How Can the Digital Economy Enable Sustainable Agricultural Development: Insights from China’s Factor Input Efficiency

Abstract

As the world’s second-largest digital economy, China plays a pivotal role in the global digital transformation. This study investigates how the digital economy contributes to sustainable agricultural development (SAD), with a focus on improving factor input efficiency. Specifically, it explores the mechanisms through which digitalization influences agricultural production, circulation, and distribution; examines regional and policy-period heterogeneity; and evaluates the nonlinear effects of digital development from 2013 to 2023. To achieve this, the study employs panel regression models and quantile regression techniques using provincial data across China. The results show that the digital economy significantly enhances SAD by improving the efficiency of key production factors—labor, capital, land, and technology. Interaction terms indicate strong positive correlations, particularly in central China. The digital economy’s impact was relatively limited prior to 2018 but intensified following major digital policy initiatives. Quantile regression results further confirm the robustness and nonlinearity of these effects, especially in more developed eastern regions. These findings underscore the importance of tailored, needs-based digital interventions to accelerate agricultural sustainability in diverse regional contexts.

1. Introduction

In 2023, China’s digital economy was valued at USD 7.74 trillion, ranking second globally and contributing over 40% of the national GDP. Since 2013, it has grown at an average annual rate of 17.22%, far outpacing overall economic growth. This expansion has catalyzed transformations across multiple sectors, including agriculture, where technologies such as big data, cloud computing, the Internet of Things, and artificial intelligence are playing an increasingly central role. In response to rising consumer demand for safer, higher-quality, and more sustainable food, digital tools are being integrated into agricultural systems to enhance productivity and sustainability.

The informatization level in agriculture reached 26.5% in 2023, supported by national strategies such as the Fourteenth Five-Year Big Data and Digital Economy Development plans. These initiatives underscore the central role of digital infrastructure in promoting high-quality development and accelerating rural revitalization. Yet, while the digital economy is widely viewed as a critical enabler of sustainable agricultural development (SAD), the empirical basis for understanding its impact remains fragmented.

Several critical gaps remain unaddressed within the current literature examining the digital economy’s role in SAD. Firstly, existing studies frequently lack clear, universally accepted definitions and standardized measurement frameworks for both the digital economy and SAD, hindering consistent analysis and comparability across contexts [1,2,3]. The complex, rapidly evolving nature of digital technologies further complicates efforts toward precise characterization and measurement, leaving a notable methodological void [1,2,4,5,6,7].

Secondly, much of the prior research emphasizes theoretical exposition, conceptual interpretation, and broad policy recommendations. However, there is a marked scarcity of empirical investigations into the detailed mechanisms through which the digital economy affects agricultural sustainability—especially mechanisms involving distinct stages of agricultural production, circulation, and distribution [8,9]. This gap limits our understanding of how digitalization practically influences agricultural efficiency and sustainability.

Thirdly, the empirical studies that do exist predominantly analyze national-level impacts, largely neglecting regional variations and differential development stages of digital economies within specific geographical contexts. This oversight obscures the nuanced regional dynamics and heterogeneous impacts critical for targeted policy interventions and localized agricultural strategies [8,9,10,11,12,13,14].

Lastly, despite the significant differences in factor endowments and resource allocation across China’s diverse regions, the literature has yet to adequately explore how digital economic transformations specifically influence agricultural factor inputs and improve their efficiency—elements fundamental to achieving sustainable agricultural growth [15,16,17].

This study aims to fill these gaps through a multipronged approach. First, we develop a comprehensive measurement framework for both the digital economy and SAD, enabling multidimensional assessment across spatial and functional dimensions. Second, we empirically examine the influence of the digital economy on SAD across production, circulation, and distribution processes. Third, we analyze how the digital economy enhances factor input efficiency and assess the extent to which this mediates its impact on agricultural sustainability, with a particular focus on regional heterogeneity and differences in policy implementation phases.

In doing so, the study contributes several key innovations. It introduces a multidimensional index to conceptualize and quantify the digital economy, accounting for infrastructure, application, and industrial development. Additionally, it repositions the role of the digital economy in improving factor input efficiency, offering fresh perspectives on optimizing agricultural labor, capital, land, and technology. Furthermore, our detailed regional and policy-phase analyses illuminate how varied contexts and policy environments shape the digital economy’s effectiveness, ultimately guiding more precise and impactful agricultural policy and practice.

The remainder of the paper is structured as follows: Section 2 offers a detailed literature review exploring existing research on digital economy and SAD and introduces our research hypotheses and conceptual framework. Section 3 outlines the methodology, variables, and data. Section 4 presents empirical findings and analyses. Section 5 discusses key findings and theoretical and practical contributions and concludes with study limitations and future directions. And Section 6 offers concluding remarks and outlines the study’s broader significance.

2. Literature Review and Research Hypotheses

2.1. Literature Review

Extensive research has explored the multifaceted nature of the digital economy and its implications for agriculture and rural development. Digital technologies, including big data, cloud computing, Internet of Things, and artificial intelligence, are recognized for their capacity to transform traditional agricultural practices, improve total factor productivity (TFP), and facilitate market responsiveness [17,18,19,20,21,22,23,24,25]. The digital economy significantly contributes to rural industrial integration, structural optimization, and innovation in agricultural production [16,26,27]. However, the inherent complexity of digital technologies poses challenges for precise characterization and quantification [2,4,5,6,7,8].

Scholars highlight various ways the digital economy transforms agriculture, such as improving factor efficiency, promoting mechanization and automation, and enabling precise resource management [28,29,30]. Due to limited agricultural resources and diverse factor endowments, enhancing factor efficiency rather than merely increasing factor input is crucial for SAD [15,16,17]. Internet-based platforms and digital infrastructures play essential roles in forecasting resource needs, optimizing resource allocation, and facilitating industrial convergence [31,32,33,34,35]. Chen et al. (2019) emphasized the role of digital technology and digital service platforms in industry integration [36].

Yet, research also acknowledges challenges associated with digital transformation in agriculture, including resource drain from rural to urban areas, rural aging, and insufficient infrastructure and talent pools [37]. Nonetheless, rural revitalization strategies and targeted policies have shown potential in reversing these negative trends, leveraging digital economy tools for rural economic rejuvenation [23,38,39,40]. Digital-economy-driven innovations in agriculture include reducing operational costs, improving output efficiency, and enabling structural and industrial optimization [31,32,33,34]. Digital technology’s rapid proliferation has spawned new economic sectors, profoundly altering industry structures and competitive landscapes [41,42,43,44,45].

