The Effect of Heterogeneous Environmental Regulations on Carbon Emission Efficiency of the Grain Production Industry: Evidence from China’s Inter-Provincial Panel Data

Abstract

Examining the impact of heterogeneous environmental regulations on the carbon emission efficiency of the grain production industry will help to provide a theoretical basis and decision-making reference for achieving the goal of carbon peaking and sustainable agricultural development. Based on the panel data of 30 provinces in China, the DEA-SBM method was used to measure the grain production industry’s carbon emission efficiency from 2011 to 2019. Separating environmental regulations into formal environmental regulations and informal environmental regulations in two parts, then the two-way fixed effect model, threshold effect model, and intermediary effect model are used to test the influence of heterogeneous environmental regulations on the grain production industry’s carbon emission efficiency. The results show that: (1) the grain production industry’s carbon emission efficiency continues to improve but still has space for improvement. (2) the relationship between formal environmental regulations and the grain production industry’s carbon emission efficiency exhibits a “U-shaped” curve; informal environmental regulations have a significantly positive effect on improving the grain production industry’s carbon emission efficiency. The conclusion is still valid after the robustness test. (3) A threshold mechanism test found that there is a single threshold effect between the formal environmental regulations and the grain production industry’s carbon emission efficiency, but it does not cross the “U” inflection point. (4) there is a “U” shaped non-linear mediating effect in the influence of formal environmental regulations on the grain production industry’s carbon emission efficiency; informal environmental regulations also have an intermediary transmission path of production agglomeration of the grain production industry. This study supplies a new perspective for promoting the grain production industry’s carbon emission efficiency.

1. Introduction

China is the largest carbon emitter in the world, which strongly affects global climate change [1]. Apart from its increasing industrial carbon emissions, agricultural carbon emissions of China are also at a high level and have become an important source of greenhouse gas emissions [2,3]. Despite China’s surging grain output in recent years, which has greatly contributed to national grain security and nutrition security [4]. However, blindly pursuing the improvement of grain production has increased the use of energy and carbon emissions, making resource waste, ecological damage, and other issues increasingly serious, which is not conducive to the Chinese government’s early realization of the goal of carbon neutrality and carbon peak [5]. Some scholars believe that from 2010 to 2050, the negative impact of grain on environmental damage may increase by 50% to 90% if there is no corresponding technology to address the serious pollution of agriculture fueled by the continuous change in population and income [6]. Anyway, there is still a gap in the sustainability of grain production between China and developed countries [7].

Organic agriculture mainly aims at producing sustainable and safe food and protecting soil and the environment [8], which is in line with the development trend of sustainable grain security in China. Organic food was famous for food safety in organic agriculture. For instance, organic wheat in Northern Europe has 25% fewer pesticide residues than conventional production [9]. With the continuous improvement of living standards, consumers’ demand for food health has increased, and at present, the supply has exceeded the demand [10], so the scale of organic grain planting has gradually expanded [11]. The carbon label on organic food can help consumers to evaluate the carbon footprint of the product at all stages of production and in turn, promote the decision-makers at the production end to have the motivation to adopt a more sustainable model to produce organic food [12,13]. In Sweden’s organic food production, producers are required to reduce input and obtain higher supply levels for the purpose of reducing greenhouse gas emissions [14]; In Italian dairy farms, the government has introduced relevant policies to reduce the use of chemical fertilizers, which can reduce the waste of the supply chain of dairy products system, thereby improving the efficiency of organic agriculture and encouraging sustainable agricultural development [15]; In the organic grain production in the United States, organic grains certified by the Ministry of Agriculture need to use organic seeds according to regulations and be planted on the land free of prohibited substances such as synthetic fertilizers [10]. These production modes provide a stable basis for agriculture or low-carbon grain production.

In this case, ensuring sustainable grain security remains an important issue in China. As land use is the major contributor to carbon emissions in the supply chain of agricultural products [16], it is urgent to stabilize grain production at a low carbon level, or in other words, improve the efficiency of carbon emissions from the grain production industry [17]. To reduce greenhouse gas emissions, mitigate global warming and promote sustainable agricultural development, the government needs to implement a series of environmental regulation policies [18]. Some scholars believe that environmental regulations include not only formal environmental regulations (supervision of pollution conducted by the government), but also informal environmental regulations (supervised by social groups, the public, and the media) [19]. In the “No.1 central document” for 2021, the government proposed “fewer chemical fertilizers but higher efficiency.” As pointed out in its plan of agricultural and suburban modernization during the 14th Five-Year Plan period (2021–2025), the government will strive to reduce emissions in agriculture and rural areas and improve the quality of arable land. The implementation of a series of government-led formal environmental regulations is intended to address the carbon emissions associated with high inputs in the grain production process. Formal environmental regulations have been studied more because of their mandatory nature [20]. Despite several government policies, there are still some farmers who do not engage in cleaner energy production practices, and sometimes informal institutions (e.g., village rules and regulations) can compensate for the limitations of formal environmental regulations [21,22]. Therefore, it is necessary to pay attention to the impact of informal institutions on the carbon efficiency of grain production.

Due to the serious shortage of agricultural resources, agricultural agglomeration has begun to appear around the world, such as the clustering of corn corps in the United States, the flower planting clusters in Dutch, and the grape planting area in French. At present, China’s grain production industry also shows a certain degree of agglomeration [23]. Both formal and informal environmental regulations can effectively drive industrial structure upgrading [24], among which industrial clustering is one of the most significant changes that could lead to carbon emission reduction, reallocation, or even an increase in total carbon emissions. Therefore, whether environmental regulations can change the carbon emission efficiency via adjusting the industrial agglomeration level of corps planting remains an open question.

The main contributions of the study are as follows: (1) Based on the environment of sustainable grain security and organic grain development, it is rare to effectively combine stable grain supply with low carbon production, that is, to measure the carbon emission efficiency level of the grain planting industry. This can present the current sustainable production capacity of the grain planting industry for the Chinese government and promote the healthy development of the organic grain industry. (2) Few scholars have studied the impact of environmental regulations on carbon emission efficiency of agriculture and even grain planting; In addition, sometimes the role of formal environmental regulations is not so obvious. It is also necessary to study the differential impact of heterogeneous environmental regulations on the carbon emission efficiency of the grain planting industry so as to provide a reference for the Chinese government to improve environmental regulation tools. (3) Furthermore, the relationship between environmental regulations and carbon emission efficiency of the grain planting industry may not be simply linear, so we used the threshold model and intermediary model proposed by Hansen [25] and Baron and Kenny [26] to deeply analyze the relationship between them and explore the mechanism of heterogeneous environmental regulations on carbon emission efficiency of grain planting industry.

This study will try to address this issue in six more sections. Section 2 combs the relevant literature and theoretical Hypothesis; Section 3 establishes the analysis model and sample selection and describes the variables according to the theoretical Hypothesis; Section 4 mainly explains the benchmark regression results, robustness tests, and regional heterogeneity results of this study; Section 5 further discusses the results of a threshold effect and intermediary effect; Section 6 summarized the conclusions of the study and puts forward relevant policy recommendations and, finally, Section 7 shows some shortcomings of the current paper, which will be the content of the follow-up study.

2. Literature Review and Theoretical Hypothesis

As carbon emissions keep surging in China’s agricultural industry, most scholars have calculated the agricultural carbon emission efficiency needed for the emission peaking target [27,28]. However, few have calculated the carbon emission efficiency for maintaining sustainable food security. Challenged by high carbon emissions and other issues, reasonable environmental regulations are urgently needed. Environmental regulations have been widely studied, including the impact of formal environmental regulations on the carbon emission efficiency at the province level [29,30] or on the industry [31] and transportation fields [32]. According to their results, the impact of formal environmental regulations on carbon emission efficiency is in “U” [29] or inverted “U” shape [33], or no significant correlation could be observed [34]. However, only a few studies shed new insights into the association between formal environmental regulations and agricultural carbon emission efficiency, as well as the carbon emission efficiency of the grain planting industry.

