The Impact of Agricultural Socialization Services on the Ecological Protection of Rice Farmland in Jianghan Plain, China

Abstract

The ecological protection of cultivated land is crucial for advancing high-quality agricultural development. This study analyzes the impact of agricultural socialization services on the ecological conservation of farmland, focusing on the reduction of chemical fertilizers and pesticides among rice farmers in the Jianghan Plain area. Utilizing data from 743 farmer household surveys conducted in 2023, the findings reveal that agricultural socialization services significantly encourage farmers to reduce chemical input usage. For every 1% increase in the level of agricultural social services, the average fertilizer use per hectare will decrease by 14% and the average pesticide use per hectare will decrease by 16.4%. The study identifies scale operation, factor substitution, and alleviation of capital constraints as mediating factors enhancing the efficacy of these services. Furthermore, heterogeneity analysis indicates that these services are more effective in reducing chemical inputs among large-scale and newer-generation farmers compared to their small-scale and older counterparts. Additionally, technology-intensive socialized services exhibit a stronger impact on reducing chemical inputs than labor-intensive services.

1. Introduction

Historically, there has been a lack of awareness among farmers about the importance of protecting cultivated land. In China, the degradation of land quality is primarily due to the widespread and often incorrect application of chemical fertilizers and pesticides. Farmers commonly apply these chemicals in excess, disregarding the recommended agronomic rates and timings. This practice significantly contributes to the deterioration of soil health. Moreover, the prevailing land use practices, which emphasize immediate crop yields at the expense of long-term soil sustainability, involve intensive cultivation without sufficient fallow periods or crop rotation. Such practices exacerbate soil depletion and erosion. As of 2019, medium- and low-grade cultivated lands accounted for 70.5% of China’s total cultivated area. The quality of these lands is approximately 30% lower than similar lands in developed countries of Europe and the United States, indicating a significant disparity in agricultural land quality that has critical implications for sustainable agricultural productivity and soil health. The degradation of land quality is further aggravated by industrial pollution and the practice of substituting poor-quality land with parcels of higher quality. Furthermore, existing land management systems frequently neglect essential practices such as Integrated Pest Management (IPM) and common soil health monitoring methods. These methods generally encompass routine soil testing to assess nutrient content and pH levels, the implementation of cover crops to improve soil structure, and the adoption of precision agriculture technologies that customize soil treatments based on detailed soil characteristic mapping. This decline in land quality poses a substantial threat to the security of agricultural products and food in China. Recognizing the urgency of the situation, the Chinese government has prioritized the ecological protection of cultivated land. Recent directives, notably the No. 1 Central Documents of 2021, 2022, and 2023, have outlined measures for straw utilization, recycling of agricultural materials, and stringent protection policies for at-risk lands. These initiatives underscore the government’s commitment to enhancing land quality and ensuring the safe use of contaminated areas. The development of agricultural social services has provided a new avenue to bolster ecological protection. Agricultural socialization involves organized, community-based services that support farmers throughout the agricultural production cycle, thereby enhancing productivity and promoting ecological sustainability. This system delivers comprehensive assistance in critical areas such as land preparation, sowing, integrated pest management, and crop rotation—essential components of sustainable agriculture. Specifically, the introduction of cover crops and green manure during land preparation enhances soil structure and fertility, diminishing the reliance on chemical inputs. In the realm of pest management, these services advocate for biological controls that maintain populations of beneficial insects, thus minimizing pesticide use. This holistic strategy embodies the principle of “hoarding food in the land and storing food in technology”, which underscores the commitment to sustainable food production. Despite their importance, there is a notable scarcity of research on the impact of these services on land ecological protection. This gap underscores the need for further investigation to optimize and expand the implementation of agricultural social services.

Current research on agricultural social services and cultivated land protection primarily examines three areas: the impact of agricultural social services on farmers’ green production, food production, and reductions in chemical fertilizer use. Firstly, agricultural socialization services facilitate green production among farmers. Studies indicate these services significantly enhance farmers’ willingness to engage in green production [1], encourage green production behaviors [2,3], and reduce agricultural carbon emissions [4,5]. Consequently, these services support the transition towards sustainable agricultural practices [6]. Secondly, agricultural socialization services contribute to enhanced food production. Research shows that these services discourage farmers from abandoning their fields [7], facilitate land transfers [8], and improve grain productivity [9]. These services also increase grain yield per unit area [10] and secure farmers’ income from grain cultivation [11,12]. Thirdly, there is ongoing debate regarding the effectiveness of agricultural socialization services in reducing chemical fertilizer use. Some studies assert that these services considerably decrease the use of chemical fertilizers in rice and wheat cultivation, with varying impacts across different service types [13,14]. Other studies noted a significant reduction in both the total and per unit area usage of chemical fertilizers, with a 1% increase in service expenditure leading to reductions of 0.055% and 0.443%, respectively [15]. Conversely, other studies argue that these services do not reduce the fertilizer used in wheat but merely lessen the disparity in fertilizer application among farmers [16]. Additionally, these services reportedly have negligible effects on fertilizer reduction among part-time farmers [17].

This paper explores three core issues of agricultural social services and their impacts on farmland ecological protection, as follows: (1) The direct impact of agricultural social services on reducing farmers’ use of fertilizers and pesticides. (2) How mediating variables such as business scale, capital constraints and factor substitution affect the effectiveness of agricultural social services in reducing fertilizer and pesticide use. (3) The differences in the impacts of agricultural social services on farmers of different farm sizes and generations. This study builds on previous literature while addressing identified gaps in the research on the impact of agricultural socialization services on the reduction of chemical fertilizers and pesticides. The shortcomings of existing studies include the following: (1) Most prior analyses of the mechanisms through which agricultural socialization services influence the reduction of chemical fertilizers and pesticides have remained theoretical and lack empirical validation. (2) While previous research has generally explored the overall effects of agricultural socialization services on the reduction of these substances, there has been limited investigation into their heterogeneous impacts.

The innovations of this paper are mainly reflected in three aspects, as follows: (1) Filling the research gap—Previous studies have mainly focused on the broad impacts of agricultural social services on green production and reducing chemical inputs, but have not studied the specific ecological impacts and the differences between different groups of farmers in depth. This study fills these gaps by studying the impact of these services on ecological protection, especially on rice cultivation, which is a topic less explored in the existing literature. (2) Methodological innovation—Unlike previous studies that usually rely on qualitative assessments or small and non-representative samples, this study conducted a comprehensive survey of 743 rice farmers. This large-scale empirical approach can provide a broader and more general understanding of the impact of agricultural social services. (3) Expanding the depth of analysis—This paper not only documents the impact, but also explores the mechanisms behind it, such as large-scale operations, factor substitution, and easing capital constraints. This multifaceted analysis enriches our understanding of the “how” and “why” behind the observed changes, making a significant theoretical contribution.

