Impacts of Green Perception Benefits and Environmental Regulation Intensity on Farmers’ Agricultural Green Production Willingness: A New Perspective of Technology Acquisition

Abstract

Agricultural green production (AGP) is a key strategy for ensuring stable and sustainable grain production in developing countries. However, from the perspective of technology acquisition, research on farmers’ willingness to adopt AGP remains limited. Based on this, a survey was conducted on 862 households in major grain-producing counties in the Huang Huai Hai Plain of China with a reliable and effective response rate of 97.44%. The aim was to employ Probit and mediation models to empirically analyze the direct impacts of green perception benefits and environmental regulation intensity on farmers’ AGP willingness, and further examine the intrinsic mechanisms of technology acquisition. The results demonstrated that both green perception benefits and environmental regulation intensity significantly enhanced farmers’ willingness to engage in AGP, with green perception benefits having a greater influence. Among the two-dimensional variables, economic benefits had a stronger promoting effect than identity benefits, with a difference of 0.044 units, while subjective regulation intensity outperformed objective regulation intensity by 0.173 units. This suggested the need to strengthen the subjective impact of AGP policies in practice. Further analysis revealed that technology acquisition mediated 5.87% of the effect of green perception benefits on farmers’ AGP willingness, with acquisition evaluation having the greatest mediating effect, followed by acquisition quality and acquisition channels. However, although the overall environmental regulation intensity did not significantly impact farmers’ willingness to engage in AGP, its two-dimensional indicators played a mediating role to varying degrees. The findings in this study provide valuable empirical evidence for promoting AGP among grain producers, contributing to grain production security and the sustainable development of developing countries. Thus, implementing environmental regulatory policies tailored to local conditions, enhancing farmers’ economic awareness and sense of responsibility, and expanding farmers’ channels for technology acquisition are reasonable policy choices.

1. Introduction

The stability of the grain production system is related to major issues such as rational utilization of arable land resources, protection of ecosystems, and effective response to climate change [1,2,3]. The current grain production activities have brought many negative impacts, including greenhouse gas emissions, soil quality degradation, water eutrophication, and a sharp decline in biodiversity [4,5]. These issues have led to the instability of the global food supply. In addition, agricultural inputs, including land, seeds, fertilizers, water resources, and other factors, will directly affect the production capacity and price of grain, thereby affecting the public’s ability to obtain it [6,7]. Therefore, promoting the green and sustainable transformation of the grain production system has become a global consensus.

Many countries or organizations, such as the European Union, the United States, and Japan, have taken measures to address the above-mentioned issues. For example, the EU have allocated 10 hectares of land on farms for crop rotation to protect soil quality [8], while encouraging cooperatives to promote “whole farm” agricultural carbon reduction [9]; The United States has encouraged farmers to voluntarily retire (fallow) farmland that produces surplus agricultural products [10], and supported farmers in coping with extreme weather to potentially increase subsidies for regenerative agriculture [11]; Japan has conducted green production practices around systems such as soil pollution remediation and green technology certification [12], and has also promoted precision agriculture technologies such as drone fertilization and intelligent irrigation [13]. The Chinese government has also introduced a series of policies related to agricultural green production (AGP). In August 2021, six departments jointly formulated the “National Agricultural Green Development Plan for the 14th Five-Year Plan”, aiming to accelerate the comprehensive green transformation of agriculture and continuously improve the rural ecological environment. In December 2024, the Ministry of Agriculture and Rural Affairs issued the “Guiding Opinions on Accelerating the Comprehensive Green Transformation of Agricultural Development and Promoting Rural Ecological Revitalization”, which clarified 10 key tasks and technical measures including agricultural water conservation, input reduction and efficiency improvement, comprehensive utilization of straw, and organic waste disposal. In 2025, the No. 1 central document of the Central Government stressed that we should stabilize the sown area of grain, promote the integration of water and fertilizer, and promote large area yield per unit area. These policy actions provided solid resources and environmental guarantees for improving the practical level of AGP and promoting comprehensive rural revitalization.

In the academic community, many scholars have explored the willingness of farmers to engage in AGP. Malabayabas et al. (2023) pointed out that farmers with higher levels of education often have a higher sense of identification and willingness to practice environmental behaviors [14]. The accumulation of household capital, such as the expansion of land scale and the establishment of social networks [15,16], could enhance their acceptance of AGP technologies. There were also studies exploring the impact of attitudes, subjective norms, and perceived behavioral control on farmers’ willingness to use green pesticides, protective agriculture, and other psychological aspects [17,18]. In addition, government subsidies and production training could effectively enhance farmers’ ecological awareness, thereby promoting their active adoption of AGP technologies [19,20]. Policy constraints reduced the possibility of excessive use of fertilizers and indiscriminate burning of straw by restricting non-AGP behaviors of farmers [21]. Furthermore, a new piece of evidence suggests that the increasing market demand for green agricultural products would incentivize farmers to adopt more environmentally friendly production methods [22], as this involves production costs and economic benefits for farmers [23]. For example, although the price of green organic fertilizer was often higher than traditional fertilizer, when farmers believed that green manure could bring better results in crop yield and quality, they were more likely to choose green manure [24].

Moreover, individuals would have emotions generated and make decisions when they were recognized and stimulated by external things. Correspondingly, farmers had internal psychological predictions about AGP practices, and their perceived green benefits would directly affect their willingness [25]. Moreover, when farmers had a strong green awareness, they were more likely to affirm the benefits of AGP, thereby protecting the agricultural ecological environment. The government could also influence farmers’ decision-making through environmental policies such as guidance or constraints [26]. From the perspective of practical results, the negative externalities of environmental pollution could lead to a certain difference between actual costs and social costs. The government needs to introduce relevant regulatory policies to regulate the balance between economic development and environmental protection [27]. Although the government is willing to implement policies related to AGP, the low awareness of farmers towards these policies in practice has, to some extent, weakened their effectiveness [4,15]. Therefore, how to strengthen the intensity of environmental regulations is also an important external factor in effectively promoting farmers’ willingness to practice AGP [19].

Although focusing on the level of internal perceived benefits and the intensity of external environmental regulations has provided us with a detailed analytical framework for deciphering how to improve farmers’ AGP willingness level, how to ensure that farmers effectively engage in AGP still depends on whether relevant technologies can play a better role. Agricultural technology and production knowledge play a crucial role in the current successful cases of global response to increasing food output [28,29]. In order to accurately adopt technology, farmers usually require more professional, practical guidance. Environmental technology training and knowledge exchange could effectively promote farmers to engage in AGP [30]. Moreover, many excellent improved production technologies have been greatly beneficial for increasing food production and improving agricultural environments in developing countries, but small farmers often found it difficult to obtain the correct application of AGP technologies. For example, there has been evidence indicating that the use of drone technology could improve the flexibility of rural farmers in emerging economies in India to access real-time information and reduce labor-intensive processes, significantly changing the productivity of rural agriculture [31]. However, some studies have found that technology subsidies had limited effectiveness. The use of agricultural extension services could significantly improve the agricultural productivity of female small-scale farmers in Uganda by providing training and information [32], thereby enhancing food security levels. These improved farming methods did not require significant upfront monetary investment, and were particularly important for farmers with limited resources.

Existing literature has mostly explored farmers’ willingness to engage in AGP from multiple factors, with a focus on a single internal or external factor, lacking in-depth exploration of the two in a unified framework. Our research had preliminarily confirmed that the combined impact of farmers’ internal perception and external environment on their willingness to engage in AGP was becoming increasingly evident [4]. Furthermore, in terms of indicator measurement, previous studies have overlooked the necessity of using multiple indicators to characterize and measure, especially in exploring the frequency and impact of environmental policy implementation. In addition, the internal mechanism, such as the role of technology acquisition, also needs to be carefully considered. At present, relevant research has confirmed that providing training and information channels related to green technology has a significant impact on improving farmers’ food productivity and safety levels [33]. The more channels for farmers to obtain agricultural technology information, the greater the impacts on their behavioral decisions [34]. Present studies have lacked an optimized combination of technology acquisition and its dimensions, as well as its mediating effect as an intrinsic mechanism. Therefore, the aim of this study was to incorporate green perception benefits and environmental regulation intensity into the analysis framework of influencing the AGP willingness of grain growers, and to explore the impact on farmers’ AGP willingness based on the transmission mechanism of technology acquisition. The novelty is summarized as follows: Firstly, unlike existing literature that has mostly focused on farmers’ willingness to engage in AGP, our research focused on grain growers in major grain-producing areas and fully considered the particularities of their production practices. Secondly, starting from the two paths of green perception benefits and environmental regulation intensity, and subdividing their dimensions, we examine the impact on the AGP willingness of grain growers. Thirdly, an innovative analysis of technology acquisition and its transmission effects in various dimensions would provide new ideas for better identifying the mechanisms that affected the AGP practices of grain growers.

The rest of this article is arranged as follows. Section 2 reviews the relevant literature, based on which research hypotheses are proposed, and a theoretical framework is constructed. Section 3 introduces the sample area and data sources, variables and their indicator measurements, as well as econometric models. Section 4 elaborates on the empirical results, involving the presentation of results on direct and mediating effects. Section 5 presents the discussion and provides some prospects for future research. Section 6 is the research conclusion and policy implications. The technical roadmap is shown in Figure 1.