Porter believes that, since the short-term environmental regulations will increase the cost of governance, the “cost effect” principle would encourage producers to innovate in clean technology amid the continuous implementation of environmental regulations. Thus, the goal of win-win economic and environmental, namely, the “innovation compensation effect” [35], is achieved. The same is true of the impact of formal environmental regulations on the carbon emission efficiency of grain production. In the short term, environmental regulations will drive agricultural producers to assign part of their input to agricultural machinery upgrading, clean facilities, and agricultural waste recycling. It can influence the cost of other production factors, which means, under the strict control of formal environmental regulations, farmers will reduce the input of pesticides, fertilizers, and agricultural film in grain planting and hence reduce the per unit yield and profits [36]. The decline in profit will further prevent producers from learning human capital, clean technologies, and advanced agricultural knowledge [37]. In conclusion, environmental regulations are not conducive to the improvement of the carbon emission efficiency level in the short term, yet in the long run, the enhancement of formal environmental regulations will raise their awareness of the high carbon emissions, drive them to change the production mode and improve the efficiency of chemical fertilizers, pesticides, agricultural film, and other production factors. In addition, the government will also introduce a series of subsidy policies to partly remove the burden on grain producers throughout the long period of shifting to technological innovation [38]. Farmers may encourage grain producers to carry out pollution control technology, machinery technology, fine varieties, and transgenic technology innovation so as to improve grain yield on the basis of low-carbon production. Formal environmental regulations follow the “innovation compensation effect,” Finally, the grain production industry’s carbon emission efficiency will be improved. Therefore, Hypothesis 1 is proposed based on the above analysis.

Hypothesis 1 (H1).

The effect of formal environmental regulations on the carbon emission efficiency of the grain planting industry is “U” shaped.

There are many studies on the impact of formal environmental regulations on carbon emission efficiency, but it ignores the role of informal environmental regulations. In recent years, as informal environmental regulations gained ground, relevant studies found that more influential and powerful informal environmental regulations facilitate local environmental governance and environmental pollution control, whether in developed countries or developing countries [39,40,41]. Hence, in China’s rural areas, where formal regulations are lagging behind [22], how informal environmental regulations affect the carbon efficiency of the grain production industry is another problem that needs to be discussed.

As proposed by Pargal, informal environmental regulations refer to the interests-driven public behavior of seeking to achieve high environmental quality through negotiation, consultation, and media exposure when the formal environmental regulations led by the government are not well supervised [42]. Informal environmental regulations rely on the public’s awareness of environmental protection. On the one hand, the public and the media can report to relevant authorities, forcing producers to carry out technological innovation to reduce environmental losses [43]. On the other hand, profit-driven individuals may overlook the public concern, and hence lead to less environmental protection behaviors [44].

According to the theory of constraints, the role of informal environmental regulations can be implemented at descriptive and prohibitive levels [45]. Descriptive constraint refers to the production behavior of conformity of producers to adapt to the economic and social environment of producers. The descriptive constraint is often not intentional. Prohibitive restraint relies on the social currency attribute of “credit and prestige” in rural society. If producers meet the local production rules, they will be recognized, while breaking the rules will be condemned. By rendering producers to pay attention to public evaluation of a series of behaviors, prohibitive restraint has a guiding role [46]. In grain production, informal environmental regulations can become the focus of constraints, improving the production behavior of grain producers through the three following aspects [47]. First is transfer internalization. Generally, grains are planted in rural areas, where the community is full of acquaintance, collectivism, and herd effect [48]. Farmers are affected by the constraint of informal environmental regulation description in grain production, so they can reasonably use chemical fertilizers and pesticides, dispose of straw, and save water resources. This kind of green production behavior and values gradually affect other people through the producer’s conformity psychology and follows the trend one after another. Second, discipline and supervision. Constrained by the prohibitions of informal environmental regulations, high input, high energy consumption, and other behaviors in grain production violate the principle of low carbon grain production. Farmers may be persuaded or even condemned by acquaintances and face a loss of reputation, trust, and so on. In addition, producers may be fined due to imposed economic penalties [49]. Third, value orientation. Descriptive constraints and prohibitive constraints of informal environmental regulations play a role together: by demonstrating the role models of green grain production and operation subjects, farmers’ eagerness for recognition and reputation stimulates their chase for green production [50]. Thus, we conclude that Hypothesis 2.

Hypothesis 2 (H2).

Informal environmental regulations can promote the carbon emission efficiency of the grain planting industry in a one-way approach.

No agreement has been obtained on whether environmental regulations can affect carbon emission efficiency by upgrading the industrial structure [51,52]. With the development of production agglomeration of China’s grain planting industry, it is particularly necessary to study the mediating role of grain production agglomeration in environmental regulations and carbon emission efficiency. The role of production agglomeration in improving agricultural carbon emission efficiency is reflected in the following three aspects. First, the production agglomeration promotes the land operation of the scale. When the operators purchase agricultural production from individual farmers, they can buy at wholesale prices with strong bargaining power. By doing so, operators save costs of labor and other production factors and hence improve grain production efficiency. Secondly, the agglomeration of the grain production industry can provide some agricultural business entities with agricultural production services for the surrounding grain growers and amortize the costs. For example, the purchased agricultural machinery can be provided for surrounding farmers when available to boost profit. Finally, after realizing the agglomeration of the grain production industry to a certain extent, with the support of the agricultural production service market, the operators of a certain scale can independently adjust the business scale, namely the scale of production and income. Hence, they gradually pay more attention to grain quality and green production, actively learn green technology and improve carbon emission efficiency. Second, the scale of operation is not limited by the land scale of different agricultural business entities. Different agricultural business entities in the grain planting cluster purchase services in a unified way to achieve the unified operation of standard specialties, such as grain production trusteeships. After the integration of service supply, the service provider can select the most appropriate technology that fits the local grain production scale. In the whole process, the grain trusteeship service is effectively combined with the use of pesticides, fertilizers, and agricultural film, further promoting the scientific fertilization, and spraying of the service provider, reducing the element input and service costs while spreading green planting technology, thereby improving grain production industry’s carbon emission efficiency. Third, grain planting is largely distributed in rural areas, where the social network is composed of relatives and neighbors. The mutual trust and valued reputation brought by this relationship can support grain production and the implementation of contracts and minimize the risk of grain production and operation and transaction costs. In clusters of grain planting, green production technology has been rapidly promoted under the leadership of emerging agricultural operators, reducing the technology investment cost per unit of grain output. In rural areas, neighbors and relatives can quickly obtain corresponding technology and information through free-riding behavior. In addition, an innovation of green high-yield technology will trigger a chain reaction, which will stimulate farmers’ intention for scientific and technological innovation again. The gathering of competitors’ grain planting industry will also promote green technology innovation and force other grain producers to carry out technological innovation [53]. The production agglomeration of the grain planting industry will also increase the income of grain farmers so that farmers have sufficient funds to participate in the environmental governance of grain planting when facing environmental regulations, thus improving the carbon emission efficiency of the grain planting industry. Hence, Hypothesis 3 is as follows.

Hypothesis 3 (H3).

The grain production agglomeration can promote the carbon emission efficiency of the grain planting industry in a one-way approach.

Formal environmental regulations are mandatory, forcing grain producers to take action to control the high pollution in the process of grain production. However, increasing the production costs of grain producers. In the early stage, due to the additional consumption of production costs caused by the purchase of clean facilities by grain producers, the income from grain planting decreased, and the opportunity cost of grain production was high. Grain producers, as rational economic people, will withdraw from grain planting in order to pursue their highest income and choose to plant some cash crops with higher net income. This is also in line with Zhong’s point of view: the difference in comparative net income of different crops will affect farmers’ enthusiasm for production to varying degrees, thus changing the production layout of crops [54]. That is, farmers grow more economic crops with higher net income, which makes it impossible for grain planting to achieve continuous planting, which is not conducive to the improvement of the production concentration level of the grain planting industry. Therefore, in the initial stage, formal environmental regulations have an inhibitory effect on the level of production agglomeration of the grain planting industry. The improvement of the grain production agglomeration level can effectively improve the carbon emission efficiency of the grain planting industry, so the initial formal environmental regulation is not conducive to improving the carbon emission efficiency of the grain planting industry by promoting the increase in production agglomeration level. However, agricultural producers will also face the constraints of formal environmental regulations when planting cash crops with higher net income. With the further strengthening of formal environmental regulations, the cost of farmers’ environmental governance will also rise, making farmers unable to meet their own demand for pursuing high interests when planting cash crops. In order to improve their income level, farmers gradually choose to work in cities, which is consistent with the background of the current increase in China’s urbanization rate, a large outflow of rural labor, and some idle land. However, in order to stabilize grain production, the government has unified the management of the remaining idle land, better revitalized the land circulation market and the new agricultural business entities and promoted the production concentration level of the grain planting industry. Therefore, the carbon emission efficiency of the grain planting industry, with the further strengthening of the formal environmental regulations, can be promoted by improving the production concentration level of the grain planting industry. Consequently, Hypothesis 4 is as follows.

