Will the Grain Imports Competition Effect Reverse Land Green Efficiency of Grain Production? Analysis Based on Virtual Land Trade Perspective

Abstract

To explore new strategies to improve the efficiency of land for grain production in China, this study empirically investigates the relationship between the grain imports competition effect and the efficiency of land for grain production from the perspective of virtual land trade based on the new-new trade theory and induced technological innovation theory. We obtained the following conclusions: (1) Overall, the efficiency of land for grain production showed a slow upward trend from 2003 to 2020, with a growth rate of 27.53%. Among them, the efficiency of land for grain production in the Huang-Huai-Hai and northeast regions has increased by 66.41% and 36.30%, respectively. (2) The grain imports competition effect reduces the profitability of domestic grain production through shocks and forces the improvement of China’s grain production land efficiency. It is caused by the trade-for-competition effect. Importing a large number of grain products with lower land use costs reduces the profits of grain producers in the domestic market through shocks, gradually eliminating producers with less efficient land use, thus forcing the improvement of land use efficiency in domestic grain production. (3) The grain imports competition effect is stronger because the developed livestock industry in the northern region requires the import of large quantities of soybeans and corn, and other virtual land content is much lower than the domestic feed grain. Also, the marginal effect of the grain imports competition effect is stronger due to the presence of more producers with less efficient land for grain production in non-major grain-producing regions, which are more likely to be eliminated by market shocks from imports. This study verifies the applicability of the trade promotion competition effect and induces the technological innovation effect in the field of grain trade, which extends the research boundary of virtual land trade in grain.

1. Introduction

1.1. Background

Grain security is a constant theme of national stability and people’s well-being and securing efficient and sustainable land use is an important prerequisite for achieving grain security [1]. China’s per capita land holdings are only 27.7% of the world average, and land shortage is a serious constraint to sustainable grain production in China [2]. This makes improving the efficiency of land for grain production a key tool to guarantee sustainable production for grain security. Moreover, due to the increase in rigid demand for grain caused by economic and population growth and the rising demand for feed grain caused by the upgrading of people’s diet structure, coupled with the constraints of domestic land endowment and utilization efficiency, it is difficult to meet the grain consumption demand by domestic grain production alone. It has become increasingly important to make full use of international grain resources and international grain markets to ensure China’s grain security [3]. According to data released by the General Administration of Customs, China’s grain imports in 2022 will be 146.87 million tons, accounting for 21% of total grain production in 2022. Indeed, grain trade has shifted from the initial transfer of surplus to large-scale imports [4].

With the gradual development of trade theory, the new-new trade theory is concerned with the fact that trade, through competitive effects, shocks lower the price of grain in the importing country, thus reducing the profitability of domestic grain production [5]. Economic agents with a low level of production efficiency are forced to withdraw from the market due to the crowding out of their market share. As for the subjects with high production efficiency levels and maintaining production in competition, the continuous increase in import competition will stimulate them to further improve the efficiency level and reduce the overall cost, thus promoting the benign development of the industry at a higher level [6]. Meanwhile, the theory of induced technological innovation points out that agricultural technology upgrading to the direction favoring the saving of relatively scarce factors and the utilization of relatively abundant factors, grain production technology will develop in the direction of land saving, thus improving the allocation efficiency of grain production land factors [7].

This study explores the mechanism of grain import competition’s effects on grain production land green efficiency from the virtual land perspective based on the new-new trade theory and induced technological innovation theory. First, the competitive effect of grain imports is measured using the cropland footprint method. Secondly, land green efficiency of grain production was measured using the DEA-GML index. Third, a panel Tobit model is used to study the impact of grain import competition effects on land green efficiency of grain production. Fourth, the instrumental variables method, quantile empirical, and subsample empirical are used to test the robustness of the benchmark empirical results. To test the applicability of the two theories in the field of grain trade, it provides a reference for policy formulation to improve the land green efficiency of grain production and to meet the need to safeguard China’s domestic grain security in the context of limited land resource endowment and increasing grain demand.

1.2. Literature Review

Literature review of welfare effects of virtual land trade. Grain trade plays an important role in alleviating the pressure on land for grain production as a link between areas with abundant and scarce land resources [8]. Previous studies have typically used comparative advantage to analyze the pattern of virtual land flows, where virtual land resources usually flow from countries or regions with higher land resource endowments to countries or regions with lower land resource endowments [9]. In reality, however, the flow of virtual land is influenced by social, policy, economic, and resource environment factors, resulting in different factors affecting virtual land resources in different regions [10]. Most studies focus on the quantification of virtual land, the influencing factors of virtual land trade, and the welfare effects of virtual land trade. On the one hand, in the virtual land trade welfare effect [11], the virtual land strategy is of great value in improving the efficiency of land resource use, securing a country’s land security, and alleviating the contradiction between land resource supply and demand, thus achieving grain security [12]. Firstly, the country’s scarce land resources will be allocated to secondary and tertiary industries to improve the economic efficiency of land resources and provide space for the expansion of urbanization and industrialization development process. Secondly, countries or regions with relatively abundant land resources can rely on export trade to obtain corresponding benefits, realize the original accumulation of capital, transform their economic and industrial structure, and continuously improve their economic strength [13].

On the other hand, the virtual land trade implicit in the grain trade process breaks the inherent and immobile nature of land resources [14], allowing them to flow in the form of trade to countries where land is relatively scarce [15]. Researchers believe that land resources in individual countries are gradually declining and that the scarcity of land resources in importing countries can be filled to some extent through virtual land trade [16]. Furthermore, some researchers have used coupling degree models and imbalance indices to measure the spatial–temporal match between land, environment, and economy and to explore the effect of virtual land trade on the match. However, existing studies have neglected the competitive effects generated by grain imports, which may have a catalytic effect on resource use efficiency in importing countries [17].

1.3. Literature Review of Land Use Efficiency

Firstly, the land use efficiency evaluation index system was studied. The measurement of land use efficiency is essentially to reveal the relationship between input and output ratios in land and crop systems. Under the given output conditions, the less the input, the higher the efficiency; and under the given input conditions, the greater the output, the higher the efficiency [18]. Currently, scholars vary greatly in the selection of indicators for measuring land use efficiency, but there is a trend of gradually changing from single-indicator input–output to multi-indicator input–output [19]. Gradually, the economic, social, and ecological benefits are incorporated into a comprehensive index system for evaluation. With the rapid development of the economy and society, environmental problems have come one after another, and the concept of sustainable development is deeply rooted in people’s minds [20]. The increase in agricultural output has been accompanied by a steady decline in land quality due to the overuse of chemical fertilizers, pesticides, and agricultural films. For this reason, scholars not only consider economic and social benefits when selecting land use efficiency indicators, but also include environmental factors as an important indicator in the measurement system, and the concept of “green efficiency” also begins to appear in the evaluation of land use effects [21]. Meanwhile, scholars have gradually measured the land use efficiency of agricultural surface source pollution as non-desired output.

