An assessment system for the digital economy was created by following data availability and scientific rigor guidelines, using the earlier research framework and previous studies as references (Wang et al., 2025; Li et al., 2024). A total of 11 indicators were selected, which primarily reflect the overall level of digital development across regions and indirectly influence agricultural digitalization. The weights of each indicator were calculated using the entropy method, with details provided in Table 1.

Drawing on existing research (Zou et al., 2024; Ma et al., 2024; Tan et al., 2024), this study constructs an evaluation indicator system for CLUE, focusing on input, expected output, and unexpected output. Table 2 lists the 11 indicators that were chosen in total.

The entropy technique circumvents the drawbacks of subjective weighing by using an objective methodology that determines weights based on the fluctuation in indicator data (Chen, 2019). Entropy is the method tested in this work to provide weights to the digital economy’s indicators and conduct a thorough evaluation of China’s level of growth in this area. Main calculation stages as in Equations 1–6:

The first step, the standardization of the raw data, the standardized data values for the positive indicators and negative indicators are as follows:

The second step, calculate the information entropy:

The third step, calculate the weight of each indicator:

The fourth step, calculate the Digital Economy Development Index:

This investigation used the super-efficiency SBM model. The efficient decision-making units are further broken down and ranked, and unexpected outcomes are included in the efficiency level assessment, improving the precision and applicability of efficiency assessments (Wu and Wen, 2025). The expression is as follows:

In Equation 7, each province represents a decision-making unit. m denotes the input for each decision-making unit, p 1 represents the expected output, p 2 represents the unexpected output, and s − , s + , s b − represent the slack variables for input, expected output, and unexpected output, respectively. x k , y k , b k represent the input, expected output, and unexpected output vectors of the decision-making unit, respectively. X, Y, B represent the input, expected output, and unexpected output matrices, respectively, and λ denotes the weight vector.

The two parts of the Coupling Coordination Degree Model are the coupling and coordination degrees. The coupling degree denotes the level of interaction among various systems, whereas the coordination degree indicates the developmental status of diverse systems according to a uniform measuring standard (Qindong et al., 2025). The CCD between the digital economy and CLUE is calculated in this study using the coupling coordination degree model. The following is the calculating formula:

In Equations 8–10: C represents the coupling degree, T represents the coordination degree, and D represents the CCD, with a value range of [0, 1]. The closer D is to 1, the greater the coordination between the two systems. α and β are the system weights, with their sum equal to 1. Based on existing research, the CCD is divided into five levels, as shown in Table 3.

Dagum (1997) suggested a decomposition method for the Gini coefficient that overcomes the drawbacks of current approaches to measuring regional differences. This approach makes finding the causes of regional discrepancies easier (Liu et al., 2025). This approach makes finding the causes of regional discrepancies easier. This study uses the Dagum Gini coefficient and its decomposition approach to thoroughly examine the general variations and causes of variations in the degree of coupling coordination regarding China’s digital economy and CLUE. The calculation formulas for the overall Gini coefficient G , intra-regional Gini coefficient G ij , and inter-regional Gini coefficient G jh are detailed in Equations 11–13:

Further, in Equations 14–16, the Gini coefficient is decomposed into the contributions of regional differences G w , inter-regional differences G nb , and super-density contributions G t :

Adjacent regions exhibit some form of spatial interaction and dependence. This work uses Moran’s I index for spatial autocorrelation analysis to make measuring their correlation easier. While the local Moran’s I index is used to examine the precise locations and extent of the aggregation, the global Moran’s I index is used to determine if spatial aggregation occurs within the region (Zhang et al., 2025). Equations 17,18 are used to calculate the global and local Moran’s I indices:

I is the global Moran’s I index. When I > 0, it indicates a positive spatial correlation in the distribution of the variable, the larger the value, the stronger the spatial correlation. When I < 0, it indicates a negative spatial correlation, and the smaller the value, the greater the spatial disparity. When I = 0, it indicates no spatial correlation of the variable. I i represents the local Moran’s I index.

This research uses the classic Markov chain and spatial Markov chain analysis methods to determine the transition probability matrix under various scenarios (Abhinav and Rajesh, 2025). The calculation formulas are as follows:

In Equations 19,20, P ij represents the transition probability, Lag denotes the spatial lag value, X represents the observed value, and W denotes the spatial weight matrix.

This model selects the CCD as the dependent variable, which has a range of values between 0 and 1, making it a limited dependent variable (Zhu et al., 2025). The influence of driving factors on the CCD may not be well reflected in the skewed regression findings that arise from using OLS to estimate a constrained dependent variable. To do regression on the driving factors, this study uses the Tobit model, which has the following formula:

In Equation 21: y it is the dependent variable, x it represents the driving factors, a is the constant term, β i is the regression coefficient, and ε it is the random error term.

