Empirical Investigation of Cultivated Land Green Use Efficiency and Influencing Factors in China, 2000–2020

Abstract

The rapid industrialization and urbanization promote socioeconomic development, but also pose a certain threat to food and ecological security. Cultivated land green use efficiency (CLGUE) is an important indictor to comprehensively reflect the coordinated relationship between cultivated land utilization and ecological protection. Therefore, it is of great practical significance to explore CLGUE to guarantee efficient and sustainable utilization of cultivated land resources. This paper thus conducts an empirical investigation of 31 provinces in mainland China during 2000–2020, aiming to measure the CLGUE level using the Super-SBM model and explore its influencing factors based on panel regression model. The data, which were mainly derived from various statistical yearbooks, together with the reference dataset, were all accurate. The results show that the average CLGUE value in China exhibited a fluctuating upward development trend, with the highest efficiency value of 0.957 in 2020 and the lowest one of 0.853 in 2003. Northeastern China had the highest efficiency value, while Central China had the lowest efficiency value. The overall ranking of CLGUE in the four major regions from high to low is Northeastern, Eastern, Western, and Central China. Spatially, there are significant diversities in CLGUE across China, which means that differentiated measures need to be taken to improve the efficiency based on regional natural conditions and the socioeconomic level. The regression model indicated that the crop diversity index, GDP per capita, urbanization level, effective irrigation rate, and fiscal support for agriculture positively influenced the CLGUE, while the proportion of natural disaster area had a negative impact. The findings had important implications for improving the CLGUE and achieving sustainable agricultural development.

1. Introduction

Cultivated land is the essence of land resources, and plays a vital role in ensuring food production, socioeconomic development, and ecological security [1,2]. The world’s cultivated land area is about 1.6 billion hectares, supporting nearly 8 billion people. It contributes an important role in alleviating the food crisis, maintaining social stability and ensuring sustainable development in the world. Since the reform and opening up in 1978, China’s agriculture has witnessed rapid development, achieving abundant harvests in both grain output and agricultural output value [3]. Grain output increased from 304.7 million tons in 1978 to 669.49 million tons in 2020, with an average growth rate of 2.85%. The total agricultural output value has also risen, from 111.75 billion yuan in 1978 to 16,690 billion yuan in 2020, with an average growth rate of 353.22%. However, the issue of food security in China cannot be ignored. The country’s cultivated land accounts for approximately 7.5% of the world’s total, but it needs to feed around 23% of the global population [4,5]. Furthermore, with the accelerated process of industrialization and urbanization, urban construction continues to encroach upon cultivated land, leading to a continuous decrease in its area [6,7]. On the other hand, as rural populations migrate to cities and there is a lack of rural labor, cultivated land is left and abandoned [8]. The pressure on cultivated land has continued to increase [9].

Cultivated land in China has being excessively exploited to some extent for its high population density and limited land resources, which may have resulted in soil degradation and decreased quality [10,11]. Additionally, the widespread use of fertilizers, pesticides, and livestock farming in rural areas has led to agriculture non-point source pollution in some regions [12]. These factors may severely constrain the sustainable utilization of cultivated land resources. To this end, in the No. 1 document of the CPC Central Committee in 2022, it was clearly proposed to strengthen the comprehensive treatment of agricultural non-point source pollution, reduce agricultural inputs, strengthen the utilization of livestock and poultry manure, agricultural film, straw and other resources, and promote the green development of agriculture and rural areas. Therefore, it is crucial for the high-efficiency and sustainable utilization of cultivated land resources to achieve social and economic sustainable development.

Cultivated land green use efficiency (CLGUE) is an important indicator for measuring the sustainable utilization of cultivated land resources [13,14,15]. Currently, research on the CLGUE mainly focuses on the measurement of cultivated land use efficiency under environmental constraints, regional differences, and its influencing factors [16,17,18,19]. In the design of indicator systems, scholars typically construct a system based on the input–output concept and incorporate carbon emissions as a non-desired output to achieve the goal of a low-carbon measurement of cultivated land utilization [5,20]. Furthermore, some scholars also consider the issue of agriculture non-point source pollution during cultivated land utilization and include it together with carbon emissions as non-desired outputs [21,22]. In terms of research methods, early studies mainly employed single-indicator and multi-indicator composite measurement methods [23,24]. Subsequently, non-parametric analysis methods such as data envelopment analysis (DEA) and parametric methods such as stochastic frontier analysis (SFA) were widely used to measure cultivated land use efficiency [25,26,27]. Among them, the Super-SBM (Slacks-Based Measure) model is capable of handling multiple inputs and outputs without the need to specify a specific functional form or pre-estimate parameters [5,28]. It effectively addresses the issue of variable relaxation in input–output analysis and has thus become one of the most popular methods in this field. As for the influencing factors, various factors at different levels can have impacts on cultivated land use efficiency. Micro-level factors include the land transfer behavior of farmers, operating scale, and household characteristics [29,30,31,32]. Macro-level factors include natural conditions, production conditions, technological inputs, and the agricultural migration population [3,17,33].