Secondly, the regional, spatial, and temporal characteristics and influencing factors of land use efficiency were studied. The analysis of spatial and temporal characteristics of land use efficiency has also been the focus of scholars and has revealed the spatial pattern and evolution pattern of land use efficiency among Chinese regions. Land use efficiency in China is still at a low level but shows a trend of gradual optimization [21], and is shrinking among the provinces in the Yangtze River basin [22]. Meanwhile, China’s land green efficiency, in general, shows a spatial distribution pattern of higher in the east and lower in the middle and west, and the efficiency growth rate is faster in the east [23]. Moreover, scholars have gradually started to pay attention to the land correlation of land use efficiency, which is found to have clustering characteristics and an increasingly obvious trend [24]. Overall, scholars have further explored the influencing factors of land use efficiency at macro and micro levels, respectively. From a macro perspective, replanting index, topographic factors, level of economic development, land size, education level, and total mechanical power have a contributing role in land use efficiency. From a micro perspective, crop types, agricultural subsidies, village economic development, and special agricultural cultivation also have a large impact on land use efficiency [25].

In summary, previous studies have explored the areas of grain import trade welfare, virtual land trade welfare, and land resource use efficiency, providing a solid theoretical foundation for this study, but there are still areas worthy of further investigation. First, studies on trade welfare in importing countries have mostly focused on the welfare effects of import trade on income and economic growth, and fewer studies have advanced them to production and resource utilization effects. In particular, few studies delve into the welfare effects of grain imports on the grain production side and its production environment. Second, most of the studies on virtual land for grain trade are at the stage of measurement and characterization and mostly analyzed at the international level, and there is less literature exploring regional differences in the impact of virtual land flows for grain trade on land use efficiency under different resource endowments. Third, the accuracy of LEGP measurement needs to be improved. Most studies only take grain production or output value as the output variable, ignoring the non-desired output of agricultural surface pollution, resulting in the measured LEGP often being overestimated. For this purpose, our study explores the regional heterogeneity of the effect of grain import competition effects on the efficiency of land for grain production from a virtual land perspective based on the new-new trade theory and induced technological innovation theory. This study innovatively verifies the applicability of the trade promotion competition effect and induced technological innovation effect in the field of grain trade, extends the research boundary of grain virtual land trade, and provides new research ideas for the study of the mechanism of reversing the domestic grain production end from the grain import end.

2. Materials and Methods

2.1. Measurement of the Grain Imports Competition Effect

2.1.1. Measurement Method of China’s Net Grain Import Virtual Land

Measurement of virtual land content per unit of grain crop. Previous studies usually measure the virtual land content of grain from both the production and consumption sides. Since this study mainly explores the impact of grain import trade on the production side, the land resources required in the grain production process in each province are measured from the production side as a measure of the virtual land content per unit of grain crop [26]. The specific equation is as follows:

$$Alan{d}_{z,it}=F{S}_{z,it}/F{P}_{z,it}$$

(1)

In Equation (1), Aland_z,it is the virtual land content per unit of grain crop and represents the virtual land content of 1 unit of z crop in area i in year t. FS_z,it denotes the area sown to crop z in region i in year t. FP_z,it denotes the yield of crop z in region i in year t. We further measured the grain import virtual land:

$$Vlan{d}_{z,it}=Alan{d}_{z,it}/I{M}_{z,it}$$

(2)

In Equation (2), Vland_z,it denotes the amount of imported virtual land for crop z in region i in year t; IM_z,it denotes the amount of imported crop z in region i in year t.

2.1.2. Measurement of the Land-Saving Effect of Net Grain Imports in China

Since our study mainly explores the impact of grain import trade on the production side, we measure the virtual land content per unit of grain crop produced in each country by measuring the land resources required for the grain production process in each province from the production side.

$$SL{D}_{ltk}=\frac{1}{Y_{ltk}}$$

(3)

In Equation (3), SLD_ltk refers to the land footprint per unit mass of k grain crops in year t of country l (m³/kg); Y_ltk refers to the yield per unit area of k grain crops in year t of country l (kg).

$$VL{I}_{tk}=SL{D}_{ltk}\times \frac{I{M}_{ltk}}{{\displaystyle \sum _{l=1}^{n}I{M}_{ltk}}}$$

(4)

In Equation (4), VLI_tk refers to the average land footprint of k grain crop imports in year t (m³/kg); IM_ltk refers to the amount of k grain crops imported from country l in year t (kg); n denotes the number of major source countries of k grain crops imported by China.

$$FIL{S}_{itk}=I{M}_{itk}\times (Vlan{d}_{itk}-VL{I}_{tk})$$

(5)

In Equation (5), FILS_itk refers to the amount of imported k grain crops saved in province i in year t (m³); IM_itk refers to the amount of imported k grain crops to province i in year t (kg). The amount of land saved by grain crop imports in province i, while capturing the number of land resources saved globally by grain imports, also happens to reflect the gap between the cost of land green efficiency of grain production and the cost of land green efficiency of grain production in importing countries. The grain imports competition effect from a resource use perspective characterizes the impact of grain imports on China’s domestic grain market.

Regarding the grain import competition effect, the differences in the amount of resources consumed in the production of grain products between different countries or regions make the virtual resource content of grain products vary between countries or regions [27]. According to the new-new trade theory, a country or region can substantially increase trade welfare by importing grain products with a virtual resource content much lower than the domestic level [28], and it can exert a trade promotion competition effect to some extent [29]. This is due to the lower cost and price of grain products produced in countries with lower virtual resource content of grain. Importing grain from these countries will crowd out the market share of the main domestic grain producers in China while depressing domestic grain prices and reducing the profitability of domestic grain production in China, which will have an impact on the domestic grain market in China [30]. The extent to which increased imports of goods or services affect economic agents through the trade promotion competition effect depends on the level of production efficiency of different economic agents [31]. Specifically, economic agents with a low level of production efficiency are forced to exit the market due to the squeezed market share. In contrast, for subjects with a high level of production efficiency and who can maintain production in the competition, the continuous increase in import competition will stimulate them to increase their efficiency level and reduce production costs, thus promoting the benign development of the industry at a higher level [32]. Accordingly, this study defines the grain import competition effect as a cost competition phenomenon that occurs when a country or region imports grain from a country or region with higher grain production efficiency (lower virtual land content) and shocks the grain market of the importing country by depressing the price of grain in the importing country and reducing the profitability of grain production in the importing country.

2.2. Measurement Method of Land Green Efficiency of Grain Production

Since the GML index can effectively balance the green development demands of maximizing desired output and minimizing undesired output and input factors, this study refers to Oh [33] to construct a GML index model to measure the changes in LEGP in 30 provincial administrative regions of China from 2003 to 2020. The specific calculations refer to Li et al. [34,35].