This study applies the entropy method to construct China’s digital economy development index from 2014 to 2022. As shown in Figure 3, the national index exhibited a fluctuating upward trend, rising from 0.0859 to 0.1998 over the study period. Regionally, the eastern area experienced rapid growth, with its index increasing from 0.1459 to 0.3156, significantly outpacing other regions. The central region’s index rose from 0.0573 to 0.1476, while the western region, starting from a lower base, increased steadily from 0.0467 to 0.1219. In 2014, most provinces had a digital economy development index between 0.05 and 0.10. By 2022, the index had grown substantially, particularly in the eastern region, where most provinces recorded values between 0.20 and 0.40. Notably, Beijing and Guangdong exceeded 0.50, indicating leading positions in digital economic development.

In summary, the growth of China’s digital economy exhibits significant regional variation. The eastern region consistently leads due to its strong technological innovation capacity and robust economic foundation. Although the central and western regions have shown positive momentum and some advantages as late developers, they continue to face challenges related to uneven development and structural constraints.

To assess CLUE in China from 2014 to 2022, the study employed the super-efficient SBM model based on unexpected output. As shown in Table 4, both national and regional CLUE demonstrated a consistent upward trend during the study period. At the national level, efficiency increased from 0.4881 to 0.8736. The central region began with a relatively high efficiency score of 0.5377 but exhibited slower growth, reaching 0.7534 by 2022. In contrast, the western region experienced the most rapid improvement, with efficiency rising from 0.4599 to 0.9214. Notably, most provinces in the western region recorded efficiency values exceeding 1.0. Although the overall efficiency in the central region remains lower than in the west, provinces such as Jilin, Hubei, and Heilongjiang surpassed the 1.0 threshold.

The steady increase in CLUE across the country and within regions can be attributed to the implementation of China’s cultivated land conservation system, strict controls on converting cultivated land to construction land, and reinforcement of the balance between cultivated land occupation and compensation. However, substantial regional disparities persist. The eastern and western regions have outperformed the central region, primarily due to advances in agricultural technology, improvements in farm management practices, and stronger policy support.

Table 5 presents a shift from mild disorder to benign coordination in the national CCD between the digital economy and CLUE, rising from 0.341 in 2014 to 0.621 in 2022. The eastern region has always been at the forefront, with the CCD increasing from 0.404 to 0.723, indicating a transition from imminent disorder to benign coordination. This sustained advantage suggests that the eastern region is better positioned to integrate and coordinate the development of the digital economy and CLUE due to its strong infrastructure and technological foundation.

In 2014, most provinces in China had relatively low CCD, primarily between 0.3 and 0.4, corresponding to the mild disorder category. By 2018, the eastern region’s coordination levels had improved further, with all provinces exceeding 0.4. Notably, coordination degrees in Shanghai, Guangdong, and Jiangsu surpassed 0.6. During the same period, the central and western regions also showed improvement, with values increasing from 0.35 to 0.50. By 2022, the CCD had significantly increased across all regions, although notable regional differences remained. In the central region, most provinces had coordination values between 0.4 and 0.6. In the western region, values ranged from 0.5 to 0.7. In the eastern region, most provinces recorded values between 0.6 and 0.9.

As shown in Figure 4, the relationship between the digital economy and CLUE in China exhibited a steady and positive growth trend from 2014 to 2022. Based on CCD levels, the study period can be divided into two distinct stages. In the first stage, from 2014 to 2018, the CCD rose from 0.341 to 0.483, progressing from the mild disorder stage to the imminent disorder stage. A key characteristic of this phase was a marked decline in the proportion of serious disorder, the initial emergence of benign coordination, and a gradual increase in the share of basic coordination. However, overall coordination remained limited, largely due to underdeveloped digital infrastructure and a shortage of skilled digital technology professionals during the early development phase. In the second stage, from 2019 to 2022, the CCD increased significantly from 0.523 to 0.621, reaching the benign coordination stage. During this period, serious disorder was eliminated, mild disorder declined sharply, and benign coordination became the dominant state. This transformation can be attributed to the Chinese government’s 2020 policy to accelerate digital development, promote the construction of a digital China, and advance smart agriculture and digital rural development. In this context, the digital transformation and upgrading of traditional agricultural sectors, along with the widespread adoption of agricultural machinery, significantly enhanced productivity and further improved the coupling coordination between the digital economy and CLUE.