Overall, existing research has provided a systematic summary of cultivated land utilization efficiency under environmental constraints. However, there is still room for further exploration. Firstly, existing research studies have mainly focused on carbon emissions from cultivated land utilization but do not consider the carbon sequestration function. Relevant studies have shown that cultivated land is not only a carbon source but also a huge carbon sink [34,35]. If the carbon sequestration function is not taken into account, the true CLGUE level cannot be accurately measured. Secondly, existing research about the study area mainly is mainly on major grain-producing regions and specific areas such as the reaches of the Yangtze River [36,37,38]. Thus, there is a lack of research that expands the scope to a national level and explores the temporal and spatial characteristics of the CLGUE and its differences in influencing factors among different regions in China. Due to the differences in resource endowments and production conditions among provinces, the interprovincial CLGUE may show different trends over time. Therefore, exploring the spatio-temporal characteristics and influencing factors of the CLGUE is of great significance in promoting the coordinated development of cultivated land utilization, agricultural economy growth, and ecological environment protection.

In this paper, we aim to use the Super-SBM model to calculate the CLGUE values based on the panel data of 31 provinces in mainland China and analyze its spatio-temporal characteristics from 2000 to 2020 by using mathematical statistics and GIS visualization methods. Then, the panel regression model is applied to explore how factors influence the CLGUE across China and its different regions from the perspective of natural conditions, socioeconomic development level, cultivated land use conditions, and agricultural policies. Finally, some policy implications are proposed to improve the CLGUE and promote sustainable agricultural development.

2. Study Area and Data Sources

2.1. Study Area

The total number of administrative divisions in China is 34 (including provinces, autonomous regions and municipalities). The country has a vast territory, which leads to a significant difference among its administrative divisions in terms of natural geography, economic development, and resource endowment [39]. Based on data availability, the research objects of this paper were selected as 31 provinces in mainland China from 2000 to 2020, excluding Hong Kong, Macao and Taiwan. In addition, according to relevant studies [40,41], we divided the administrative divisions of China into four regions: Northeastern, Eastern, Central and Western China (Figure 1). Specifically, Northeastern China consists of Liaoning (LN), Jinlin (JL), and Heilongjiang (HLJ). Eastern China consists of Beijing (BJ), Tianjin (TJ), Hebei (HEB), Shanghai (SH), Jiangsu (JS), Zhejiang (ZJ), Fujian (FJ), Shandong (SD), Guangdong (GD), and Hainan (HAN). Central China consists of Shanxi (SX), Anhui (AH), Jiangxi (JX), Henan (HEN), Hubei (HUB), and Hunan (HUN). Western China consists of Tibet (TI), Qinghai (QH), Ningxia (NX), Xinjiang (XJ), Shaanxi (SAX), Gansu (GS), Chongqing (CQ), Inner Mongolia (IMG), Sichuan (SC), Guizhou (GZ), Yunnan (YN), and Guangxi (GX).

2.2. Data Sources

In this paper, we used the input–output data of cultivated land utilization in mainland China during 2000–2020. The data were mainly derived from various statistical yearbooks, including the China Statistical Yearbook, China Environmental Statistical Yearbook, China Energy Statistical Yearbook, China Water Resource Bulletin, and China Rural Statistical Yearbook. Some missing data were obtained using interpolation or trend extrapolation. The above basic data were collated and summarized in Excel 2019. Finally, a panel dataset on CLGUE measurement and its influencing factors was formed.

3. Methodology

3.1. Variable Selection of Measuring the CLGUE

Essentially, the CLGUE means to obtain as large a desired output and as small an undesirable output as possible with as little resource input as possible, which comprehensively reflects the coordinated relationship between cultivated land utilization and ecological protection [17,23]. Based on existing research [5,15,17,19] and considering data availability, we constructed a measurement indicator system of CLGUE (

Table 1. Evaluation index system of the CLGUE.