Empirical studies suggest digital technology significantly enhances agricultural productivity, promotes economies of scale, and fosters industry specialization and aggregation [17,23,46]. However, gaps remain in current research, particularly regarding universally accepted definitions and comprehensive measurement standards for both digital economy and SAD [1,2,3]. Most studies have concentrated on theoretical frameworks or localized empirical impacts, lacking robust empirical evidence and global analytical perspectives [8,9,10,11,12,13,14].

Addressing these gaps, this study advances research by developing precise metrics for digital economy and SAD, investigating their empirical linkages across production, circulation, and distribution, and exploring mechanisms to enhance factor input efficiency. Furthermore, it accounts for regional heterogeneity, elucidates marginal contributions across provincial scales, and analyzes the evolving impact of digital economy development stages on SAD.

2.2. Research Hypotheses

The digital economy is reshaping the structure and performance of modern agriculture through both direct and indirect pathways. This study formulates two central hypotheses to guide the empirical investigation of its impact on SAD.

2.2.1. Direct Effect of the Digital Economy on SAD

SAD cannot be separated from the optimization of factor inputs and technological efficiency. By calculating technological efficiency from the view angle of inputs, it is seen that, ceteris paribus, producers can reduce factor input volume yet maintain consistent output volume through technological efficiencies. To achieve optimal levels of productivity, it is imperative that inefficient producers tap into their internal potential by enhancing their technological efficiency and removing any constraints hindering their performance [39]. At the same time, SAD entails diversification that involves using agricultural resources as a basis and developing related secondary and tertiary industries, such as recreational agriculture, rural e-commerce, installment agriculture, etc. The establishment and growth of emerging enterprises, models, and formats are contingent upon leveraging digital technology as a fundamental basis for the expansion of supplementary applications [47].

Digital productivity is a type of advanced productivity, and it serves as a critical catalyst for the emergence of novel frameworks for economic development. However, it brings huge challenges to the traditional economy [24]. Digital is a high-level factor of production that enters into production, mobility, consumption, and virtually all other system components. As agriculture industrializes, agricultural outputs penetrate larger markets, achieve broader distribution, and become integrated with other phases of production, processing, and marketing. This requires the continuous optimization of agricultural e-commerce platforms and further accelerates the circulation of agricultural products [8,47,48].

Cost is a critical factor in the growth of SAD. Improved agricultural efficiency and enhanced agricultural competitiveness have driven SAD costs down. The cost of traditional agriculture in China is still relatively high, and the digital transformation of agriculture has helped accelerate agricultural modernization by lowering production costs and creating multifunctional advantages [15]. Digital technology has made labor inputs more productive and has provided technological support for development, thus transforming agricultural production into large-scale production by reducing input costs, bolstering production and operational and circulation efficiency, expanding sales channels, and increasing network markets [42,49]. The digital economy enables supply chain synergies that are often lacking in traditional agriculture, where information asymmetry between buyers and sellers, relatively high transaction and circulation costs, and inefficient marketing of agricultural products all add cost. The digital economy makes it easier for producers and their supply chain partners to obtain information on demand about agricultural products and adjust accordingly. In sum, the digital economy reduces production and circulation costs, enables modern large-scale agricultural production, streamlines agricultural distribution channels, ensures the quality of agricultural products, and changes industrial relations [50]. Drawing upon the preceding theoretical analysis, we advance the subsequent hypothesis:

H1. 

The digital economy directly facilitates SAD by positively influencing agricultural production, product circulation, and income distribution.

2.2.2. Indirect Effects via Factor Input Efficiency

Moreover, except for these direct effects, the digital economy possesses indirect effects on SAD. Wu (2018) indicated that the digital economy has brought about widespread uptake of digital infrastructure and substantial improvement in digital technologies, further expedited the dissemination of information, and optimized the allocation of resources, so as to consolidate the foundation of industrial development and promote industrial transformation [35]. Lin and Mao (2022) posited that the digital economy has substantially elevated total factor productivity, especially in China’s central region [51], as new data components have modernized agricultural production modes, transformed traditional agriculture through greater mechanization, greening, and standardization, enhanced agricultural quality and productivity, and helped boost revenues [16,52]. The digital economy’s spatial spillover effect has disrupted traditional economic models and affected total factor agricultural productivity by upgrading of industrial structures and optimizing factor markets [53].

The impact of the digital economy on SAD is evidenced by indirect effects. First, digital intelligence and digital technology have contributed to greater labor productivity. At the same time, digital applications and the use of intelligent machines have enhanced agricultural productivity through more effective allocation of resources, including human resources [54]. Second, the digital economy enables diversified integration, open sharing, and decentralized organization of traditional production factors, including but not limited to capital and labor. In sum, inputs such as capital, labor, knowledge, managerial skill, and other production factors are effectively leveraged. The construction of digital infrastructure affects factor productivity through the degree of capitalization and per capita R&D level [29,55]. Third, the digital economy helps target and leverage capital investments by identifying regional industrial advantages and new growth opportunities [56,57]. Fourth, the digital economy contributes to factor, land, and utilization efficiency through large-scale agricultural production. Thus, modern digital information technology has freed agriculture from many traditional limitations, made markets more transparent, improved resource allocation, and brought about new industries and new formats [58]. Based on the aforementioned theoretical analysis, the following hypothesis is proposed:

H2. 

The digital economy indirectly promotes SAD by improving the efficiency of labor, capital, land, and technology inputs.

Figure 1 presents a conceptual framework illustrating the pathways through which the digital economy influences SAD. The digital economy exerts its impact through both direct and indirect channels. The direct effects (H1) arise from the development of digital infrastructure and digital technologies, which immediately enhance agricultural productivity. These technologies support modernized production systems, mass-scale operations, and industrial chain extensions, leading to improved circulation and sales of agricultural products and increased farmer incomes. The indirect effects (H2) operate through the reconfiguration and optimization of agricultural input factors. Digital transformation enables more efficient allocation and use of labor, capital, land, and technological resources. This improved factor input efficiency contributes to SAD by enhancing resource use, lowering costs, and promoting ecological balance and rural resilience. This dual-pathway framework underscores the need to assess both the structural (infrastructure/technology) and functional (input efficiency) mechanisms through which the digital economy fosters agricultural sustainability.

3. Method

3.1. Model Construction

Input factors are key to the level and quality of SAD. High-quality industrial development must take into account both the homogeneity and the heterogeneity of input factors, as optimizing and improving the efficiency of the factor allocation structure is essential to SAD. The digital economy plays a crucial role in enhancing the efficiency of agricultural inputs, optimizing resource flows, and driving structural adjustments [45,59,60].

The digital economy relies upon the foundation of digital infrastructure and data, which synergistically enable optimal utilization of information and communication technology to facilitate improvements in efficiency and optimization of structures [21]. However, the digital economy is inextricably linked with conventional factors, including labor and capital [44].