Hypothesis 4 (H4).

The relationship between formal environmental regulations and the level of production agglomeration of the grain planting industry is “U” shaped; There is a “U” type nonlinear intermediary effect in the impact of formal environmental regulations on the carbon emission efficiency of the grain planting industry.

The role of informal regulations in promoting grain production agglomeration mainly depends on the interpersonal trust to cooperate between farmers [55]. This kind of trust mainly refers to the trust between relatives and neighbors in rural areas. First of all, under the supervision of informal environmental regulations, farmers tend to display more obvious pro-environmental behavior to have wider access to agricultural information, such as the utilization of agricultural wastes such as straw. In this way, farmers promote industrial clustering by establishing an information-sharing mechanism with neighbors who trust each other in a way to reduce the search cost of cooperative objects and reduce the information asymmetry in the game with the government [56]. Secondly, under the supervision of informal environmental regulations, the neighborhood relationship with “acquaintance trust” as the core can establish a cooperation mechanism, increase mutual cooperation expectations, promote the production agglomeration of the grain planting industry, share the cost of environmental governance, and achieve the principle of mutual benefit. In addition, since every resident is well connected in a suburban area, the realization of concentrated production enables farmers to shoulder environmental responsibility and possible risks. Finally, to reduce the uncertainty of information disclosure on environmental issues, reduce the cost of information search and reduce risks, the public would tighten supervision on the agricultural business entities [57], which will also promote the agglomeration of the grain planting industry. It is precise because the production agglomeration of the grain planting industry can improve the carbon emission efficiency, so informal environmental regulations can improve the carbon emission efficiency level of the grain planting industry by promoting the production agglomeration of the grain planting industry. Therefore, Hypothesis 5 is given according to the analysis.

Hypothesis 5 (H5).

Informal environmental regulations can promote the production agglomeration level of the grain planting industry; There is a linear intermediary effect in the impact of informal environmental regulations on the carbon emission efficiency of the grain planting industry.

In order to more clearly show the different impact mechanisms of heterogeneous environmental regulations on the carbon emission efficiency of the grain planting industry, a flow chart is drawn, as shown in Figure 1.

3. Research Methods and Data

In previous studies, few scholars conducted theoretical discussions and empirical tests on the relationship between heterogeneous environmental regulations and the carbon emission efficiency of the grain planting industry. However, we must clarify the mechanism between the two if we achieve the goal of sustainable grain security in China. Referring to previous scholars’ research, adopting the fixed effects model to study the relationship between the heterogeneous environmental regulations and the grain production industry’s carbon emission efficiency [58]. In addition, this paper uses the mediator model and the threshold model to further explore the nonlinear impact and indirect mechanism of heterogeneous environmental regulations on the grain production industry’s carbon emission efficiency. The specific model construction, variable selection, and relevant data sources are as follows.

3.1. Model Construction

3.1.1. Benchmark Model

In this study, we use 30 provinces’ panel data in China. In view of the different situations in each province, there may be missing variables that do not change over time, and in order to reduce the endogeneity of the model, we try to establish a two-way fixed effect model to explore the influence of formal environmental regulations and informal environmental regulations on the carbon emission efficiency of China’s grain production industry.

On modeling the effects of formal environmental regulations and informal environmental regulations on the efficiency of carbon emissions from grain production industry.

$${CE}_{\mathit{it}}{=\alpha }_0{+\alpha }_1{\mathit{fen}}_{\mathit{it}}{+\alpha }_2{\mathit{fen}}_{\mathit{it}}^2{+\alpha }_{3}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(1)

$${\mathit{CE}}_{\mathit{it}}{=\alpha }_0{+\alpha }_1{\mathit{inen}}_{\mathit{it}}{+\alpha }_{2}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(2)

In Equations (1) and (2), CE_it is the dependent variable, which indicates the different carbon emission efficiency levels of each province; fen_it and inen_it are the core explanatory variables, expressing different levels of formal environmental regulations and informal environmental regulations of each province; Control_it expresses the control variables; i indicates the province; t means the time; α₀ is a constant; μ_i is the individual fixed effect that does not change with time; v_i is the time fixed effect; ε_it is a random perturbation. In order to verify the “U” theory in Hypothesis 1, the square term of formal environmental regulations is introduced in Equation (1) to represent a high level of formal environmental regulation, and the explanatory variables are treated by logarithm. We centralized the variable ‘the formal environmental regulations’ to mitigate the error caused by the potential multicollinearity problem caused by the square term of formal environmental regulation.

3.1.2. Threshold Effect Model

In the above theoretical analysis, the formal environmental regulations maybe have a “U” effect on the grain production industry’s carbon emission efficiency. In order to accurately control the optimal degree of environmental regulations implementation and enhance the efficiency of environmental regulations, analyzing the non-linear relationship between formal environmental regulations and the grain production industry’s carbon emission efficiency by adopting Hansen’s panel threshold model and constructing the following model [25].

$${\mathit{CE}}_{\mathit{it}}{=\theta }_0{+\theta }_1{\mathit{fen}}_{\mathit{it}}\times \mathrm{I}\left(\mathit{fen}\le {\gamma }_1\right){+\theta }_2{\mathit{fen}}_{\mathit{it}}\times \mathrm{I}\left({\mathit{fen}>\gamma }_1\right){+\theta }_{3}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(3)

In Equation (3), I(·) is the indicative function, γ₁ is the threshold value, and θ are the coefficients of the effect of environmental regulations on the grain production industry’s carbon emission efficiency in different threshold intervals.

3.1.3. Intermediary Effect Model

According to the above theoretical analysis, formal environmental regulations may have an indirect influence on the grain production industry’s carbon emission efficiency through mediating variables. Constructing a mediating mechanism model in order to verify the above analysis. Since this intermediary effect may be nonlinear, the traditional three-step regression method for testing linear intermediary effect cannot accurately reveal the action path of intermediary variables between independent variables and dependent variables. Therefore, this paper uses the regulation path analysis method proposed by Edwards & Lambert to test [59]. First, the intermediary variable as the dependent variable was regressed, then use formal environmental regulations to regress the intermediary variable in Equation (4). Then intermediate variable grain planting production agglomeration and the interaction term between the independent variable formal environmental regulation and the intermediate variable are introduced into the new equation. In order to mitigate the errors caused by the potential multicollinearity problems caused by interaction items, the two variables of formal environmental regulations and production agglomeration are centralized. The specific form is shown in Equation (5):

$$M_{\mathit{it}}={\beta }_0+{\beta }_1{\mathit{fen}}_{\mathit{it}}+{\beta }_2{\mathit{fen}}_{\mathit{it}}^2+{\beta }_{3}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(4)

$${\mathit{CE}}_{\mathit{it}}={\gamma }_0+{\gamma }_1{\mathit{fen}}_{\mathit{it}}+{\gamma }_2{\mathit{fen}}_{\mathit{it}}^2+{\gamma }_3{M}_{\mathit{it}}+{\gamma }_4{\mathit{fen}}_{it}\times M_{\mathit{it}}+{\gamma }_{5}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(5)

Based on Hypothesis 5, informal environmental regulations may also indirectly affect the grain production industry’s carbon emission efficiency, the same as formal environmental regulations. Referring to the method of Baron and Kenny [26], we used the traditional three-step test method for regression. Firstly, regress the intermediary variable is the dependent variable, then informal environmental regulations are used to regress the mediating variable. Finally, take the mediating variable into model (2) as a control variable for regressing. The specific forms are shown in Equations (6) and (7).

$$M_{\mathit{it}}={\beta }_0+{\beta }_1{\mathit{inen}}_{\mathit{it}}+{\beta }_{2}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(6)

$${\mathit{CE}}_{\mathit{it}}= {\gamma }_0+{\gamma }_1{\mathit{inen}}_{\mathit{it}}+{\gamma }_2{M}_{\mathit{it}}+{\gamma }_{3}contro{l}_{it}{+u}_i{+\nu }_t{+\epsilon }_{\mathit{it}}$$

(7)

In Equations (4)–(7), M_it indicates the mediating variable; other explanatory and control variables are the same as in Equations (1) and (2); The intermediary effect share is calculated using “β₁ × γ₁/α₁“ and “β₂ × γ₂/α₂”.