Factor Layer	Indicator Layer	Variables	Abbreviations	Units
Inputs	Land	The sown area of cultivated land	SACL	1000 hectares
	Labors	Number of employees in the primary industry	NEPI	10,000 peoples
	Means of production	Total power of agricultural machinery	TPAM	gigawatt
		Usage of agricultural films	UAF	ton
		Usage of pesticides	UP	ton
		Usage of chemical fertilizers	UCF	10,000 tons
Expected outputs	Economic output	Total value of agricultural outputs	TVAO	100 million yuan
	Social output	Grain output	GO	10,000 tons
	Environmental output	Total carbon sinks	TCS	ton
Unexpected outputs	Non-point source pollution	Total emissions of COD, TN and TP	NPSP	1010 m3
	Carbon emission	Total carbon emissions	TCE	10,000 tons

). The input variables included three parts, i.e., land, labor, and the production materials of cultivated land utilization, such as the research of Xie et al. (2018) [15] and Yang et al. (2021) [17]. Among them, land and labor were characterized by the sown area of cultivated land and the number of employees in the primary industry, respectively. The production materials were represented by the total power of agricultural machinery, and the usage of agricultural films, pesticides, and chemical fertilizers, respectively. The expected outputs included three aspects: economic output, social output, and environmental output. The total value of agricultural output reflects the economic benefits brought by various crops in the cultivated land utilization process, providing an important basic support for the economy and industrial development. So, it was used as the economic output of the CLGUE. The grain output reflects the supply capacity of cultivated land utilization for the staple food of local residents, which is of great significance for ensuring food security and social stability. Therefore, it can be used as the social output. Crops planted on cultivated land can absorb carbon dioxide through photosynthesis, serving as a carbon sink [34,42]. Thus, the total carbon sinks were selected as the environmental output. The formula was as follows:

$$Ca={\sum }_{n=1}^m{Ca}_n={\sum }_{n=1}^m\left\{\left[{\delta }_n\times Q_n\times (1-{\epsilon }_n)\right]/{\sigma }_n\right\}$$

(1)

where $Ca$ represents the total carbon sink of crops grown on cultivated land, ${Ca}_n$ represents the carbon absorption of the nth crop, $m$ is the type of crops planted, including wheat, corn, rice, sorghum, soybean, millet, and potatoes. ${\delta }_n$ is the carbon content of the nth crop, $Q_n$ is the yield of the nth crop, ${\epsilon }_n$ represents the moisture coefficient of the nth crop, and ${\sigma }_n$ is the economic coefficient of the nth crop. The economic coefficient, water coefficient and carbon absorption rate are seen as in

Table 2. Economic coefficient, water coefficient and carbon absorption rate of seven kinds of food crops. Unit: %.

Crops	Economic Coefficient	Water Coefficient	Carbon Absorption Rate
Rice	45.00	12.00	41.40
Wheat	40.00	12.00	48.50
Corn	40.00	13.00	47.10
Millet	35.00	15.00	44.60
Sorghum	35.00	15.00	44.60
Beans	34.00	13.00	45.00
Potato	70.00	70.00	42.30

[43,44].

The agricultural non-point source pollution (NPSP) and total carbon emission (TCE) were selected as the unexpected outputs of CLGUE. The calculation method of the NPSP was as follows. Firstly, the total nitrogen (TN), total phosphorous (TP), and chemical oxygen demand (COD) outputs generated by pesticides, fertilizers, and agricultural waste in farmland use were calculated. Then, the standard emissions of NPSP can be obtained by multiplying the output of each type of pollutant and their respective emission assessment coefficients. According to existing research [45,46], the emission coefficients of COD, TN, and TP were set at 20 mg/L, 1.0 mg/L, and 0.2 mg/L, respectively. The TCE were calculated based on an integrated assessment method for agricultural carbon emission coefficients. The formula was as follows:

$$Ce=\sum {(E}_i\times {\theta }_i)$$

(2)

where $Ce$ is the total carbon emissions, $E_i$ is the total amount of various carbon emission sources, and ${\theta }_i$ is the emission coefficient for each carbon emission source. The main carbon emission sources include fertilizers, plastic mulch film, diesel fuel consumption in agricultural machinery use, tillage, and agricultural irrigation. The carbon emission coefficients are shown in

Table 3. Carbon emission coefficients of main carbon sources in cultivated land utilization.