2.3. Variable Selection

In this study, the corresponding indicators were selected to measure the independent variable, dependent variable, instrumental variable, and control variable based on the previous studies. The specific variable indicators and measurement methods are shown in

Table 1. Variables and calculation methods.

Variable Category	Variable	Calculation Method	Unit
Independent variable	Land green efficiency of grain production	Measured by the Global Malmquist–Luenberger Index	-
Dependent variable	Competitive effects of grain imports	Statistics	×105 hectares
Instrumental variable	Farming scale	Milk production + beef production + poultry production + lamb production + pork production	×108 tons
Control variable	Technical environment	Technology market turnover/internal R&D expenditure	-
	Financial support for agriculture	Expenditure on agriculture, forestry, and water affairs in financial expenditure	×1010
	Agricultural machinery	Total power of agricultural machinery/number of employees in primary industry	Kilowatt/per
	Specialization of crop cultivation	Herfindahl index for wheat, rice, corn, beans, and potatoes	-
	Environmental regulation	Investment in environmental pollution control as a share of GDP	%
	Urbanization level	Share of urban population	%
	Percentage of village land	(Current land area of villages in each province/current land area of villages nationwide) × 100%	%
	Industrial and economic structure	Effective irrigated area/crop sown area	-

.

Independent variable: land green efficiency of grain production. In this study, concerning Zhi et al. [36], input–output indicators were chosen and the GML index method was employed. These were then combined with the study of Buchthal et al. [37] to include non-desired outputs, and the ecological value of grain cultivation was added to the desired output indicators. Our research aims to maximize the economic and ecological values of grain production while minimizing agricultural surface pollution, carbon emissions, and other input factors, to more scientifically, accurately, and reasonably reflect LEGP.

Dependent variable: competitive effects of grain imports. The virtual land content of the same agricultural product produced in various regional years causes differences. The majority of grain commodities imported from the global market are soybeans and maize, both of which contain less virtual land than domestically produced virtual land. In this study, the “competitive effects of grain imports” are calculated by multiplying the difference between the virtual land content of China’s domestic grain production and the virtual land content of imported grain by the volume of grain imports.

Instrumental variable: farming scale. This study uses “the sum of China’s milk, beef, poultry, lamb, and pork production” as a proxy for the variable “farm scale”. This is because the production of feed requires large quantities of soybeans and corn, whereas the average virtual land content of imported soybeans and corn is much lower than domestic. According to the trade promotion competition effect on the domestic grain market, soybeans with a higher virtual land content can have a greater impact. The “competitive effects of grain imports” are primarily characterized by the “land-saving effect of grain imports such as soybeans and corn.” Therefore, “farming scale” is theoretically correlated with the endogenous variable “grain import competition effect” and is not directly correlated with the random error term and LEGP in the current period. Therefore, “farm scale” is chosen as the instrumental variable for this investigation.

Control variable: Based on previous research on the factors influencing LEGP [38], eight variables were chosen as control variables for this study model: technical environment, financial support to agriculture, agricultural machinery, specialization of crop cultivation, environmental regulation, urbanization level, percentage of village land and industrial and economic structure [39].

2.4. Empirical Model Design of Competitive Effects of Grain Imports Influencing Land Green Efficiency of Grain Production

Since the LEGP data type is censored data, the Tobit empirical model was used to test the following equation:

$$Legp=\left\{\begin{array}{ll}{Legp}_{it}^{*}=\sigma +{\alpha }_1{Filc}_{it}+{\alpha }_2{X}_{z,it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}& {Legp}_{it}^{*}>0\\ 0& {Legp}_{it}^{*}\le 0\end{array}\right.$$

(6)

$$Legp=\left\{\begin{array}{ll}{Legp}_{it}^{*}=\sigma +{\alpha }_1{Filc}_{i(t-1)}+{\alpha }_2{X}_{z,it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}& {Legp}_{it}^{*}>0\\ 0& {Legp}_{it}^{*}\le 0\end{array}\right.$$

(7)

$$Legp=\left\{\begin{array}{ll}{Legp}_{it}^{*}=\sigma +{\alpha }_1{Filc}_{it}+{\alpha }_2{Filc}_{it}^2+{\alpha }_3{X}_{z,it}+{\mu }_i+{\phi }_t+{\epsilon }_{it}& {Legp}_{it}^{*}>0\\ 0& {Legp}_{it}^{*}\le 0\end{array}\right.$$

(8)

In Equations (6)–(8), LEGP^*_it is the explanatory variable, indicating the LEGP in region i in year t. Filci_(t−1) is the first-order lag term of competitive effects of grain imports. X_z,_it are control variables representing other factors affecting LEGP in region i in year t, z = 1, 2, ..., 8 denote the eight control variables of technological environment, level of financial support to agriculture, level of agricultural machinery, degree of crop cultivation specialization, environmental regulation, level of urbanization, percentage of village land, and industrial economic structure, respectively. Equation (6) is the baseline model for this study, which is used to test the linear relationship between the competitive effects of grain imports on LEGP. Equation (7) adds the competitive effects of grain imports lag term for testing the lag of the effect of competitive effects of grain imports on LEGP. Equation (8) is added to the quadratic term of competitive effects of grain imports to test the nonlinear effect of the effect of competitive effects of grain imports on LEGP.

To eliminate the influence of confounding effects such as extreme values and error terms on the estimation results and to more objectively and comprehensively describe the stage differences in the effects of competitive effects of grain imports on LEGP at different quartiles, the following panel quantile empirical model was developed:

$${Legp}_{it\tau }^{*}=\sigma +{\beta }_{1\tau }{Filc}_{it\tau }+{\beta }_{2\tau }{X}_{z,it\tau }+{\mu }_{i\tau }+{\phi }_{t\tau }+{\epsilon }_{it\tau } {Legp}_{it\tau }^{*}>0$$

(9)

The τ in Equation (9) denotes the quantile. In this study, quantile empiricals were performed at 10%, 20%, ..., and 90% quartile.

3. Results

3.1. Measurement Results of Competitive Effects of Grain Imports

3.1.1. China’s Grain Production Virtual Land Content Time Variation Characteristics

Figure 1 demonstrates that the national average virtual land content of the four main grain productions exhibits the following characteristics of variation. Overall, the national average virtual land content of the four most important grain productions exhibits a consistent downward trend, as a result of the continuous development in the technical level of grain production and the progressive decline in grain production per unit area. Regarding agricultural products specifically, the average virtual land content of soybeans, wheat, rice, and corn is declining steadily. Soybean production fell from 0.641 ha/ton in 2003 to 0.485 ha/ton in 2020, a decrease of up to −24.31%. The wheat yield per hectare decreased from 0.374 ha/ton in 2003 to 0.277 ha/ton in 2020, a decrease of −25.76%. Among them, China’s wheat production fell sharply in 2018 due to extreme weather and other natural disasters, resulting in a sudden drop in wheat production per unit area and a sudden increase in the average value of arable land consumed to produce each unit of wheat. The rice yield per hectare decreased from 0.212 ha/ton in 2003 to 0.168 ha/ton in 2020, a decrease of up to −20.70%. From 0.226 ha/ton in 2003 to 0.177 ha/ton in 2020, maize production decreased by up to −20.25%. Overall, due to the continuous improvement of grain production technology, the yield of grain per unit area is increasing, and the amount of arable land consumed to produce each ton of grain is gradually decreasing. However, China’s relatively small soybean production and low degree of large-scale, specialized production have resulted in the consumption of far more arable land resources per ton of soybean production than for other grain crops.