3.2. Variable Definition

3.2.1. Explained Variables: Carbon Emissions Efficiency (CE)

The grain production industry has both “desired outputs” and “undesired outputs.” The traditional DEA-CCR model cannot consider the effect of “slack variables” and “undesired outputs” on efficiency, which leads to biased results. Therefore, it is more reasonable to use the DEA-SBM model proposed by Tone when considering the efficiency of carbon emissions of the grain production industry [60,61]. Therefore, we use the DEA-SBM model to calculate the grain production industry’s carbon emission efficiency. It is assumed that there is n decision-making units (DMUs) in the grain growing process, all with n input indicators x, s₁ desired output indicators y_g, and s₂ non-desired output indicators y_b. The definition matrix:

$$\begin{gathered} {X=(x}_1{,x}_2{,\dots ,x}_n)ϵR^{(m\times n)} \\ {Y}^g{=(y}_1^g{,y}_2^g{,\dots ,y}_n^g)ϵR^{{(s}_1\times n)} \\ {Y}^b{=(y}_1^b{,y}_2^b{,\dots ,y}_n^b)ϵR^{{(s}_2\times n)} \end{gathered}$$

(8)

In Equation (8),$XϵR_{+}^n,yϵR_{+}^m,zϵR_{+}^k$.

Then DEA-SBM model can be expressed as:

$$\begin{gathered} {\rho }^{*}=\min \frac{1-\frac{1}{m}{{\displaystyle \sum }}_{i=1}^m\frac{\overline{s_i}}{x_{i0}}}{1+\frac{1}{s_1{+s}_2({{\displaystyle \sum }}_{i=1}^{s_1}\frac{s_i^g}{y_{i0}^g+{{\displaystyle \sum }}_{i=1}^{s_2}\frac{s_i^b}{y_{i0}^b}})}} \\ s.t{.x}_0 {=X\lambda +S}^{-} \\ {y}_0^g {=Y}^g{\lambda -s}^g \\ {y}_0^b {=Y}^b{\lambda +s}^b \end{gathered}$$

(9)

In Equation (9), λ ≥ 0, s⁻ ≥ 0, s^g ≥ 0, s^b ≥ 0. s⁻, s^g, and s^b represent the input, desired output, and slack variables of non-desired output in the grain production industry, respectively. λ is the weight variable, and ρ* is the objective function value, which takes values in the range of [0,1]. When ρ* = 1, s⁻, s^g, s^b are all 0, which indicates that the decision unit is valid, and when ρ* < 1, s⁻, s^g, s^b are not all 0, which indicates that the decision unit is invalid at this time.

The carbon emission and efficiency level of the grain production industry is calculated by using the existing research methods, and considering the availability of data, the following six input-output variables are selected.

Expected output factors: referring to the research of Tian [62,63,64], the total grain output y^g is taken as the expected output of the DEA-SBM model, and the output is the annual grain output of each province.

Input factors: Input variables select four basic factors of the grain production industry, including labor, fertilizer, land, and machinery. The land input factor x₁ selects the grain planting area of each province (unit: khm²); Referring to Chen and Yao [65,66], the labor input factor x₂ separates the number of employees in the first industry from the weight A (unit: 10,000 people). Wherein weight A = (gross agricultural output value/total value of agricultural, forestry, husbandry, and fishing) × (grain sown area/crop sown area); Fertilizer input element x₃ is expressed by the amount of agricultural fertilizer applied for grain production (unit: 10,000 t), which is separated from the total amount of fertilizer applied for agricultural production by weight B. Wherein, weight B = grain sown area/total crop sown area; The machinery input factor x₄ is calculated by separating the total power of agricultural machinery in each province from the weight B.

Unexpected output factors: select the carbon emissions generated in the process of grain production as the unexpected output factor y^b (unit: kg/hm²) and the carbon emission measurement with carbon emission coefficient. The calculation method is as follows:

$$E={\displaystyle \sum }{E}_i={\displaystyle \sum }{T}_i{*\delta }_i$$

(10)

In Equation (10), E refers to the carbon emissions of grain production, E_i indicates the carbon emissions of the i-th carbon source in the grain production industry, and T_i expresses the use of the i-th carbon source of the grain production industry, δ_i denotes the carbon emission coefficient of type i carbon source. Dividing the main carbon sources of the grain production industry into six parts: chemical fertilizer, pesticide, agricultural film, agricultural diesel, agricultural sowing, and agricultural irrigation. Among them, the input elements of various carbon sources for the grain production industry are separated by weight B according to agricultural input elements. According to the practices recommended in the ‘Guidelines for the Preparation of Provincial Greenhouse Gas Inventories in 2011′ by the China Development and Reform Commission issued by IPCC (2006, 2007) and the research results of Li and Wu [67,68,69].

Table 1. Carbon source emission coefficient of the grain production industry.

Carbon Source	Carbon Emission Coefficient	Reference Source
Chemical Fertilizer	0.8956 kg·kg−1	Oak Ridge National Laboratory
Pesticides	4.9341 kg·kg−1	Oak Ridge National Laboratory
Agricultural Film	5.1800 kg·kg−1	IPCC
Agricultural Diesel	0.5927 kg·kg−1	Institute of Resource, Ecosystem and Environment of Agriculture, Nanjing Agricultural University
Agricultural irrigation	25 kg·ha	Dubey
Agricultural sowing	3.1260 kg·ha	College of Biology and Technology, China Agricultural University

shows various carbon sources’ carbon emission coefficients.

3.2.2. Core Explanatory Variables: Environmental Regulations

At present, there is no uniform standard in academia regarding the measurement of environmental regulations in agriculture. Existing research mainly adopts the prices of chemical fertilizer, total investment in environmental pollution control, and the number of environmental policies implemented at the end of the year to indicate formal environmental regulations in agriculture. However, the above variables can only be obtained up to 2017 in China, which is sort of timeliness. Therefore, according to previous studies and the issue in the number of years for which data are available, the level of economic development GDP by modified with adjustment coefficients is selected as the variable of formal environmental regulations (fen_it) [70]. The reason is that the level of local economic development can represent, to a certain extent, the investment in environmental governance of provincial governments, the importance of emphasis on environmental protection, and the demand for a high-quality environment. The higher the degree, the stronger the formal environmental regulations and supervision. In addition, the adjustment coefficient is adjusted using the inverse of the distance from the regional center of each province to the interior of the border. With the same intensity of informal environmental regulations, the larger the regional area, the larger the area needed to be covered by the relevant formal environmental regulations’ degree of provinces, the further away from the central region, the lower the enforcement intensity. Grain is generally planted in rural areas, far away from the central regions of each province, and the average intensity of informal environmental regulations will be lower. Therefore, the product of the two is used to better reflect the strength of formal environmental regulations in each province. The specific calculation method is as follows:

$$\mathit{fen}=\mathrm{GDP}\times \frac{1}{\frac{2}{3}\times \sqrt{\frac{\mathit{area}}{\pi }}}$$

(11)

In Equation (11), GDP is the economic development level of each province, the area is the provincial area of each province, and π is the circumference of the circle.

With the continuous development of the Internet, the mass media has played an extraordinary role in environmental risks, awareness, and solutions. According to the 2020 Statistical Report on the Development of the Internet in China, the utilization rate of Internet users who record public search behavior has reached 82.50% in China. The public’s concern for the environment is consistent with environmental protection action. Some scholars used Google Trends and Google Search to measure public concern about environmental issues [71,72]. However, Google withdrew from mainland China in April 2010. Since then, the public’s use of Google has declined, which will affect the consistency of the index. As the largest Chinese search engine in China at this stage, Baidu has the advantages of wide coverage and high data availability, which can make a detailed analysis of the data of different provinces in China according to the search frequency and location statistics [73]. Baidu Index is a product launched by Baidu in 2011. It is based on the search volume of Internet users in the Baidu engine in various regions and takes keywords as the statistical object. When Internet users use the Baidu search engine to search for environmental information, the background will use scientific algorithms to calculate the weight of a keyword in the search frequency of Baidu pages in a certain period of time. Therefore, Baidu Index can better reflect the public’s concern for keywords whose main advantage is that it can represent the relative popularity of keywords in designated areas without being affected by Internet users or total search volume. According to Yang [74], the informal environmental regulations variable (inen_it) uses the Baidu index with “environmental pollution” as the keyword to measure the degree of public concern about pollution in various regions. According to Core [75,76], the proportion of each enterprise’s output value in the total industrial output value is combined with the public’s concern about the environmental pollution of a certain area to measure the public’s concern about the environmental pollution of different enterprises in a certain area. So, using the idea for reference and taking advantage of the weight A and the proportion of the first industry in GDP to modify it as the informal environmental regulations variable of the grain production industry. The specific calculation method is as follows:

$$\mathit{inen}=\mathit{baidu}\times \mathrm{A}\times \frac{\text{GDP} \text{of} \text{the} \text{first} \text{industry}}{\text{GDP}}$$

(12)

In Equation (12), baidu refers to the public’s concern about environmental pollution in each province, weight A indicates the proportion of grain production industry in the total agricultural output value, and GDP represents the economic development level of each province.