2. Materials and Methods

2.1. Theoretical Analysis

2.1.1. The Direct Impact of Agricultural Socialization Services on the Ecological Protection of Farmers’ Farmland

The ecological protection of farmers’ cultivated land can be approached in two primary ways; on one hand, through self-execution, and on the other hand, through indirect execution. Self-execution involves farmers independently purchasing and utilizing green production factors, managing all stages of cultivation—from land preparation and sowing to pest control and harvesting—without external assistance. In this model, farmers serve as the principal agents of ecological protection. Conversely, indirect execution occurs when farmers engage agricultural socialization services. These organizations invest in green production factors and manage various farming activities, including sowing, pest control, and harvesting. In this model, agricultural socialization service organizations act as the agents of ecological protection [18]. Indirect execution offers three primary advantages over self-execution, as follows: reduced costs of green production factors, enhanced adoption of green production technologies, and more efficient operation and management. Agricultural socialization service organizations, due to their large-scale purchases, have significant bargaining power, which allows them to procure inputs at costs below market prices, thereby lowering production costs for farmers. Furthermore, as leaders in adopting new agricultural technologies, these organizations have access to sufficient capital, government support, and expert resources to advance the use of green technologies such as precision irrigation systems, integrated pest management (IPM) practices, and soil health monitoring tools. Finally, their modern organizational structures and management practices enable them to deliver comprehensive, high-quality services efficiently, fostering the more effective ecological protection of cultivated land. Based on this analysis, the following hypothesis is proposed:

Hypothesis 1.

Agricultural social services significantly enhance the ecological protection of farmers’ cultivated land.

2.1.2. The Mechanism of Agricultural Socialization Services in Ecological Protection of Farmers’ Cultivated Land

Agricultural social services encompass a spectrum of activities, including land preparation, sowing, management, harvesting, and market support [5]. These services are facilitated through government backing, cooperative organizations, and partnerships with local institutions [3]. Designed to mitigate farmers’ labor requirements, enhance access to technology and markets, and foster better ecological conditions, these services enable farmers to utilize agricultural technology more efficiently [4]. This, in turn, boosts productivity and sustainability. Financial constraints significantly hinder the participation of small farmers in agricultural modernization. Due to limited resources, low returns, and high risks, these farmers often lack the funds necessary for adopting modern agricultural practices [19]. The highly specific nature of agricultural machinery and equipment across different stages of production exacerbates this issue, making it challenging for small farmers to afford necessary machinery for all activities, thereby reinforcing their financial limitations [20]. Agricultural social service organizations can mitigate these financial barriers by providing agricultural machinery services at reduced costs. This support not only enables small farmers to access essential machinery without the need for large capital investments but also decreases the cost and risk associated with such purchases. Furthermore, by alleviating the capital constraints associated with the specificity of agricultural machinery and equipment, these organizations facilitate small farmers’ investments in funds and the adoption of agricultural technologies. This support not only enhances agricultural production efficiency but also increases the likelihood of farmers adopting technologies for the ecological protection of cultivated land, thereby promoting sustainable agricultural practices. Based on these considerations, the following hypothesis is proposed:

Hypothesis 2.

Agricultural social services enhance the ecological protection of farmers’ cultivated land by easing financial constraints.

The promotion of factor substitution mechanisms is crucial in advancing the ecological protection of cultivated land. This mechanism can be broken down into two primary categories: agricultural labor substitution and technological substitution. First, agricultural labor substitution addresses the severe shortage of agricultural labor, which has been exacerbated by the rapid urbanization and industrial migration of young and middle-aged rural workers. Agricultural socialization services facilitate the mechanization of key farming processes such as plowing, sowing, pest control, and harvesting. This mechanization not only compensates for the lack of human labor but also ensures the standardized application of agricultural production inputs like fertilizers and pesticides, thereby reducing the likelihood of their excessive use and mitigating the adverse effects of labor shortages on cultivated land ecological protection. Second, technological substitution is pivotal for enhancing cultivated land ecological protection. Despite the critical role of agricultural science and technology in this regard, limited financial resources among farmers and scarce governmental support for agricultural technology extension often hinder access to advanced agricultural technologies. Agricultural socialization services bridge this gap by introducing advanced technologies and additional high-value funds into the agricultural sector. This facilitates easier access for farmers to cutting-edge cultivated land ecological protection technologies, subsequently encouraging the adoption of environmentally sustainable practices. Based on the analysis, the following hypothesis is proposed:

Hypothesis 3.

Agricultural socialization services enhance the ecological protection of farmers’ cultivated land by promoting factor substitution.

The mechanism for expanding business scale through agricultural socialization services plays a pivotal role in the ecological protection of cultivated land. This approach operates through two main avenues: mitigating land abandonment and facilitating land transfers. On one hand, agricultural socialization services counteract the abandonment of cultivated land. In our country, the small and highly fragmented nature of per capita cultivated land complicates agricultural production. This challenge is exacerbated by the significant loss of agricultural laborers and the serious aging of the rural workforce, resulting in a shortage of labor and substantial abandonment of cultivated land. By mechanizing the entire process of plowing, sowing, pest control, and harvesting, agricultural socialization services compensate for the lack of family labor, boost farmers’ enthusiasm for grain cultivation, and help mitigate the abandonment of cultivated land [21]. On the other hand, these services promote land transfers. Agricultural socialization services alleviate capital, labor, and technological constraints faced by farmers. Those with planting experience and financial capability are encouraged to seek economies of scale by transferring more land, thereby expanding their operations. As farmers’ business scales grow, the efficiency of agricultural social services improves. This efficiency leads to a significant reduction in the use of chemical fertilizers and pesticides per unit area, lowers farming costs, and further incentivizes farmers to engage in the ecological protection of their cultivated land [22]. Based on this analysis, the following hypothesis is proposed:

Hypothesis 4.

Agricultural social services enhance the ecological protection of farmers’ cultivated land by promoting the expansion of business operations.

2.2. Research Design

2.2.1. Data Sources

The data for this study were obtained through a questionnaire survey conducted among rice farmers in the Jianghan Plain during August and September 2023. The Jianghan Plain was selected for several reasons: it is a major rice-producing region in the country, it has seen significant development in agricultural socialized services, and it has been the focus of extensive pilot projects in this area. Additionally, the Jianghan Plain boasts unique advantages in terms of cultivated land resources and is a focal point of agricultural production, highly prioritized by the government. It also plays a critical role in the rural revitalization efforts in Hubei Province. The sampling strategy involved a combination of stratified and random sampling methods to select representative farmers for household surveys. Based on the development level of agricultural socialization services and the characteristics of cultivated land resources in the region, eight counties and cities within the Jianghan Plain were chosen. From each county, three to four towns were randomly selected, and from each town, two to three villages were chosen. Within each village, 20–25 farmers were identified as sampling points. Ultimately, approximately 860 farmers were randomly selected as subjects for this study. Of these, 743 valid questionnaires were returned, yielding a questionnaire validity rate of 86.39%.

The sample size of 743 farmers was determined to balance statistical power with practical feasibility specific to the Jianghan Plain. This number was targeted to ensure adequate representation across varying farm sizes and generations, enabling robust subgroup analyses. Such a sample size provides a 95% confidence level with a 5% margin of error, taking into account the variability noted in preliminary studies. The sampling methodology employed a multistage stratified random sampling technique to optimize representativeness. Initially, counties in the Jianghan Plain were categorized based on the intensity of agricultural socialization services. Subsequently, villages within these strata were randomly selected, and from these, farming households were also randomly chosen. This approach effectively manages variability and minimizes sampling bias across different layers of the population. Despite the thoroughness of the sampling strategy, potential limitations remain. Non-response bias is a concern, as some farmers selected for the study may have opted not to participate. Additionally, the representativeness of the results could be compromised if the chosen villages exhibit atypical agricultural practices that do not accurately reflect those of the broader Jianghan Plain region.