To better analyze the impact of the digital economy on SAD, this study employs an existing SAD research model as a reference and establishes a baseline model, as depicted in Formula (1),

$$SA{D}_{it}={\alpha }_0+{\alpha }_1{D}_{it}+{\alpha }_2{L}_{it}+{\alpha }_3{K}_{it}+{\alpha }_4{R}_{it}+{\alpha }_5{T}_{it}+{\mu }_{it}$$

(1)

where $SA{D}_{it}$ refers to the SAD level for province $i$ in year $t$; $D_{it}$ is the digital economic development level for province $i$ in year $t$; $L_{it}$, $K_{it}$, $R_{it}$, $T_{it}$ refer to input factors including labor, capital, land, and technology; ${\mu }_{it}$ is the random error term. ${\alpha }_0$ is a constant, ${\alpha }_1$, ${\alpha }_2$, ${\alpha }_3$, ${\alpha }_4$, ${\alpha }_5$ are the coefficients of the digital economy, labor, capital, land, and technology on SAD.

To capture the nuanced relationship between the digital economy and SAD, we employ a multifaceted econometric strategy designed to reveal both average effects and underlying heterogeneity across contexts. Recognizing the potential for complex interactions between the digital economy and key production factors—such as labor quality, land use, and capital inputs—we adopt a cross-term regression model. This allows us to probe the transmission pathways through which the digital economy influences SAD, identifying not only whether an effect exists but also through which channels it operates, as shown in Formulas (2)–(5).

$$SA{D}_{it}={\beta }_1+{\beta }_2{L}_{it}+{\beta }_3{K}_{it}+{\beta }_4{R}_{it}+{\beta }_5{T}_{it}+\chi D_{it}\times L_{it}+{\epsilon }_{it}$$

(2)

$$SA{D}_{it}={\gamma }_1+{\gamma }_2{L}_{it}+{\gamma }_3{K}_{it}+{\gamma }_4{R}_{it}+{\gamma }_5{T}_{it}+\kappa D_{it}\times K_{it}+{\delta }_{it}$$

(3)

$$SA{D}_{it}={\varphi }_1+{\varphi }_2{L}_{it}+{\varphi }_3{K}_{it}+{\varphi }_4{R}_{it}+{\varphi }_5{T}_{it}+\rho D_{it}\times R_{it}+{\upsilon }_{it}$$

(4)

$$SA{D}_{it}={\lambda }_1+{\lambda }_2{L}_{it}+{\lambda }_3{K}_{it}+{\lambda }_4{R}_{it}+{\lambda }_5{T}_{it}+\pi D_{it}\times T_{it}+{\zeta }_{it}$$

(5)

In this model, the cross-terms of labor, capital, land, technology factors, and the digital economy are denoted as $D_{it}\times L_{it}$, $D_{it}\times K_{it}$, $D_{it}\times R_{it}$, $D_{it}\times T_{it}$, respectively. The random error terms are represented by $\epsilon ,\delta ,\upsilon ,\xi$.

To further explore distributional effects and move beyond average treatment estimates, we incorporate quantile regression analysis. This method is particularly well-suited to our data, as it relaxes the assumption of normality and is robust to outliers and heteroscedasticity. By estimating the conditional effects of the digital economy across different points in the SAD distribution, quantile regression enables us to assess whether impacts are more pronounced in regions with low, median, or high levels of agricultural sustainability—offering a more complete picture than mean-based methods alone.

Finally, to account for institutional and regional heterogeneity, the analysis is disaggregated by geographic region and by political periods characterized by distinct policy regimes. This differentiated approach enables us to examine whether the influence of the digital economy on SAD is context-dependent and whether institutional transitions (e.g., shifts in development priorities or digital policy frameworks) condition its effectiveness. Such granularity enhances the interpretive power and policy relevance of the findings.

3.2. Interpreted Variable and Core Explanatory Variable

In this model, the interpreted variable is SAD. Sustainable agricultural development (SAD) encompasses more than increases in agricultural output. It involves the optimization of agricultural structures, extension of value chains, adoption of environmentally sound technologies, and realization of agriculture’s multifunctionality—including ecological protection, rural vitality, and food system resilience. SAD spans the entire agricultural process, from production and processing to circulation and distribution, and requires the simultaneous pursuit of economic viability, social equity, and environmental sustainability. This variable includes comprehensive indicators in combination with China’s rural production factor mobility, productivity development, and industrial integration policies [34]. An indicator system was developed for sustainable agricultural development that reflects the development process and the degree of development in each of the production, circulation, and distribution links. This illustrates how SAD positively affects the income of farmers and contributes to shifts in agricultural production practices and the integration of rural industries. Therefore, the indicators construct for SAD involves circulation, production, and distribution. The circulation link includes the circulation and sale of agricultural products. The production link involves agricultural production methods, agricultural production technologies, and large-scale production. The distribution link includes the income of farmers. The indicators of SAD in this paper are all positive indicators, signifying that changes in these indicators are consistent with the changes in SAD. Therefore, this study posits that SAD is impacted by changes in agricultural production, agricultural product circulation, and farmers’ incomes [60].

The digital economy serves as a core explanatory variable in this study. The digital economy refers to a broad spectrum of economic activities that are enabled and driven by digital technologies. It is characterized by the use of data as a core production factor, modern information and communication technologies (ICTs) as the principal enablers, and digital infrastructure and platforms as the primary channels through which value is created, exchanged, and optimized. At its core, the digital economy functions through the integration of digital tools—such as big data analytics, cloud computing, the Internet of Things, and artificial intelligence—into traditional sectors, thereby improving productivity, restructuring economic systems, and fostering innovation-driven growth. This variable cannot be replaced by a single indicator, and the construct of the digital economy ought to be developed from a multidimensional perspective [56,61]. Considering the different links of the digitization process and digital indicators in the sustainable development goals [62,63], combining the development process of the digital economy, the digital index setting should be related to digital knowledge, information, technology, and other aspects. At the same time, it is important to consider the degree of digitization of production, sales, infrastructure, and other components so as to build a safe, high-quality digital industry throughout the industrial chain [61]. This study’s digital economy indicator system is constructed by considering the digital infrastructure construction, application, and industry development perspectives. Digital infrastructure construction is assessed using six indicators, including switch capacity, delivery routes, and computer usage. Digital application is measured by ten indicators, such as Internet penetration and the presence of information enterprises, while digital industry development is evaluated through nine indicators, including software business revenue and information security income. This design effectively reflects the development process of the digital economy while also acknowledging the significant impact of the digital economy on the digital industry’s growth [60]. The digital economy indicators are all positive indicators, which implies that changes in these indicators correspond to changes in the digital economy. Therefore, accelerating construction of digital infrastructure, expanding digital applications, and fostering the development of the digital industry will boost the digital economy.