3.2.3. Threshold Variable

There is one threshold variable in this paper: the informal environmental regulations(fen_it). The variable is the same as those in (11).

3.2.4. Intermediary Variable

There is one intermediary variable in this paper: the production agglomeration level of the grain production industry (jj_it). According to the production characteristics of grain crops and the availability of data, the location entropy index that can represent the concentration and specialization of grain crops in each province is used to evaluate the production agglomeration level of the grain production industry. According to the previous research methods [77], the calculation formula is as follows:

$$j{j}_{it}=\frac{E_{it}/E}{A_{it}/A}$$

(13)

In Equation (13), jj_it represents the locational entropy index of grain crops in a province; E_i is the grain crops’ sown area in i province, E is the crops’ sown area in each province, A_i is the sown area of grain crops in China, and A is the crops’ sown area in China. It indicates that the production agglomeration level of the grain production industry in this province is higher than the national average level if jj_it is greater than 1, and it shows that the production agglomeration level of the grain production industry in this province is equal to the national average level if jj_it is equal to 1; it expresses that the level of production agglomeration of the grain production industry in this province is less than the national average if jj_it is less than 1.

3.2.5. Control Variables

So as to analyze the impact of heterogeneous environmental regulations on the grain production industry’s carbon emission efficiency, different variables are controlled as follows: (1) Urbanization rate (czhl_it), measured by (1—rural population/population at the end of the year); (2) Electricity consumption (yd_it), since it is not directly recorded in the statistical yearbook, the rural electricity consumption is used as the weight A for separation. (3) Per capita planting industry output value (rnc_it), measured by the ratio of planting industry output value to agricultural population; (4) Institutional factors (zd_it), measured by the proportion of agricultural, forestry, and water expenditure in local financial support for agriculture. (5) Income factor (sr_it), measured by the per capita income level of rural residents.

3.3. Data Sources and Descriptive Statistics

During the “Twelfth Five Year Plan” period (2011–2015), “ecological civilization construction” was proposed in the 18th National Congress of the Communist Party of China vigorously for the first time. Since environmental regulations can be used as one of the effective means of ecological civilization governance, the research period of this paper starts in 2011. In view of the fact that the formal environmental regulations variables in the following text cannot obtain data for 2020 (The administrative land area of 30 provinces in 2020 is not given in the ‘China Urban Statistical Yearbook 2021′), thus take 2011–2019 as this paper’s sample period, and 30 provinces’ (excluding Xizang, Hong Kong, Macao, and Taiwan) data are selected for the study. We obtained relevant data in this study from the ‘China Statistical Yearbook (2012–2020)’, ‘China Rural Statistical Yearbook (2012–2020)’, ‘China Urban Statistical Yearbook (2012–2020)’, ‘China Population and Employment Statistical Yearbook (2012–2020)’, Baidu index obtained from ‘Baidu search engine.’ We used linear interpolation to fill in the individual missing values while each explanatory variable was logarithmically processed. We have carried out centralized trend analysis, dispersion analysis, and analysis on whether the distribution of the population belongs to normal distribution. The descriptive statistics of the main variables are as follows in

Table 2. Descriptive statistical analysis of main variables.

Variable	Average	Standard Deviation	Maximum	Minimum
Carbon emission efficiency	0.696	0.191	1.000	0.383
Formal environmental regulations	214.109	204.984	1273.921	20.691
Informal environmental regulations	3.308	2.116	13.305	0.079
Production agglomeration level	0.943	0.200	1.405	0.511
Electricity consumption	291.485	421.100	1949.100	4.100
Per capita agricultural output value	4.788	2.111	13.728	1.014
Income level of rural residents	12,380.890	8030.438	112,950.400	3909.400
Financial support for agriculture	0.114	0.032	0.190	0.041
Urbanization rate	57.582	12.322	89.785	36.091

.

4. Empirical Result

4.1. Analysis of Carbon Emission Efficiency Results

The grain production industry’s carbon emission efficiency in China was measured using MAX-DEA software, and the results are shown in Figure 2. From 2011 to 2019, China’s grain production industry’s carbon emission efficiency showed an overall upward trend, which is inseparable from the concept of green development advocated for many years in China. Further dividing the whole research period into two stages (the “Twelfth Five Year Plan” period and the “13th Five Year Plan” period), it can be found that during the “Twelfth Five Year Plan” period (2011–2015), the Ministry of Agriculture issued the “Implementation Plan for the Comprehensive Utilization of Crop Straw “and other laws and regulations, which have achieved remarkable results, significantly improving grain production industry’s carbon emission efficiency. In the “13th Five Year Plan” period (from 2016 to 2019, in view of the unavailability of formal environmental regulation data in 2020, only the year 2019 will be discussed), the “13th Five Year Plan Work Plan for Controlling Greenhouse Gas Emissions” emphasized that a series of measures, such as vigorously developing low-carbon agriculture and promoting straw returning to farmland will help to further reduce the carbon emission content in the process of grain production. Compared with the “Twelfth Five Year Plan” period, the overall grain production industry’s carbon emission efficiency has been improved to a new stage. However, overall, the current average carbon emission efficiency of the national grain production industry is 0.696, and there is still much room for improvement.

Furthermore, measuring the production carbon emission efficiency of China’s four major grain crops (rice, corn, wheat, and soybean; the planting area of these four crops accounts for 80% of the total food crops) from 2011 to 2019 and the results are also shown in Figure 2. It can be seen from the figures that in recent years, the carbon emission efficiency of rice, wheat, and soybean has had an upward trend, but the carbon emission efficiency of corn has decreased year by year. In 2017–2019, the carbon emission efficiency of corn planting was the lowest and far lower than the other three staple crops.

Dividing the 30 provinces according to the three major economic zones. The results at the regional level show that the carbon emission efficiency of grain production shows significant differences between regions. The highest carbon emission efficiency of grain production (0.806) is in the central region, followed by the eastern region (0.653) and the western region (0.644) is the lowest. The main reason is that the current grain production industry is mainly concentrated in the central region, and the focus of R&D tends to be on the agricultural sector, while the economic development of eastern provinces is dominated by the tertiary industry, which does not pay much attention to carbon emissions of the grain production industry. In addition, due to the relatively backward grain production conditions and green production technologies, the western region lacks the construction of ecological environment infrastructure, leading to the lowest grain production industry’s carbon emission efficiency.

At the provincial level, as shown in Figure 3, less than half (13) of the 30 provinces are above the overall carbon emission efficiency level of the grain production industry. Among them, Shandong, Anhui, Hebei, Yunnan, and other major grain-producing provinces are below the average carbon emission efficiency of the national grain production industry. Therefore, we need to pay attention to the balance between grain yield and grain production environmental protection.

At the same time, the change in carbon emission efficiency among different provinces of the four major grain crops was explored again, and the results in 2011 and 2019 were plotted as shown in Figure 4. Jiangxi, Hunan, and Anhui are the main rice-producing areas, but their carbon emission efficiency has been low level in recent years. Tianjin, Guizhou, Zhejiang, and Gansu provinces have improved their rice carbon emission efficiency. Heilongjiang, and Neimenggu, two major corn-producing areas, have maintained a high carbon emission efficiency level for many years, while Shandong, Henan, and Gansu need to further improve their corn carbon emission efficiency levels. Even to the extent that Hebei, Henan, Shandong, Gansu, etc., as major corn-producing areas, have shown a decline in carbon emissions. Overall, the carbon emission efficiency of wheat production has an upward trend in different provinces. Among them, Heilongjiang and Neimenggu are the highest, while coastal provinces are lower. The carbon emission efficiency of soybean production also has an upward trend in different provinces. Heilongjiang and Neimenggu are the highest. However, Jilin, as the main soybean-producing area, has shown a downward trend. As the main soybean-producing areas, Anhui and Henan also need to further improve the carbon emission efficiency of soybean production.