2.2.2. Model Construction

To assess the influence of agricultural socialization services on the ecological protection of farmers’ cultivated land, this study builds on existing literature and employs the ordered logit (ologit) model. The model is structured as follows:

$$Ecologica{l}_i=\beta +{\beta }_{i}Servic{e}_i+{\mathrm{z}}_k{X}_i+{\epsilon }_i$$

(1)

In Formula (1), the variable Ecological_i represents the ecological protection of cultivated land for the i-th farmer, quantified by the average fertilizer application rate (expressed in jin per hectare) and the average pesticide application rate per hectare. The variable Service_i denotes the level of engagement with agricultural socialization services, measured by the number of service links utilized by the farmer. The model includes X_k, a set of control variables, which account for demographic and socio-economic factors such as age, gender, educational level, health status, size of the agricultural labor force, whether the individual holds a cadre position, and whether they are a member of a cooperative. Lastly, ε_i represents the random error term.

To examine the influence of agricultural socialization services on the ecological protection of farmers’ cultivated land, this study employs a mediating effect model. This model is based on the methodologies proposed by Baron and Kenny and further refined by Wen Zhonglin and colleagues [23,24], as follows:

$$M_i=a+a_{1}Servic{e}_i+z_k{X}_i+{\epsilon }_i$$

(2)

$$Ecologica{l}_i=\gamma +{\gamma }_{2}Servic{e}_i+{\gamma }_2{M}_i+{\mathrm{z}}_k{X}_i+{\epsilon }_i$$

(3)

Among them, Formula (1) is the total effect of agricultural socialization services on the ecological protection of farmers’ cultivated land (amount of chemical fertilizers and pesticides applied), Formula (3) ${\gamma }_1$ is the direct effect of agricultural socialization services on the ecological protection of farmers’ cultivated land, and ${\alpha }_1{\gamma }_2$ is the agricultural. The relative size of the intermediary effect of socialized services on the ecological protection of farmers’ cultivated land is the ratio of the intermediary effect to the total effect. M_it is the intermediary variable, which mainly includes capital constraints, factor substitution, and business scale. The remaining dependent variables, independent variables, and control variables are the same as before.

Statistical analyses ${\beta }_1$ were conducted using Stata15.0 software, selected for its advanced regression capabilities and comprehensive support for mixed-effects models, crucial for handling the multilevel data typical in agricultural research. Additionally, Stata’s robust graphical features were utilized to generate high-quality visuals that effectively illustrate our findings.

2.2.3. Variable Setting

Dependent Variable. The dependent variable in this study is the ecological protection of cultivated land, which focuses on maintaining the ecological balance of such land. This encompasses efforts to prevent soil pollution—specifically from chemical fertilizers, pesticides, and heavy metals—as well as combating land degradation phenomena including salinization, desertification, and soil erosion. Excessive use of chemical fertilizers and pesticides has been identified as a primary cause of soil pollution. Consequently, reducing the application of these substances is considered a critical strategy for enhancing the ecological protection of cultivated land [25]. Drawing on previous studies, this study uses the amount of fertilizer and pesticide applied by farmers as indicators to assess the impact of agricultural socialization services on the ecological protection of cultivated land [26,27].

Core Explanatory Variables. Agricultural socialization services constitute the core explanatory variables in this study. These services encompass a range of activities provided by agricultural socialization service organizations, designed to support farmers throughout the agricultural production process. The services include land preparation, sowing, seedling transplanting, fertilization, pest and disease control, irrigation, harvesting, transportation, and drying, among other production-related tasks. The adoption of agricultural socialization services, as well as the extent to which farmers utilize various service links, significantly impacts the ecological protection of cultivated land. To quantify this impact, this article selects seven critical service links: land preparation, sowing, raising and transplanting, fertilization, pest control, irrigation, and harvesting, based on previous research [14,22]. The level of engagement with agricultural socialization services is measured by the number of service links utilized by farmers, with a possible range from 0 to 7.

Mechanism Variables. Based on the theoretical analysis presented earlier, this study identifies that agricultural social services influence the ecological protection of farmers’ cultivated land through three primary mechanisms: capital constraints, factor substitution, and business scale. Consequently, this paper selects these three aspects as the focal mechanism variables. Research by Yang Zi et al. indicates that the specific nature of agricultural machinery and equipment assets significantly contributes to the capital constraints experienced by small-scale farmers [20]. In this context, capital constraints are measured by the acquisition of agricultural machinery, coded as “1 = yes, 0 = no”. Furthermore, agricultural social services primarily facilitate the replacement of agricultural labor. Thus, factor substitution is assessed based on the capacity of these services to alleviate family labor shortages, denoted as “1 = yes, 0 = no”. Additionally, under the prevailing institutional framework, the extent of land scale management is most directly reflected by changes in land transfer among farmers, as noted by Jiang Song et al. [28]. Therefore, business scale is quantified by the occurrence of land transfers, indicated by “1 = yes, 0 = no”.

Control Variables. In order to account for potential confounders that may influence the effectiveness of agricultural socialization services and the ecological protection of cultivated land, this study incorporates several control variables based on insights gleaned from existing research. These variables include the farmers’ age, gender, education level and health status, the size of the agricultural labor force, the cadre status, and membership in cooperatives. The selection of these variables is aimed at isolating the specific effects of agricultural socialization services on ecological outcomes by controlling for demographic and socio-economic factors that could also impact these results. The detailed measurement methods for each of these variables are presented in

Table 1. Descriptive statistical analysis of variables.

Variable	Measurement and Assignment
Agricultural social services	The number of production links that farmers adopt in socialized services, the value ranges from 0 to 7
Fertilizer application amount	Amount of chemical fertilizer applied per acre (catties)
Pesticide application amount	Quantity of pesticides and fertilizers per hectare (jin)
Changes in fertilizer application rates	Changes in the amount of chemical fertilizer applied per acre: 1 = significantly reduced, 2 = reduced, 3 = no change, 4 = increased, 5 = significantly increased
Changes in pesticide application rates	Changes in pesticide application rate per acre: 1 = significantly reduced, 2 = reduced, 3 = no change, 4 = increased, 5 = significantly increased
Agricultural socialization service expenditures	Total service cost per hectare in all links (unit: CNY thousand/hectare)
Fertilization services	Purchase service fee per hectare during fertilization (unit: CNY thousand/hectare)
Pest and disease control services	Purchase service cost per hectare for pest control (unit: thousand yuan/hectare)
Business scale	Whether the farmer transferred the land: 1 = yes, 0 = no
Financial constraints	Whether farmers purchase agricultural machinery: 1 = yes, 0 = no
Element substitution	Can agricultural socialization services alleviate family labor shortage? 1 = can, 0 = cannot
Whether to adopt agricultural socialization services	1 = yes, 0 = no
Whether to adopt fertilization services	1 = yes, 0 = no
Whether to adopt pest control services	1 = yes, 0 = no
Does your village provide agricultural socialization services? 1 = provided; 0 = not provided	1 = yes, 0 = no
Is it convenient for you to obtain agricultural socialization services?	1 = yes, 0 = no
Age	Actual age
Gender	1 = male, 0 = female
Culture	1 = Primary school and below, 2 = Junior high school, 3 = High school, 4 = College and above
Healthy	1 = very poor, 2 = poor, 3 = average, 4 = good, 5 = good
Agricultural labor	Number of family agricultural laborers
Is it a cadre?	1 = yes, 0 = no
Total household income	actual income
Cultivated area	actual area
Soil fertility	1 = very poor, 2 = poor, 3 = average, 4 = better, 5 = very good
Whether to join a cooperative	1 = yes, 0 = no
Is it agricultural technology training?	1 = yes, 0 = no
Disaster frequency	1 = hardly ever, 2 = occasionally, 3 = frequently
Whether to buy agricultural insurance	1 = yes, 0 = no

.