3.1.2. China’s Virtual Land Content of Grain Production

Table 2. Virtual land content of four major grain products in different regions of China.

Region	Soybean Virtual Land	Maize Virtual Land	Rice Virtual Land	Wheat Virtual Land
Huang-Huai-Hai region	0.594	0.201	0.185	0.184
Northeast region	0.545	0.329	0.164	0.165
South China	0.475	0.498	0.243	0.190
Southwest China	0.585	0.383	0.216	0.197
Middle and lower reaches of the Yangtze River	0.436	0.344	0.208	0.159
Northwest	0.686	0.272	0.161	0.223
National	0.554	0.338	0.197	0.187

Note: All units of grain products involved in the table are hectares/ton.

displays the fictitious land use for grain production in China from 2003 to 2020. With a national average of 0.554 ha/ton, soybean production in China has the highest virtual land content overall. The virtual land content of rice production is the lowest, with a national average of 0.187 ha/ton, or around one-third that of soybeans. Conversely, the national averages for virtual land content for wheat and corn are 0.338 ha/ton and 0.197 ha/ton, respectively. The northwest and Huang-Huai-Hai regions had the highest virtual land content of soybean production, with regional differences in virtual land content of soybean production being significant (0.436–0.686 ha/ton). South China and the middle and lower reaches of the Yangtze River regions, in contrast, are at lower levels, as are the southwest and northeast. South China and southwest areas both had high levels of the virtual land content of wheat production (0.201–0.498 ha/ton), which showed significant regional disparities. The middle and lower reaches of the Yangtze River were all at medium levels in the northeast zone, whilst they were at low levels in the northwest and the Huang-Huai-Hai region. South China and southwest regions had the highest levels of virtual land content per ton of maize produced (0.161–0.243 ha/ton). The middle and lower reaches of the Yangtze River, the Huang-Huai-Hai region, and the northeast and northwest regions were at lower levels. Both the northwest and the southwest had high levels of virtual land content per ton of rice produced (0.159–0.223 ha/ton). The northeast, middle, and lower reaches of the Yangtze River regions were at low levels, whereas South China and Huang-Huai-Hai regions were at medium levels. Regional differences are primarily influenced by the level of grain production technology and the production per unit area.

3.1.3. The Changing Characteristics of China’s Competitive Effects on Grain Imports

China’s grain import land-saving effect increased from 21,699.037 km² in 2003 to 105,400.713 km² in 2020, an increase of 385.74% (Figure 2). The land-saving effect of importing China’s four most important cereals is increasing at a rate of approximately 9.18% per year. In addition, the scale variation characteristics of the import land-saving effect vary considerably more.

First, the magnitude of the wheat import saving effect encountered a rapid increase followed by a decline. It increased from −156.397 km² in 2003 to 1130.807 km² in 2013 before falling to −2900.517 km² by 2020. Second, the magnitude of rice’s import-saving effect demonstrates a general trend of rapid decline followed by sluggish recovery, increasing from −279,558 km² in 2003 to −2798.367 km² in 2017 and then decreasing to −1300.772 km² in 2020. Third, compared to wheat and rice, the maize import land-saving effect is growing at a faster rate, from 257.308 square kilometers in 2003 to 6881.148 square kilometers in 2020, a 26.743% increase. Fourth, soybean has the largest magnitude of imported land-saving effect among grain products, with a rapid increase from 21,877.684 km² in 2003 to 108,952.683 km² in 2015, followed by a decline to 81,275.331 km² in 2018 and a return to 102,720.853 km² in 2020.

3.2. Results Analysis of Land Green Efficiency of Grain Production

Figure 3 depicts the evolution of China’s LEGP from 2003 to 2020. The majority of sample values for LEGP are distributed between 0.5 and 1.5, and the distribution curve of the kernel density function primarily exhibits a peak with high intensity and short span, indicating that the level of LEGP in China does not vary significantly. LEGP is decreasing in intensity and increasing in differentiation over time, and the maxima continue to move to the right, indicating that the overall LEGP in China shows an upward trend from 2003 to 2020 and that the increase in some regions is greater than the increase in others. This indicates that China’s overall LEGP will increase between 2003 and 2020, with some regions experiencing a rate of increase greater than 1.5. This is partially attributable to the gradual improvement of grain production technology, which has increased the total factor productivity of grain, and partly attributable to the expansion of high-standard farmland construction, which has increased LEGP by increasing grain yield per unit area.

The specific situation of LEGP in each region from 2003 to 2020 is depicted in Figure 4, which reveals glaring differences between regions as a result of the influence of economic and social development and the availability of water resources. The national average LEGP from 2003 to 2020 exhibits a gradual upward trend; the average growth rate for each region is 27.53%. The period 2003–2011 exhibits a fluctuating and slightly decreasing trend of 3.31%, while the period 2013–2020 exhibits a fluctuating and increasing trend of 32.11%. In particular, the LEGP in the Huang-Huai-Hai region and the northeast region is at a high level and shows a rapidly increasing trend, with the LEGP in the Huang-Huai-Hai region increasing by 66.41% from 2003 to 2020 and the LEGP in the northeast region increasing by 36.30% over the same period. From 2003 to 2020, the agricultural production land green efficiency in Northeast China increased by 36.30%. In South China and Southwest China, the LEGP is at a medium level and demonstrates a decreasing trend, followed by a rapid and slow trend. Specifically, the agricultural production land green efficiency in South China decreased by 3.6% between 2003 and 2010 and increased by 18.42% between 2011 and 2020. From 2003 to 2020, the grain production land efficacy in Southwest China increased by 23.68%.

The LEGP in the middle and lower reaches of the Yangtze River and the northwest region is low and demonstrates a rapid decline followed by a rapid rise. From 2003 to 2013, LEGP in the middle and lower reaches of the Yangtze River region decreased by 12.68% but increased by 36.48% from 2014 to 2020. In the northwest, agricultural production land green efficiency decreased by 42.58 % between 2003 and 2011 and then increased by 40.77% between 2012 and 2020. The regional distribution demonstrates that regions with lower land resource endowments are capable of driving LEGP improvement, whereas regions with higher land resource endowments have fewer drivers of LEGP improvement.

3.3. Baseline Empirical of Competitive Effects of Grain Imports Affecting Land Green Efficiency of Grain Production

Based on Equation 6,

Table 3. The effect of land competition on land use efficiency in grain production.