The relationship between the digital economy and SAD is inherently synergistic. Digital technologies provide powerful tools to improve resource allocation, enhance factor input efficiency, and enable real-time decision making, all of which are critical to achieving sustainable agricultural outcomes. Through mechanisms such as smart farming, e-commerce platforms, and digital supply chains, the digital economy supports the transformation of agriculture into a more integrated, efficient, and adaptive system. As such, the digital economy is not merely an enabler but a strategic driver of SAD in the modern era.

Given that the digital economy and SAD are composite indicators, they require integration into a single indicator. This paper employs the entropy method, based on relevant literature, to address the incorporation of the digital economy and SAD, which was divided into four steps.

In the first step, an assessment matrix was constructed, as shown in Formula (6).

$$X_{m\times n}={\left[X_{ij}\right]}_{m\times n}$$

(6)

where $X_{m\times n}$ is a matrix with $m$ column and $n$ rows.

In the second step, the assessment matrix was standardized using the maximum–minimum method, as shown in Formula (7).

$$x_{m\times n}={\left[x_{ij}\right]}_{m\times n}={\left[{\displaystyle \frac{X_{ij}-\underset{1\le i\le m,1\le j\le n}{X_{\min }}}{\underset{1\le i\le m,1\le j\le n}{X_{\max }}-\underset{1\le i\le m,1\le j\le n}{X_{\min }}}}\right]}_{m\times n}$$

(7)

where $i=1,2,\dots ,m$, $j=1,2,\dots ,n$, $x_{m\times n}$ is a standardized matrix.

In the third step, we computed the weight of the relevant indicators using the information entropy weighting (IEW), as presented in Formula (8).

$${\omega }_j={\displaystyle \frac{G_j}{{\displaystyle \sum _{j=1}^n{G}_j}}}={\displaystyle \frac{1+H_j}{n-{\displaystyle \sum _{j=1}^n{H}_j}}}$$

(8)

where $H_j=-\left({\displaystyle \sum _{i=1}^m{f}_{ij}Ln{f}_{ij}}\right)$, $f_{ij}={\displaystyle \frac{1+x_{ij}}{{\displaystyle \sum _{i=1}^m\left(1+x_{ij}\right)}}}$.

$G_j=1-H_j$ is the difference coefficient of column $j$. ${\omega }_j$ is information entropy weight of column $j$.

In the final step, the comprehensive value of the relevant indicators was computed.

$$C={\displaystyle \sum _i{\omega }_j{x}_{ij}}$$

(9)

where ${\displaystyle \sum _j^n{\omega }_{ij}}=1$.

3.3. Control Variables

SAD is not only influenced by the digital economy and factor input efficiency but is also inextricably linked to factor inputs. Therefore, different factor inputs were processed as control variables, including labor, capital, land, and technology, and details are as follows:

As the principal component of SAD, labor carries a big weight in agricultural production, processing, and marketing [64], as well as in promoting SAD. The migration of transitory workers and the return of college students to their hometowns are important to the development of rural industries. These labor factors are measured by rural employees.

Capital is essential to SAD. Notwithstanding the vital role of agricultural enterprises in the economy, the associated operational risks, substantial investments, and long-term return horizons are typically prohibitive for many potential investors. At present, decentralized farming operations are still common. To develop and expand rural operations, active farmer participation is imperative for the production of agricultural goods [65]. The capital component is assessed based on the amount of completed investments in fixed assets of rural households.

Land forms the material basis for SAD. Farmers, governments, enterprises, and others form linkage mechanisms through land holdings and transfers. Their participation in the interest distribution fosters the inflow of production factors towards agriculture and enhances land use efficiency. However, issues such as poor returns, obsolete thinking modes, ineffective circulation processes, and simple disinterest often hamper such processes [66]. The land factor is quantified by the amount of agricultural land, which serves as a crucial input in agricultural production and influences its overall efficiency and sustainability.

As the power source of SAD, technological inputs enable small-scale farmers to integrate and engage in the progress of contemporary agriculture. At the same time, technological penetration and functional complementarity of agricultural industry development promote the formation of new technological forms and product features [44]. Therefore, we quantify technological factors in agriculture by evaluating the total power of agricultural machinery utilized in the production process.

3.4. Data Source and Statistical Analysis

This study draws on data spanning 2013 to 2023, primarily sourced from the China Statistical Yearbook (2014–2024), with supplementary data from the China Rural Statistical Yearbook and relevant Provincial Statistical Yearbooks to address missing values and ensure completeness. Due to persistent data gaps, Tibet is excluded from the analysis.

The selection of the 2013–2023 time window is both empirically and historically motivated. From 2013 onward, China’s rural economy underwent a strategic transformation marked by the vertical and horizontal integration of agriculture into secondary and tertiary sectors—such as rural tourism, leisure agriculture, facility farming, and cultural agriculture. This marked the beginning of a policy-driven shift toward constructing modern agricultural industrial systems, laying the foundation for SAD.

Simultaneously, the digital economy—while still nascent prior to 2013—began to expand significantly across rural China. The introduction of national digital development strategies and rural informatization initiatives contributed to the accelerated penetration of digital technologies into agricultural systems. This trend became particularly pronounced after 2018, following the release of key digital economy policy frameworks, including the Fourteenth Five-Year Plan for the digital economy. These policies catalyzed infrastructure development, digital inclusion programs, and innovation ecosystems in rural regions, resulting in a nonlinear but increasingly evident influence of the digital economy on SAD.

To account for these structural inflection points, the analysis distinguishes between two phases: 2013–2017 and 2018–2023, corresponding to pre- and post-policy intensification periods. This temporal decomposition enables a more nuanced understanding of how the digital economy’s role in agriculture has evolved in response to institutional and technological changes.

Descriptive statistics, panel regression models, interaction terms, and quantile regression techniques are employed to analyze the data, test the robustness of findings, and capture regional heterogeneity in the digital economy’s effect on SAD. The statistical analysis was performed using Stata 16.0, with appropriate controls for endogeneity and regional fixed effects.

As shown in

Table 1. Statistical description of Variables.

	Mean	Max	Min	St. dev.	Med
Digital economy	0.238	0.826	0.041	0.176	0.178
SAD	0.311	0.781	0.021	0.188	0.266
Labor	0.072	0.209	0.002	0.051	0.063
Capital	0.031	0.097	0.000	0.022	0.027
Land	0.556	1.530	0.009	0.387	0.552
Technology	0.350	1.335	0.009	0.293	0.265

, the mean value of the digital economy is 0.238, with a standard deviation of 0.176. The average level of SAD is 0.311, with a standard deviation of 0.188. Thus, the levels of China’s digital economy and SAD are still relatively low. However, substantial regional disparities exist with regard to the progress made in the digital economy, revealing an uneven development of this sector. It appears that enhancing the digital economy and SAD is imperative.