3. Results and Discussion

3.1. Baseline Regression Results

Table 2. The impact of agricultural socialization services on the amounts of chemical fertilizers and pesticides used by farmers.

	(1)	(2)	(3)	(4)
	Fertilizer Application Amount	Pesticide Application Amount	Fertilizer Application Amount	Pesticide Application Amount
Agricultural social services	−0.144 ***	−0.177 ***	−0.140 ***	−0.164 ***
	(−4.38)	(−5.41)	(−4.00)	(−4.72)
Gender			−0.080	0.063
			(−0.52)	(0.41)
Age			0.005	−0.007
			(0.62)	(−0.90)
Culture			0.017	−0.012
			(0.75)	(−0.50)
Healthy			−0.105	−0.026
			(−1.52)	(−0.37)
Agricultural labor			0.047	−0.122
			(0.63)	(−1.64)
Is it a village cadre?			−0.266 **	−0.113
			(−2.14)	(−0.84)
Total household income			−0.012	−0.042 *
			(−0.52)	(−1.79)
Cultivated area			−0.001	−0.001
			(−0.18)	(−0.35)
Soil fertility			−0.113 *	0.003
			(−1.89)	(0.09)
Whether to join a cooperative			0.371 ***	0.031
			(2.69)	(0.22)
Is it agricultural technology training?			−0.208	−0.144
			(−1.62)	(−1.08)
Disaster frequency			0.039	0.129
			(0.43)	(1.41)
Whether to buy agricultural insurance			0.019	−0.117
			(0.22)	(−1.36)
N	743	744	736	737

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

presents the effects of agricultural socialization services on the usage of chemical fertilizers and pesticides by farmers. The table is divided into four columns: Columns (1) and (2) display the estimation results without control variables, while Columns (3) and (4) include control variables in the estimations. The findings indicate a significantly negative coefficient for agricultural socialization services in Columns (1) and (2), with statistical significance at the 1% level. This suggests that agricultural socialization services lead to a substantial reduction in the application of chemical fertilizers and pesticides. Upon incorporating control variables in Columns (3) and (4), the magnitude of the coefficient decreases but remains significantly negative at the 1% level, reinforcing the initial findings. Specifically, a 1% increase in the level of agricultural socialization services is associated with a 14% decrease in chemical fertilizer usage and a 16.4% decrease in pesticide usage per hectare. These results confirm that agricultural socialization services contribute positively to the ecological protection of farmers’ cultivated land, as evidenced by the validation of Hypothesis 1 of this study. Compared to traditional methods of manual fertilization and pesticide application, agricultural socialization service organizations employ advanced reduction technologies and equipment. This specialized, intensified, and scaled approach enhances the precision, speed, and efficiency of applications, improving the utilization rate of chemical inputs and effectively reducing their overall usage.

3.2. Endogeneity Analysis

Considering potential issues of bidirectional causality and omitted variables, this study addresses the problem of endogeneity. The bidirectional causality could manifest as follows: on one hand, excessive fertilization and pesticide application by farmers can lead to soil pollution and a deteriorating ecological environment, prompting increased governmental efforts to protect and manage the agricultural ecosystem and encourage the adoption of agricultural social services. On the other hand, given the high costs and low efficiency of manual fertilization and pesticide application, farmers, acting as rational economic agents, may opt for agricultural social services to reduce costs and enhance production efficiency. To address the bias resulting from endogeneity, this study employs an instrumental variable approach, as suggested by Zhang Anran et al. [29]. The instrumental variable chosen is the accessibility of agricultural socialization services. It is posited that farmers with easier access to these services are more likely to adopt them compared to those for whom access is difficult. However, the ease of access to agricultural socialization services does not necessarily correlate with a reduction in the use of chemical fertilizers and pesticides. Therefore, while the convenience of obtaining agricultural socialization services may directly influence the adoption of such services, it does not directly affect the volume of fertilizers and pesticides used.

Table 3. Endogeneity test.

	(1)	(2)	(3)	(4)
	Stage 1	Stage 2	Stage 1	Stage 2
	Agricultural Social Services	Fertilizer Application Amount	Agricultural Social Services	Pesticide Application Amount
Agricultural social services		−0.2252 *		−0.2667 *
		−0.133		−0.156
Instrumental variable	0.7690 ***		0.7896 ***
	−0.186		−0.186
Gender	0.188	−0.134	0.184	0.010
	−0.166	−0.095	−0.166	−0.115
Age	0.0142 *	0.000	0.0142 *	−0.005
	−0.008	−0.005	−0.008	−0.006
Culture	0.041	0.002	0.041	−0.010
	−0.026	−0.015	−0.026	−0.018
Healthy	−0.088	−0.037	−0.087	0.009
	−0.075	−0.043	−0.075	−0.052
Agricultural labor	0.085	0.020	0.086	−0.0971 *
	−0.080	−0.045	−0.080	−0.054
Is it a village cadre?	0.087	−0.1757 **	0.089	−0.100
	−0.139	−0.079	−0.139	−0.095
Total household income	−0.014	−0.005	−0.014	−0.0305 *
	−0.025	−0.014	−0.025	−0.017
Cultivated area	0.001	−0.001	0.001	−0.001
	−0.003	−0.002	−0.003	−0.002
Soil fertility	0.017	−0.0487 **	0.018	−0.002
	−0.041	−0.023	−0.041	−0.028
Whether to join a cooperative	0.110	0.1928 **	0.096	−0.012
	−0.152	−0.086	−0.151	−0.103
Is it agricultural technology training?	0.190	−0.1588 *	0.200	−0.121
	−0.141	−0.082	−0.141	−0.100
Disaster frequency	0.3116 ***	−0.009	0.3051 ***	0.009
	−0.091	−0.064	−0.091	−0.075
Whether to buy agricultural insurance	−0.6548 ***	0.098	−0.6511 ***	0.026
	−0.089	−0.099	−0.089	−0.116
Cragg-Donald Wald F-statistic	20.699		41.572
Anderson canon. corr. LM statistics	29.475 (p = 0.0000)		36.685 (p = 0.0000)
N	731	731	732	732

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

presents the results of the endogeneity test for the impact of agricultural socialization services on the usage of chemical fertilizers and pesticides. The first-stage regression results in Columns (1) and (3) of Table 3 reveal that the coefficient of the instrumental variable (“Is it convenient for you to obtain agricultural socialization services?”) is significantly positive at the 1% level, aligning with expectations. This significant relationship indicates a strong correlation between the instrumental variable and the endogenous explanatory variables. The Cragg–Donald Wald F statistic exceeds the 10% critical value of 16.38, effectively addressing concerns regarding the weakness of the instrumental variable. Furthermore, the p-value of the Anderson canon. corr. LM statistic is 0.0000, leading to the rejection of the hypothesis of insufficient instrumental variables, thereby validating the appropriateness of the instrumental variable selected for this analysis. The second-stage regression results, displayed in Columns (2) and (4) of Table 3, show that the coefficient of agricultural socialization services is significantly negative at the 10% level, confirming that agricultural socialization services substantially reduce the application of chemical fertilizers and pesticides. Additionally, when comparing these results with those in Table 2, the absolute value of the coefficient for agricultural socialization services in the instrumental variable estimation (Columns (2) and (4) of Table 3) is notably larger. This suggests that the negative impacts of agricultural socialization services on chemical fertilizer and pesticide usage are underestimated when endogeneity issues are not considered.

