen

(2024). The effect of agricultural insurance participation on rural households' economic resilience to natural disasters: Evidence from China. Journal of Cleaner Production, 434, 140123. https://doi.org/10.1016/j.jclepro.2023.140123

67.

J. L.

(2013). Income gap, labour migration and the food production. Population and Development, 19(03), 21–28.

68.

Wang

(2023). The impact of digital financial inclusion on the level of agricultural output. Sustainability, 15(5), 4138. https://doi.org/10.3390/su15054138

69.

Van Ranst

(2009). Production scenarios and the effect of soil degradation on long-term food security in China. Global Environmental Change, 19(4), 464–481. https://doi.org/10.1016/j.gloenvcha.2009.06.002

70.

Zhang

Song

(2022). Analysis of the threshold effect of agricultural industrial agglomeration and industrial structure upgrading on sustainable agricultural development in China. Journal of Cleaner Production, 341, 130818. https://doi.org/10.1016/j.jclepro.2022.130818

71.

Zhang

(2025). The impact of digital inclusive finance on agricultural carbon emissions at the city level in China: The role of rural entrepreneurship and agricultural innovation. Journal of Cleaner Production, 512, 145469. https://doi.org/10.1016/j.jclepro.2025.145469

72.

Zhao

Zhang

Cai

Liu

(2022). The impact of digital finance use on sustainable agricultural practices adoption among smallholder farmers: An evidence from rural China. Environmental Science and Pollution Research, 29, 39281–39294. https://doi.org/10.1007/s11356-022-18939-z

73.

Zins

Weill

(2016). The determinants of financial inclusion in Africa. Review of Development Finance, 6(1), 46–57. https://doi.org/10.1016/j.rdf.2016.05.001

Food Security and Digital Inclusive Finance: Evidence From China

Abstract

Keywords

Introduction

Literature Review

Theoretical Analysis and Research Hypothesis

Research Design

Food Security Index Calculation and Information Mining

Model Construction

Data Source and Variable Setting

Empirical Analyses

Index Analysis of Food Security Level

Time Evolution Analysis

Spatial Distribution

Convergence Analysis

Analysis on the Effect of DIF on Food Security

Benchmark

Robustness Test

Endogenous Test

Mechanism Analysis

Heterogeneity Analysis

Social Factors

Natural Factors

Further Analyses

Conclusions and Policy Recommendations

Footnotes

Disclaimer

ORCID iDs

Ethical Considerations

Consent to Participate

Funding

Declaration of Conflicting Interests

Data Availability Statement

References