4. Empirical Study

4.1. Basic Regression Analysis

First, regression of Equation (1) was performed utilizing data collected between 2013 and 2023. Model (1) in

Table 2. Results of the direct effect of digital economy on SAD.

Interpreted Variable		SAD	SAD	Production Link	Circulation Link	Distribution Link
	Model	(1)	(2)	(3)	(4)	(5)
Variable
Constant		0.115 *	0.012	−0.012 *	−0.026 *	0.049 *
$D$		0.826 *	0.726 *	0.215 *	0.345 *	0.150 *
$L$			−0.816 *	−0.397 *	0.334 *	−0.546 *
$K$			1.171 *	0.009	0.650 *	0.481 *
$R$			0.069 *	0.023 **	0.059 *	0.022 *
$T$			0.316 *	0.234 *	0.031 ***	0.028 ***
Fitting degree		0.597	0.898	0.841	0.844	0.473

Note: *, ** and *** are significance levels of 1%, 5% and 10%.

shows results of the impact of the digital economy on SAD at the national level. According to Model (1), the impact coefficient is positive, suggesting that the digital economy has a promoting effect on SAD, which is consistent with the preliminary hypothesis. Model (2) introduces other factors as control variables to verify the stability of the effect. In comparison to Model (1), Model (2) highlights a positive and stable impact coefficient, providing strong evidence for the promoting effect. Models (3)–(5) illustrate the effects of the digital economy on various dimensions of SAD, including production, circulation, and distribution. Models (3)–(5) demonstrate that the digital economy promotes agricultural production, agricultural product circulation, and income distribution, which is consistent with H1. In this process, capital, land, and technology elements all have a positive impact on SAD. The lateral verification shows that improved technology drives greater land and capital efficiency, thus strengthening SAD. However, labor plays a negative role in SAD for two principal reasons. First, with the implementation of mechanized production, a large portion of the labor force has been liberated and no longer participates in agricultural production, making the agricultural labor force and SAD incompatible. Second, improvements in agricultural technology have led to the emergence of facility agriculture, digital agriculture, and other advancements which require higher-quality and better skilled labor inputs to improve production efficiency.

4.2. Inspection of Relevant Mechanisms

As shown in

Table 3. Results of the indirect effect of digital economy on SAD.

Interpreted Variable		SAD	SAD	SAD	SAD	SAD
	Model	(6)	(7)	(8)	(9)	(10)
Variable
Constant		0.012	0.178 *	0.170 *	0.178 *	0.178 *
$D$		0.726 *
$L$		−0.816 *	−3.015 *	−0.479 **	−1.676 *	−1.290 *
$K$		1.171 *	2.327 *	−3.612 *	2.547 *	2.407 *
$R$		0.069 *	0.075 *	0.072 **	−0.086 *	0.293 *
$T$		0.316 *	0.267 *	0.292 *	0.135 *	−0.336 *
$D\times L$			7.657 *
$D\times K$				17.012 *
$D\times R$					1.297 *
$D\times T$						1.771 *
Fitting degree		0.898	0.779	0.763	0.842	0.781

Note: * and ** are significance levels of 1% and 5%.

, Models (6)~(10) reveal that the cross-term of the digital economy and other elements fosters SAD and suggest that, while the digital economy depends on these traditional elements, it bolsters their efficiency. Thus, consistent with H2, the digital economy ameliorates the efficiency of labor, capital, land, and technology. First, knowledge, information, and data are based on inputs of labor, land, and capital investment, and digital technology uses these inputs to improve the effectiveness of labor and the utilization of capital to achieve effective resource allocation. Second, the digital economy is characterized by the syncretism of digital technology, labor, capital, and other factors and is the core element of the production factor system in the digital economic era [24]. The high permeability of the digital economy enables accurate and effective matching of its elements, leading to optimized industrial structure configuration. Third, investment in the digital economy increases the proportion of high-value input factors and accelerates the integration and replacement of traditional factors, such as labor and capital, stimulating demand for high-value labor and thus improving the overall efficiency of capital, land utilization, technological inputs, and production methods. Fourth, the information and networking features of the digital economy enable rapid and accurate assessment of supply and demand factors, thus streamlining and improving the economic input/output ratios.

4.3. Regional Heterogeneity Analysis

To reveal the relation between digital economy and SAD, this study examines China as a case study and categorizes it into three areas, eastern, central, and western, to verify how the digital economy can enable SAD in each region. Specific results are shown in

Table 4. Impact estimates for Eastern China.

Interpreted Variable		SAD	SAD	SAD	SAD	SAD
	Model	(11)	(12)	(13)	(14)	(15)
Variable
Constant		0.077 *	0.238 *	0.239 *	0.238 *	0.233 *
$D$		0.515 *
$L$		−2.664 *	−6.084 *	−3.118 *	−2.592 *	−2.409 *
$K$		1.948 *	2.724 *	−4.696 *	2.612 *	2.581 *
$R$		0.619 *	0.724 *	0.724 *	0.157	0.680 *
$T$		0.017	0.025	0.023	0.042	−0.521
$D\times L$			6.601 *
$D\times K$				16.195 *
$D\times R$					1.075 *
$D\times T$						1.158 *
Fitting degree		0.909	0.872	0.868	0.865	0.822

Note: Eastern China comprises eleven provinces and cities. * is significance levels of 1%.

,

Table 5. Impact estimates for Central China.

Interpreted Variable		SAD	SAD	SAD	SAD	SAD
	Model	(16)	(17)	(18)	(19)	(20)
Variable
Constant		−0.051 *	0.209 *	0.191 *	0.162 *	0.172
$D$		1.618 *
$L$		−0.726 *	−4.416 *	−0.648 **	−0.759 *	−0.536 **
$K$		−0.315	0.974	−6.619 *	0.774	0.506
$R$		0.083 *	0.114 *	0.101 *	−0.072 *	0.153 *
$T$		0.193 *	0.124 **	0.135 *	0.046	−0.367 *
$D\times L$			17.821 *
$D\times K$				38.313 *
$D\times R$					1.520 *
$D\times T$						2.392 *
Fitting degree		0.948	0.901	0.881	0.912	0.901

Note: Central China comprises eight provinces. * and ** are significance levels of 1% and 5%.

and

Table 6. Impact estimates for Western China.