3.3. Robustness Check

To evaluate the robustness of the baseline regression results, this study implements several robustness tests, each employing different methodologies—substituting the dependent variable, altering the core explanatory variable, and excluding specific samples. These approaches are detailed as follows: ① Substituting the Dependent Variable—The study substitutes the initial dependent variables (amounts of chemical fertilizer and pesticide application) with their respective changes over time. The robustness of these substitutions is assessed, with the results presented in Columns (1) and (2) of

Table 4. Robustness check.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	Changes in Fertilizer Application Rates	Changes in Pesticide Application Rates	Fertilizer Application Amount	Pesticide Application Amount	Fertilizer Application Amount	Pesticide Application Amount	Cut Out Samples	Cut Out Samples
Agricultural social services	−0.558 ***	−0.658 ***					−0.517 ***	−0.641 ***
	(−14.38)	(−16.86)					(−6.68)	(−8.49)
Agricultural socialization service expenditures			−0.319 **	−0.494 ***
			(−1.98)	(−3.10)
Whether to adopt agricultural socialization services					−0.847 ***	−1.719 ***
					(−5.53)	(−10.75)
Gender	0.042	−0.114	−0.088	0.061	−0.009	−0.097	−0.032	−0.207
	(0.28)	(−0.76)	(−0.57)	(0.39)	(−0.06)	(−0.65)	(−0.09)	(−0.54)
Age	−0.010	0.008	0.006	−0.006	−0.012 *	0.005	0.018	0.009
	(−1.49)	(1.15)	(0.81)	(−0.85)	(−1.77)	(0.74)	(0.54)	(0.24)
Culture	−0.041 *	0.004	0.020	−0.009	−0.049 **	−0.002	0.140	−0.063
	(−1.90)	(0.21)	(0.85)	(−0.36)	(−2.15)	(−0.10)	(0.67)	(−0.31)
Healthy	−0.107	0.082	−0.115 *	−0.043	−0.043	0.113 *	−0.004	0.096
	(−1.58)	(1.20)	(−1.66)	(−0.62)	(−0.64)	(1.67)	(−0.03)	(0.76)
Family labor	−0.034	0.025	0.049	−0.122	−0.051	0.007	0.158	0.138
	(−0.47)	(0.36)	(0.66)	(−1.64)	(−0.71)	(0.11)	(1.20)	(1.08)
Is it a village cadre?	0.142	−0.117	−0.246 **	−0.115	0.174	−0.091	0.727	−0.922 *
	(0.95)	(−0.95)	(−1.96)	(−0.87)	(1.16)	(−0.69)	(1.54)	(−1.96)
Total household income	−0.057 ***	−0.018	−0.010	−0.038	−0.058 **	−0.030	−0.120 **	0.005
	(−2.61)	(−0.81)	(−0.44)	(−1.60)	(−2.42)	(−1.39)	(−2.41)	(0.11)
Cultivated area	0.003	0.006 **	0.000	−0.001	0.002	0.006 *	0.002	0.0108 **
	(1.16)	(2.25)	(−0.15)	(−0.33)	(0.74)	(1.96)	(0.47)	(2.57)
Soil fertility	−0.025	0.050	−0.113 *	0.008	−0.032	0.029	−0.207	−0.347 **
	(−0.78)	(1.44)	(−1.81)	(0.25)	(−0.98)	(0.88)	(−1.19)	(−2.14)
Whether to join a cooperative	−0.117	0.111	0.387 ***	0.052	−0.135	−0.022	−0.317	−0.224
	(−0.86)	(0.81)	(2.80)	(0.37)	(−0.99)	(−0.16)	(−1.12)	(−0.79)
Agricultural technology training	0.011	−0.104	−0.179	−0.095	−0.114	−0.209 *	−0.098	−0.338
	(0.08)	(−0.84)	(−1.39)	(−0.72)	(−0.89)	(−1.69)	(−0.39)	(−1.46)
Disaster frequency	−0.173 **	−0.009	0.073	0.166 *	−0.250 ***	−0.154 *	−0.115	−0.182
	(−2.20)	(−0.11)	(0.78)	(1.81)	(−3.12)	(−1.79)	(−0.90)	(−1.42)
Whether to buy agricultural insurance	0.002	−0.063	−0.033	−0.162 *	0.220 ***	0.072	0.011	−0.026
	(0.02)	(−0.77)	(−0.39)	(−1.90)	(2.67)	(0.89)	(0.07)	(−0.17)
N	737	737	736	737	737	737	502	502

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

. ② Altering the Core Explanatory Variables—This study adopts the methodologies of prior research and utilizes agricultural social service expenditure as a proxy variable for agricultural social services to assess the robustness of the results. Additionally, inspired by Zhang Mengling et al. (2023) [14], the study considers the adoption of agricultural socialization services as another proxy variable. The outcomes of these substitutions are documented in Columns (3) to (6) of Table 4. ③ Excluding Samples—Considering that farmers over 65 years old may have a diminished capacity for agricultural labor and potentially less enthusiasm for ecological protection, samples from this age group are excluded to verify the stability of the findings. The re-estimated results are shown in Columns (7) and (8) of Table 4.

The estimation results presented in Columns (1)–(8) of Table 4 consistently demonstrate that regardless of the robustness test employed, agricultural socialization services exert a significant negative impact on both the amount of and change in chemical fertilizer and pesticide applications. This finding indicates that agricultural socialization services effectively promote reductions in chemical fertilizer and pesticide usage, corroborating the baseline regression results. These outcomes affirm that the conclusions of this study possess a considerable degree of robustness, underscoring the reliability of the research findings across different analytical scenarios.

3.4. Mechanism Inspection

Previous research has demonstrated that agricultural social services contribute to the reduction in farmers’ use of chemical fertilizers and pesticides. Building on these findings, this section delves into the specific mechanisms through which agricultural socialization services facilitate this reduction. Drawing from earlier theoretical analyses, the investigation will proceed by examining three distinct mechanisms: business scale, financial constraints, and factor substitution.