4.2. Benchmark Regression

In order to reduce the problems of autocorrelation and heteroscedasticity in the model, the “xtscc,fe” command was used to correct the impact of heteroscedasticity and autocorrelation on the regression results [60]. The software Stata/SE 15.1 is used for the measurement and regression of Equations (1) and (2). The detailed results of benchmark regressions of the impact of formal environmental regulations and informal environmental regulations on the carbon emission efficiency of China’s grain production industry are as follow in

Table 3. Benchmark regression results.

Variables	(1)	(1)	(1)	(2)	(2)	(2)
fen	−0.210 ***	−0.121 **	−0.226 ***
	(0.022)	(0.060)	(0.030)
fen2	0.038 ***	0.011	0.029 **
	(0.010)	(0.017)	(0.012)
inen				0.034 **	0.076 **	0.075 ***
				(0.012)	(0.038)	(0.018)
yd	−0.059	−0.007	−0.059	−0.084 *	−0.038 *	−0.076 *
	(0.047)	(0.026)	(0.041)	(0.043)	(0.022)	(0.036)
rnc	0.118 **	0.083 **	0.076 **	0.062	0.020	0.004
	(0.047)	(0.039)	(0.024)	(0.038)	(0.031)	(0.021)
sr	0.049 *	0.045 ***	−0.040 ***	0.024	0.047 ***	0.032 **
	(0.023)	(0.011)	(0.003)	(0.015)	(0.012)	(0.010)
zd	0.035	−0.057	−0.027	0.005	−0.062	−0.051
	(0.055)	(0.053)	(0.043)	(0.048)	(0.050)	(0.029)
czhl	0.651 ***	0.259	0.589 ***	0.232 *	0.195	0.202 **
	(0.166)	(0.193)	(0.170)	(0.124)	(0.147)	(0.082)
Constant term	−2.229 **	−1.003	−1.985 ***	−0.163	−0.562	−0.235
	(0.782)	(0.827)	(0.592)	(0.654)	(0.586)	(0.265)
Province FE	YES		YES	YES		YES
Year FE		YES	YES		YES	YES
R-squared	0.2318	0.3227	0.3563	0.1974	0.3226	0.3321
F-value	300.95	171.19	458.81	99.89	262.52	125.46

Note: *, **, *** are significant at 10%, 5%, and 1% levels, respectively. The following table is the same. (1) and (2) respectively represent the regression results of provincial fixed effect, time fixed effect and two-way fixed effect in Equations (1) and (2).

.

As shown in (1), the low level of formal environmental regulations has a significant negative influence on the grain production industry’s carbon emission efficiency, while the high level of formal environmental regulations can significantly improve the grain production industry’s carbon emission efficiency. The relationship between the two is “U” shaped, which confirms Hypothesis 1. In (2), informal environmental regulations are significant at the level of 1%, expressing that informal environmental regulations have a certain role in promoting the improvement of the grain production industry’s carbon emission efficiency, which confirms Hypothesis 2. The relationship between them is verified.

4.3. Robustness Test

We used the method of replacing variables to regress Equations (1) and (2) again for testing the robustness of the above benchmark regression results. According to previous research methods, we replaced formal environmental regulations with the proportion of total investment in environmental governance in GDP [78]. Since only the data from 2005 to 2017 can be obtained in the ‘China Environmental Statistics Yearbook,’ we selected the above available data for regression, and also expanded the sample size. To some extent, it solves the possible time lag of formal environmental regulations policies. In addition, the ‘China Environmental Statistics Yearbook’ have begun to count the agricultural “environmental pollution,” “water pollution,” “air pollution,” “solid waste,” and other pollution emissions in agriculture since the Twelfth Five Year Plan period, which also proved that the agricultural environmental pollution problem in recent years should not be underestimated. In terms of information disclosure, the mass media plays an extraordinary role in environmental risks, awareness, and solutions. The main forms of mass media mainly include newspapers and the Internet, and the above regress has used the “Baidu Index” to express the role of the Internet. Therefore, according to the research methods of previous scholars, we used the most influential newspapers in the local area of each province (mainly the regional daily newspaper), and the keywords of “environmental pollution,” “water pollution,” “air pollution,” and “solid waste” are used to search respectively [79,80]. The number of news reports obtained is summed up to show the informal environmental regulations variables. The inspection results are shown in

Table 4. Robustness test.

Variables	(1)	(2)
fen	−0.006
	(0.008)
fen2	0.034 **
	(0.014)
inen		0.013 *
		(0.007)
yd	0.014	−0.086 *
	(0.019)	(0.042)
rnc	−0.023	0.032 *
	(0.032)	(0.016)
sr	−0.003	0.022 *
	(0.010)	(0.011)
zd	0.009	−0.038
	(0.031)	(0.032)
czhl	−0.024	0.215 **
	(0.032)	(0.076)
Constant term	0.775 ***	−0.121
	(0.086)	(0.258)
Province FE	YES	YES
Year FE	YES	YES
R-squared	0.2361	0.3266
F-value	5.17	954.52

.

From the test results in the above table, the core explanatory variables are basically consistent with the results of benchmark regression. It proves that the relationship between formal environmental regulations and the grain production industry’s carbon emission efficiency is “U” shaped, and informal environmental regulations can positively promote the grain production industry’s carbon emission efficiency again.

4.4. Regional Heterogeneity

As the natural environmental resources and economic levels of different regions in China have distinctions, in order to verify whether the empirical results differ across regions, according to China’s economic development level and policy implementation effect, the study area is divided into the eastern region and the central and western regions. The economic development level of the eastern region is high, but the economic drive of the grain planting industry is not obvious; The economic development level of the central and western regions is relatively low, and they are the main grain-producing regions. The grain planting industry has led the economy significantly. Moreover, China’s grain policy has played a better role in the central and western regions because grain production is mainly in the central and western regions, such as subsidy policy for grain production. Then discussing the regional heterogeneity impact of heterogeneous environmental regulations on the grain production industry’s carbon emission efficiency. The inspection results are shown in

Table 5. Heterogeneity regression results.

Variables	Eastern Regions		Central and Western Regions
	(1)	(2)	(1)	(2)
fen	−0.343 ***		−0.273 ***
	(0.078)		(0.069)
fen2	0.073 **		0.034 *
	(0.028)		(0.016)
inen		0.133 ***		−0.021
		(0.027)		(0.045)
yd	−0.083 **	−0.087 **	0.015	−0.042 **
	(0.034)	(0.027)	(0.017)	(0.016)
rnc	0.017	−0.062 *	0.084	0.083 **
	(0.044)	(0.029)	(0.054)	(0.029)
sr	0.497 ***	0.268 **	0.038 ***	0.018
	(0.083)	(0.113)	(0.010)	(0.020)
zd	−0.142 *	−0.143 **	−0.021	0.001
	(0.064)	(0.059)	(0.049)	(0.051)
czhl	0.137	−0.318 *	0.820 ***	0.195
	(0.277)	(0.140)	(0.179)	(0.117)
Constant term	−4.334 ***	−0.342	−3.227 ***	−0.092
	(1.154)	(0.636)	(0.604)	(0.428)
Province FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
R-squared	0.4062	0.4079	0.3969	0.3352
F-value	178.03	101.94	81.79	58.46

.

It can be seen from Table 5 that formal environmental regulations have a “U” impact on the carbon emission efficiency of the grain planting industry, which is still established in both the eastern and the central and western regions, but different from the national results, informal environmental regulation is significant in the eastern region, while it is not significant in the central and western regions. At the same time, the current economic level of the central and western regions is lower than that of the eastern regions. The public still has an awareness of sacrificing the environment for economic improvement. The lack of formal government regulations cannot significantly improve the carbon emission efficiency of the grain planting industry.

5. Further Analysis

5.1. Threshold Mechanism Test

According to the above analysis, the formal environmental regulations maybe have here a “U” shaped nonlinear impact on the grain production industry’s carbon emission efficiency. Then, to what extent can formal environmental regulations be implemented to effectively promote the grain production industry’s carbon emission efficiency? In order to control the optimal degree of informal environmental regulations implementation, we further used Hansen’s panel for analyzing the nonlinear effect.