Carbon Emission Sources	Fertilizers	Plastic Film	Diesel Fuel Consumption	Tillage	Agricultural Irrigation
Emission coefficients	0.8956 kg·kg−1	5.18 kg·kg−1	0.5927 kg·kg−1	312.60 kg·km−2	25.00 kg·hm−2

[43,47].

3.2. Super-SBM with Undesired Output

As for measuring the cultivated land use efficiency, the traditional data envelopment analysis (DEA) model and slack-based measure (SBM) model are currently the most widely used methods. However, the efficiency measured by the DEA model only includes proportionate improvements, which cannot solve the issue of inefficiency caused by the input–output relaxation and may lead to bias in the efficiency evaluation of decision-making units [48]. To solve this problem, Tone (2002) [49] proposed the Super-SBM model, which not only includes non-desirable outputs, but also can directly incorporate relaxation variables into the objective function. Thus, we applied this model to measure the CLGUE values. The Super-SBM model with undesired output was as follows.

$$\begin{array}{c}\mathrm{M}\mathrm{i}\mathrm{n}\mathsf{\rho }=\frac{1-\frac{1}{\mathrm{N}}{\sum }_{\mathrm{n}=1}^{\mathrm{N}}\frac{{\mathrm{s}}_{\mathrm{n}}^{\mathrm{x}}}{{\mathrm{x}}_{\mathrm{n}0}}}{1+\frac{1}{\mathrm{M}+1}(\sum _{\mathrm{m}=1}^{\mathrm{M}}\frac{{\mathrm{s}}_{\mathrm{m}}^{\mathrm{y}}}{{\mathrm{y}}_{\mathrm{m}0}}+\sum _{\mathrm{i}=1}^{\mathrm{I}}\frac{{\mathrm{s}}_{\mathrm{i}}^{\mathrm{b}}}{{\mathrm{b}}_{\mathrm{i}0}})}\\ \mathrm{s}.\mathrm{t}.{\displaystyle \sum _{\mathrm{k}=1}^{\mathrm{K}}}{\mathsf{\lambda }}_{\mathrm{k}}{\mathrm{x}}_{\mathrm{n}\mathrm{k}}+{\mathrm{s}}_{\mathrm{n}}^{\mathrm{x}}={\mathrm{x}}_{\mathrm{n}0},\mathrm{n}=1,2,\dots \mathrm{N}\\ {\displaystyle \sum _{\mathrm{k}=1}^{\mathrm{K}}}{\mathsf{\lambda }}_{\mathrm{k}}{\mathrm{y}}_{\mathrm{m}\mathrm{k}}-{\mathrm{s}}_{\mathrm{m}}^{\mathrm{y}}={\mathrm{y}}_{\mathrm{m}0},\mathrm{m}=1,2,\dots \mathrm{M}\\ {\displaystyle \sum _{\mathrm{k}=1}^{\mathrm{K}}}{\mathsf{\lambda }}_{\mathrm{k}}{\mathrm{b}}_{\mathrm{i}\mathrm{k}}+{\mathrm{s}}_{\mathrm{i}}^{\mathrm{b}}={\mathrm{b}}_{\mathrm{i}0},\mathrm{i}=1,2,\dots \mathrm{I}\\ {\displaystyle \sum _{k=1}^K}{\lambda }_k=1;{\lambda }_k\geqq 0;s_n^x\geqq 0;s_m^y\geqq 0;s_i^b\geqq 0\end{array}$$

(3)

where $\mathsf{\rho }$ represents the CLGUE under constant returns to scale, and $0\le \mathsf{\rho }\le 1$; $\mathrm{K}$ is the amount of unit; $\mathrm{N}$, $\mathrm{M}$, and $\mathrm{I}$ represent the number of inputs, expected outputs, and unexpected outputs, respectively; ${\mathrm{s}}_{\mathrm{n}}^{\mathrm{x}}$, ${\mathrm{s}}_{\mathrm{m}}^{\mathrm{y}}$, and ${\mathrm{s}}_{\mathrm{i}}^{\mathrm{b}}$ represent slack variables for inputs, expected outputs, and unexpected outputs, respectively; ${\mathrm{x}}_{\mathrm{n}0}$, ${\mathrm{y}}_{\mathrm{m}0}$, and ${\mathrm{b}}_{\mathrm{i}0}$ are the output value of inputs, expected outputs, and unexpected outputs for each unit; ${\mathsf{\lambda }}_{\mathrm{k}}$ is the weight vector.