2. Literature Analysis, Hypothesis Development and Conceptual Model

2.1. Green Perception Benefits and Farmers’ AGP Willingness

Studies have shown that farmers will weigh the pros and cons based on the consideration of maximizing benefits when expressing their willingness to take a certain action, and ultimately make the optimal decision [4]. For most farmers, AGP is a new sustainable model, and in terms of increasing their willingness to adopt new technologies, farmers’ green perception benefits are crucial [5]. Previous studies in behavioral economics have laid a solid foundation for the above achievements, believing that farmers will consider green inputs such as organic fertilizers or biopesticides as additional expenditure costs [6,35]. Only when the cost risk of such expenditure is significantly lower than the economic benefits obtained can farmers trigger changes in behavioral decisions. As stated by Guo et al. (2022) [36], farmers will only adopt new technologies when they recognize that they can bring more economic benefits under certain conditions. In addition, social capital theory suggested that social networks such as cooperatives and neighborhood relationships are key channels for the dissemination of environmental awareness [37]. Farmers are likely to respond to policies related to environmental interests under such collective environmental actions. A study using field survey data from apple growers around the Bohai Sea in 2022 found that the environmental benefits of adopting green pest control technologies vary among different groups. Among them, economic benefits and operational management factors play a significant role in promoting the adoption of green pest control technologies [38]. There are also studies that define and assign values to green perception benefits from three aspects: ‘Can bring considerable economic income to the family’, ‘Beneficial for reducing water and soil pollution’, and ‘Responding to the government’s call will gain social recognition’ [4], and explore their positive effects on farmers’ willingness to engage in AGP. The above ideas coincide with the attitude elements in the theory of planned behavior, which internalizes environmental responsibility into a sense of moral obligation [39], thereby achieving sustainability in external stimuli, turning towards identity recognition to a certain extent. Scholars have also pointed out that when the green perception cost exceeds expected benefits or conflicts with other priorities, farmers often resist changes [40]. Green perception benefits can motivate individuals to engage in specific behaviors, partly because perceived benefits greatly help farmers understand the environmental degradation caused by non-AGP behaviors [31]; On the other hand, the positive results and abundant harvests can help enhance the willingness to engage in AGP [41]. Thus, the green perception benefits will directly affect farmers’ AGP willingness. In other words, the higher the perceived benefits of farmers, the better their evaluation of the benefits obtained from engaging in AGP, which will directly have a positive impact on farmers’ AGP willingness. Therefore, the following hypothesis is proposed:

H1: 

Green perception benefits and its dimensions, including economic benefits, environmental benefits, responsibility benefits, and identity benefits, have a significant positive impact on farmers’ AGP willingness.

2.2. Environmental Regulation Intensity and Farmers’ AGP Willingness

Agricultural environmental pollution has significant negative externalities, and environmental regulation essentially internalizes the environmental costs brought about by these negative externalities [42]. In the process of agricultural production, the government regulates and punishes farmers’ environmental pollution behaviors through policy measures and legal regulations, which can standardize the production, distribution, and application effects of chemical inputs by farmers. For example, Cheng et al. (2025) [43] conducted a study based on questionnaire data from 60 townships in Heilongjiang Province, China, and found that policy subsidies and penalized restraint policies have a significant impact on farmers’ intention to apply straw returning technology. The straw technology application in these policies plays an important role in reducing agricultural production costs and increasing farmers’ income. This is consistent with the viewpoint of a study that when local environmental policies are not comprehensive or strict enough, farmers may tend to overuse chemical inputs such as pesticides and fertilizers or dispose of agricultural material waste at will because they do not have to bear additional environmental costs [44]. These views are particularly critical for smallholder farmers in developing countries. For example, in Bangladesh, the number of times a farmer receives technology extension significantly affects the amount of nitrogen fertilizer used by the farmer [45]. In Ethiopia, the establishment of public social networking platforms, cooperative organizations, and mutual aid institutions within communities is crucial to support the negative impacts of agricultural technology adoption among older farmers [46]. Moreover, the theory of social information processing proposes that farmers collect relevant environmental information, use experience to comprehensively analyze and judge the information, and adjust their behavioral decisions accordingly [47]. Therefore, the ways in which environmental regulations can be effective include at least two aspects: the impacts on farmers’ subjective cognition and the intensity of objective implementation. From the perspective of the impacts on subjective cognition, environmental regulations can help improve farmers’ sense of identity in implementing AGP, enhance AGP performance, promote farmers’ self-regulation, increase their satisfaction with regulatory policies, and thus increase farmers’ willingness level [48]. However, objectively strengthening the frequency of environmental inspections, such as constraints, incentives, guidance, and contracts that the government conducts on farmers’ production behavior each year, can greatly change the expected cost and benefit levels of farmers engaging in AGP, thereby affecting their AGP willingness [49]. Moreover, binding measures such as standards, regulations, or penalties, as well as guiding measures like subsidies, rewards, or investments, can indirectly reflect the impact of environmental policy pressure on farmers’ AGP willingness. The same applies to contractual measures like verbal commitments or written agreements on petitions, resources, or the environment [27]. Therefore, the following hypothesis is proposed:

H2: 

Environmental regulation intensity and its dimensions, including objective regulation intensity, subjective regulation intensity, have a significant positive impact on farmers’ AGP willingness.

2.3. Mediation Effects of Technology Acquisition

The diffusion of innovation theory holds that the lack of equipment, networks, or skills not only increases the difficulty for behavioral subjects to acquire new technologies but also directly affects their evaluation of complexity or compatibility [50,51]. These sources of acquisition mainly include three categories: institutional information support [52,53], information and communication-based learning [54], and social interaction processes [28,44]. This study defines technology acquisition as the category, quality, and overall evaluation of channels through which farmers can obtain AGP technologies in agricultural production. Among them, the acquisition channels mainly involve objective-level technology acquisition channels, including agricultural cooperatives, agricultural technology promotion departments, agricultural material enterprises or retailers, traditional media and new media, etc. For example, Li et al. (2024) [55], using 656 survey data of apple growers in Shaanxi and Gansu provinces of China, found that obtaining national policy information and agricultural production information from the Internet channels is very helpful for farmers’ information accessibility, and the information acquisition ability has a positive impact on farmers engaged in green production. Acquisition quality is also an important factor in improving technology adoption, and therefore, the quality of technology acquisition may affect farmers’ AGP willingness. It mainly involves the subjective level of technology acquisition quality, including information validity, relationship closeness, cost savings, etc. [56]. Yang et al. (2024) [57] argue that participating in technical training can improve farmers’ understanding of technology with high quality, thereby promoting the adoption of fertilizer reduction and efficiency enhancement technologies by farmers. Although the above results have obvious advantages in information acquisition, the quality evaluation of these channels by farmers may vary depending on their personal technical proficiency, information screening ability, and trust in the media. This study categorizes this type of influence as acquisition evaluation. It mainly involves farmers’ evaluations of technology acquisition, including evaluations of channels such as interpersonal communication, media platforms, and agricultural technology promotion [58,59]. Farmers’ feedback or understanding of technology affects their perception of AGP, and therefore, their evaluation of information acquisition may also affect their willingness. Therefore, the following hypothesis is proposed:

H3: 

Technology acquisition and its dimensions, including acquisition channels, acquisition quality, and acquisition evaluation, have a significant positive impact on farmers’ AGP willingness.

Through agricultural technology, farmers can also learn how to adopt greener farming methods without sacrificing yield. First, in terms of knowledge and cognitive enhancement, technology acquisition enhances farmers’ understanding and awareness of environmentally friendly agricultural practices through education and training activities, thereby promoting farmers’ awareness of the importance of AGP methods. This viewpoint is supported by the research conducted by Mack et al. (2024) [60] on agricultural regulatory policies for farmers. There is also a study suggesting that the use of organic agriculture and biopesticides not only helps protect the environment but also improves soil quality and crop yields [61]. Moreover, the higher the level of awareness of AGP technology, the more significant the increase in farmers’ AGP willingness. Second, in terms of economic benefits and cost perception, technology acquisition reduces the initial investment and learning costs of adopting AGP technologies, which makes farmers more willing to try and adopt these new methods. For example, the higher the perceived expected benefits of straw returning technology among farmers, the more likely it is to encourage them to spend more time and energy browsing or watching online images, videos, and knowledge related to air, soil, and water pollution caused by harmful behaviors such as straw burning [62]. Thus, the following hypothesis is proposed:

H4: 

Green perception benefits have a significant positive impact on farmers’ AGP willingness through technology acquisition.

H4-1: 

Green perception benefits have a significant positive impact on farmers’ AGP willingness through acquisition channels.

H4-2: 

Green perception benefits have a significant positive impact on farmers’ AGP willingness through acquisition quality.

H4-3: 

Green perception benefits have a significant positive impact on farmers’ AGP willingness through acquisition evaluation.

Although previous studies have explored the impact of environmental regulation intensity on enhancing farmers’ AGP willingness, some scholars have suggested that the key to the effectiveness of policies is to make farmers perceive the accessibility of technology [63]. In particular, providing targeted technical information can better assist farmers in adopting it correctly. For example, Li et al. (2022) [64] emphasized the importance of improving measurement errors in non-land agricultural inputs (such as fertilizers, pesticides, seeds, etc.) in agricultural surveys, and proposed how to correct farmers’ misconceptions about policies to improve input allocation [65]. Wang and Zhe (2024) [34] conducted a study based on survey data of grain growers in Hebei Province, China, and found that the number of technical information channels plays a mediating role between socialized organizations with strong government support and the adoption of water-saving irrigation technologies by farmers. Therefore, obtaining technology to promote farmers’ understanding of environmental policies related to AGP also plays an important role. In addition, when farmers realize that adopting AGP practices can provide policy incentives such as economic subsidies and environmental rewards from the government, they are more likely to adjust their actions and choose to lean towards policy goals [35]. Some studies have also pointed out that the greater the information acquisition ability of farmers through more channels, the more significant the impact they can have on the adoption of fertilizer reduction techniques by farmers [54]. On the contrary, when farmers face environmental regulatory fines for not adopting a certain type of technology as required by the government, they are willing to adjust their production decisions to meet regulatory goals out of rational economic considerations [34]. Therefore, the following hypothesis is proposed:

H5: 

Environmental regulation intensity has a significant positive impact on farmers’ AGP willingness through technology acquisition.

H5-1: 

Environmental regulation intensity has a significant positive impact on farmers’ AGP willingness through acquisition channels.