In order to test whether there is a threshold effect of formal environmental regulations on the grain production industry’s carbon emission efficiency, this paper takes formal environmental regulations as a threshold-dependent variable and uses the minimum residual sum of squares as the limiting condition to determine the threshold value. By using the bootstrap iterative sampling method to set multiple thresholds for estimating the thresholds of formal environmental regulations level, test results are shown in

Table 6. Threshold estimates and test results.

Variables	Threshold	F-value	p-Value	Critical Value			Threshold Values	Confidence Interval
fen	Single threshold	11.69 *	0.072	14.5640	17.2429	25.0750	6.5987	[6.5229,6.6288]

.

It can be seen from Table 6 that there is a single threshold effect between formal environmental regulations and the grain production industry’s carbon emission efficiency. The threshold value is e^6.5987 and falls within the interval of [e^6.5229, e^6.6288] at a 95% confidence level, indicating that the threshold value has passed the valid test.

As can be seen from

Table 7. Threshold model regression results.

Variables	(3)
fen × I(fen ≤ e6.5987)	−0.180 ***
	(0.055)
fen × I(fen > e6.5987)	−0.158 **
	(0.054)
yd	−0.094 ***
	(0.028)
rnc	0.131 ***
	(0.039)
sr	0.046 *
	(0.028)
zd	0.019
	(0.049)
czhl	0.531 ***
	(0.177)
Constant term	−0.670
	(0.593)
Province FE	YES
Year FE	YES
R-squared	0.2543
F-value	11.35

, the significance levels and coefficient directions of core explanatory variables and control variables are basically consistent with the fixed-effects model previously estimated, which indicates that the model estimation results are robust. When the level of formal environmental regulations is less than e^6.5987, the impact coefficient of formal environmental regulations on the grain production industry’s carbon emission efficiency is −0.180; When the formal environmental regulations level is greater than e^6.5987, the coefficient value changes to −0.158. This shows that the influence of formal environmental regulations on the grain production industry’s carbon emission efficiency has been negative in the research interval. However, this negative impact is weakening, which means that the “innovation compensation effect” brought by the current formal environmental regulations has partially offset the “cost effect,” but it has not crossed the “U” inflection point yet, which cannot bring significant “innovation compensation effect” to improving grain production industry’s carbon emission efficiency effectively.

The negative impact of formal environmental regulations on the carbon emission efficiency of the grain planting industry gradually weakens, indicating that the “cost effect” of formal environmental regulation on the grain production industry’s carbon emission efficiency is greater than the “innovation compensation effect” in recent years, which also shows that the government has invested a lot of pollution control costs in controlling carbon emission pollution and its determination to promote carbon emission reduction. In addition, although the “innovation compensation” effect of formal environmental regulations is small, it has not offset the negative effects of pollution control costs. However, the negative impact on the grain production industry’s carbon emission efficiency is diminishing with the enhancement of the intensity of formal environmental regulations, suggesting that the “innovation compensation” effect is increasing.

In some cases, informal environmental regulations fill the blank of formal environmental regulations in terms of controlling the grain production industry’s carbon emission efficiency. Sometimes the formal environmental regulations guided by the government lack the knowledge of farmers’ cognitive level when implementing the green production behavior policy, which makes it difficult for farmers to understand the true connotation of green production. In addition, the unilateral promotion of green production practices lacks interaction and communication, which makes it difficult for farmers to obtain information about its effectiveness and income increase. In contrast, informal environmental regulations in terms of forms through media and other means avoid the complexity of formal environmental regulations in monotony language. With the help of easy-to-understand language, which publicizes the good evaluation of low-carbon production behavior, and through the herd mentality and interaction of farmers, farmers can clarify low-carbon production behavior and improve the grain production industry’s carbon emission efficiency. Although informal environmental regulations have played a direct positive and significant role in the grain production industry’s carbon emission efficiency, they are weak regulations. Therefore, formal environmental regulations are also required to complement, and farmers are urged to carry out low-carbon production with the force of laws and regulations so as to make up for the defects of informal environmental regulations. Therefore, formal environmental regulations and informal environmental regulations complement each other to improve the grain production industry’s carbon emission efficiency, and both are indispensable.

5.2. Intermediary Mechanism Test

Models (4)–(7) test the mediating effects of production agglomeration. According to the theoretical analysis, next, we will further test whether heterogeneous environmental regulations can indirectly promote the improvement of the grain production industry’s carbon emission efficiency through the path of the grain planting industry production agglomeration level. The inspection results are shown in

Table 8. Intermediary effect regression results.

Variables	(4)	(5)	(6)	(7)
	jj	CE	jj	CE
fen	−0.165 ***	−0.155 ***
	(0.028)	(0.021)
fen2	0.023 ***	0.042 **
	(0.002)	(0.015)
jj		0.084 **		0.350 ***
		(0.035)		(0.093)
fen*jj		0.323 ***
		(0.079)
inen			0.135 ***	0.028
			(0.015)	(0.024)
yd	−0.070 ***	−0.024	−0.071 ***	−0.052
	(0.019)	(0.037)	(0.015)	(0.032)
rnc	−0.038	0.088 ***	−0.130 ***	0.049 **
	(0.023)	(0.019)	(0.014)	(0.016)
sr	0.009	0.033 ***	0.014	0.027 *
	(0.012)	(0.007)	(0.008)	(0.013)
zd	−0.043	0.000	−0.083 **	−0.022
	(0.028)	(0.042)	(0.032)	(0.029)
czhl	0.225 *	0.398 ***	−0.077	0.229 **
	(0.104)	(0.111)	(0.105)	(0.090)
Constant term	−0.841 ***	−1.257 **	0.317	−0.346
	(0.227)	(0.479)	(0.287)	(0.321)
Province FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES
R-squared	0.3516	0.4111	0.4629	0.3621
F-value	8.13	320.34	12.03	8.51

.

It can be seen from Table 8 in Equations (5) and (7) the increase in the production agglomeration level of the grain production industry has a significant effect on improving the grain production industry’s carbon emission efficiency, which confirms Hypothesis 3.

In Equations (4), formal environmental regulations inhibit the production agglomeration level of the grain planting industry at a low level, but the production agglomeration level of the grain planting industry can be significantly improved with the strengthening of formal environmental regulations. As Hypothesis 4: the relationship between the two is “U” shaped. In Equations (6), informal environmental regulations are beneficial to the improvement of the production agglomeration level of the grain planting industry to confirm Hypothesis 5. In general, the production agglomeration level of the grain planting industry plays an intermediary role in the effect of heterogeneous environmental regulations on the carbon emission efficiency of grain production. Among them, there is a “U” type intermediary effect in the impact of formal environmental regulations on the grain production industry’s carbon emission efficiency, and the intermediary effect accounts for 22.56% and 24.51% of the total effect, respectively. Informal environmental regulations can promote the production agglomeration level of the grain planting industry to improve the grain production industry’s carbon emission efficiency, and this path accounts for 63% of the total effect.

In the long run, safeguarding national grain security is not only reflected in the security of “quantity” and “quality,” but also needs to focus on many aspects of grain production, such as “resource security” and “ecological security.” Reasonable environmental regulations have played important roles in improving the production agglomeration level of the grain planting industry and thus improving the carbon emission efficiency of the grain planting industry: on the one hand, reasonable environmental regulations have promoted improvement of, the production agglomeration level of grain planting industry, spawned the scale effect, reduced production costs, improved the utilization rate of green planting technology of the grain production industry, and thus improved grain production industry’s carbon emission efficiency; On the other hand, the centralized management of the grain production industry and the utilization of production facilities have promoted the pollution control of the grain production industry. In addition, the production agglomeration of grain planting will also have a division effect on carbon emission efficiency. This division of labor effect can not only improve the consistency of grain crop planting but also facilitate service outsourcing so as to achieve carbon emission reduction; each link of grain planting can be refined to improve the specialization of pesticide and fertilizer application, reduce unnecessary waste of production factors to improve the production environment.