Columns (1)–(3) of

Table 5. Mechanism of action test.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	Business Scale	Fertilizer Application Amount	Pesticide Application Amount	Financial Constraints	Fertilizer Application Amount	Pesticide Application Amount	Element Substitution	Fertilizer Application Amount	Pesticide Application Amount
Business scale		−0.434 ***	−0.134 **
		(−2.62)	(−1.92)
Financial constraints					−0.219 *	−0.101 **
					(−1.78)	(−2.62)
Element substitution								−0.714 ***	−0.204 **
								(−3.46)	(−2.26)
Agricultural social services	0.216 ***	−0.137 *	−0.144	−0.122 **	−0.138 ***	−0.160 ***	0.148 ***	−0.142 ***	−0.168 ***
	(2.84)	(−1.69)	(−0.43)	(2.54)	(−3.92)	(−4.58)	(3.22)	(−4.03)	(−4.81)
Gender	0.293	−0.392	−0.387	0.325	−0.073	0.071	0.131	−0.080	0.064
	(0.84)	(−1.02)	(−1.04)	(1.51)	(−0.47)	(0.46)	(0.62)	(−0.52)	(0.41)
Age	−0.042 **	0.015	−0.005	−0.024 **	0.004	−0.007	−0.011	0.004	−0.007
	(−2.38)	(0.79)	(−0.30)	(−2.52)	(0.49)	(−0.93)	(−1.10)	(0.61)	(−0.91)
Culture	0.504 **	0.345	−0.197	−0.087	0.015	−0.013	−0.016	0.017	−0.012
	(2.21)	(1.44)	(−0.81)	(−1.35)	(0.64)	(−0.54)	(−0.48)	(0.74)	(−0.50)
Healthy	0.202	0.084	−0.089	0.072	−0.102	−0.025	−0.165 *	−0.106	−0.030
	(1.33)	(0.53)	(−0.58)	(0.74)	(−1.47)	(−0.36)	(−1.75)	(−1.53)	(−0.43)
Labor force	0.101	0.445 **	−0.113	0.196 **	0.053	−0.118	0.092	0.049	−0.118
	(0.66)	(2.49)	(−0.67)	(1.98)	(0.71)	(−1.58)	(0.91)	(0.65)	(−1.59)
Is it a village cadre?	−0.427	−0.144	0.306	−0.115	−0.273 **	−0.117	−0.133	−0.267 **	−0.116
	(−1.32)	(−0.40)	(0.93)	(−0.58)	(−2.20)	(−0.87)	(−0.73)	(−2.15)	(−0.86)
Household income	0.037	−0.036	−0.019	0.009	−0.012	−0.042 *	0.014	−0.012	−0.0407 *
	(0.90)	(−0.83)	(−0.46)	(0.30)	(−0.50)	(−1.77)	(0.45)	(−0.52)	(−1.73)
Cultivated area	0.233 ***	0.586	0.369	0.941 **	−0.137	−0.075	0.155 ***	−0.275	−0.426
	(5.92)	(1.60)	(0.11)	(2.37)	(−0.00)	(−0.28)	(3.45)	(−0.10)	(−0.16)
Soil fertility	−0.322	−0.998 ***	−0.380	0.044	−0.114 *	0.004	−0.022	−0.113 *	0.003
	(−1.31)	(−4.34)	(−1.62)	(0.98)	(−1.84)	(0.11)	(−0.36)	(−1.90)	(0.08)
Whether to join a cooperative	0.485	0.327	−0.308	0.374 **	0.383 ***	0.049	−0.132	0.369 ***	0.026
	(1.50)	(0.96)	(−0.91)	(2.03)	(2.77)	(0.35)	(−0.66)	(2.67)	(0.18)
Agricultural technology training	−0.157	−0.777 **	0.076	−0.012	−0.209	−0.152	−0.032	−0.208	−0.144
	(−0.50)	(−2.45)	(0.24)	(−0.06)	(−1.63)	(−1.15)	(−0.18)	(−1.61)	(−1.09)
Disaster frequency	0.513 *	0.555 *	0.573 *	−0.098	0.038	0.137	−0.206	0.037	0.125
	(1.85)	(1.92)	(1.94)	(−0.68)	(0.41)	(1.50)	(−1.40)	(0.39)	(1.37)
Whether to buy agricultural insurance	−0.394 **	−0.386 *	−0.400 *	−0.123	0.019	−0.125	−0.143	0.019	−0.122
	(−1.99)	(−1.85)	(−1.92)	(−1.01)	(0.22)	(−1.45)	(−1.19)	(0.21)	(−1.41)
N	150	150	150	736	736	736	(3)	736	737

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

explore the mechanism of operation scale in relation to agricultural socialization services. Column (1) shows a significantly positive coefficient for agricultural social services, suggesting that these services effectively encourage farmers to expand their business operations. Columns (2) and (3) reveal that the coefficients for business scale are significantly negative, while the coefficients for agricultural socialization services are significantly and insignificantly positive, respectively. These results indicate that business scale significantly mediates the relationship between agricultural socialization services and the reduction in chemical fertilizer and pesticide usage. The mediation effect is such that agricultural socialization services reduce pesticide and fertilizer applications by facilitating the expansion of operational scale. This finding confirms Hypothesis 4 of this study. The primary reason for this effect is that agricultural socialization services boost farmers’ enthusiasm for crop production, which leads to an expansion in production scale. This expansion, in turn, enhances the efficiency of chemical fertilizer and pesticide utilization, thereby reducing their overall consumption.

Columns (4)–(6) of Table 5 assess the role of financial constraints as mediated by agricultural socialization services. In column (4), the significantly negative coefficient associated with agricultural socialization services suggests that these services effectively alleviate farmers’ financial constraints. The coefficients in columns (5) and (6) for both capital constraints and agricultural socialization services are significantly negative, demonstrating that capital constraints function as a mediator in the relationship between agricultural socialization services and the reduction in chemical fertilizer and pesticide use. Specifically, agricultural socialization services facilitate this reduction by mitigating financial barriers, thereby validating Hypothesis 3. The underlying mechanism is that agricultural social services not only lower production costs for farmers but also diminish the substantial financial requirements for purchasing agricultural machinery, thereby motivating farmers to decrease their usage of chemical inputs.

Columns (7)–(9) of Table 5 delve into the mechanism of factor substitution facilitated by agricultural socialization services. In column (7), the coefficient associated with agricultural socialization services is significantly positive, indicating that these services effectively promote factor substitution among farmers. In columns (8) and (9), both the factor substitution coefficient and the agricultural socialization service coefficient are significantly negative, demonstrating that factor substitution acts as a mediator between agricultural socialization services and the reduction in the use of chemical fertilizers and pesticides. This mediation suggests that agricultural socialization services contribute to the decreased application of these chemicals by enhancing factor substitution. Consequently, Hypothesis 2 of this study is supported. The primary mechanism is that agricultural social services can effectively replace traditional agricultural labor, allowing for the more precise application of fertilizers and pesticides.

3.5. Heterogeneity Analysis

To further investigate the potential heterogeneity in the impact of agricultural social services on the reduction in chemical fertilizers and pesticides among farmers, this study examines several dimensions: farm scale, generational differences, and variation in service linkages. The research specifically aims to determine whether agricultural socialization services are more effective in reducing chemical inputs for large-scale and new-generation farmers compared to their smaller-scale and older counterparts. The study will explore these effects across different operational scales, generational cohorts, and the specific services utilized in the agricultural process.

Table 6. The impact of agricultural socialization services on fertilizer and pesticide variables, based on different scale perspectives.

	Small-Scale Farmers		Medium-Sized Farmers		Large-Scale Farmers
	(1)	(2)	(3)	(4)	(5)	(6)
	Fertilizer Usage	Pesticide Usage	Fertilizer Usage	Pesticide Usage	Fertilizer Usage	Pesticide Usage
Agricultural social services	−0.120 **	−0.137 ***	−0.140 ***	−0.164 ***	−0.309 **	−0.342 *
	(−2.53)	(−2.89)	(−4.00)	(−4.72)	(−2.17)	(−1.82)
Gender	−0.074	−0.117	−0.080	0.063	−0.655	6.703 ***
	(−0.36)	(−0.56)	(−0.52)	(0.41)	(−0.52)	(2.91)
Age	−0.006	−0.011	0.005	−0.007	0.105	0.016
	(−0.63)	(−1.08)	(0.62)	(−0.90)	(1.55)	(0.19)
Culture	−0.023	−0.065	0.017	−0.012	−0.502	−0.237
	(−0.69)	(−1.38)	(0.75)	(−0.50)	(−0.71)	(−0.30)
Healthy	−0.195 **	0.105	−0.105	−0.026	0.354	0.484
	(−2.05)	(1.08)	(−1.52)	(−0.37)	(0.65)	(0.73)
Labor force	0.013	−0.178 *	0.047	−0.122	0.296	0.589
	(0.13)	(−1.67)	(0.63)	(−1.64)	(0.42)	(0.75)
Is it a village cadre?	−0.203	−0.180	−0.266 **	−0.113	2.403	0.783
	(−1.38)	(−1.08)	(−2.14)	(−0.84)	(1.43)	(0.33)
Household income	−0.051	−0.108 ***	−0.012	−0.042 *	−0.016	−0.027
	(−1.39)	(−2.91)	(−0.52)	(−1.79)	(−0.14)	(−0.21)
Cultivated area	0.011	0.120 ***	−0.001	−0.001	−0.009	0.010
	(0.28)	(3.05)	(−0.18)	(−0.35)	(−1.16)	(1.02)
Soil fertility	−0.092	−0.007	−0.113 *	0.003	−1.191 *	0.221
	(−1.64)	(−0.19)	(−1.89)	(0.09)	(−1.78)	(0.25)
Whether to join a cooperative	0.506 ***	0.085	0.371 ***	0.031	0.072	1.102
	(2.94)	(0.48)	(2.69)	(0.22)	(0.07)	(0.82)
Agricultural technology training	−0.171	−0.144	−0.208	−0.144	−2.264	26.470
	(−0.98)	(−0.79)	(−1.62)	(−1.08)	(−1.21)	(0.01)
Disaster frequency	0.005	0.034	0.040	0.129	0.827	−5.179 ***
	(0.04)	(0.32)	(0.43)	(1.41)	(1.04)	(−2.73)
Whether to buy agricultural insurance	0.072	−0.035	0.019	−0.117	−0.299	−0.111
	(0.63)	(−0.31)	(0.22)	(−1.36)	(−0.54)	(−0.17)
N	416	416	736	737	31	31