	(1)	(2)	(3)	(4)	(5)	(6)
Competitive effects of grain imports		0.007 ***	0.171 ***		0.008 ***	0.026 ***
		(2.956)	(2.895)		(2.730)	(5.895)
Competitive effects of grain imports first-order lag				0.012 ***
				(6.739)
Competitive effects of grain imports square						−0.001 ***
						(−5.042)
Technical environment	−0.001	−0.001	−0.006	0.001	0.029	−0.003
	(−0.115)	(−0.087)	(−0.147)	(0.098)	(1.469)	(−0.282)
Financial support for agriculture	0.010 **	0.009 *	−0.022	0.005	0.002	0.005
	(2.248)	(1.959)	(−1.115)	(1.119)	(0.644)	(1.053)
Agricultural machinery	0.005 *	0.005 *	0.004	0.004 *	0.010 ***	0.006 **
	(1.954)	(1.944)	(0.416)	(1.687)	(6.344)	(2.082)
Specialization of crop cultivation	0.071	0.073	0.200	0.067	−0.296 ***	0.037
	(0.939)	(0.968)	(0.685)	(0.907)	(−5.292)	(0.499)
Environmental regulation	−0.008	−0.003	0.122 **	0.006	−0.001	0.001
	(−0.686)	(−0.276)	(2.051)	(0.500)	(−0.069)	(0.079)
Urbanization level	−0.419 *	−0.369*	−0.028	−0.373 *	0.607 ***	−0.234
	(−1.831)	(−1.673)	(−0.029)	(−1.666)	(6.592)	(−1.186)
Percentage of village land	0.003	0.005	0.125	0.009	0.037 ***	0.011
	(0.224)	(0.400)	(1.490)	(0.726)	(9.485)	(0.952)
Industrial and economic structure	−0.234 *	−0.187	0.981	−0.142	−0.374 **	−0.099
	(−1.649)	(−1.316)	(1.490)	(−0.999)	(−2.317)	(−0.706)
Time fixed effect	Yes	Yes	Yes	Yes	No	Yes
Individual fixed effects	Yes	Yes	Yes	Yes	No	Yes
Phase I F-statistic values			13.33
Wald’s test value			124.04 ***
Constant term	1.339 ***	1.260 ***	−0.566	1.248 ***	−0.029	1.107 ***
	(8.022)	(7.687)	(−0.567)	(7.462)	(−0.231)	(7.251)
Observed values	540	540	540	510	540	540

Note: ***, **, * represent significance at p < 1%, p < 5%, and p < 10% levels, respectively; the numbers in parentheses are z-values.

presents the results of the model estimation of the effects of competitive effects of grain imports on LEGP [40]. The empirical results in column (1) serve as a benchmark for other empirical models. According to the empirical results in column (2), the effect of competitive effects of grain imports on LEGP are significant at the p < 1% level with a coefficient of 0.008, indicating that an increase of 1 unit in the effect of competitive effects of grain imports will increase to 0.008 units in LEGP. Taking into account the prospective endogeneity problem of the model, this research employs the instrumental variables method to address the endogeneity problem. The results of the two-stage empirical indicate that the empirical coefficient of competitive effects of grain imports on LEGP is significant at the P5% level with a coefficient of 0.198, indicating that competitive effects can effectively improve LEGP. The control variables in the meantime also had an impact on the results. First, financial support for agriculture improves the technical efficiency of grain production by investing in the construction of agricultural infrastructure and promotes technological progress in grain production by investing in special funds for the research and development of grain production technology, thereby reducing the amount of arable land resources required per unit of grain production, which in turn has an impact on the efficiency of water use in grain production. Second, a higher level of agricultural mechanization will effectively reduce the human error of factor inputs in the process of grain cultivation, thus reducing the redundancy of factor inputs of arable land resources for grain production. Thus, the level of agricultural machinery may also have an impact on land use efficiency for grain production.

As seen in column (6), the effect of the squared term of competitive effects of grain imports on LEGP passes the significance test at the p < 1% level with a coefficient of −0.001, indicating that the relationship between competitive effects of grain imports and LEGP has an inverted U-shaped relationship (based on Equation (8)).

3.4. Quantile Empirical of Competitive Effects of Grain Imports Affecting Land Green Efficiency of Grain Production

A more objective and thorough explanation of the competitive impacts of grain imports on LEGP at different quartiles (based on Equation (9)) is offered to reduce the impact of extreme values and error terms on estimation findings. To test the baseline empirical findings and investigate regional heterogeneity, quantile empiricals were performed at 10%, 20%, …, and 90% quantile points, corresponding to columns (1)–(9) of

Table 4. Quantile empirical results of the effect of land competition on land use efficiency in grain production.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
Competitive effects of grain imports	0.004	0.008 **	0.008 **	0.007 *	0.010 *	0.012 *	0.023 ***	0.022 **	0.030 ***
	(1.089)	(2.382)	(2.523)	(1.815)	(1.941)	(1.659)	(2.940)	(2.449)	(2.991)
Technical environment	−0.030 ***	−0.037 ***	−0.040 ***	−0.030	−0.017	−0.001	0.006	0.018	0.114 **
	(−2.976)	(−3.565)	(−2.634)	(−1.446)	(−0.790)	(−0.025)	(0.244)	(0.449)	(2.083)
Financial support for agriculture	0.008	0.005	0.003	−0.002	−0.002	−0.007	−0.011	0.005	0.028 *
	(1.413)	(0.970)	(0.593)	(−0.271)	(−0.264)	(−0.859)	(−1.178)	(0.383)	(1.861)
Agricultural machinery	0.007	−0.005	0.006	0.005	0.011 **	0.013 ***	0.015 **	0.025 **	0.022 *
	(0.899)	(−0.621)	(0.775)	(0.800)	(2.217)	(2.641)	(1.978)	(2.343)	(1.912)
Specialization of crop cultivation	−0.278 ***	−0.257 ***	−0.269 ***	−0.311 ***	−0.325 ***	−0.316 ***	−0.379 ***	−0.456 ***	−0.369 ***
	(−5.836)	(−5.676)	(−5.014)	(−5.204)	(−6.545)	(−3.625)	(−3.416)	(−3.328)	(−2.720)
Environmental regulation	0.018	0.016	−0.007	−0.009	−0.016	−0.016	−0.003	−0.009	−0.005
	(1.264)	(1.118)	(−0.435)	(−0.603)	(−1.155)	(−1.051)	(−0.178)	(−0.518)	(−0.244)
Urbanization level	0.479 ***	0.584 ***	0.566 ***	0.621 ***	0.577 ***	0.550 ***	0.515 ***	0.722 ***	0.660 **
	(5.139)	(5.621)	(5.612)	(5.510)	(6.818)	(6.031)	(3.073)	(2.796)	(2.280)
Percentage of village land	0.037 ***	0.036 ***	0.034 ***	0.040 ***	0.036 ***	0.038 ***	0.032 ***	0.026 ***	0.014
	(6.758)	(6.639)	(5.222)	(5.525)	(6.630)	(6.561)	(3.661)	(2.812)	(1.320)
Industrial and economic structure	−0.301*	−0.277*	−0.288 *	−0.258	−0.353*	−0.322	−0.232	−0.160	−0.132
	(−1.810)	(−1.951)	(−1.914)	(−1.477)	(−1.877)	(−1.551)	(−0.922)	(−0.536)	(−0.466)
Constant term	0.902 ***	0.895 ***	0.948 ***	0.934 ***	1.053 ***	1.055 ***	1.010 ***	0.935 ***	0.926 ***
	(5.859)	(6.880)	(6.800)	(6.685)	(7.566)	(7.184)	(5.497)	(6.425)	(6.384)