3.3. Panel Regression Model

In this paper, the dynamic panel regression model was applied to analyze the influencing factors of CLGUE in China and its four regions. The model was as follows.

$$ln{Y}_{\mathrm{i}\mathrm{t}}={\beta }_0+\sum _{t=1}^n{\beta }_{k}ln{x}_{it}{+\epsilon }_{it}$$

(4)

where $Y_{\mathrm{i}\mathrm{t}}$ represents the dependent variable, $x_{it}$ is the independent variable, ${\beta }_0$ is the intercept, ${\beta }_k$ represents the regression coefficients, and ${\epsilon }_{it}$ is the error term.

Existing research has shown that natural conditions, socioeconomic development level, cultivated land use conditions, and agricultural policies can influence cultivated land use efficiency [17,19,33,50]. Based on the existing research and cultivated land condition of China, we selected the variables that characterize influencing factors of the CLGUE (

Table 4. Selection of influencing factors for the CLGUE.

Variables	Abbreviations	Calculating Methods	Units
Crop diversity index	Diversity	The rate of sown area and the total area of cultivated land	%
GDP per capita	GPC	The rate of total GDP and total population	%
Urbanization level	Urbanization	The rate of urban population and total population	%
Effective irrigation rate	EIR	The rate of total irrigation area and the total area of cultivated land	%
Proportion of natural disaster area	Disaster	The rate of natural disaster area and the total area of a region	%
Fiscal support for agriculture	FSA	The rate of fiscal support for agriculture and the total fiscal support	%

). They were the crop diversity index, GDP per capita, urbanization level, effective irrigation rate, proportion of natural disaster area, and fiscal support for agriculture, respectively. Therefore, the specific model investigating the influencing factors of CLGUE can be set as follows.

$$ln{Y}_{\mathrm{i}\mathrm{t}}={\beta }_0+{\beta }_{1}lnDiversity+{\beta }_{2}lnGPC+{\beta }_{3}lnUrbanization+{\beta }_{4}lnEIR+{\beta }_{5}lnDisaster+{\beta }_{6}lnFSA+{\epsilon }_{it}$$

(5)

4. Results

4.1. Changes of Input and Output Variables of the CLGUE

Figure 2 shows the changes in input indicators for average value of 31 provinces in China between 2000 and 2020. We can find that all indicators have increased to some extent in the study period, except for the number of employees in primary industry (NEPI). The sown area of cultivated land (SACL) showed a fluctuating upward trend. It fluctuated greatly from 2000 to 2007, continued to increase from 2007 to 2016, and remained relatively stable between 2017 and 2020. The NEPI decreased year after year during the study period, with the value of 11.1768 million in 2000 decreasing to 5.7168 million in 2020. The input of the total power of agricultural machinery (TPAM) steadily increased from 2000 to 2015, with an average increase of 127.0 gigawatts. However, there was a brief decline in 2016, but it gradually increased thereafter. The usage of agricultural films (UAF) showed a continuing increasing tendency between 2000 and 2016, with the average value increasing from 43,078.91 tons in 2000 to 83,955.16 tons in 2016. However, it gradually decreased from 2016 to 2020. The usage of pesticides (UP) and chemical fertilizers (UCF) all showed a trend of first increasing and then decreasing. The UP increased from 41,275.30 tons in 2000 to 58,287.70 tons in 2014, and gradually decreased from 2015 to 2020. The UCF increased significantly from 2000 to 2015, with an increase of 45.22% in its amount over the 16 years. However, it gradually decreased between 2016 and 2020. It can be found that the amounts of UAF, UP, and UCF all showed a decreasing trend in the past five years, indicating that these means of production were gradually regulated in China and their usages were continuously decreasing.

Figure 3 exhibits the changes in output indicators for average value of 31 provinces in China during 2000–2020. The total value of agricultural output (TVAO) and grain output (GO) all showed a steady upward trend in the study period. The TVAO increased from 44.69 billion yuan in 2000 to 231.45 billion yuan in 2020, with an increase rate of 418.80%. The GO increased from 14.909 million tons in 2000 to 21.596 million tons in 2020, with an increase rate of 44.86%. The total carbon sinks (TCS) showed a gradual increase from 2000 to 2016, from an amount of 2,693.50 tons in 2000 to 4,088.26 tons in 2016. The peak was reached in 2016, but there was a noticeable decrease in 2020. The non-point source pollution (NPSP) showed a tendency to continuously increase from 2000 to 2015. However, it decreased steadily between 2016 and 2020. This corresponds to the decreasing trend of UAF, UP, and UCF, indicating that enhancing the cultivated land environment protection can contribute to a reduction in pollution from agricultural activities. The total carbon emissions (TCE) showed a trend of first increasing and then decreasing during the study period. It appeared a continuous upward trend from 2000 to 2015, then began to slowly decrease up to 2020.