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

assesses the impact of agricultural socialization services on the reduction of chemical fertilizers and pesticides among farmers of varying sizes. Following the classification method used by Liu Sha and Zhu Mande, farmers are categorized based on cultivated land area into three groups: small-scale (less than 10.5 acres), medium-scale (10.5–49.55 acres), and large-scale (49.55 acres or more) [30,31]. The results in Table 6 reveal that the coefficients for agricultural socialization services across columns (1)–(6) are consistently and significantly negative for all size groups, indicating that agricultural socialization services have substantially facilitated the reduction in chemical fertilizers and pesticides among small-, medium-, and large-scale farmers. A comparative analysis of the absolute values of the coefficients suggests notable differences in the effectiveness of these services across different farm sizes. The impact is most pronounced for large-scale farmers, followed by medium-scale, with small-scale farmers experiencing the least impact. This pattern suggests that the effectiveness of agricultural socialization services in reducing chemical input use increases with the scale of farming operations. Furthermore, the coefficient magnitude for large-scale farmers is approximately twice that observed for medium- and small-scale farmers. This indicates that for farms exceeding 49.55 acres, the efficacy of agricultural socialization services in reducing chemical inputs is notably enhanced, achieving the most substantial reduction in chemical fertilizer and pesticide usage.

Table 7. The impact of agricultural socialization services on fertilizer and pesticide variables, based on the perspective of intergenerational differences.

	New-Generation Farmers		Old-Generation Farmers
	(1)	(2)	(3)	(4)
	Fertilizer Usage	Pesticide Usage	Fertilizer Usage	Pesticide Usage
Agricultural social services	−0.196 ***	−0.392 ***	−0.128 ***	−0.140 ***
	(−3.30)	(−2.64)	(−3.45)	(−3.81)
Gender	0.512	0.487	−0.128	0.028
	(0.84)	(0.84)	(−0.78)	(0.17)
Age	−0.011	0.026	0.008	−0.003
	(−0.20)	(0.53)	(0.84)	(−0.30)
Culture	−0.053	0.020	−0.029	−0.134
	(−0.94)	(0.40)	(−0.28)	(−1.24)
Healthy	−0.704 *	0.352	−0.095	−0.027
	(−1.71)	(0.93)	(−1.33)	(−0.38)
Labor force	0.566 *	0.252	0.017	−0.162 **
	(1.76)	(0.90)	(0.21)	(−2.04)
Is it a village cadre?	0.558	−0.046	−0.290 **	−0.084
	(1.12)	(−0.10)	(−2.22)	(−0.58)
Household income	−0.008	−0.015	−0.023	−0.041
	(−0.14)	(−0.26)	(−0.89)	(−1.58)
Cultivated area	−0.014	−0.012	0.000	−0.001
	(−1.45)	(−1.07)	(−0.04)	(−0.20)
Soil fertility	−0.547	−0.187	−0.099 *	0.011
	(−1.24)	(−0.43)	(−1.79)	(0.32)
Whether to join a cooperative	1.926 ***	0.007	0.265 *	−0.024
	(2.67)	(0.01)	(1.81)	(−0.16)
Agricultural technology training	−0.920	−0.865	−0.168	−0.117
	(−1.26)	(−1.23)	(−1.25)	(−0.84)
Disaster frequency	−0.038	−0.311	0.059	0.153
	(−0.08)	(−0.54)	(0.59)	(1.63)
Whether to buy agricultural insurance	−0.258	−0.910 *	0.013	−0.099
	(−0.48)	(−1.75)	(0.15)	(−1.13)
N	58	58	666	667

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

assesses the impact of agricultural socialization services on the reduction in chemical fertilizers and pesticides among farmers, differentiated by generational cohorts. Following the methodologies used by existing studies, this study classifies farmers into two groups based on their year of birth: those born before 1980 are categorized as first-generation or older generation farmers, while those born after 1980 are considered new-generation farmers [32,33]. The results displayed in Table 7 indicate that the coefficients of agricultural socialization services across columns (1)–(4) are significantly negative for both new- and old-generation farmers, suggesting that agricultural socialization services have effectively facilitated the reduction in chemical fertilizers and pesticides across generational lines. A comparison of the absolute values of the coefficients in columns (1)–(2) with those in columns (3)–(4) reveals that the impact is more pronounced among new-generation farmers than among their older counterparts. This suggests a stronger marginal impact of agricultural socialization services on chemical input reduction within the newer generation. The greater responsiveness of new-generation farmers can be attributed to their higher receptivity to advanced agricultural technologies and a deeper understanding of agricultural socialization services. They are more likely to embrace these services, whereas older-generation farmers, who often have a stronger attachment to traditional farming practices and the land itself, may prefer to manage their operations independently if capable. Consequently, the adoption of agricultural socialization services is considerably more limited among older farmers compared to those of the newer generation. This generational difference underscores the need for tailored approaches in promoting agricultural socialization services that accommodate the specific preferences and capacities of different farmer groups.

Table 8. The impact of agricultural socialization services on fertilizer and pesticide variables, based on the perspectives of different links.

	(1)	(2)	(3)	(4)
	Fertilizer Usage	Pesticide Usage	Fertilizer Usage	Pesticide Usage
Technology-intensive social services	−0.958 ***	−1.409 ***
	(−14.19)	(−18.51)
Labor-intensive social services			−0.894 ***	−0.825 ***
			(−12.27)	(−11.92)
Gender	0.016	−0.131	0.045	−0.127
	(0.11)	(−0.88)	(0.30)	(−0.84)
Age	−0.011 *	0.008	−0.011	0.004
	(−1.65)	(1.23)	(−1.60)	(0.57)
Culture	−0.036	0.014	−0.050 **	−0.007
	(−1.57)	(0.67)	(−2.34)	(−0.33)
Healthy	−0.104	0.080	−0.086	0.091
	(−1.54)	(1.16)	(−1.28)	(1.33)
Labor force	−0.007	0.088	−0.065	−0.018
	(−0.10)	(1.26)	(−0.90)	(−0.25)
Is it a village cadre?	0.147	−0.143	0.126	−0.123
	(0.98)	(−1.16)	(0.85)	(−0.99)
Household income	−0.061 ***	−0.030	−0.048 **	−0.007
	(−2.77)	(−1.38)	(−2.20)	(−0.33)
Cultivated area	0.003	0.006 **	0.003	0.006 **
	(1.10)	(2.37)	(0.99)	(2.08)
Soil fertility	−0.037	0.044	−0.015	0.047
	(−1.14)	(1.29)	(−0.48)	(1.41)
Whether to join a cooperative	−0.182	0.000	−0.051	0.140
	(−1.33)	(−0.00)	(−0.38)	(1.02)
Agricultural technology training	−0.004	−0.091	−0.018	−0.149
	(−0.03)	(−0.73)	(−0.14)	(−1.21)
Disaster frequency	−0.159 **	0.019	−0.203 ***	−0.087
	(−2.02)	(0.23)	(−2.63)	(−0.99)
Whether to buy agricultural insurance	0.080	−0.011	0.029	0.007
	(0.98)	(−0.14)	(0.35)	(0.08)
N	737	737	737	737