In recent years, the public’s attention to environmental issues has obviously increased. Take the Baidu index searched by the keyword “environmental pollution” in 2011, 2017, and 2019 as an example: in 2011, the Baidu index was 68.90, while in 2017 and 2019, it was 126.42 and 105.58, respectively, nearly twice that of 2011. The sown area of grain increased from 110573 thousand hectares in 2011 to 117989 thousand hectares in 2017 and then gradually decreased to 116063.6 thousand hectares in 2019. To some extent, the grain sown area reflects the level of production agglomeration, which represents the degree of specialization on China’s limited land. The changing trend of public concern is the same as that of grain production agglomeration, indicating that informal environmental regulation has a positive effect on the production agglomeration of the grain planting industry. For a long time, China’s large-scale agricultural operation has been the focus of rural economic system reform. Over the years, the continuous deepening of the “three rights division” reform of agricultural land has changed the nature of social association among farmers and has also laid a foundation for the exertion of informal environmental regulations, promoting farmers to continue the specialized division of labor and cooperative production, and realizing the scale operation of farmers’ grain production [81]. In the survey of Hubei Province, grain producers mainly engaged in agricultural income have a more positive response to informal environmental regulations. They expand their scale by engaging in concentrated production so as to better participate in green production and hope to obtain rewards and returns [82]; This is similar to Yami’s view: in rural Uganda, the formal institutions are weak in guiding and controlling the input market, while the informal institutions can promote farmers to share knowledge, take collective action and achieve labor force sharing, ultimately achieving the goal of Sustainable Crop Awareness [83].

According to the estimation results, the impact of control variables on the grain production industry’s carbon emission efficiency mainly includes the following five aspects. Firstly, the electricity consumption in the grain production process negatively inhibits the improvement of the grain production industry’s carbon emission efficiency at the level of 1%. The large consumption of electric energy has aggravated the carbon emission in the grain production industry, which is not conducive to improving the grain production industry’s carbon emission efficiency. Secondly, the improvement of the economic level in the planting industry can significantly improve the grain production industry’s carbon emission efficiency at the level of 1%. The improvement of the economic level is conducive to the upgrading of production technology of the grain production industry, improving the comprehensive production capacity of grain and making grain farming industrialized. The grain production industry’s carbon emission efficiency has been improved with the low-carbon and efficient production mode. Thirdly, the per capita income level of rural residents is significantly positive at the level of 10%. With higher income levels in rural areas, more citizens pay attention to the improvement of quality of life, so they are more sensitive to environmental issues. To a certain extent, it has played a supervisory role in the grain production industry and promoted the improvement of the carbon emission efficiency level of the grain production industry. Fourthly, local financial support for agriculture has a positive but insignificant effect on the level of the grain production industry’s carbon emission efficiency. The support of financial funds for the production of the grain planting industry is conducive to the optimization and upgrading of the production of the grain planting industry, but blindly investing will not reduce the production cost of the grain planting industry at the root. Farmers should be encouraged to carry out green innovation from the mode of production to improve the carbon emission efficiency of grain production. Finally, the urbanization level significantly promotes the improvement of the grain production industry’s carbon emission efficiency at the level of 1%. The improvement of urbanization level will further expand the scale of the city, which will lead to the reduction in the land scale of agriculture. Therefore, the grain production industry will be affected, leading to the reduction in carbon emissions, thus having a positive impact on improving the grain production industry’s carbon emission efficiency.

6. Conclusions and Suggestions

6.1. Conclusions

In the context of achieving “double carbon,” the implementation of reasonable environmental regulations are important means to promote low carbonization of grain planting and high-quality agricultural development. Therefore this paper takes the panel data of 30 provinces in China from 2011 to 2019 as the research sample, uses the SBM-DEA model to measure the carbon emission efficiency of China’s provincial grain production industry, analyzes the impact of formal environmental regulations and informal environmental regulations on the grain production industry’s carbon emission efficiency based on the two-way fixed effect model, and uses the panel threshold model and the intermediary effect model to further explore the mechanism. The research results show that: the carbon emission efficiency of China’s grain production industry continued to improve from 2011 to 2019, with the highest in the central regions, the second in the eastern regions, and the lowest in the western regions. There is a significant single-threshold effect of formal environmental regulations on the grain production industry’s carbon emission efficiency. The “innovation compensation effect” is gradually offsetting the negative impact of the “cost effective” with the enhancement of the intensity of formal environmental regulations. However, at present, only relying on formal environmental regulations cannot have a significant “innovation compensation effect” and cannot cross the “U” inflection point. Informal environmental regulations can improve the grain production industry’s carbon emission efficiency effectively, but it is not obvious in the central and western regions. The production agglomeration level of the grain planting industry has a “U” type intermediary role in the carbon emission efficiency of formal environmental regulations in the grain planting industry, with the intermediary effects accounting for 22.56% and 24.51%, respectively; Informal environmental regulations can also promote the level of production agglomeration of the grain planting industry to improve grain production industry’s carbon emission efficiency, and the proportion of this path is 63% respectively in the total effect. The electricity consumption for grain planting has a negative impact on the grain production industry’s carbon emission efficiency, while the per capita income level of rural residents, urbanization rate, and economic level of the planting industry has a positive impact on the grain production industry’s carbon emission efficiency.

6.2. Suggestions

Based on this, this paper puts forward the following policy recommendations.

First of all, improve the formal environmental regulations led by the government. Firstly, the government needs to formulate stricter and clearer environmental protection laws and regulations for the grain planting industry and establish a more scientific and complete environmental protection law enforcement performance evaluation system. Secondly, it is also necessary to strengthen the investment in pollution control of the grain planting industry and give certain subsidies to grain producers to control the environment. Finally, the carbon label should be pasted on the organic food produced when it is sold, and the selling price should be appropriately raised. In addition, the insurance support for planting grain should be strengthened, and the premium subsidy policy for grain planting insurance should be implemented to ensure the income of grain producers.

Next, attach importance to the benefits of informal environmental regulations to environmental governance. First of all, regular training on green production technology should be carried out in rural areas to improve the overall quality of farmers and make green food production a spontaneous awareness and initiative. Secondly, setting up a special media column to report on green grain production and broaden the public’s supervision and governance channels for low-carbon grain production. Finally, build high-quality grain planting bases in rural areas, such as standardized production bases for green grain raw materials and organic agricultural products.

Thirdly, improve the production agglomeration level of the grain planting industry. First of all, we should appropriately expand the grain sown area while stabilizing the grain sown area. Taking the sown area as a binding indicator, compact local responsibilities, and paying close attention to grain production. Secondly, we should strengthen the construction of high standard farmland, strengthen land protection, and improve infrastructure construction such as irrigation projects; Standardize regional management and service, and provide a good external environment for the production equipment of grain planting industry. Thirdly, strengthening investment in agricultural science and technology and striving to improve the integrated promotion and application of high-quality and high-yield grain varieties, fertilizer and drug dual control technology, mechanized sowing, and other technologies. Finally, optimize the grain quality structure, carry out the “three products and one standard” action in the grain planting industry, further promote high-quality special varieties of grain, and expand the planting area of special corn and high oil and high protein soybeans.

Fourthly, improve the level of agricultural scientific and technological innovation. First of all, we should formulate a reasonable agricultural technology innovation plan, agricultural technology innovation projects, financial subsidies, and talent incentives to encourage continuous innovation in agricultural science and technology. Secondly, agricultural production and operation entities should strengthen information exchange to achieve mutual benefits and win-win results. Finally, increasing investment in green innovative agricultural technologies to ensure the development of green grain planting and the organic grain industry.

Finally, vigorously develop green and organic grain planting. First of all, we will promote standardized grain production, start the pilot work of standardization of the whole industrial chain of modern agriculture, and accelerate the establishment of an agricultural standard system. Secondly, create a green brand of grain. Actively promote the creation of green, organic, and geographical indication grains, and establish green and high-quality grain brands. Finally, build a grain quality and safety traceability platform. Establish a grain quality traceability system to make the source and flow of grain products clear, the grain production information can be queried, the production responsibility can be investigated, and the grain safety of residents can be guaranteed.

7. Insufficient

The carbon emission efficiency of the grain planting industry is an important indicator to measure the low carbon and high yield of grain, which is conducive to maintaining sustainable grain security and the development of organic agriculture around the world. There are still some problems worthy of further discussion in the research: (1) This paper only analyzes the relationship between environmental regulations and the carbon emission efficiency of the whole grain production industry, and different kinds of crops may have different production conditions, so the follow-up research will further explore the differences between different crops. (2) It is important work to accurately measure the carbon emission efficiency of the grain planting industry. Grain crop production not only releases carbon emissions but also absorbs a certain amount of carbon emissions as a carbon sink. Therefore, it is of great significance to accurately measure the carbon emission content and carbon emission efficiency of grain crop production. (3) This paper only makes a preliminary study on the impact of China’s regional environmental regulations on the carbon emission efficiency of the grain planting industry. However, the grain production of different countries is not the same, so the differences between different countries will be explored later to provide a reference for maintaining global grain security and people’s happy life.