From the temporal variation of the CCD, it is evident that in 2014, 2018, and 2022, China as a whole was in different stages of coordinated development. These years were selected as characteristic points for visualization analysis using ArcGIS 10.8.1, as shown in Figure 5. Throughout the study period, a distinct spatial differentiation pattern was visible in the CCD between China’s digital economy and CLUE. In 2018, a “multi-core” growth pole, centered around Guangdong, Shanghai, and Jiangsu—three provinces with benign coordination—radiated outward, driving the expansion of the coastal eastern region toward the inland central region. This ultimately formed a development pattern of “eastern lead, central moderate, and western rise.” Even though the coordination in the west and center of the country got better across the research, these areas still faced problems like uneven policy support, limited benefits from technological advancements, and basic industrial setups. As a result, the digital economy and CLUE failed to interact benignly, and a gap remained compared to the eastern region. Although regional disparities remained, the nation’s overall CCD had significantly improved by 2022. The eastern and some central areas were home to the majority of the high-coordination provinces. In 2014, 57.8% of provinces in the central and western regions were in a state of serious disorder. By 2022, the coordination types in these areas gradually shifted to basic coordination, with provinces such as Sichuan, Henan, Hubei, Guizhou, Chongqing, and Shaanxi improving to a benign coordination state.

This study uses Stata 17.0 to calculate the global Moran’s index of the CCD for the period from 2014 to 2022. As shown in Table 6, all estimated values of the Moran’s index are positive. In 2018, the p-value is greater than 0.05, indicating that the result is not statistically significant. In contrast, the p-values for the other years are all less than 0.05, suggesting statistically significant spatial autocorrelation. These results indicate that provinces with similar levels of coupling coordination, whether high or low, tend to cluster spatially. Moreover, the global Moran’s index follows a generally upward trend, rising from 0.0960 in 2014 to 0.1418 in 2022. This pattern shows that the spatial agglomeration of the CCD has gradually intensified over time.

Using the economic geographic distance matrix, the local Moran index for 2014 and 2022 was calculated, and the local Moran scatter plots for these key years were generated. As illustrated in Figure 6, the “high-high” (HH) areas in the first quadrant and the “low-low” (LL) areas in the third quadrant are the main patterns of how the digital economy and CLUE are connected, and this connection remains fairly stable over time. The “High-High” (HH) type agglomeration areas in the first quadrant are mainly concentrated in the central and eastern regions, indicating that the provinces in these areas have strong spillover effects and act as important growth poles driving coordinated development in other regions. On the other hand, provinces in the western regions mostly appear in the “Low-Low” (LL) type agglomeration areas in the third quadrant. The number of provinces exhibiting the “Low-High” (LH) and “High-Low” (HL) clustering patterns is small and dispersed across different regions, suggesting weaker spatial heterogeneity in these areas. Ultimately, a spatial association pattern with “homogeneous features as the main characteristic and heterogeneous features as the supplementary” is formed. However, among the four spatial correlation patterns, the majority of provinces fall under the “Low-Low” (LL) type homogeneous agglomeration pattern in the third quadrant.

Figure 7a illustrates that from 2014 to 2022, the Gini coefficient of regional coupling coordination exhibited a declining trend, indicating that disparities in internal coupling coordination levels across regions have gradually narrowed. The average Gini coefficient in the central region was lower than that in the eastern region, which in turn was lower than that in the western region. This suggests that the central provinces exhibited less variation in coordination development than the eastern region, while the western region showed the most pronounced disparities. Notably, the central region experienced the largest decrease in intra-regional differences, with its Gini coefficient declining from 0.173 in 2014 to 0.076 in 2022, marking a 56.07 percent reduction.

These differences reflect imbalances in technological innovation, economic development, and other structural factors among provinces during the promotion of the digital economy and the enhancement of CLUE. The implementation of the Central Region Revitalization Plan in 2016 provided targeted policy support, leading to industrial upgrading, improved infrastructure connectivity, and the integration of underdeveloped areas into broader regional economic systems. These initiatives significantly fostered coordinated development across eastern, central, and western regions, thereby narrowing the internal development gap within the central region.

As illustrated in Figure 7b, inter-regional differences generally declined over time; however, disparities between the eastern and western regions have gradually widened since 2019. The average Gini coefficient between the eastern and western regions is 0.138, which is higher than the national average and thus warrants special attention. These differences may stem from uneven development in economic capacity, capital accumulation, and technological progress. The eastern region benefits from geographic advantages such as flat terrain and access to international trade through major ports, making it a hub for industrial clustering and global commerce. Foreign investment and advanced technologies are highly concentrated in this region, enabling the transformation of western natural resources into high value-added products through technological upgrading. This dynamic has created a pattern of “resource transfer westward, profit flow eastward,” which has also contributed to labor siphoning from the west to the east. As a result, technological talent has continuously migrated eastward, slowing the western region’s progress in digital economy development and CLUE.