Note: ***, **, * indicate significant at the 1%, 5% and 10% levels respectively; the robust standard errors are in brackets.

explores the impact of agricultural socialization services on farmers’ reduction in chemical fertilizers and pesticides, analyzing the effects associated with different service links. Drawing from the research of Zhang Mengling, this study categorizes agricultural socialization services into seven distinct links, dividing them based on their reliance on agricultural production factors into technology-intensive socialized services (including seedling cultivation, fertilization, prevention, and control) and labor-intensive socialized services (such as land preparation, sowing, harvesting, and irrigation) [14]. The results in Table 8 reveal that technology-intensive socialized services show a positive and significant coefficient at the 1% level in columns (1) and (2), indicating a substantial contribution to the reduction in chemical fertilizers and pesticides. Similarly, labor-intensive socialized services also demonstrate a positive and significant impact at the 1% level in columns (3) and (4), contributing to the reduction in these chemical inputs. However, when comparing the absolute values of the coefficients between technology-intensive and labor-intensive services, the coefficients for technology-intensive services in columns (1)–(2) exceed those for labor-intensive services in columns (3)–(4), suggesting that technology-intensive services have a more pronounced marginal impact on the reduction in fertilizers and pesticides. The primary reason for these differences is that labor-intensive services primarily substitute agricultural labor with mechanization, focusing on enhancing production speed with less emphasis on the application of fertilizers and pesticides during the service process. Consequently, the impact of labor-intensive services on reducing chemical inputs is relatively smaller. On the other hand, technology-intensive services, which replace manual labor with biochemical technologies, are geared towards increasing crop yields and ensuring crop health. These services typically involve more intensive fertilization and pesticide application strategies, thereby exerting a greater influence on reducing chemical inputs.

3.6. Discussion of Related Results

This paper analyzes data from 743 rice farmers in Hubei Province, collected in 2023, to explore how agricultural social services contribute to reducing the use of fertilizers and pesticides, thereby enhancing ecological protection. The results indicate a significant decrease in chemical inputs: for each 1% increase in service levels, fertilizer and pesticide applications per hectare declined by 14% and 16.4%, respectively. This pattern aligns with that in the results of Gao et al. [15], who observed similar trends across various agricultural settings, suggesting universal patterns across different geographic and cultivation contexts. This consistency bolsters the reliability and relevance of our findings. The study confirms that scale operation, factor substitution, and alleviation of financial constraints play intermediary roles in linking agricultural social services to reduced chemical input usage. These findings corroborate the theoretical model proposed by Yang Zi et al. [20] and are empirically supported by Lin, Y. et al. [4], who found that efficiency improvements in large-scale operations decrease the per-unit-area demand for chemical inputs. Furthermore, our analysis shows that agricultural social services substantially reduce the use of chemical inputs by expanding operational scales, enhancing factor substitution, and easing financial burdens. Zhang and Chen [21,22] support this, highlighting the role of technological interventions in promoting sustainable practices by decreasing dependence on chemical inputs. Heterogeneity analysis reveals significant effects across all farm sizes, particularly among large-scale operations. Additionally, these services markedly decrease chemical input usage among both new and established generations of farmers, with more pronounced effects observed in the newer generation. As Zhang et al. [29] suggest, this generational difference is vital, indicating that future policies should tailor strategies to different demographic groups to optimize the uptake of sustainable practices. Both technology-intensive and labor-intensive social services contribute significantly to reducing chemical inputs, though the impact of technology-intensive services is more pronounced. This study not only validates existing theories but also addresses a crucial research gap by quantifying the impacts of specific mediators and highlighting differences across generations.

4. Conclusions

4.1. Research Conclusions

This study presents robust evidence of the role agricultural social services play in reducing dependency on fertilizers and pesticides, thereby enhancing the ecological sustainability of agriculture. It demonstrates that by expanding operational scales, enhancing factor substitution, and alleviating financial constraints, sustainable agricultural practices can be significantly improved. These findings underscore the importance of developing targeted strategies within future policymaking to optimize the adoption of sustainable practices across diverse agricultural groups. Specifically, the study highlights the following: (1) Fertilizer reduction—Agricultural social services decrease fertilizer usage by 14% per unit area for every 1% increase in service level, achieved through optimized resource allocation and service integration. (2) Pesticide reduction—Similarly, a 1% increase in service level results in a 16.4% reduction in pesticide usage per unit area, underscoring the efficacy of these services in minimizing chemical inputs. (3) Impact of mediating variables—The study identifies operational scale, factor substitution, and financial relief as key mediators that facilitate the translation of agricultural social services into positive ecological outcomes. These insights are invaluable for policymakers and agricultural managers aiming to craft more effective environmental protection strategies. (4) Differences among subgroups—The research investigates variations in service impact across different scales of agricultural operations and among various generations of farmers, offering detailed insights that inform policymaking in diverse agricultural settings in China.

The principal contributions of this paper are threefold: (1) Empirical evidence of ecological benefits—Through extensive data analysis, this paper not only confirms the effectiveness of agricultural social services in reducing fertilizer and pesticide usage, but also establishes a direct correlation between enhanced service levels and reduced chemical inputs. (2) Understanding of mediating factors—The paper thoroughly explains how the expansion of operational scale, increased factor substitution, and the alleviation of financial pressures act as conduits linking agricultural social services with ecological outcomes. (3) Analysis of heterogeneity in effects—This elucidates the variable impacts of agricultural social services across different scales of operations and generations of farmers, providing a foundation for the development of tailored agricultural policies.

4.2. Policy Recommendations

The policy recommendations derived from this article are as follows: Promote the expansion of farming scale and develop high-standard farmland—The research findings indicate that the larger the farming scale, the more effective the ecological protection of farmland facilitated by agricultural socialization services. Notably, once the planting scale exceeds 49.55 acres, the effectiveness of agricultural socialization services in reducing chemical fertilizer and pesticide use doubles. Future initiatives should therefore focus on enhancing the farmland of small and medium-sized farmers. This can be achieved by encouraging the consolidation of farmland through mergers and replacements, promoting the development of large-scale and high-standard farmland. Such strategies will maximize the benefits of agricultural socialization services in reducing chemical inputs. Increase financial support and enhance social service levels—Currently, the cost for farmers to adopt agricultural socialization services remains prohibitive, significantly diminishing planting income and hindering the widespread adoption and application of these services. To address this, it is recommended that the government increases subsidies for farmers who engage in agricultural socialization services. Modifying the subsidy framework from one based on cultivated land area to one based on service adoption levels could encourage farmers to utilize multiple aspects of agricultural socialization services, thereby advancing agricultural modernization.