Interpreted Variable		SAD	SAD	SAD	SAD	SAD
	Model	(21)	(22)	(23)	(24)	(25)
Variable
Constant		−0.032 *	0.031 *	0.034 *	0.033 *	0.035 *
$D$		0.790 *
$L$		−0.922 *	−1.261 *	−0.706 *	−0.776 *	−0.726 *
$K$		1.794 *	1.938 *	0.375	1.879 *	1.777 *
$R$		0.124 *	0.121 *	0.142 *	0.047	0.145 *
$T$		0.267 *	0.342 *	0.285 *	0.316 *	0.098
$D\times L$			3.917 *
$D\times K$				11.065 *
$D\times R$					0.706 *
$D\times T$						1.503 *
Fitting degree		0.940	0.887	0.892	0.901	0.904

Note: Western China comprises eleven provinces, cities, and autonomous regions. * is significance levels of 1%.

. The impact of the digital economy on SAD in the eastern region is analyzed through Models (11) to (15). Analysis using Model (11) demonstrates a clear and meaningful positive effect in this specific geographic area, which suggests that enhancements to the digital economy will continue to boost SAD. Analysis using Models (12)~(15) suggests a noteworthy positive effect in the eastern region, which aligns with the results observed at the national level. The eastern region consistently exhibits higher levels of both the digital economy and SAD compared to the national average. However, the digital economy’s impact on SAD in this region is weaker than at the national level, with the interaction effects between the digital economy and key production factors—labor, capital, land, and technology—being lower than those observed nationwide.

The impact of the digital economy on SAD in central China is examined through Models (16) to (20). Model (16) highlights a significant positive impact in this region. Additionally, the positive contribution of the cross-terms of the digital economy and other factors to SAD is demonstrated in Models (17) to (20). The interaction terms between the digital economy and labor, capital, land, and technology are significant at the 1% level. This may be attributed to the central region’s predominantly plain geography, which supports the cultivation of traditional crops. In these agricultural provinces, large-scale farming is prevalent, agricultural industry clusters are well-developed, and the digital economy enhances the efficiency of labor, capital, land, and technology.

In the analysis of the impact on SAD in western China, Models (21) to (25) were employed. According to Model (21), the digital economy in the western part has exerted a positive effect on promoting SAD. The cross-terms of the digital economy and other factors are also shown to have a positive role in SAD, as indicated by Models (22) to (25). The interaction terms between the digital economy and labor, capital, land, and technology are statistically significant at the 1% level.

The digital economy has had a positive impact on SAD throughout China, but the regional impacts differ significantly. Participation in the digital economy has helped boost labor efficiency and has had a positive impact on SAD, notably in central and eastern China.

Eastern China has a relatively large and highly skilled labor population. The eastern region has experienced a boost in labor efficiency by capitalizing on digital tools and drawing on labor resources from the central and western parts. The central part of China contains the majority of the country’s significant agricultural provinces, where large-scale and modern production is more extensive. In terms of agricultural capital efficiency, the eastern region has exhibited comparatively high performance when compared to the central and western parts. In terms of agricultural land efficiency, the central region ranks highest, followed by the eastern region, with the western part lagging behind. In terms of technological efficiency, the central region stands out as particularly prominent, while the eastern and western regions exhibit less marked progress.

4.4. Heterogeneity Analysis of Policy Periods

Chinese national policies and strategies have underscored the significance of developing the digital economy since the 18th National Congress. The influence of this sector on SAD may be linked to the varying policy periods. The 19th National Congress report stated clearly that China’s forthcoming digital economic development will be guided by strategic promotion of integration of digital economy and substantial economy. The 20th National Congress report emphasized the acceleration of constructing “Digital China” and presented strategic initiatives to expedite the digital economy growth. These reports demonstrate the Party Central Committee’s accurate grasp of the characteristics of China’s economic development at different stages and its insightful understanding of the digital economy. The government work report of 2017 marked the first instance in which the term “digital economy” was introduced. Based on national digital economy policy guidance, this paper was adopted in 2017 as the reference point for division. We divide the policy period into two stages, 2013–2017 and 2018–2023, for testing. As exhibited in

Table 7. Results of policy period heterogeneity analysis.

Interpreted Variable		SAD
		2013–2017
	Model	(26)	(27)	(28)	(29)	(30)	(31)
Variable
Constant		0.097 *	0.001	0.161 *	0.157 *	0.165 *	0.165 *
$D$		0.819 *	0.730 *
$L$			−0.686 *	−2.489 *	−0.340	−1.315 *	−0.862 *
$K$			1.232 *	2.439 *	−3.096 *	2.576 *	2.169 *
$R$			0.086 *	0.087 **	0.092 **	−0.099 *	0.160 *
$T$			0.249 *	0.181 *	0.195 *	0.108 **	−0.357 *
$D\times L$				6.700 *
$D\times K$					16.006 *
$D\times R$						1.239 *
$D\times T$							1.749 *
Fitting degree		0.645	0.917	0.787	0.788	0.836	0.777
		2018–2023
	Model	(32)	(33)	(34)	(35)	(36)	(37)
Variable
Constant		0.130 *	0.020 ***	0.187 *	0.172 *	0.187 *	0.184 *
$D$			0.828 *	0.702 *
$L$			−0.593 *	−3.370 *	−0.203	−1.928 *	−1.394 *
$K$			1.208 *	2.418 *	−3.838	2.651 *	2.620 *
$R$			0.020	0.035	0.008	−0.082 **	0.193 *
$T$			0.377 *	0.332 *	0.376 *	0.141 *	−0.322 *
$D\times L$				8.903 *
$D\times K$					17.633 *
$D\times R$						1.344 *
$D\times T$							1.762 *
Fitting degree		0.568	0.894	0.798	0.764	0.850	0.789

Note: *, ** and *** are significance levels of 1%, 5% and 10%.

, Models (26) and (27) are intended to analyze the effect of digital economy on SAD during the period of 2013–2017. Model (26) does not contain other variables, but Model (27) includes other variables. While the estimated coefficient of digital economy in Model (27) is lower than in Model (26), the positive regression coefficient suggests that the model remains robust. Models (32) and (33), which are based on the results from 2018 to 2023, indicate that the digital economy has yielded a favorable impact on SAD. The impact of the digital economy on SAD from 2013–2017 was weaker than from 2018–2023. Prior to 2018, the influence was still in its early stages. However, after 2018, it grew and strengthened, indicating that the effect was greatly affected by the implementation of timely strategic digital economic policies. From the viewpoint of factor input efficiency, the impact coefficients of the intersection of the digital economy and labor, capital, land, and technology on SAD from 2013–2017 were 6.700, 16.006, 1.239, and 1.749. From 2018 to 2023, the impact coefficients of the interaction between the digital economy and labor, capital, land, and technology on SAD were 8.903, 17.633, 1.344, and 1.763, respectively. Compared to 2013–2017, these coefficients increased significantly, indicating that the digital economy has progressively enhanced factor input efficiency and exerted a deepening influence on SAD.