H5-2: 

Environmental regulation intensity has a significant positive impact on farmers’ AGP willingness by acquisition quality.

H5-3: 

Environmental regulation intensity has a significant positive impact on farmers’ AGP willingness through acquisition evaluation.

Based on the above analysis, this study constructed a conceptual model as shown in Figure 2. The difference between this model and existing research is that it includes two key explanatory variables, namely green perception benefits, environmental regulation intensity and their various dimensions. It also includes mediating variables, namely technology acquisition and its various dimensions, and the dependent variable, namely the AGP willingness of grain growers.

3. Materials and Methods

3.1. Research Area and Data Sources

The research area is located in the major grain-producing counties of Hebei, Henan, and Shandong Province in the Huang Huai Hai Plain, as shown in Figure 3. The region belongs to the warm temperate monsoon climate, with abundant solar and thermal resources and flat terrain. Most of the precipitation occurs from July to September, suitable for the growth of various crops. The crops grown in this region are mainly planted in two seasons a year. Winter wheat is sown around early October and harvested from late May to early June. In addition, the region is renowned for its highly complex agricultural production and advanced farming methods employed by farmers. These conditions make it an important agricultural production area in China, which is of great significance for ensuring national food security and agricultural product supply.

In addition, the grain planting and production situation in the three provinces plays an important role in the development of the national agricultural economy. As reported in the China Statistical Yearbook 2024, Henan, Shandong, and Hebei provinces produced 66.24, 56.55, and 38.10 million tons of grain in 2023, accounting for 9.53%, 8.13%, and 5.48% of China’s total grain production, respectively. These figures underscore the critical role of these provinces in ensuring national food security. Nonetheless, the agricultural ecological environment in the region is still not optimistic. Due to long-term overexploitation and utilization, the land, surface water, and biological resources in the region have been severely damaged, seriously affecting the yield and quality of crops. Thus, while promoting increased grain production, these three provinces have also brought serious negative impacts, especially the production environment pollution and ecosystem damage caused by long-term overuse of fertilizers and pesticides.

This study adopted a multi-stage stratified sampling survey and conducted household visits and field data collection from March to April and June to July 2023. The specific investigation plan is as follows: a total of 8 major grain-producing counties were selected, namely Ningjin County in Xingtai City and Daming County in Handan City, Qihe County in Dezhou City and Cao County in Heze City, Taikang County and Luyi County in Zhoukou City, Yongcheng City and Xiayi County in Shangqiu City. The selection of these 8 counties as sample points is primarily due to their distinct natural conditions and strong advantages in grain production, characterized by extensive planting areas and high yields. Additionally, these counties benefit from annual special funding support from the central government. However, they face challenges such as water scarcity and declining farmland fertility, necessitating an urgent shift toward sustainable grain production practices. Thus, their selection as research focal points is highly representative.

Furthermore, taking into account the geographical location of major grain-producing counties and the convenience of actual investigations, a list of all townships in each sample county was obtained. Meanwhile, based on the grain planting area of each township, 2 townships were selected from each sample county; then, a list of all administrative villages in each sample township was obtained, and 2 administrative villages from each township were selected based on the number of administrative villages and grain planting households. Finally, a list of actual farming households from the village committee (excluding non-farmers whose land has been fully transferred) was obtained, and 35–40 grain growers from each administrative village were selected based on the actual farming area. There are a total of 16 townships and 32 administrative villages. It is worth mentioning that the interviewers are professionally trained university teachers and graduate students in related fields. Specifically, the research team conducted a 2-day training on the logic of questionnaire contents, agricultural green production terminology, sensitive question questioning techniques, handling of abnormal situations, standardized signing of informed consent forms, and signing of data confidentiality agreements. During the survey process, two supervisors promptly reviewed the collected data. Moreover, the investigator conducted one-on-one interviews with the subjects and completed the survey questionnaire on site. A total of 1250 questionnaires were distributed, and after excluding various types of unqualified questionnaires, a total of 1218 valid samples were obtained, with an effective rate of 97.44%. The questionnaire mainly includes personal and family characteristics of grain growers, grain green production situation, green perception benefits and environmental regulation intensity variables.

3.2. Variable Selection

The explained variable is the farmers’ AGP willingness. AGP willingness refers to the behavioral and psychological intentions exhibited by farmers towards engaging in AGP, which can be divided into very unwilling to implement AGP and very willing to do so. Therefore, based on the actual agricultural production situation of sample farmers and their families, and drawing on the research results of Guo et al. (2022) [36] and Charatsari et al. (2024) [7], the following questions were set in the field survey to measure farmers’ AGP willingness: “Is your family willing to implement agricultural green production?”. The scale ranges from five levels of “very unwilling to very willing” and is assigned values of 1, 2, 3, 4, and 5, respectively.

The key explanatory variables involve green perception benefits and environmental regulation intensity. Firstly, based on the content of this study and the actual research situation, drawing on the research of [61,66], multiple indicators were used to characterize the measurement items of green perception benefits. In this study, green perception benefits refer to the judgments and evaluations made by farmers regarding all perceived economic, environmental, responsibility, and identification benefits in engaging in AGP. It can be divided into economic benefits, environmental benefits, responsibility benefits, and identity benefits. Secondly, drawing on the achievements of Huang et al. (2024) [67] and Bolognesi and Pflieger (2024) [68], multiple indicators are used to characterize the measurement items of environmental regulation intensity, starting from the frequency and impact intensity of policy measures implemented by government departments. The environmental regulation intensity in this study refers to the specific level or pressure of government departments in implementing relevant environmental regulations during the process of AGP. This mainly includes objective and subjective aspects. Among them, objective regulation intensity refers to the number of times government departments take relevant regulatory measures (such as supervision, incentives, and guidance) on AGP each year; subjective regulation intensity refers to the degree to which regulatory measures related to AGP (such as supervision, incentives, and guidance) affect farmers annually.

It is worth mentioning that SPSS 25.0 software was used to test the measurement indicators of various dimensions of green perception benefits and environmental regulation intensity. The KMO values for economic benefits, environmental benefits, responsibility benefits, and identity benefits were found to be 0.901, 0.814, 0.763, and 0.872, respectively. After using the maximum variance method for factor rotation, the cumulative variance contribution rates of the four common factors based on eigenvalues > 1 were 76.416%, 67.211%, 60.196%, and 66.375%, respectively. In addition, the KMO values of objective regulatory intensity and subjective regulatory intensity obtained through testing were 0.876 and 0.698, respectively. The cumulative variance contribution rates of the two common factors based on eigenvalues > 1 are 68.203% and 61.282%, respectively. The above results indicate that factor analysis on the dimensions of green perception benefits and environmental regulation intensity is applicable and yields good results.

Drawing on the related research [53,69,70], some demographic factors such as gender, age, education level, years of cultivation, household cadre situation, number of household labor force, planting scale, and whether to join cooperatives were used as control variables. The definitions and assignments of each variable are shown in

Table 1. Definition and assignment of variables.

Variables	Number	Definition and Assignment	Mean Value	Standard Deviation
Explained Variable
Green production willingness	GPW	Is your family willing to implement green production behaviors? 1 = Very unwilling; 2 = Not very willing; 3 = General; 4 = More willing; 5 = Very willing	0.806	0.395
Key explanatory variables
Green perception benefits	Gpb	Obtained through factor analysis of Gpb1–Gpb5	0	1
Economic benefits	Gpb1	Agricultural products associated with implementing AGP can generate relatively considerable income; 1 = strongly disagree, 2 = somewhat disagree, 3 = average, 4 = somewhat agree, 5 = strongly agree	0	1
Environmental benefits	Gpb2	Reasonable use of chemical inputs such as fertilizers and pesticides can reduce environmental pollution, such as water and soil; 1 = strongly disagree, 2 = somewhat disagree, 3 = average, 4 = somewhat agree, 5 = strongly agree	0	1
Responsibility benefits	Gpb4	Unreasonable use of chemical inputs such as fertilizers and pesticides has polluted the environment, and farmers should bear some responsibility; 1 = strongly disagree, 2 = somewhat disagree, 3 = average, 4 = somewhat agree, 5 = strongly agree	0	1
Identity benefits	Gpb5	The rational use of chemical inputs such as fertilizers and pesticides has protected the environment and gained social recognition; 1 = strongly disagree, 2 = somewhat disagree, 3 = average, 4 = somewhat agree, 5 = strongly agree	0	1
Environmental regulation intensity	Eri	Obtained through factor analysis of Eri1–Eri2	0	1
Objective regulation intensity	Eri1	The number of regulatory measures (such as supervision, incentives, and guidance) taken annually for AGP	0	1
Subjective regulation intensity	Eri2	The impact of regulatory measures related to AGP (such as supervision, incentives, and guidance) on me every year	0	1
control variable
Gender	Sex	Male = 1, Female = 0	0.634	0.482
Age	Age	What is your age? Actual age	55.230	11.595
Educational level	Edu	How many years of education have you received? Actual years	5.213	3.691
Years of cultivation	Cry	How many years have you been engaged in agricultural planting? Actual years	30.019	14.907
Household cadre situation	Sca	Do you have any public officials or village cadres in your family? Yes = 1, No = 0	0.077	0.266
Number of household labor force	Lab	How many employees aged 16 and above are there in your family? Actual number of people	3.567	1.606
Planting scale	Ara	What is the actual arable land area (including self-owned land, contracted land, leased and subcontracted land) in your household?	8.544	12.382
Whether to join cooperatives	Org	Has your family joined an agricultural cooperative? Yes = 1, No = 0	0.192	0.394

.