4.2. The Spatial-Temporal Patterns of the CLGUE

Figure 4 shows the temporal trend of CLGUE from 2000 to 2020 in China. During the study period, the average CLGUE value exhibited a fluctuating upward development trend, indicating that some provinces with lower efficiency tended to converge towards the high levels. The highest efficiency value was 0.957 in 2020, and the lowest one was 0.853 in 2003, with an average value of 0.905. Overall, the CLGUE in China was at a relatively high level, and the gaps among different provinces were gradually narrowing. Due to regional development imbalances, there were regional disparities in the endowment of factors for agricultural land production. As a result, there were variations in CLGUE among different regions in China.

Figure 5 illustrates the temporal evolution of CLGUE in different regions of China from 2000 to 2020. The overall ranking of CLGUE in the four major regions from high to low was Northeastern, Eastern, Western, and Central China. Northeastern China had the highest CLGUE level, with an average efficiency value of 1.129 during the study period. The efficiency showed a steady increasing tendency, rising from 1.013 in 2000 to 1.272 in 2020, with a growth rate of 25.54%. Eastern China had an average efficiency value of 0.965 and it exhibited a slow upward trend, with an increase from 0.914 in 2000 to 1.005 in 2020. Western China had an average CLGUE value of 0.923 during the study period. It showed a certain degree of decline, indicating there is need for further improvement in the future. Central China had the lowest CLGUE value, with an average efficiency of 0.657, and exhibited a characteristic of wave-like change. Therefore, there was significant room for improvement in efficiency in the future.

Based on the results of the CLGUE measurement in China, the efficiency level was divided into four categories from low to high: low level (<0.50), low–medium level [0.50–0.80), medium–high level [0.80–1.00), high level (≥1.00). In order to visually represent the spatial pattern of CLGUE in China, we imported the results of provincial-level measurements into the ArcGIS 10.2 software and created a spatiotemporal differentiation map of CLGUE (Figure 6). From the overall perspective, there were significant spatial differentiation characteristics in CLGUE across China, with some differences in efficiency between the 31 provinces. Most provinces in China experienced a slight increase in CLGUE overall during 2000–2020. Spatially, the provinces of Heilongjiang, Jilin, Shanghai, Guizhou, Tianjin, Guangdong, and Hainan generally had higher CLGUE values, while Gansu, Shanxi, Yunnan, Guangxi, Hubei, and Hunan had lower ones. To this end, the provinces with the lower efficiency should pay more attention to cultivated land conservation in the process of cultivated land utilization. Efforts should be made to minimize the negative impacts, while improving agricultural output and grain production, e.g., reducing the use of pesticides, chemical fertilizer, and other chemical productions, and controlling agricultural non-point source pollution and carbon emissions.

4.3. The Influencing Factors of the CLGUE

Table 5. Regression analysis of influencing factors of CLGUE in China and its four regions.

Variables	China	Northeastern China	Eastern China	Central China	Western China
Diversity	0.074 *	0.136 **	0.085 *	−0.064 **	0.057 *
GPC	0.016 **	0.009 **	−0.002 *	−0.003 **	0.024 **
Urbanization	0.008 *	0.003 *	−0.011	−0.005 ***	0.007 *
EIR	0.026 ***	−0.004 *	0.006 **	0.031 *	−0.018 **
Disaster	−0.046 *	−0.035	−0.052	−0.041	−0.038 **
FSA	0.387 *	0.514 *	0.224 **	0.417 ***	0.482 **
${\beta }_0$	3.254 *	4.547 **	2.214	3.185 *	3.201
Adjusted R	0.768	0.821	0.865	0.921	0.746
F-statistic	95.124	36.214	49.524	53.165	185.457
Prob.(F)	0.000	0.000	0.000	0.000	0.000

Note: *, **, and *** denote the significance at the 10%, 5%, and 1% levels, respectively.