This is due to two reasons: first, the level of China’s digital economy from 2013–2017 was relatively low. The scope of China’s digital economy comprised only 4.4% of the country’s GDP in 2013, but it has been increasing since then. Second, China’s digital economy is mainly oriented towards manufacturing and service industries and was targeted mainly at large enterprises from 2013 to 2017. It was not widely used in agricultural production, agricultural product circulation, or farm product income distribution, and its effect on the agricultural sector had not been fully realized. The performance from 2018–2023 was different. As of 2018, China’s digital economy expanded to CNY 31.3 trillion, comprising 34.8% of the country’s GDP. Digital economy applications in agriculture are gradually expanding. Digital technology has penetrated rural areas and been integrated into agriculture. The digital transformation of agriculture is inevitable for agricultural modernization. In 2023, the scope of China’s digital agricultural economy accounted for 8.9% of agricultural added value and the influence of the digital economy on agricultural development is increasingly profound. The utilization of digital technology has enhanced the efficiency of agricultural factor inputs, thereby propelling the progress of SAD.

4.5. Robustness Analysis

Models (2) and (3) incorporate interaction terms to examine the impact of the digital economy on enhancing the input efficiency of labor, capital, land, and technology in relation to SAD. The significantly positive coefficients of these interaction terms suggest a strong interrelationship between the digital economy and these production factors. To better capture the nonlinear dynamics between variables, we employ quantile models to assess the heterogeneous effects of the digital economy on SAD across different quantiles, mitigate the influence of outliers, and enhance robustness. The results are presented in

Table 8. Quantile regression results across regions.

	Nationwide			Eastern China
Interpreted Variable	SAD	SAD	SAD	SAD	SAD	SAD
Quantile	0.25	0.50	0.75	0.25	0.50	0.75
Constant	−0.027 *	−0.002	−0.038 *	0.024	0.101 *	0.071 *
$D$	0.676 *	0.783 *	0.903 *	0.589 *	0.628 *	0.599 *
$L$	−0.710 *	−0.634 *	−0.888 *	−2.559 *	−2.786 *	−2.708 *
$K$	1.899 *	1.018 *	0.609 ***	1.435 **	1.334 *	1.523 *
$R$	0.086 *	0.052 **	0.079	0.584 *	0.455 *	0.494 *
$T$	0.236 *	0.314 *	0.295 *	0.067	0.174 **	0.139 *
Fitting degree	0.662	0.695	0.747	0.683	0.750	0.756
	Central China			Western China
Interpreted Variable	SAD	SAD	SAD	SAD	SAD	SAD
Quantile	0.25	0.50	0.75	0.25	0.50	0.75
Constant	−0.054 *	−0.042 **	−0.035	−0.038 *	−0.042 *	−0.035 *
$D$	1.593 *	1.590 *	1.531 *	0.669 *	0.836 *	0.885 *
$L$	−1.014 *	−0.656 ***	−0.633 **	−0.779 *	−0.939 *	−1.142 *
$K$	0.277	−0.227	−0.237	1.987 *	2.156 *	1.962 *
$R$	0.103 *	0.079 *	0.074	0.098 ***	0.102 **	0.170
$T$	0.141 *	0.168 *	0.235 *	0.257 ***	0.289 *	0.242 *
Fitting degree	0.775	0.787	0.782	0.775	0.751	0.751

Note: *, ** and *** are significance levels of 1%, 5% and 10%.

.

Table 8 reveals that the impact of the digital economy on SAD exhibits nonlinear characteristics. At the national level, the effect is relatively small at lower percentiles but increases at higher percentiles, suggesting that provinces with higher levels of SAD benefit more prominently. In the eastern region, the impact coefficient initially rises with increasing quantiles before declining, indicating that the digital economy exerts a stronger influence on SAD in provinces with lower levels. In the central region, the impact coefficient decreases steadily across quantiles, implying that the effect of the digital economy is weaker in provinces with higher levels of SAD than in those with lower levels. In contrast, in the western region, the impact coefficient increases with quantiles, suggesting that the digital economy plays a more significant role in provinces with higher SAD levels.

4.6. Endogeneity Test

Endogeneity concerns—arising from omitted variable bias, measurement error, reverse causality, or sample selection—pose a threat to the internal validity of the model. In particular, the possibility that regions with more advanced sustainable agricultural development (SAD) may be more likely to foster a digital economy raises the issue of reverse causation. To mitigate this, we employ a two-step generalized method of moments (GMM) approach, using the share of value added by secondary and tertiary industries in GDP as an instrumental variable for the digital economy. This instrument is plausibly exogenous to SAD but strongly correlated with the development of the digital economy, satisfying the relevance and exclusion restrictions. The GMM estimates yield a coefficient of 0.527 for the effect of the digital economy on SAD, significant at the 1% level, reinforcing the robustness of the main findings.

5. Discussion

5.1. Pathways and Dynamics of Digital Economy in Shaping SAD

This study leverages panel data from 2013 to 2023 to empirically explore how digitalization shapes SAD through improvements in factor input efficiency. By employing a multidimensional measurement framework for both the digital economy and SAD, and by decomposing results by region and policy phases (2013–2017 and 2018–2023), our analysis offers new insights into the mechanisms underpinning agricultural transformation in China.

Our findings clarify several critical pathways. First, we confirm that the digital economy significantly promotes SAD across production, circulation, and distribution dimensions, substantiating hypothesis H1. Although aligned with earlier studies highlighting digital infrastructure’s beneficial role in agriculture modernization [39], our analysis extends these findings by quantifying how digitalization systematically influences integrated agricultural value chains, using a more comprehensive and nuanced measure of SAD.

Second, the analysis demonstrates that digital economy advancements drive SAD primarily by increasing the efficiency of critical production inputs. Notably, labor and capital efficiencies emerge as particularly responsive, reflected by the highest interaction coefficients, confirming hypothesis H2. These results move beyond conventional productivity-centric narratives prevalent in prior literature, providing concrete evidence that digital technologies actively optimize factor allocation and use. Regional analysis further highlights pronounced effects in central China, enriching existing scholarship on regional disparities [67].

Third, a clear policy-phase effect emerges from our temporal analysis. The digital economy’s influence on SAD intensified significantly following the implementation of targeted national digital economy strategies post-2018. This temporal dimension, often overlooked in existing cross-sectional analyses, underscores the vital role institutional frameworks and policy environments play in determining the efficacy of digital interventions [68,69].

Fourth, our analysis identifies a complex, nonlinear, and regionally differentiated relationship between digital economy development and SAD. The eastern region exhibits an inverted U-shaped pattern, reflecting initial rapid gains followed by diminishing marginal returns, likely indicating saturation effects. Conversely, the central region displays a decreasing trend over time, potentially signaling structural limitations in translating digital initiatives into sustainable agricultural outcomes. The western region shows sustained positive impacts, indicative of ongoing benefits accruing from recent investments in digital infrastructure. These region-specific findings diverge significantly from previous studies that typically portray regional effects as static and uniform [70,71].