The mediation variable is technology acquisition. It refers to the channels through which farmers obtain AGP technology during the agricultural production process, the quality of AGP technology obtained through these channels, and how it helps agricultural production activities, aiming to encourage farmers to actively engage in AGP behaviors. In general, the more channels of AGP technology available to farmers, the better the quality of the technology carried in the channels, and the greater the help of technology to production activities, the more conducive it is to improving farmers’ willingness to engage in AGP. Therefore, referring to the relevant measurement indicators in Ayalew et al. (2022) [52] and Chen et al. (2024) [21], this study divides technology acquisition into three dimensions: acquisition channels, acquisition quality, and acquisition evaluation. The definition and assignment of technical acquisition indicators are shown in

Table 2. Definition, assignment and descriptive statistics of the technology acquisition variable.

Variables	Number	Definition and Assignment	Mean	Standard Deviation
Technology acquisition	TA	Obtained through factor analysis of Tac, Taq and Tae	0.000	1.000
Acquisition channels	Tac	Obtained through factor analysis of Tac1–Tac6	0.000	1.000
	Tac1	Is it through agricultural professional cooperatives to obtain green production technology? Almost no = 1, less often = 2, generally = 3, more often = 4, many times = 5	2.982	1.276
	Tac2	Is it through communication with other farmers in the surrounding area that green production technology is obtained? Almost no = 1, less often = 2, generally = 3, more often = 4, many times = 5	3.489	1.045
	Tac3	Did you obtain green production technology through the agricultural technology promotion department? Almost no = 1, less often = 2, generally = 3, more often = 4, many times = 5	3.213	1.126
	Tac4	Is green production technology obtained through agricultural input enterprises or retailers? Almost no = 1, less often = 2, generally = 3, more often = 4, many times = 5	3.145	1.291
	Tac5	Is green production technology obtained through traditional media such as television, radio, newspapers, and magazines? Almost no = 1, less often = 2, generally = 3, more often = 4, many times = 5	2.776	1.526
	Tac6	Is it through new media such as network TV, digital video, electronic magazine, official account, etc., that we obtain green production technology? Almost no = 1, less often = 2, generally = 3, more often = 4, many times = 5	2.638	1.520
Acquisition quality	Taq	Obtained through factor analysis of Taq1–Taq3	0.000	1.000
	Taq1	How effective is obtaining information on agricultural green production technology? Very poor = 1, poor = 2, average = 3, good = 3, very good = 5	3.514	0.875
	Taq2	How closely is the relationship between obtaining agricultural green production technology? Very poor = 1, poor = 2, average = 3, good = 3, very good = 5	3.295	1.131
	Taq3	How cost-effective is it to acquire agricultural green production technology? Very poor = 1, poor = 2, average = 3, good = 3, very good = 5	3.461	0.886
Acquisition evaluation	Tae	Obtained through factor analysis of Tae1 and Tae3	0.000	1.000
	Tae1	Overall, how does the green production technology obtained through interpersonal communication channels help with production activities? No help = 1, less help = 2, average help = 3, more help = 3, very much help = 5	3.758	0.796
	Tae2	Overall, how does the agricultural green production technology obtained through media platforms help with production activities? No help = 1, less help = 2, average help = 3, more help = 3, very much help = 5	3.187	1.235
	Tae3	Overall, how does the green production technology obtained through agricultural technology promotion channels help with production activities? No help = 1, less help = 2, average help = 3, more help = 3, very much help = 5	3.638	1.056

.

SPSS 25.0 software was used to test the measurement indicators of each dimension obtained by the technology. The KMO values of the acquisition channel, acquisition quality, and acquisition evaluation were 0.842, 0.715, and 0.660, respectively, indicating that the factor analysis results were good. According to Bartlett’s sphericity test, the approximate chi-square values for obtaining channels, obtaining quality, and obtaining evaluation are 2539.688, 762.599, and 319.730, respectively, all of which are significant at the 1% level. Moreover, after using the maximum variance method for factor rotation, the cumulative variance contribution rates of the three common factors based on eigenvalues > 1 were 64.60%, 73.52%, and 60.28%, respectively. This indicates that exploratory factor analysis of various dimensions of technology acquisition yields good results.

3.3. Econometric Model

3.3.1. Benchmark Model

Farmers’ AGP willingness is a multi-class ordered variable. When analyzing this type of discrete dependent variable with multiple orderliness, using ordered Probit or ordered Logit models for result estimation is more ideal. However, compared to the latter, the former assumes more rigorous conditions. In addition, there are some nested data structures in this study, but mixed logit needs to capture individual heterogeneity, and the calculation process is relatively complex. Furthermore, hierarchical modeling also requires a sufficient number of high-level groups. In this study, the ordered Probit model not only effectively estimates the probability changes of each level, but also can be applied to analyze ordinal data of a single decision level. Therefore, drawing on the research methods of Bellizzi et al. (2018) [71], this study assumed that farmers’ AGP willingness satisfied normal distribution conditions and used a multivariate ordered Probit model that was effective in sorting for analysis. The equation for the model is as follows:

$${GPW}^{*}={\beta }_1{N}_i+{\beta }_2{F}_i+{\epsilon }^{*}$$

(1)

In Equation (1), ${GPW}^{*}$ is an unobservable latent variable, namely farmers’ AGP willingness. $N_i$ is the explanatory variable, representing green perception benefits and environmental regulation intensity, as well as their various dimensions and control variables, including perception benefits and environmental policies. $F_i$ is the control variable, including gender, age, education level, years of cultivation, household cadre situation, number of household labor force, planting scale, and whether to join cooperatives. ${\beta }_1,{\beta }_2$ are coefficients to be estimated, and ${\epsilon }^{*}$ is a random error term that follows a standard normal distribution. ${GPW}_i$ is the dependent variable with a range of $\{1,2, 3, 4, 5\}$. The relationship between the observable score of farmers ${GPW}_{i }$ and the unobservable latent variable ${{GPW}_i}^{*}$ is as follows:

$$GPW=\left\{\begin{array}{c}1 (\mathrm{Very} \mathrm{unwilling}), \mathrm{if} {GPW}^{*}\le c_1\\ 2 (\mathrm{Not} \mathrm{very} \mathrm{willing}), \mathrm{if} c_1<{AGPW}^{*}\le c_2\\ 3 \left(\mathrm{General}\right), \mathrm{if} c_3<{GPW}^{*}\le c_4\\ 4 (\mathrm{More} \mathrm{willing}), \mathrm{if} c_4<{GPW}^{*}\le c_5\\ 5 (\mathrm{Very} \mathrm{willing}), \mathrm{if} {GPW}^{*}\ge c_5\end{array}\right.$$

(2)

In Equation (2), $c_i(i=1,2,3,4,5)$ is the critical value for farmers to evaluate their green production willingness at different levels, estimated simultaneously with parameters and estimated coefficients, and $c_1<c_2<c_3<c_4<c_5$. From this, it can be concluded that the probabilities P of different scores for farmers’ AGP willingness are as follows:

$$P(GPW=1\mid \mathrm{X})=\mathrm{\varnothing }(c_1-{\beta }_1{N}_i-{\beta }_2{F}_i)$$

(3)

$$P(GPW=2\mid \mathrm{X})=\mathrm{\varnothing }(c_2-{\beta }_1{N}_i-{\beta }_2{F}_i)-\mathrm{\varnothing }(c_1-{\beta }_1{N}_i-{\beta }_2{F}_i)$$

(4)

$$P(GPW=3\mid \mathrm{X})=\mathrm{\varnothing }(c_3-{\beta }_1{N}_i-{\beta }_2{F}_i)-\mathrm{\varnothing }(c_2-{\beta }_1{N}_i-{\beta }_2{F}_i)$$

(5)

$$P(GPW=4\mid \mathrm{X})=\mathrm{\varnothing }(c_4-{\beta }_1{N}_i-{\beta }_2{F}_i)-\mathrm{\varnothing }(c_3-{\beta }_1{N}_i-{\beta }_2{F}_i)$$

(6)

$$P(GPW=5\mid \mathrm{X})=\mathrm{\varnothing }(c_3-{\beta }_1{N}_i-{\beta }_2{F}_i)-\mathrm{\varnothing }(c_4-{\beta }_1{N}_i-{\beta }_2{F}_i)$$

(7)

In Equations (3)–(7), $\mathrm{\varnothing }$ is the cumulative density function of the standard normal distribution, and the model uses the maximum likelihood estimation method for estimation.

It is worth mentioning that in the Probit model, the coefficient $\beta$ can only reflect significance and sign direction, and cannot directly represent marginal effects. Therefore, this study uses marginal effects to further analyze the empirical results. The marginal effect represents the impact of the unit change of $X_k$ on $p\left(\gamma =j\right)$. The equation expressed is as follows:

$${P(\gamma }_j)={\displaystyle \frac{𝜕P(\gamma =j|x)}{𝜕X_j}}={\beta }_k[\mathrm{\varnothing }\left({\mu }_{j-1}-\mathrm{X}\beta \right)-\mathrm{\varnothing }({\mu }_j-\mathrm{X}\beta )]$$

(8)

In Equation (8), $j=1,2,3,4,5,$ ${P(\gamma }_j)$ is the marginal effect of variable ${\gamma }_j$.

3.3.2. Mediation Effect Model

The mediator variable refers to the mediator between the independent variable and the dependent variable. If the independent variable $N_i$ affects the dependent variable $AGPW$ through variable $M_i$, then variable $M_i$ is considered a mediator variable. As mentioned earlier, farmers’ AGP willingness is a multivariate and ordered dependent variable.

Therefore, this study draws on relevant research results, and uses hierarchical regression to test the mediating effect and conduct an analysis. The equation for the testing model is as follows:

$$GPW=c{N}_i+{\epsilon }_1$$

(9)

$$M_i=a{N}_i+{\epsilon }_2$$

(10)

$$PW=c^{\prime }{N}_i+b{M}_i+{\epsilon }_3$$

(11)

In Equations (9)–(11), $GPW$ represents farmers’ AGP willingness. $N_i$ represents the green perception benefits and its various dimensions, as well as the environmental regulation intensity and its various dimensions. $M_i$ represents technology acquisition and its various dimensions, including acquisition channels, acquisition quality, and acquisition evaluation. ${\epsilon }_1$, ${\epsilon }_2 \mathrm{a}\mathrm{n}\mathrm{d} {\epsilon }_3$ are random error terms.