shows the influencing factors of the spatiotemporal disparities of CLGUE in China and its four regions. Diversity has a positive impact on CLGUE. The greater the Diversity, the higher the level of CLGUE. For every 1% increase in Diversity, the CLGUE in China will improve by 0.074%. From a regional perspective, except for Central China, Diversity elicits positive effects on the CLGUE in Northeastern, Eastern, and Western China. Generally, a higher crop diversity index corresponds to a higher land use intensity, resulting in higher expected outputs such as grain production and agricultural value per unit area. This is beneficial for improving the CLGUE to some extent. However, as the Diversity increases, there is also an increasing investment in agricultural machinery, as well as in the usage of fertilizers, pesticides, and plastic films. This may result in excessive carbon emissions and agricultural non-point source pollution, which have some negative impacts on CLGUE, for example the Central China.

GPC and Urbanization have positive impacts on CLGUE. For every 1% increase in GPC and Urbanization, the CLGUE in China will improve by 0.016% and 0.008%, respectively. From a regional perspective, the two variables both have positive impacts on CLGUE in Northeastern and Western China, while they have certain negative impacts on CLGUE in Eastern and Central China. With rapid industrialization and urbanization in China, a large number of rural populations are continuously leaving the countryside and flocking to urban areas. Consequently, this leads to the abandonment of substantial amounts of cultivated land, resulting in a negative impact on CLGUE to some extent, for example, in Eastern and Central China. Meanwhile, Urbanization can also have an impact on cultivated land utilization by promoting a shift from extensive farming to a more intensive and efficient utilization mode. This, in turn, may have a positive effect on CLGUE by promoting more sustainable and efficient cultivated land-use practices.

EIR has a positive impact on CLGUE. For every 1% increase in EIR, the CLGUE will rise by 0.026% in China. It directly affects the growth condition and yield of crops, improving soil health and protecting ecological environment [51,52]. The higher the EIR, the higher the CLGUE level. From the regional perspective, EIR has a positive impact on the CLGUE in Eastern and Central China, while it tends to have negative impacts in Northeastern and Western China. This may be because there are abundant water resources and better agricultural irrigation infrastructure in Eastern and Central China, which causes EIR to have a positive impact on CLGUE. On the other hand, Northeastern and Western China have relatively poor agricultural infrastructure and face water scarcity, resulting in lower EIR, which limits the CLGUE level to some extent.

Disaster has a negative impact on the CLGUE in China and its four regions. For every 1% increase in Disaster, CLGUE will decrease by 0.046%. Natural disasters, such as floods, droughts, landslides, and mudslides, pose a threat to regional land use and agriculture production, severely limiting the increase in grain production and agricultural output. It also negatively impacts the production and living environment of rural areas, which is detrimental to the sustainable use of regional cultivated land.

FSA has a positive impact on the CLGUE in China and its four regions. The higher the FSA, the higher the level of CLGUE. For every 1% increase in it, the CLGUE will improve by 0.387%. Some FSA policies such as direct subsidies for grain production, agricultural tax reforms, and subsidies for the purchase of agricultural machinery are beneficial for controlling the abandonment of cultivated land and alleviating the financial burden on farmers in terms of agricultural production. They can stimulate farmers’ enthusiasm for production, and play a positive role in increasing investment and improving expected outputs. In addition, FSA can help to promote agricultural science and technology development and rural environmental governance, which could have a positive effect on the green and low-carbon production of cultivated land.

5. Discussion

5.1. Comparison with Previous Studies

In this study, we conducted an empirical investigation of CLGUE and influencing factors in China from 2000 to 2020. Overall, we found the average CLGUE value exhibited a fluctuating upward trend, from the lowest value of 0.853 in 2003 to the highest efficiency value of 0.957 in 2020. And Northeastern China had the highest efficiency value, while Central China had the lowest one. Similarly, Xie et al. (2018) [15] also measured the green efficiency of arable land use in China during the period 1995–2013 by using a non-radial directional distance function approach. They found that the green efficiency of arable land use had been declining from 1995 to 2000, then increasing from 2001 to 2013. It is accordant with our study that the average CLGUE value in China showed an overall upward trend from 2000 to 2020. In addition, Xie et al. (2018) [15] also demonstrated that the most efficient region was in Northeast China, while the lowest-efficiency region was in central China. Kuang et al. (2020) [5] explored the cultivated land use efficiency in China by considering carbon emissions. They found that the average value of cultivated land use efficiency in China declined from 2000 to 2003, then increased from 2004 to 2017. Hou et al. (2019) [50] took the Yangtze River Economic Belt of China as a case study, and discussed how urbanization influenced the eco-efficiency of cultivated land utilization. They demonstrated that driving and feedback effects had positive impacts on efficiency, while agglomeration and barrier effects showed negative impacts. Summarily, different indicator evaluation systems and models of the CLGUE will lead to different measurement results. For instance, some studies regard agricultural non-point source pollution as an undesirable output, emphasizing the negative impacts of environmental pollution on the CLGUE [3,15,53]. Some studies regard agricultural carbon emissions as an undesirable output, which focus on the low-carbon utilization and clean production of cultivated land [5,17,19,21]. Our study comprehensively considers the environmental effects as the outputs, i.e., carbon emissions, carbon sinks and agricultural non-point source pollution, which is conducive to more accurately evaluating the CLGUE level.