Note: ***, **, * represent significance at the p < 1%, p < 5%, and p < 10% levels, respectively; the numbers in parentheses are z-values. Models (1)–(9) represent empirical results at 10%, 20%, ..., and 90% quartiles, respectively.

. The coefficients and significance of the competitive effects of grain imports on LEGP vary at different quartiles, but they are significant and stable from the 10% to 90% quartiles, confirming the baseline empirical results.

Figure 5 shows quantile empirical graphs to illustrate how grain imports affect LEGP in different quartiles. Grain’s coefficient of competitive effects is 10–90%. Grain import competition impacts LEGP growth fast from 0.004 to 0.030. It shows that as grain imports grow, their favorable impact on LEGP increases. The trade promotion competition effect and the induced technological innovation effect increase LEGP when grain imports are stronger in regions with greater differences in virtual land content of grain production at home and abroad.

3.5. Heterogeneity Analysis of Competitive Effects of Grain Imports Affecting Land Green Efficiency of Grain Production

3.5.1. Heterogeneity Analysis of North–South Regions in China

This study separated the sample into northern and southern areas based on geographical characteristics (

Table 5. Heterogeneity in the impact of land competition on land use efficiency in grain production in the north and south.

	Northern Region				Southern Region
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Competitive effects of grain imports		0.010 ***	0.152 *			−0.001	−0.006
		(2.652)	(1.786)			(−0.363)	(−0.071)
Competitive effects of grain imports first-order lag				0.010 ***				0.004
				(2.783)				(1.565)
Technical environment	−0.035 ***	−0.031 **	0.014	−0.031 **	−0.042 **	−0.042 **	−0.035 *	−0.040 *
	(−2.601)	(−2.338)	(0.296)	(−2.353)	(−2.108)	(−2.111)	(−1.681)	(−1.851)
Financial support for agriculture	0.034 ***	0.032 ***	−0.002	0.028 ***	0.008 *	0.008 *	0.008	0.006
	(5.047)	(6.675)	(−0.069)	(6.154)	(1.877)	(1.903)	(0.392)	(1.298)
Agricultural machinery	−0.030 ***	−0.028 ***	−0.004	−0.032 ***	0.009 ***	0.009 ***	0.009 ***	0.009 ***
	(−3.950)	(−3.737)	(−0.141)	(−6.290)	(6.366)	(6.374)	(2.924)	(6.292)
Specialization of crop cultivation	−0.340	−0.283	0.218	−0.203	0.002	0.002	0.064	0.024
	(−1.416)	(−1.184)	(0.288)	(−0.812)	(0.026)	(0.025)	(0.478)	(0.385)
Environmental regulation	−0.032 **	−0.023	0.118	−0.024	0.007	0.006	−0.006	0.021
	(−2.259)	(−1.558)	(1.296)	(−1.610)	(0.399)	(0.375)	(−0.131)	(1.261)
Urbanization level	−3.269 ***	−3.023 ***	−0.277	−3.086 ***	0.946 ***	0.953 ***	1.456 **	0.951 ***
	(−9.726)	(−8.655)	(−0.134)	(−8.852)	(6.305)	(6.308)	(2.389)	(6.165)
Percentage of village land	−0.012	−0.008	0.062	−0.007	0.026*	0.026*	−0.006	0.026*
	(−0.488)	(−0.353)	(0.765)	(−0.305)	(1.782)	(1.741)	(−0.055)	(1.795)
Industrial and economic structure	−0.006	0.054	0.866	0.052	−0.580 ***	−0.584 ***	−0.612	−0.490**
	(−0.039)	(0.330)	(1.301)	(0.322)	(−2.975)	(−2.992)	(−1.471)	(−2.559)
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Individual fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Phase I F-statistic values			16.38				12.07
Wald’s test value			23.22 ***				165.26 ***
Constant term	2.800 ***	2.584 ***	−0.075	2.669 ***	0.996 ***	0.998 ***	0.913	0.920 ***
	(9.481)	(8.592)	(−0.040)	(8.762)	(5.572)	(5.574)	(1.392)	(5.107)
Observed value	270	270	270	255	270	270	270	255

Note: ***, **, * represent significance at p < 1%, p < 5%, and p < 10% levels, respectively; the numbers in parentheses are z-values.

). In column (1), control factors on LEGP are regressed. As shown in columns (2)–(4), competitive effects of grain imports have a significant positive effect on LEGP, similar to the full sample empirical. The empirical results for the sample from the southern region are depicted in columns (5) through (8), with column (5) displaying the empirical results of control variables on the efficacy of grain production land. The results of the baseline empirical model, the inclusion of instrumental variables, and the inclusion of the first-order lag term of the effects of grain imports indicate that the effects of grain imports on LEGP are not statistically significant, contrary to the results of the full sample empirical (columns (6)–(8)).

3.5.2. Heterogeneity Analysis of Sub-Grain Production Areas

This study also divided the sample into main and non-main grain production areas for testing based on their characteristics (

Table 6. Heterogeneity in the effect of land competition on land use efficiency in grain production in different production areas.