Furthermore, the steps for testing the mediating effect are as follows: Step 1, check whether the regression coefficient $c$ of the independent variable $N_i$ in Equation (9) is significant. If the coefficient $c$ is significant, proceed to the next step of the test. If the coefficient $c$ is not significant, stop the test, indicating that variable $M_i$ has not produced any mediating effect. Step 2, test whether the regression coefficients $a$ and $b$ of the independent variable $X$ in Equations (10) and (11) are significant. If both coefficients $a$ and $b$ are significant, continue to test whether coefficient $c^{\prime }$ is significant. If coefficient $c^{\prime }$ does not pass the significance test, it indicates that variable $M_i$ has a complete mediating effect. If coefficient $c^{\prime }$ passes the significance test, it is considered that variable $M_i$ has a partial mediating effect. Step 3, if at least one of the coefficients a and b is significant, the Sobel test needs to be continued on the basis of Step 2. If the Sobel test variable $M_i$ is not significant, it indicates that there is no mediating effect. Otherwise, it indicates that there is a mediating effect, and the mediating effect ratio is the proportion of the mediating effect of the independent variable $N_i$ on the dependent variable $GPW$ through variable $Z$ to the total effect of the independent variable $N_i$ on the dependent variable $GPW$, which is ${\displaystyle \raisebox{1ex}{$ab$}\!\left/ \!\raisebox{-1ex}{$c$}\right.}$.

3.3.3. Further Causal Relationship Analysis and Endogeneity Solution

According to the previous literature analysis, technology acquisition and its dimensions (acquisition channels, acquisition quality, acquisition evaluation) are important factors that affect farmers’ AGP willingness. However, farmers who are more interested in AGP may be more motivated to seek information, such as access channels. It is also possible that farmers with more access to information, such as channels, may be more willing to engage in AGP. In this case, there may be a bidirectional causal relationship between technology acquisition and farmers’ willingness to engage in AGP, leading to endogeneity issues and resulting in biased results.

To eliminate the possible impact of this bias, this study draws on the causal relationship handling method proposed by Jiang Ting (2022) [72], and uses the “average number of contacts between sample farmers and technology promoters in the same village except for individuals ($ANC$)” as the identification variable to effectively solve the endogeneity problem mentioned above. The selection criteria are as follows: Firstly, rural areas are often a society of acquaintances, and farmers’ decision-making often shows a strong same-group effect. The more times farmers in the same village come into contact with technology promoters, the more likely they are to develop a willingness psychology, thereby increasing their willingness to engage in AGP; secondly, since the channels and quality of technology obtained by other farmers in the same village are not related to their personal AGP willingness. Therefore, it does not directly affect the willingness of individual farmers to engage in AGP. Thirdly, the technology acquisition by fellow villagers will not directly affect farmers’ AGP, as the intensity of their technology acquisition is determined by the higher-level agricultural extension department based on macro policies such as regional planning and experimental projects. This is an external environment for farmers, which can only be influenced by neighborhood effects on farmers’ technology acquisition, and then affect their AGP decisions. The above explanation satisfies the three conditions of correlation, exogeneity, and exclusivity constraints required for identifying variables in selection.

The above relationship is shown in Figure 4. Among them, $Y$ is the dependent variable $GPW$, $D$ is the identification variable $ANC$, and $M_i$ is the ab mediator variable, which refers to the TA and its various dimensions mentioned earlier. (1) The formula indicates that $D$ has a causal effect on $Y$; (2) The equation represents that $M_i$ has a causal effect on $Y$, thus establishing a causal chain of $D$ → $D$ → $Y$. It also indicates that $D$ may independently affect $Y$ in addition to $M_i$; Equation (3) represents that $D$ has a causal effect on $M_i$.

4. Results

This study examined the influence of green perception benefits and environmental regulation intensity on grain growers’ willingness to adopt agricultural green practices (AGP). It begins with a benchmark estimation (Section 4.1) to establish a foundational analysis framework. Section 4.2 further explores the mediating role of technology acquisition in transmitting these effects, highlighting its influence on shaping farmers’ AGP willingness.

4.1. Benchmark Estimation

We conducted a multicollinearity test before estimating the model. This mainly includes two indicators: variance inflation factor (VIF) and tolerance (1/VIF). From the test results, it can be seen that the maximum value of VIF in the key explanatory variable is 1.46 < 10 (green perception benefits), the maximum value of VIF in the control variable is 2.01 < 10 (years of cultivation), and the mean VIF is 1.40 < 10. This indicates that the explanatory variables in empirical analysis are within an acceptable range, meaning there is no issue of multicollinearity.

4.1.1. The Direct Impacts of Green Perception Benefits and Environmental Regulation Intensity

From

Table 3. Benchmark regression results of influencing factors of behavioral willingness.

Variable	Model 1	Standard Error	Model 2	Standard Error
Green perception benefits	0.212 **	0.088
Economic benefits			0.178 **	0.094
Environmental benefits			0.031	0.046
Responsibility benefits			0.066	0.062
Identity benefits			0.134 ***	0.048
Environmental regulation intensity	0.103 **	0.043
Objective regulatory intensity			0.168 **	0.093
Subjective regulatory intensity			0.341 ***	0.041
Gender	−0.152 *	0.092	−0.309 *	0.173
Age	−0.007	0.004	0.006	0.005
Educational level	0.024 ***	0.012	0.220 ***	0.069
Years of cultivation	0.011	0.037	0.002	0.004
Household cadre situation	0.297 **	0.163	0.524 *	0.298
Number of household labor force	−0.049	0.056	−0.031	0.101
Whether to join cooperatives	0.059 *	0.232	0.007 *	0.000
Pseudo-R2	0.013	—	0.024	—
Log-likelihood	−801.077	—	−796.025	—
Prob > chi2	0.003	—	0.001	—
LR-test	28.71		44.680
Observations	1218	—	1218	—

*, **, *** indicate significance at the statistical levels of 10%, 5%, and 1%, respectively.

, it can be seen that this study first explored the impact of green perception benefits and environmental regulation intensity on farmers’ AGP willingness, and the estimated results are shown in Model 1. Secondly, two-dimensional variables, namely green perception benefits and environmental regulation intensity indicators, were introduced to further analyze the impact of different dimensions on farmers’ AGP willingness. The estimated results are shown in Model 2. Finally, from the logarithmic likelihood ratio, chi-square value, and pseudo value, it can be seen that the overall fitting effect of each model is good.

From the estimation results of Model 1 and Model 2, it can be found that green perception benefits have a significant positive impact at the 5% statistical level, with a regression coefficient of 0.212, indicating that green perception benefits can enhance farmers’ AGP willingness. Specifically, economic benefits and identity benefits have a significantly positive impact on farmers’ AGP willingness at the 5% and 1% statistical levels, respectively. Furthermore, from the regression coefficient, the promoting effect of economic benefits is 0.044 units greater than that of identity benefits. This indicates that, all other conditions being equal, for every unit increase in economic benefits, the effect of enhancing farmers’ willingness to engage in AGP is 0.044 units greater than for every unit increase in identity benefits. While environmental benefits and responsibility benefits do not have a significant impact on farmers’ willingness. Therefore, H1 is partially confirmed. It is assumed that before participating in AGP activities, farmers first consider whether they can obtain considerable benefits. Moreover, farmers’ expectations for economic benefits, such as labor conservation, high productivity, and high returns, are more likely to encourage them to engage in environmentally friendly activities [5]. Furthermore, farmers recognize that reducing the overuse of pesticides, fertilizers, and their associated packaging waste, while adopting high-quality, environmentally sustainable crop varieties, implementing deep loosening techniques for soil moisture retention, and recycling crop straw, not only conserves resources but also enhances their willingness to adopt AGP practices.

In addition, the environmental regulation intensity has a significantly positive impact at the 5% statistical level, with a regression coefficient of 0.103, indicating that the environmental regulation intensity can enhance farmers’ AGP willingness. In terms of the impact of various dimensions of environmental regulation intensity, objective and subjective regulatory intensity are significantly positive at the 5% and 1% statistical levels, respectively, with regression coefficients of 0.168 and 0.341, indicating that both objective and subjective regulatory intensity can positively affect farmers’ AGP willingness, and the impact of subjective regulatory intensity is relatively large. Thus, H2 is confirmed. This is largely due to the significant influence of government policies and initiatives related to AGP behaviors. And it has shaped the production practices of farmers and their families, thereby fostering a genuine and heartfelt commitment to actively participate in AGP. Moreover, there may be some psychological cognitive biases in the government’s transmission of regulatory policy signals, and the cost impacts of this bias are higher than that of revising objective regulatory policies themselves. This result is consistent with the research findings of Wang et al. (2023) [54], which emphasized that the environmental regulation intensity promotes the enforcement of environmental behaviors by farmers in aquaculture farms.

4.1.2. Robust Test

To test the robustness of the benchmark estimation results, this study drew on existing research methods [6], deleted sample data that were too old or too small, and compared the ordered Logit model with the ordered Probit model mentioned above (

Table 4. Robust test results.

Variable	Model 1	Standard Error	Model 2	Standard Error
Green perception benefits	0.209 **	0.069
Economic benefits			0.108 **	0.037
Environmental benefits			0.036	0.028
Responsibility benefits			0.055	0.036
Identity benefits			0.139 ***	0.022
Environmental regulation intensity	0.125 **	0.067
Objective regulatory intensity			0.162 ***	0.023
Subjective regulatory intensity			0.378 ***	0.041
Control variables	Yes	—	Yes	—
Pseudo-R2	0.021	—	0.365	—
Log-likelihood	−886.192	—	−723.042	—
Prob > chi2	0.003	—	0.001	—
LR-test	20.88	—	51.328	—
Observations	866	—	866	—

**, *** indicate significance at the statistical levels of 5%, and 1%, respectively.

). It was found that the impact of green perception benefits and environmental regulation intensity on farmers’ AGP willingness was almost consistent with the benchmark regression results in Table 3. This indicates that the robustness of the model was validated.