5.2. Theoretical Contributions

This study advances the theoretical discourse on the digital economy and SAD in three substantive ways. First, it proposes an integrated, multidimensional framework for defining and measuring both the digital economy and SAD. Most existing literature has relied on fragmented or single-dimensional metrics, limiting analytical scope and theoretical coherence. In contrast, this study constructs a systematic index encompassing digital infrastructure, digital application, and digital industrialization, thereby offering a structural representation of digital economic evolution. On the SAD side, the framework transcends production-centric definitions by incorporating product circulation and income distribution—thus aligning more closely with holistic sustainability paradigms and rural transformation goals. This dual-conceptual expansion lays a foundation for more robust theorizing about the interactions between digitalization and agrarian development.

Second, the study contributes a theoretically grounded mechanism model that distinguishes between the direct and indirect pathways through which the digital economy influences SAD. Direct effects arise from enhanced production processes, while indirect effects are mediated through improvements in factor input efficiency. This distinction enables a more granular understanding of how digitalization reshapes agricultural systems across value-chain stages. By formally modeling and empirically validating these pathways, the study extends prior theoretical frameworks that have largely treated digital agriculture as a black box, and it offers a replicable model for future inquiries into digital-sector linkages.

Third, the research introduces factor input efficiency—including labor, land, capital, and technology efficiency—as a central mediating construct in the digitalization–SAD nexus. This perspective moves beyond conventional productivity measures and emphasizes structural transformation in input allocation and resource optimization. The study further theorizes spatial heterogeneity as an integral dimension of digital transformation, showing how regional disparities mediate the effectiveness of digital interventions. In doing so, it reframes digital development not as a uniform catalyst but as a context-sensitive process contingent on local conditions and institutional readiness.

5.3. Practical Implications

The findings of this study underscore the transformative potential of the digital economy in driving SAD, particularly through improvements in factor input efficiency and industrial integration. To harness this potential, several policy and investment priorities are recommended.

First, it is recommended to consolidate digital infrastructure and inclusive technology access. The deepening integration of digital technologies into rural industries remains uneven, with notable disparities in infrastructure, human capital, and service accessibility. While big data and smart technologies hold promise for improving agricultural productivity and sustainability, their diffusion is hampered by limited availability and uptake among smallholder farmers. Policymakers should prioritize building robust, inclusive digital infrastructures that bridge urban–rural divides, especially in central and western China, where the digital economy exhibits a rising and region-specific influence on SAD. Incentives should be introduced to support the development and scaling of tailored digital services for small and medium agricultural actors, ensuring that technological benefits are equitably distributed across demographic and geographic boundaries.

Second, the findings suggest the need for embedding digital technologies across the agricultural value chain. To enhance the quality, safety, and traceability of agricultural products, digital technologies must be embedded across the entire agricultural value chain—from production and processing to distribution and market intelligence. Investment should focus on integrating interoperable digital systems, improving standards for data sharing, and facilitating the adoption of advanced tools such as AI-powered analytics and IoT-based monitoring. These innovations can reduce inefficiencies, increase transparency, and optimize resource allocation, creating conditions for higher value-added agriculture.

Third, it is important to align digital development with regional strengths and stages. The nonlinear and regionally heterogeneous effects of the digital economy on SAD point to the need for differentiated development strategies. In the east, where digital infrastructure is relatively mature, the focus should shift to optimizing applications and improving the return on digital investment. In contrast, the central and western regions require foundational investments in digital literacy, infrastructure, and institutional support to build digital capacity and leverage their emerging advantages. Region-specific pathways should guide the allocation of resources and the design of supportive policies that match local strengths and stages of digital transformation.

Finally, it is essential to enhance human capital and governance for digital transition. Sustained progress in digital-driven agricultural transformation requires coordinated investments in human capital, particularly digital skills training for rural populations. Policies should promote the flexible flow of talent into agriculture and incentivize digital entrepreneurship in rural areas. Additionally, institutional mechanisms for evaluating digital performance, protecting data rights, and fostering innovation ecosystems are crucial to ensure the long-term scalability and governance of digital agriculture.

In summary, the digital economy represents not only a technological evolution but also a structural opportunity to reimagine rural development. For both policymakers and investors, a strategic approach to digital integration—grounded in empirical evidence and responsive to regional contexts—will be essential to unlocking inclusive, efficient, and sustainable agricultural futures.

5.4. Limitations and Future Directions

While this study provides robust macro-level evidence on the role of the digital economy in promoting SAD, several limitations remain. The absence of micro-level data limits insight into behavioral mechanisms at the farm or enterprise level, underscoring the need for future research incorporating household surveys or case studies. Additionally, while regional and temporal heterogeneity are explored, effectively visualizing these multidimensional dynamics remains challenging, and future work could benefit from advanced spatial–temporal visualization techniques. Methodologically, although the study employs panel and quantile regression models, these approaches rest on assumptions that may not fully capture the complexity of causal relationships, and some endogeneity may persist despite the use of instrumental variables. Moreover, the quantification of both the digital economy and SAD, while systematic, simplifies inherently multidimensional constructs. Future research should refine these metrics, explore nonlinear modeling, and extend the temporal scope to capture emerging digital trends and technologies.

6. Conclusions

This study presents comprehensive empirical assessments of how the digital economy drives SAD in China. By employing panel data from 2013 to 2023 and constructing a multidimensional indicator system, we provide new evidence that the digital economy significantly promotes SAD—both directly through enhanced production, circulation, and distribution, and indirectly via improvements in factor input efficiency. Our findings reveal that the digital economy’s influence is not uniform: it varies substantially across regions and over time. Central China exhibits the strongest interaction between digital development and input efficiency, while post-2018 digital policies have substantially amplified national-level impacts. Moreover, the relationship between digitalization and SAD is nonlinear, underscoring the need for differentiated, context-specific strategies. Theoretically, this study contributes a new conceptual framework linking digital infrastructure, digital industrial development, and digital applications to sustainable outcomes in agriculture. It deepens our understanding of how digital technologies reshape agricultural systems through input optimization and sectoral integration. Practically, our results offer targeted guidance for policy design—highlighting the importance of inclusive infrastructure, regionally tailored interventions, and investments in digital talent and governance systems. Ultimately, the digital economy represents more than a technological shift—it is a structural force capable of transforming rural development. As digital transformation accelerates, ensuring that its benefits reach across geographies, production systems, and communities will be essential to achieving an inclusive and sustainable agricultural future.