	Major Grain-Producing Areas				Non-Grain-Producing Areas
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Land cost competition effectiveness		0.002	0.096			0.007 **	0.202 **
		(0.533)	(0.412)			(2.337)	(2.471)
Competitive effects of grain imports first-order lag				0.004 *				0.016 ***
				(1.673)				(6.065)
Technical environment	−0.012	−0.010	0.051	−0.011	−0.000	−0.001	−0.072	0.004
	(−0.646)	(−0.566)	(0.283)	(−0.630)	(−0.011)	(−0.090)	(−1.091)	(0.223)
Financial support for agriculture	0.001	0.001	−0.005	0.002	−0.004	−0.003	0.030	−0.006
	(0.189)	(0.171)	(−0.171)	(0.311)	(−0.516)	(−0.334)	(0.881)	(−0.729)
Agricultural machinery	0.009	0.008	−0.044	0.001	0.003	0.004	0.016	0.004
	(1.233)	(1.088)	(−0.363)	(0.121)	(1.020)	(1.134)	(1.160)	(1.228)
Specialization of crop cultivation	−0.405 **	−0.402 **	−0.319	−0.125	0.149	0.156*	0.388	0.121
	(−2.409)	(−2.392)	(−0.567)	(−0.661)	(1.585)	(1.698)	(0.964)	(1.311)
Environmental regulation	0.000	0.004	0.232	0.016	−0.007	−0.005	0.045	0.005
	(0.020)	(0.208)	(0.425)	(0.898)	(−0.428)	(−0.301)	(0.706)	(0.299)
Urbanization level	−0.581	−0.602	−2.417	−0.644 *	−0.173	−0.105	0.998	−0.098
	(−1.538)	(−1.586)	(−0.741)	(−1.761)	(−0.549)	(−0.389)	(0.643)	(−0.351)
Percentage of village land	0.028 **	0.028 **	0.085	0.032 **	0.002	0.002	−0.187	0.006
	(1.984)	(2.000)	(0.653)	(2.323)	(0.064)	(0.076)	(−0.820)	(0.234)
Industrial and economic structure	−0.259 *	−0.248	0.492	−0.157	−0.215	−0.176	0.159	−0.182
	(−1.725)	(−1.641)	(0.289)	(−1.122)	(−0.853)	(−0.706)	(0.157)	(−0.711)
Time fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Individual fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Phase I F-statistic values			0.199				11.32
Wald’s test value			0.97				126.43 ***
Constant term	1.467 ***	1.461 ***	1.053	1.335 ***	1.205 ***	1.129 ***	0.493	1.157 ***
	(6.590)	(6.556)	(0.729)	(5.822)	(3.957)	(6.076)	(0.383)	(6.018)
Observed value	234	234	234	221	306	306	306	289

Note: ***, **, * represent significance at p < 1%, p < 5%, and p < 10% levels, respectively; the numbers in parentheses are z-values.

). Column (1) shows the empirical control factors on LEGP. Columns (2)–(4) show that grain imports do not affect LEGP, unlike the full sample empirical. The empirical results for the sample of non-grain-producing regions are displayed in columns (5) through (8), with column (5) displaying the empirical results of the control variables on the LEGP. The empirical results for either the baseline empirical model, the inclusion of instrumental variables, or the inclusion of a first-order delayed term for the effects of grain imports indicate that the effects of grain imports on grain production land are negative. The empirical results for the baseline empirical model, the inclusion of the instrumental variables, or the inclusion of the first-order lagged term for the effects of grain imports indicate a significant positive effect of competitive effects of grain imports on LEGP, which is more comparable to the results of the full sample empirical.

3.5.3. Mechanisms of Competitive Effects of Grain Imports to Reduce Domestic Grain Profits to Force Land Green Efficiency Improvements

Based on hierarchical empirical analysis, the mechanism by which the competitive effects of grain imports reverse the improvement in LEGP by decreasing the profitability of grain production is examined. The effect of competitive effects of grain imports on grain production profit meets the significance test at the 1% level and the empirical coefficient is −0.006 as shown in column (1) of

Table 7. Results of testing the mediating mechanism of the impact of water competition effect on water use efficiency of grain production.

	(1)	(2)	(3)	(4)
Competitive effects of grain imports	−0.006 *		0.007 ***	0.006 ***
	(−1.790)		(2.956)	(2.662)
grain production profit		−0.121 **		−0.097 *
		(−2.119)		(−1.688)
Technical environment	−0.031 **	−0.005	−0.001	−0.004
	(−2.466)	(−0.444)	(−0.087)	(−0.349)
Financial support for agriculture	−0.003	0.010 **	0.009 *	0.008 *
	(−0.647)	(2.134)	(1.959)	(1.878)
Agricultural machinery	0.000	0.005 *	0.005 *	0.005 *
	(0.056)	(1.957)	(1.944)	(1.949)
Specialization of crop cultivation	−0.294 ***	0.051	0.073	0.058
	(−5.146)	(0.663)	(0.968)	(0.768)
Environmental regulation	0.032**	−0.004	−0.003	−0.001
	(2.285)	(−0.336)	(−0.276)	(−0.048)
Urbanization level	−1.023 **	−0.523 **	−0.369 *	−0.445 *
	(−2.297)	(−2.105)	(−1.673)	(−1.863)
Percentage of village land	−0.031 **	0.001	0.005	0.004
	(−2.132)	(0.049)	(0.400)	(0.268)
Industrial and economic structure	0.024	−0.225	−0.187	−0.184
	(0.226)	(−1.596)	(−1.316)	(−1.300)
Time fixed effect	Yes	Yes	Yes	Yes
Individual fixed effect	Yes	Yes	Yes	Yes
Constant term	0.788 ***	1.410 ***	1.260 ***	1.316 ***
	(3.397)	(7.947)	(7.687)	(7.563)
Observed value	540	540	540	540

Note: ***, **, * represent significance at p < 1%, p < 5%, and p < 10% levels, respectively; the numbers in parentheses are z-values.

. It demonstrates that the competitive effects of grain imports hurt the profit of domestic grain production. With a coefficient of −0.121, column (2) indicates that the profitability of grain production has a significant influence on the LEGP. It demonstrates that the profitability of grain production has a negative influence on the land productivity of grain production. The effect of competitive effects of grain imports on LEGP in column (3) is statistically significant at the 1% level, with an empirical coefficient of 0.007. The addition of the grain production profit variable in column (4) reveals that the coefficient of competitive effects of grain imports on LEGP passes the significance test at the 1% level; the coefficient is 0.006 and the coefficient is decreasing. This indicates that the profitability of grain production partially mediates the effect of grain imports’ competitive effects on the LEGP.

4. Discussion and Conclusions

4.1. Discussion of Empirical Results for Competitive Effects of Grain Imports Affecting Land Green Efficiency of Grain Production

Grain manufacturing has factor input redundancy compared to optimal output. The magnitude of grain production efficiency losses depends on factor resource endowments and agricultural technologies [41]. In a non-fully efficient market, grain production’s social and economic advantages lead to the over-input of production elements, which burdens resources and the environment and reduces resource efficiency. However, by improving resource use and efficiency, grain total factor productivity can be increased, reducing the need for land, pesticides, and fertilizers. Market mechanisms can affect the local grain sales market and push domestic grain production efficiency by importing grain goods with comparative disadvantages from the country to the international market. New-new commerce and induced technological innovation theories explain this.

The trade promotion competition effect in the new-new trade theory indicated that import competitiveness boosts domestic industry productivity [5]. The competitive effect of higher imports on economic agents relies on production efficiency. Due to market share compression, economic agents with low production efficiency must leave the market. In contrast, subjects with a high level of production efficiency and who maintain production in the competition will be stimulated to improve their efficiency and reduce production costs, promoting the industry’s benign development. Grain producers’ low profitability makes grain production efficiency crucial to their sustained production and development. As a major limitation on grain production, land resources must be used more efficiently. According to the theory of induced technological innovation, agricultural production is limited by scarce factors, but improving agricultural technology saves scarce factors and uses abundant factors, inducing technological change.