4.1.3. The Marginal Effect Test Results of Influencing Factors of Behavioral Willingness

As mentioned earlier, the regression coefficients of the ordered Probit model can only reflect significance and sign direction. To reflect the actual effects of green perception benefits and environmental regulation intensity on farmers’ AGP willingness, this study calculated the marginal effects of green perception benefits and environmental regulation intensity at the mean level. The results are shown in

Table 5. Marginal effect results of influencing factors of behavioral willingness.

Variables	Very Unwilling	Not Very Willing	General	More Willing	Very Willing
	dy/dx	dy/dx	dy/dx	dy/dx	dy/dx
Green perception benefits	−0.008	−0.018 **	0.030 **	0.005 *	0.062 ***
Economic benefits	−0.001	−0.003	−0.005 *	0.012 **	0.023 **
Environmental benefits	0.002	−0.008	0.027 *	0.003	0.013
Responsibility benefits	−0.002	−0.007	−0.012	0.003	0.014
Identity benefits	0.045 **	0.017 ***	0.027 ***	0.007 *	0.056 ***
Environmental regulation intensity	−0.004	−0.016 *	0.026 **	0.007	0.054 ***
Objective regulatory intensity	−0.001	−0.005	0.003 *	0.012 ***	0.020 **
Subjective regulatory intensity	−0.006 *	−0.022 **	0.009 *	0.036 ***	0.073 ***
Control variables	Controlled	Controlled	Controlled	Controlled	Controlled

*, **, *** indicate significance at the statistical levels of 10%, 5%, and 1%, respectively.

.

From Table 4, it can be seen that green perception benefits reduce the probability of farmers choosing “Not very willing” by 1.8%, while the probabilities of choosing “General”, “More willing”, and “Very willing” increase by 3%, 0.5%, and 6.2%, respectively, with the highest increase in probability observed in the “Very willing” scenario. This further indicates that green perception benefits can significantly increase farmers’ AGP willingness. Moreover, dimensional variables such as economic benefits and identity benefits can increase the probability of farmers choosing “More willing” or “Very willing”. The possible reason is that farmers who perceive the benefits of AGP are more likely to identify with the rational use of chemical inputs such as pesticides and fertilizers, and are willing to implement measures to protect the agricultural ecological environment or ensure the quality and safety of food products.

Meanwhile, environmental regulation intensity reduces the probability of farmers choosing “Not very willing” by 1.6%, while the probability of choosing “General” and “Very willing” increases by 2.6% and 5.4%, respectively. This overall effect mainly comes from the marginal effects of objective and subjective regulatory intensity in the internal dimensions, especially the probability values of farmers choosing “More willing” and “Very willing”. Moreover, subjective regulatory intensity increases the probability of farmers making choices significantly, by 3.6% and 7.3%, respectively. This is because the greater the specific implementation of government measures related to constraints, incentives, and guidance for AGP, the greater the impacts on farmers, and the complementary advantages of different regulatory measures are also good, which is beneficial for enhancing farmers’ AGP willingness.

4.1.4. Further Examination of Endogeneity Issues Caused by Reverse Causality

The more benefits farmers perceive from AGP, the higher their willingness to engage in AGP. The increase in willingness to engage in AGP may have a feedback mechanism, where farmers who are more willing to engage in AGP may perceive more benefits. However, this reverse causal relationship may lead to instability in the estimated coefficients, thereby affecting the accuracy of the research results. Therefore, this study draws on the instrumental variable method proposed by Jiang Ting (2023) [73] and Zhao et al. (2024) [74] to address endogeneity issues. At the same time, the average perceived benefits of AGP by other farmers in the same village (excluding individuals themselves) were selected as the instrumental variable for green perception benefits.

Firstly, in rural societies with close geographical relationships and limited information channels, frequent experience exchange and imitative behaviors among farmers can significantly influence individuals’ own judgments on the benefits of AGP based on others’ perceptions. Thus, there is a significant correlation between the average benefit perception of other farmers and the benefit perception of individual farmers. Secondly, as the constructed instrumental variables exclude the individual’s own perception and are only based on the average values of other farmers in the same village, it effectively avoids the possibility of individual AGP willingness, having a reverse impact on green perception benefits and thus affecting the instrumental variables. Meanwhile, the subjective judgments of other farmers mainly stem from their own experiences and environmental perceptions, and are not directly influenced by individual farmer behaviors; thus, they can be regarded as exogenous variables. Finally, the impacts of this instrumental variable on whether farmers engage in AGP are mainly achieved through the influence path of their green perception benefits. The perception of other farmers does not directly intervene or determine individual farmers’ AGP willingness, nor do they have other potential channels to influence their willingness. Therefore, this instrumental variable conforms to the exclusivity constraint of ‘only indirectly acting on the dependent variable through the dependent variable’.

The estimation results of the instrumental variable method are shown in

Table 6. Results of instrumental variable method estimation.

Variable	Phase One (Green Perception Benefits)		Phase Two (Farmers’ AGP Willingness)
Green perception benefits	—	—	0.301 **	0.026
The average perceived benefits of AGP by other farmers in the same village (excluding individuals themselves)	0.248 **	0.035	—	—
Control variables	Yes		Yes
Observations	1218		1218
Wald	19.38 ***

**, *** indicate significance at the statistical levels of 5% and 1%, respectively.

. The green perception benefits significantly enhance farmers’ AGP willingness, and the Wald value of the weak instrumental variable test is 19.38, significant at the 1% level, indicating the effectiveness of instrumental variables and the reverse causal relationship leading to the resolution of endogeneity issues.

4.2. The Mediation Effect Test of Technology Acquisition

This study further investigated the intrinsic mechanisms of technology acquisition and its multidimensional variables to determine their mediating role in shaping farmers’ willingness to adopt AGP. Specifically, it examines how green perception benefits, environmental regulation intensity, and their respective dimensions influence this willingness through technology acquisition. It should be noted that in this section, farmers’ AGP willingness is represented by Y1, while the green perception benefits and the environmental regulations intensity are represented by X1 and X2, respectively. The total index of technology acquisition and its various dimensions (acquisition channels, acquisition quality, and acquisition evaluation) are represented by M1, M11, M12, and M13, respectively.

4.2.1. The Mediation Effect of Technology Acquisition in Green Perception Benefits on Farmers’ AGP Willingness

Figure 5a presents the test results of the mediation effect of the total technology acquisition index and its dimensions on the relationship between green perception benefits and farmers’ AGP willingness. Specifically, regression analyses were performed in two steps: first, examining the direct impact of green perception benefits on farmers’ AGP willingness, and second, exploring the relationship between green perception benefits and the overall technology acquisition index. The results show that green perception benefits (X1) had a significantly positive impact on farmers’ AGP willingness (Y1) and the overall index of technology acquisition (M1) at the 1% statistical level, indicating that the test coefficients for c and a were significant. In the third step, both the green perception benefits and the total technology acquisition index were included in the mediation effect test model. The coefficient value of green perception benefits (X1) was significant at the 1% statistical level, but the coefficient of the total technology acquisition index (M1) was not significant; that is, c’ was significant (Z = 2.93, SE = 0.043) and b was not significant. Subsequently, based on the above results, a Sobel test was conducted to determine that the total index of technology acquisition (M1) is significant at the 10% statistical level. Meanwhile, the Bootstrap (95% CI, 5000 self-sampling) method was used again for validation, and the indirect effect value was obtained as 0.164, 95% CI [0.109, 0.312]. This interval does not contain 0. This indicates that the overall index of technology acquisition plays a mediating role in the impacts of green perception benefits on farmers’ AGP willingness.

As for the acquisition channels of two-dimensional variables, the regression results of green perception benefits and farmers’ AGP willingness, as well as green perception benefits and acquisition channels in the first and second steps, showed that green perception benefits (X1) had significantly positive and negative effects on farmers’ AGP willingness (Y1) and acquisition channels (M11) at the 1% and 10% statistical levels, respectively (Figure 5b). This indicates that the test coefficients of c and a were significant. In the third step, both green perception benefits and acquisition channels (M11) were included in the mediation effect test model. The coefficient values of green perception benefits (X1) and acquisition channels (M11) were found to be significant at the 5% statistical level; that is, c’ was significant and b was significant. This indicates that there is a partial mediating role of access channels in the impacts of green perception benefits on farmers’ AGP willingness.

In terms of the acquisition quality of two-dimensional variables, regression analysis was conducted in the first and second steps on the relationship between green perception benefits and farmers’ AGP willingness, as well as between green perception benefits and acquisition quality (Figure 5c). The results show that green perception benefits (X1) had a significantly positive impact on farmers’ AGP willingness (Y1) and acquisition quality (M12) at the 1% statistical level, indicating that the test coefficients for c and a were significant. In the third step, both green perception benefits and acquisition quality (M12) were included in the mediation effect test model. The coefficient value of green perception benefits (X1) was not significant, while the coefficient value of acquisition quality (M12) was significant at the 5% statistical level, indicating that c’ was not significant and b was significant. This indicates that obtaining quality plays a fully mediating role in the impacts of green perception benefits on farmers’ AGP willingness.

As for the acquisition evaluation of two-dimensional variables, regression analysis was conducted in the first and second steps between green perception benefits and farmers’ AGP willingness, as well as between green perception benefits and acquisition evaluation (Figure 5d). It was found that green perception benefits (X1) had a significantly positive impact on farmers’ AGP willingness (Y1) and acquisition evaluation (M13) at the 1% statistical level, indicating that the test coefficients for c and a were significant. In the third step, both green perception benefits and acquisition evaluation were included in the mediation effect test model. The coefficient value of green perception benefits (X1) was not significant, while the coefficient value of acquisition evaluation (M13) was significant at the 5% statistical level, meaning that c’ was not significant and b was significant. This indicates that acquisition evaluation plays a completely mediating role in the impacts of green perception benefits on farmers’ AGP willingness.