5.2. Policy Implications

Some policy implications were proposed based on this study. Firstly, the government should strengthen the formulation and implementation of relevant policies to promote the green utilization of cultivated land. We should increase the publicity of cultivated land green utilization, improve incentive mechanisms for ecological utilization modes, and encourage the development of circular agriculture and biofertilizers. Only in this way, can a low-carbon-emission and green assessment system of cultivated land be established and provide policy supports for the CLGUE. Secondly, we should explore regional differentiation paths for enhancing the CLGUE in China. There are significant differences in resource endowments and economic development levels among different regions in China. For the economically developed and resource-abundant regions, such as Central and Eastern China, it is crucial to vigorously promote land circulation and large-scale operations, develop modern agriculture, implement clean production, and promote cultivated land green utilization. For the relatively economically underdeveloped regions, such as Northeastern and Western China, it is important to actively formulate pro-agriculture policies related to cultivated land utilization. This includes continuously absorbing more agricultural labor force while promoting the optimization of cultivated land functions. Furthermore, efforts should be made to enhance the capacity of cultivated land to reduce carbon emissions and increase carbon sequestration. Thirdly, the government needs to increase financial investment in agriculture. It is of significance to continuously strengthen financial investment in agriculture, with a particular focus on controlling non-point source pollution from agriculture. This includes strict control over the use of chemical fertilizers, pesticides, and plastic films. Additionally, through financial subsidies, we can encourage farmers to adopt green and low-carbon practices in cultivated land utilization. This will not only can lead to a demonstration effect but also drive the comprehensive improvement in the CLGUE level.

5.3. Limitations

This paper also has some limitations. First, we mainly focused on the estimation of CLGUE at the provincial level in China. The research scale was large and there was no detailed analysis on CLGUE at the city level. Second, this study explored China’s CLGUE change, but included a lack of global changes on cultivated land and its use efficiency. In future research, we will adopt prefecture-level cities as the research units to estimate the CLGUE and investigate the underlying impact mechanism on agricultural economic growth. Furthermore, our study will investigate global changes on cultivated land utilization and analyze the differences in CLGUE in different countries in the world.

6. Conclusions

This paper used the Super-SBM model to measure the CLGUE values of 31 provinces in mainland China during 2000–2020 and analyze its influencing factors based on a panel regression model. We found that the average CLGUE in China exhibited a fluctuating upward development trend, with the highest efficiency value of 0.957 in 2020 and the lowest one of 0.853 in 2003. Northeastern China had the highest efficiency value, with an average of 1.129, while Central China had the lowest efficiency value, with an average efficiency of 0.657. Spatially, there are significant differentiations in CLGUE across China. Heilongjiang, Jilin, Shanghai, Guizhou, Tianjin, Guangdong, and Hainan generally had higher CLGUE values, while Gansu, Shanxi, Yunnan, Guangxi, Hubei, and Hunan had lower ones. The regression model indicated that Diversity, GPC, Urbanization, EIR, and FSA positively influenced the CLGUE, while Disaster had negative impacts. In addition to improving the CLGUE and achieving the sustainable utilization of cultivated land resources, some attention should be paid to establishing a low-carbon emission and green assessment system of cultivated land and continuously strengthening financial investment in agriculture, with a particular focus on controlling agricultural carbon emissions and non-point source pollutions.

Our study customized an empirical investigation into the CLGUE in China. Through our research, we offered valuable policy recommendations that can effectively promote sustainable agricultural development. To further improve the depth of CLGUE, future research should adopt prefecture-level cities as the research units to estimate the CLGUE level in China and investigate global changes on cultivated land utilization and analyze the differences of CLGUE of different countries in the world.