As a limitation on grain production, land resources must be used efficiently. China’s main imported grain products (soybeans, maize, etc.), have lower virtual land content than local self-production. Due to the trade promotion competition effect, as China’s grain imports increase, larger grain-growing entities are more likely to survive with higher grain production land resource use efficiency and lower production costs, while smaller grain-growing entities are more likely to be eliminated from the market. The virtual land content of domestic and overseas grain production differs according to land resource efficiency. Grain imports also reduce domestic grain producers’ market share, requiring them to improve land resource use efficiency. This increases grain production efficiency, saves land resources lost to changing production methods, and reduces pollutant emissions. China’s per capita land resource possession is only 27.7% of the world average, so land resource elements are relatively scarce, which makes technological changes in grain production spontaneously develop in the direction of saving land resources and promoting grain production efficiency.

Further, considering the time lag and dynamics of grain import trade affecting LEGP, grain imports in the first half of the year may affect domestic LEGP in the second half and the following year through the market mechanism response, while the structure of grain imports in the second half affects the following year [42]. Thus, grain import commerce might alter grain cultivation structure, resource pressure, resource usage efficiency, and land green efficiency in the following year. Thus, LEGP must account for the time lag and dynamic character of grain import trade. Thus, land green efficiency and grain production time lag must be considered.

Domestic and international grain production is optimal. Regarding the competition from grain imports, when grain imports save 1.3 million hectares of land, land cost becomes competitive, resulting in the highest LEGP. LEGP may drop beyond this threshold. Grain imports initially have little influence on the local market since the market self-regulates by removing inefficient grain producers. However, when the impact exceeds a certain level, the market’s self-regulatory mechanism will fail, and grain producers with relatively high land green efficiency will be squeezed out of the market, resulting in a decrease in LEGP.

4.2. Discussion and Conclusions of Sub-Regional Empirical Results

Northern areas benefit more from competitive grain imports on LEGP than southern ones. Livestock production is more established in the north, therefore feed crops like soybeans and corn are in more demand. With the progressive development of the import scale, a large number of grain items with lower virtual land content enter the domestic market, which has a tremendous influence on the northern grain market and strengthens trade promotion competition. The scale expansion effect reduces marginal grain production land factor inputs in the north due to grain import competition. Thus, it improves grain production efficiency and reduces land resource waste owing to shifting production operations. However, Southern China has a lot of mountainous and hilly terrain, so most grain production is in northern provinces, and land scale limits LEGP in southern regions. The southern area has more developed secondary and tertiary sectors than the north, and most citizens do not depend on agricultural output for their income, therefore there is a significant incentive for land to change to non-grain production. Thus, while imported grain has a competitive influence on the southern region, it does not greatly enhance LEGP.

Non-main producing locations benefit more from grain imports’ competitive impacts on LEGP. The trade promotion competition effect weakens the marginal promotion effect of grain imports on large-scale cultivation in the main grain-producing regions because their grain production and operation scale is higher. The marginal contribution of competitive impacts of grain imports on LEGP in the northern regions is similarly less. Nevertheless, non-major grain-producing regions will have a stronger impact on their grain markets by importing a large number of grain products with lower virtual land content, which can effectively eliminate producers with lower grain production scale and efficiency in those regions and help promote large-scale and efficient production. Large-scale cultivation also promotes production service outsourcing in non-main grain-producing regions, horizontal and vertical division of labor in grain production, precise factor input, and reduced land resource factor redundancy. Thus, grain imports increase land green efficiency in non-primary producing regions more than primary production areas.

4.3. Discussion of the Mediating Mechanism of Competitive Effects of Grain Imports Affecting Land Green Efficiency of Grain Production

According to the new trade theory, due to the trade promotion competition effect, China’s grain imports have increased, bringing grain with low land costs from the international grain market into the domestic grain market, lowering grain prices, and shrinking domestic grain producers’ profits. Due to land green efficiency and reduced production costs, larger grain providers are more likely to thrive [43]. Smaller grain producers with poorer land green efficiency and greater production costs are more likely to be eliminated from the market. The virtual land content of domestic and international grain production is affected by land green efficiency [44]. Grain imports also reduce domestic grain costs, earnings, and market share, driving local LEGP to improve. According to induced technological innovation, agricultural technology is improved to save limited factors and use plentiful elements. China’s land constraint has driven grain production technology toward land conservation, improving land green efficiency.

4.4. Conclusions

Based on virtual land trade, new-new trade theory, and induced technical innovation theory, this study experimentally analyzes the route mechanism of competitive impacts of grain imports compelling LEGP improvement. (1) From 2003 to 2020, LEGP increased 27.53% nationally. LEGP in the Huang-Huai-Hai region and northeast region is high and expanding rapidly, at 66.41% and 36.30%, respectively. (2) Grain imports can drive China to enhance LEGP. The non-linear effect test demonstrates an “inverted U-shape” between the two: LEGP is highest when grain imports save 1.3 million hectares. The non-linear effect test also shows an “inverted U-shape”: when imported grain saves 1.3 million hectares, LEGP is strongest, but beyond this threshold, it may decrease. (3) Quantile empirical findings showed that LEGP rose as the competitive impacts of grain imports grew. (4) The sub-sample empiricals found that the positive effect of competitive effects of grain imports on LEGP was stronger in northern regions than southern regions and stronger in non-main grain production regions than main grain production regions. Grain imports improve LEGP more in non-major grain production regions than in main production areas. (5) The mechanism test indicated that grain imports’ competitive impacts can increase land use efficiency by lowering local grain production’s profitability.

Our findings provide new ways to boost LEGP in China. First, the government should increase investment in research and development of innovative land-saving technologies in regions with high virtual land content in grain production, such as the northwest and Huang-Huai-Hai regions, and continue to reduce this virtual land content. Second, since the increase in land-saving effect can force the improvement of domestic LEGP, the government should encourage grain enterprises to import more grain from countries with relatively lower virtual land content, i.e., higher LEGP. To increase domestic LEGP, the government could encourage grain firms to import more grain from nations with lower virtual land content. Third, increase grain imports from regions with lower LEGP, such as the middle and lower reaches of the Yangtze River and the northwest region, to strengthen trade promotion competition and induce technological innovation effects to improve LEGP in these regions.

This study does not further consider in depth the heterogeneity of the impact of grain import competition effects on the water use efficiency of grain producers at different scales. Differences in the impact of the competitive effects of grain imports on the water use efficiency of grain producers of different scales were not explored separately due to space constraints. An attempt can be made in the future to collect additional data related to information on grain producers of different sizes to further explore the heterogeneity of the impact of grain imports on the water use efficiency of grain producers at different scales.