Overall, it can be concluded that the expected impacts of the total index of technology acquisition in H4 are validated, and some of its dimensional indicators are validated. In addition, to further refine the magnitude of mediating and direct effects, this section draws on the research method of Wang et al. (2023) [54], which calculates the proportion of mediating effects using the formula ab/c. The calculations indicate that the mediating effect of technology acquisition accounts for 5.87% of the total influence of green perception benefits on farmers’ AGP willingness. Because farmers have a natural conservatism towards new technologies, improving the value of technology acquisition channels through low-cost intervention or high-yield implementation still has practical significance in increasing farmers’ AGP willingness. Specifically, the mediating effect of acquisition channels, as a two-dimensional variable, contributes 8.50% to the total effect. Meanwhile, acquisition quality mediates 4.89%, while acquisition evaluation plays the most significant role, accounting for 14.72%. This is because, on the one hand, farmers perceive that AGP can bring some benefits, such as higher prices of agricultural products, and they will be more proactive in acquiring green technologies. On the other hand, technology acquisition plays a bridging role in it. Without the intermediate link, it may be difficult to directly translate green perceived benefits into farmers’ production intentions.

4.2.2. The Mediation Effect of Technology Acquisition in Environmental Regulation Intensity on Farmers’ AGP Willingness

Figure 6a reveals the test results of the mediation effect of the total technology acquisition index and its dimensions on the relationship between environmental regulation intensity and farmers’ AGP willingness. In terms of the overall index of technology acquisition, regression analysis was conducted on the relationship between environmental regulation intensity and farmers’ AGP willingness, as well as the relationship between environmental regulation intensity and the overall index of technology acquisition in the first and second steps. The results show that environmental regulation intensity (X2) had a significantly positive impact on farmers’ AGP willingness (Y1) and the overall index of technology acquisition (M1) at the 1% statistical level, indicating that the test coefficients for c and a were significant. Then, in the third step, both the environmental regulation intensity and the total index of technology acquisition were included in the mediation effect test model. The coefficient value of environmental regulation intensity (X2) was significant at the 1% statistical level, but the coefficient of the total index of technology acquisition (M1) was not significant; that is, c’ was significant (z = 5.11, SE = 0.043) and b was not significant. Subsequently, a Sobel test was conducted based on the above results, and it was found that the results were not significant. Meanwhile, the Bootstrap (95% CI, 5000 self-sampling) method was used again for validation, and the indirect effect value was obtained as 0.102, 95% CI [−0.689, 0.145]. This interval contains 0. This suggests that the overall index of technology acquisition does not mediate the impacts of environmental regulation intensity on farmers’ AGP willingness.

Regarding acquisition channels as a two-dimensional variable, the regression results from the first and second steps—examining the relationships between environmental regulation intensity and farmers’ AGP willingness, as well as between environmental regulation intensity and acquisition channels—reveal that environmental regulation intensity (X2) has a significant positive effect on both farmers’ green production willingness (Y1) and acquisition channels (M2) at the 1% and 5% statistical levels, respectively (Figure 6b). This indicates that the test coefficients c’ and a are significant. Then, in the third step, both the environmental regulation intensity and the acquisition channel (M11) were included in the mediation effect test model. The coefficient values of environmental regulation intensity (X2) and acquisition channel (M11) were found to be significant at the 1% and 5% statistical levels, respectively; that is, c’ was significant and b was significant. This indicates that there is a partial mediating role of access channels in the impacts of environmental regulation intensity on farmers’ AGP willingness.

With regard to acquisition quality as a two-dimensional variable, the regression results from the first and second steps—analyzing the relationships between environmental regulation intensity and farmers’ AGP willingness, as well as between environmental regulation intensity and acquisition quality—demonstrate that environmental regulation intensity (X2) has a significant positive effect on both farmers’ AGP willingness (Y1) and acquisition quality (M12) at the 1% statistical level (Figure 6c). This confirms that the test coefficients c’ and a are significant. Then, in the third step, both environmental regulation intensity (X2) and acquisition quality (M12) were included in the mediation effect test model. The coefficient values of environmental regulation intensity (X2) were not significant, while the coefficient values of acquisition quality (M12) were significant at the 1% and 5% statistical levels, respectively, indicating that c’ was significant and b was significant. This indicates that acquisition quality plays a partial mediating role in the impacts of environmental regulation intensity on farmers’ AGP willingness.

In terms of acquisition evaluation of two-dimensional variables, regression analysis was conducted on the relationship between environmental regulation intensity and farmers’ AGP willingness, as well as the relationship between environmental regulation intensity and acquisition evaluation in the first and second steps (Figure 6d). The results showed that environmental regulation intensity (X2) had a significantly positive impact on farmers’ AGP willingness (Y1) and acquisition evaluation (M13) at the 1% and 5% statistical levels, respectively, indicating that the test coefficients for c and a were significant. Then, in the third step, both the environmental regulation intensity and the acquisition evaluation were included in the mediation effect test model of this section. The coefficient value of environmental regulation intensity (X2) was significant at the 1% statistical level, while the coefficient value of evaluation obtained (M13) was not significant; that is, c’ was significant (z = 4.83, SE = 0.043) and b was not significant. Next, the Sobel test was conducted based on the above results, and it was found that the results were significant. Meanwhile, the Bootstrap (95% CI, 5000 self-sampling) method was used again for validation, and the indirect effect value was obtained as 0.136, 95% CI [0.087, 0.228]. This interval does not contain 0. This indicates that there is a mediating effect of acquisition evaluations on the impacts of environmental regulation intensity on farmers’ AGP willingness.

Overall, it can be concluded that the expected impacts of the overall index of acquisition technology in H5 have not been validated, but some of its dimensional indicators are validated. As mentioned above, this section further elaborates on the size of the mediating and direct effects, and the formula for calculating the proportion of mediating effects is ab/c. The calculations indicate that the mediating effect of acquisition channels, as a two-dimensional variable, accounts for 8.24% of the total impact of environmental regulation intensity on farmers’ AGP willingness. Meanwhile, acquisition quality plays a more substantial mediating role, contributing 6.83% to the total effect. Additionally, acquisition evaluation mediates 4.29%.

4.2.3. Endogeneity Test Results of Bidirectional Causal Relationship

Table 7. Endogeneity test results of the relationship between technology acquisition and farmers’ AGP willingness.

Step	Standardized Equation	Regression Coefficient Test
Step 1	Y1 = 0.178D	SE = 0.024, Z = 2.932 ***
	Y1 = 0.178D	SE = 0.024, Z = 2.932 ***
	Y1 = 0.178D	SE = 0.024, Z = 2.932 ***
	Y1 = 0.178D	SE = 0.024, Z = 2.932 ***
Step 2	Y1 = 0.161D	SE = 0.079, Z = 6.031 **
	+0.126M1	SE = 0.064, Z = 2.035
	Y1 = 0.149D	SE = 0.023, Z = 2.614 ***
	+0.198M11	SE = 0.065, Z = −2.181 **
	Y1 = 0.166D	SE = 0.064, Z = 2.814 *
	+0.113M12	SE = 0.045, Z = 4.281 **
	Y1 = 0.146D	SE = 0.049, Z = 4.672 **
	+0.155M13	SE = 0.028, Z = 1.861 ***
Step 3	M1 = 0.132D	SE = 0.075, Z = 2.484 **
	M11 = 0.148D	SE = 0.072, Z = 4.233 *
	M12 = 0.106D	SE = 0.024, Z = 2.712 **
	M13 = 0.208D	SE = 0.052, Z = 4.071 ***

*, **, *** indicate significance at the statistical levels of 10%, 5%, and 1%, respectively.

reveals the endogeneity test results between technology acquisition and farmers’ AGP willingness. Meanwhile, Appendix A (Figure A1) provides a detailed calculation process for the mediation effects of bidirectional causal relationships. The various dimensions of technology acquisition, including acquisition channels, acquisition quality, and acquisition evaluation, can play a positive mediating role in identifying variables and farmers’ willingness to engage in green agricultural production. Among them, the total effect of technology acquisition is 0.178, and the indirect effects of acquisition channels, acquisition quality, and acquisition evaluation are 16.46%, 6.73%, and 18.11%, respectively. This indicates that there are indeed some endogeneity issues in the previous regression, but when considering endogeneity issues by identifying variables, the technical acquisition and its mediating effects in various dimensions are robust and reliable.

5. Discussion

This study refined the dimensions of green perception benefits and environmental regulation intensity, utilizing technology acquisition and its multifaceted variables as mediators. It employed a mediation effect model to analyze their impact mechanisms on farmers’ willingness to adopt AGP. Additionally, we integrate practical factors and psychological characteristics that influence farmers’ AGP willingness, explaining how green perception benefits—across economic, environmental, responsibility, and identity dimensions—directly shape farmers’ willingness and how these effects can be mediated through technology acquisition. Furthermore, this study examines the real-world influence of environmental regulation intensity. Existing research mostly explores the policy effects from the perspectives of policy implementation, such as the implementation area of environmental regulation [75], specific implementation content [67], while ignoring the specific impact of farmers as the target of regulation. To address this gap, we incorporate both subjective and objective environmental regulation intensity into a unified research framework, providing a comprehensive analysis of its direct and indirect effects on farmers’ AGP willingness.