Figure 7c reveals that inter-regional differences have consistently contributed the most to the total Gini coefficient, maintaining a dominant share of approximately 47%. In contrast, intra-regional differences remain relatively stable, typically ranging between 27 and 30%. Additionally, the declining contribution rate of hypervariable density suggests that developmental disparities among provinces are gradually narrowing. Given that inter-regional discrepancies represent the largest share of overall inequality, future efforts to address spatial imbalances in the coupling coordination between the digital economy and CLUE should prioritize narrowing regional gaps.

Several elements influence the degree of cooperation between China’s digital economy and the efficiency of cultivated land use. Based on relevant studies, this paper takes the CCD as the dependent variable and analyzes the impact of five factors as independent variables: urbanization, economic level, industry structure, technological innovation level, and urban–rural development gap, with the variable definitions provided in Table 8. A multicollinearity test is performed on the driving factors to guarantee the independence of each variable and the dependability of the regression findings (De La Puente Pacheco et al., 2025). Further analysis may be carried out as there is no multicollinearity among the variables, as indicated by the VIF of all driving factors being less than 10.

The driving factors of the CCD from 2014 to 2022 were examined by including the degree of coordination between China’s digital economy and CLUE into a Tobit regression model with random effects panel using StataMP 17 software. The regression results are shown in Table 9. As seen from Table 2, there are significant differences among the various driving factors, with the degree of impact ranked as follows: urbanization > technological innovation level > economic level > urban–rural development gap > industry structure.

Economic development, technological innovation, and urbanization have a positive impact on the coordination between China’s digital economy and CLUE. The spatial concentration of population, capital, and technology during urbanization accelerates the spread of digital technologies into agriculture, promoting a shift from scattered to intensive land use. Around cities, cultivated land is consolidated into contiguous farming areas through reclamation and restoration efforts. This aggregation of production factors improves land use efficiency, reduces the cost of technology adoption, and promotes the widespread application of digital tools such as land monitoring systems and intelligent irrigation technologies. Technological innovation is the fundamental driving force. The development of smart agricultural equipment and the integration of digital technologies enhance the precision of land resource management. Agricultural big data platforms combine meteorological, soil, and market information to optimize crop rotation plans, reduce pollution from fertilizers and pesticides, improve soil quality, and increase land use efficiency. These improvements help resolve the traditional conflict between agricultural efficiency and environmental protection. Economic development provides the material foundation for digital research and application. Economically developed regions are more likely to invest in digital infrastructure and promote green technologies, thereby supporting the intensive and ecological transformation of cultivated land and facilitating the integration of digital technologies with land resources.

In contrast, the urban–rural development gap and industrial structure negatively affect coordination. To enhance the synergy between the digital economy and land use efficiency, it is necessary to adjust the industrial structure and narrow urban–rural disparities. The government should focus on guiding industries toward high value-added activities that enable the diffusion of digital capabilities into the agricultural sector, reshaping cultivated land use models. If the urban–rural gap remains large, it may lead to the outflow of rural labor and capital, widen the technological divide, and concentrate public resources in urban areas. As a result, rural areas may become overly reliant on intensive land development to compensate for economic disadvantages, increasing ecological pressure and weakening ecosystems. Addressing these imbalances, promoting integrated urban and rural development, and improving the coordination between the digital economy and land use efficiency are essential for sustainable progress.

The driving characteristics of the CCD between the digital economy and CLUE in various areas are analyzed. Table 9 columns (2) through (4) display the findings. First, with values of −0.6740, −0.4490, and −0.3590, respectively, the industry structure coefficients in the eastern, central, and western areas all passed the significance test, suggesting that the influence of industry structure varies by region. Second, urbanization has a significant impact on the eastern and western regions. This shows that urbanization promotes the efficient integration of the digital economy and agricultural resources through the concentration of population, capital, and technology in cities, thereby improving CLUE. Third, economic levels have a positive impact on the central and western regions. Economic growth can drive the development of the digital economy, thereby creating conditions for improving CLUE. Fourth, the urban–rural development gap has a negative impact on central and western regions. The urban–rural development gap leads to a long-term one-way flow of capital, technology, and talent toward cities, creating a digital divide. Rural areas lack digital infrastructure and talent, weakening the synergistic potential between the digital economy and farmland efficiency. Finally, technological innovation has a significant positive effect on the eastern region, but not on other regions. This shows that the eastern region may be more likely to convert science and technology into productivity and improve CLUE due to its high level of science and technology transformation. In contrast, the digital infrastructure level in the central and western regions is relatively backward, and the cost of technology diffusion is too high, thereby inhibiting and offsetting the promotional effect that technological innovation should have on the CCD.