From the perspective of direct effects, green perception benefits positively influence farmers’ willingness to adopt AGP. Previous studies have highlighted the environmental improvements associated with encouraging farmers to implement AGP [76,77]. However, the realization of this effect depends on farmers’ willingness and ability to adopt AGP technologies effectively. There are also studies attempting to explore the perceived factors that affect farmers’ AGP willingness, but these studies mostly focus on single factors such as value perception [78], risk perception [79], and technology perception [80]. Moreover, the willingness of farmers to engage in AGP is largely related to their perceived expected economic benefits, such as premiums, subsidies, etc. [4]. In addition, the government’s environmental regulation policies will increase explicit costs and implicit punishment measures, especially the uncertainty of the frequency and timing of spot checks. For farmers dominated by economic rationality, their green perception benefits can provide stable positive incentives, such as purchase prices with basic guarantees, grain subsidies, etc. [6]. The results of this study further reveal the strength of the perceived impacts on the four aspects of economy, environment, responsibility, and identity. Economic and identity benefits have a significant impact on farmers’ AGP willingness, while environmental and responsibility benefits do not show a significant impact. On the one hand, when farmers perceive a significant increase in economic value, they are often more willing and capable of using chemical inputs such as fertilizers and pesticides appropriately. As stated by Faccioli et al. (2024) [49], it is crucial to provide farmers with more knowledge about long-term investment returns if they perceive economic income as the main obstacle to adopting AGP. On the other hand, it is equally important to enhance farmers’ sense of social responsibility and environmental identity in adopting AGP technologies. Only when farmers have a strong sense of environmental protection identity can their willingness to engage in AGP be effectively enhanced.

From the perspective of indirect effects, although the expected impacts of the overall index of technology acquisition have not been verified, its dimension significantly mediates the relationship between green perception benefits and farmers’ AGP willingness. Among them, acquisition quality and acquisition evaluation play a key role in the impacts of green perception benefits on farmers’ AGP willingness. This is because, on the one hand, farmers perceive that AGP can bring some benefits, such as higher prices of agricultural products, and they will be more proactive in acquiring green technologies. On the other hand, technology acquisition plays a bridging role in it. Without the intermediate link, it may be difficult to directly translate green perception benefits into farmers’ production intentions. Moreover, the ranking of the proportion of each dimension in the total effect is acquisition evaluation > acquisition quality > acquisition channels. Generally speaking, the higher the trust of farmers in new production models, the higher the probability of service-related practices being adopted [81]. The results of this study enrich this viewpoint, as for farmers, the key is not to acquire more types of AGP technologies, but whether this technology meets production needs and can effectively save costs; that is, improving the acquisition quality is more important than the acquisition channels. In addition, we explained the two-dimensional variable of acquisition evaluation through three indicators: interpersonal communication, media platforms, and agricultural technology promotion. The results showed that acquisition evaluation plays a completely mediating role in the impacts of green perception benefits on farmers’ AGP willingness. There have been studies exploring the impacts of technology acquisition sources on farmers’ adoption of AGP technologies [60,61]. However, most of these studies only focus on the importance of technology intermediaries, but ignore the specific practical effects of technology acquisition sources. This study focuses on asking farmers the following question during the data collection stage: “How do these media help with production activities?”, which helps to enrich the understanding of the impact of acquisition evaluation on farmers’ AGP willingness.

In addition, empirical testing of the impacts of environmental regulation intensity on farmers’ AGP willingness provides a reference basis for formulating agricultural environmental policies. The results show that environmental regulation intensity and its two dimensions have a significant effect on improving farmers’ AGP willingness, with subjective regulation intensity having a relatively greater impact. From a subjective perspective, a study on social psychology suggests that psychological characteristics are the primary factor influencing farmers’ AGP willingness [5,25]. It is necessary to strengthen advanced concepts of AGP by deepening farmers’ subjective understanding of environmental regulatory policies, thereby enhancing their willingness [22]. In particular, based on information obtained through field research, farmers are more concerned about the subjective impacts of economic incentive regulations on them, rather than constraints or punitive regulations. This is mainly because AGP is a new agricultural production model, and farmers pay more attention to environmental regulatory policies that can help them directly cope with economic losses. From an objective perspective, the number of times the research area adopts relevant regulatory measures (such as supervision, incentives, and guidance) for AGP each year can effectively quantify the actual effect of environmental regulatory policies. This result is supported by Feng et al. (2023) [75] and Dong et al. (2024) [27], which emphasize that the formulation of government regulations is conducive to leveraging the environmental improvement effects of AGP technologies.

We also used a mediation effect model to examine how the environmental regulation intensity and its two dimensions affect farmers’ AGP willingness through technology acquisition. The environmental regulation intensity mainly affects farmers’ AGP willingness through the intermediary effects of acquisition channels, acquisition quality, and acquisition evaluation. Its proportion in the total effect is ranked as follows: acquisition quality > acquisition channels > acquisition evaluation. Compared to existing research [52,62], the results of this study not only reveal the differences in their impact willingness but also indicate that acquisition quality is the key to the effectiveness of AGP technology. Here, we choose to measure the quality indicators of obtaining information effectiveness, relationship closeness, and cost savings of AGP technologies. Despite the government’s attempts to widely disseminate the advantages of AGP technology, farmers have not actively adopted the technology, indicating that farmers are influenced in their adoption decisions. We attempt to reveal the mechanism by which acquisition quality affects farmers’ AGP willingness through the above indicators. Zhang et al. (2024) [82] pointed out that the effectiveness of topdressing technology depends on a reasonable amount of fertilizer, and theoretically, the average profit return can be as high as 36%. However, farmers often fail to achieve this profit in practice, mainly because the technical recommendations of the Kenyan Ministry of Agriculture ignore the conditions of small-scale farms. Based on this study, we emphasize that when the government promotes AGP technology, it should pay attention to the effectiveness, closeness, and frugality of technical information.

Some limitations that exist also need to be acknowledged. Firstly, the findings of this study, particularly the dimension division and impacts of environmental regulation intensity, as well as the mechanism and effects of the technology acquisition dimension, provide universal promotion solutions for developing countries with similar experiences to China. However, although this study provides some informative references, regional specificity should also be taken into account when promoting in developing countries, as our data mainly comes from northern China. Secondly, the data in this study are cross-sectional data obtained through field investigations, which may lead to potential social desirability bias. Therefore, future research can consider using dynamic panel data, which will enable us to track the changes of the same object at different time points, thereby revealing long-term trends and dynamic processes. Thirdly, based on the limitations mentioned above, we can use dynamic data in the future to more accurately evaluate the sustainability of AGP willingness and track the long-term implementation effects of environmental regulations, providing valuable references for formulating more reasonable policies and incentive measures.

6. Conclusions and Implications

This study used micro-farmer data from major grain-producing counties in the Huang Huai Hai Plain of China to empirically analyze the direct impacts of green perception benefits and environmental regulation intensity on farmers’ AGP willingness, and further examine the intrinsic mechanisms of technology acquisition.

The main research conclusions are as follows: (1) Both green perception benefits and environmental regulation intensity positively influence farmers’ AGP willingness. Among the two-dimensional variables, economic benefits and identity benefits have a significant positive impact, with economic benefits exerting the greatest relative effect (0.178), followed by identity benefits. In contrast, environmental benefits and responsibility benefits do not show a significant impact. (2) Both objective and subjective regulatory intensity positively affect farmers’ AGP willingness, with subjective regulatory intensity having a more pronounced effect (0.341). (3) The overall technology acquisition index plays a mediating role in the impacts of green perception benefits on farmers’ AGP willingness. Among the dimensions of technology acquisition, the mediating effect of green perception benefits on farmers’ AGP willingness follows the order acquisition evaluation > acquisition channels > acquisition quality. In comparison, the overall technology acquisition index does not play a mediating role in the impacts of environmental regulation intensity on farmers’ AGP willingness ranks as follows: acquisition channels > acquisition quality > acquisition evaluation. This study reveals that economic benefits and subjective regulation intensity are the core levers driving farmers’ AGP willingness, and the contribution ranking of various dimensions of technology acquisition is reversed due to policy sources, providing a dimension-level basis for precise intervention.

Based on the findings, the following implications are proposed:

(1)

For farmers and families: The first is to strengthen farmers’ awareness of economic benefits and their sense of responsibility, which can effectively increase their willingness to adopt green production practices, thereby facilitating the widespread application of green technologies. It is crucial to emphasize both economic incentives and value recognition among farmers, fostering their environmental responsibility. This can be achieved through initiatives such as environmental information dissemination, technology adoption subsidies, and promotion of service reforms, which improve farmers’ environmental awareness and skills. The second is to expand technology acquisition channels and strengthen quality evaluation. Broadening access to green production technologies and ensuring proper evaluation of their quality can help farmers effectively learn about and comprehend relevant policies. This not only enhances their sense of identity with green production practices but also deepens their understanding of information related to sustainable food production.

(2)

In terms of the analyzed field: On the one hand, researchers in related fields need to have a deeper understanding of the formation mechanism of green perception benefits, the threshold effects of environmental regulation intensity, as well as tracking the dynamic evolution path of farmers’ recognition of economic benefits to environmental value under policy intervention. On the other hand, the researchers can explore the integration of interdisciplinary fields such as agricultural economics and policy science. This can not only better coordinate the role of technology promotion and policy incentives, but also more comprehensively quantify the correlation logic between environmental regulatory policies and farmers’ behavioral decisions.

(3)

As for the policymakers: Primarily, government departments should enhance environmental regulatory policies for green technologies. Strengthening environmental regulatory policies related to green technology promotion is essential. This can drive significant agricultural technological innovation while minimizing input factor mismatches and information asymmetry in production. Additionally, the government should strengthen the training of agricultural technology extension personnel, improve their professional competence and service level, and establish a scientific and reasonable green production technology evaluation system for technology. Meanwhile, environmental regulation intensity should be tailored to local conditions, taking into account regional variations in agricultural non-point source pollution and farmers’ existing green production practices. Finally, government departments should also leverage environmental regulation to encourage green production. Strengthening environmental regulation intensity can incentivize farmers to participate in green production and allow them to experience the environmental improvements that result from active engagement in sustainable agricultural practices.