Next Article in Journal
Sustainability of Local Labour Market in South Africa: The Implications of Imports Competition from China
Next Article in Special Issue
From Biotechnology to Bioeconomy: A Review of Development Dynamics and Pathways
Previous Article in Journal
The Effect of Environmentally Sustainable Practices on Customer Citizenship Behavior in Eco-Friendly Hotels: Does the Green Perceived Value Matter?
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

How Agricultural Extension Services Improve Farmers’ Organic Fertilizer Use in China? The Perspective of Neighborhood Effect and Ecological Cognition

1
Management School, Hainan University, Haikou 570100, China
2
School of Finance, Nanjing University of Finance and Economics, Nanjing 210023, China
3
School of Economics, Hainan University, Haikou 570100, China
*
Authors to whom correspondence should be addressed.
Sustainability 2022, 14(12), 7166; https://doi.org/10.3390/su14127166
Submission received: 25 April 2022 / Revised: 1 June 2022 / Accepted: 8 June 2022 / Published: 11 June 2022

Abstract

:
Improving farmers’ usage of organic fertilizer is critical for the green and high-quality development of China’s agriculture. Based on 492 mango farmers’ survey data in Hainan Province, this paper uses an endogenous switching regression (ESR)model, empirically analyzes the impact of agricultural extension services on farmers’ willingness to use organic fertilizer, and further investigates the mediating role of ecological cognition and the moderating role of neighborhood effect in the influence mechanism. Results show agricultural extension services have a significant positive effect on farmers’ willingness to use organic fertilizer, ecological cognition has a partial mediating effect in the influence mechanism, which accounts for 17.84% of the total effect. The neighborhood effect has a positive moderating effect in the influence mechanism of ecological cognition on farmers’ willingness to use organic fertilizer. These results imply that agricultural extension services play a significant role in China’s sustainable agricultural development and by improving their awareness and taking advantage of the neighborhood effect, we can stimulate farmers’ willingness to green production. The study also puts forward policy recommendations on further promoting farmers’ use of organic fertilizer.

1. Introduction

Chemical fertilizers have made significant contributions to China’s agricultural economic growth and national food security [1]. However, excessive fertilizer usage, one of the primary drivers of non-point source pollution from Chinese agriculture, has resulted in a series of environmental and ecological problems, such as soil hardening, water pollution, and air quality deterioration, severely undermining the sustainability of agriculture [2,3]. The condition of China’s farmed land has deteriorated in recent years. Between 2009 and 2015, the proportion of high-quality cultivated land in China declined by 3%, while the proportion of medium- and low-quality cultivated land climbed by 3% [4]. Due to China’s limited agricultural land resources, it is urgent to protect these cultivated lands and maintain their quality. The Ministry of Agriculture developed a national action plan in 2015 to achieve zero growth in fertilizer use by 2020. It emphasized the importance of promoting the transformation of fertilization methods and utilization of organic fertilizer resources, as well as encouraging and guiding farmers to increase organic fertilizer application. As an environmentally-friendly agricultural technique, organic fertilizer can effectively alleviate agricultural pollution and retain the land output benefit [5,6]. In this context, it is imperative to promote organic fertilizer to replace chemical fertilizer.
China is a traditional agricultural country with a long history of organic fertilizer application, and farmers are experienced in organic fertilizer application. However, farmers’ application of organic fertilizer is still insufficient. According to data from the Chinese Ministry of Agriculture and Rural Affairs’ Agricultural Technology Promotion Center, the proportion of organic fertilizer usage to overall fertilizer usage decreased from 25% in 2003 to 8–10% in 2017 [7]. Additionally, empirical research at the micro-level also indicates that the proportion of farmers in China who use an organic fertilizer is low [8,9]. As the end-users of agricultural fertilizers and stakeholders of the land ecological environment, farmers’ production behavior decisions significantly impact the promotion of organic fertilizer substitution [10]. Therefore, it is critical to understand factors affecting farmers’ use of organic fertilizers to develop strategies to improve their use and formulate relevant policies and measures.
Academics have conducted extensive research on farmers’ organic fertilizer use behavior, mainly focusing on the influence of individual and family characteristics, psychological characteristics, economic income level, livestock and breeding, and operation scale [11,12,13,14]. Meanwhile, some scholars argue that external support can serve as a technology adoption incentive for farmers to make adoption decisions based on their characteristics. Currently, the government agricultural extension service, which has a dominant position and leading role in China’s agricultural technology extension system, is an important channel for farmers to accept new technologies. Agricultural extension service emphasizes the government’s intervention, control, and institutional links, and plays a major role in promoting agricultural technology. But for a long time, extension services have been difficult to adapt to the diversified technical needs of farmers, which leads to the short supply of agricultural technology and the low efficiency of extension services. As a result, whether government extension services are still effective in promoting the adoption of environmentally friendly technologies by farmers in the current market environment is a question that deserves exploration.
Additionally, farmers’ decisions are influenced by their cognitive level and the environment they live in [15,16]. The cognition of the benefit of a friendly environment may affect farmers’ decision to use organic fertilizer. Agricultural extension services can increase farmers’ cognition levels, promoting farmers’ organic fertilizer use. Besides, as farmers are not only “economical people” but also “social people,” their behavior tends to be more rational under the combined effect of the two roles [17]. As the main characteristic of social people, farmers’ behavior is easily influenced by the external environment, including interpersonal relationships, social networks, and cultural traditions [18]. The complexity and uncertainty of the agricultural production environment make it inevitable that farmers encounter problems during the long-term use of organic fertilizers. The likelihood of continued use of organic fertilizer increases when village officials can be consulted for solutions or learn from the experience of those around them to improve their farming capacity. In fact, as a typical relational society in rural China, the prevalence of social relationships influences farmers’ behavior [19,20]. Family, friends, neighbors, and agricultural technicians are essential components of farmers’ social networks [21]. However, the impact of neighborhood effects continues to be overlooked in existing agricultural extension research. For this reason, the impact of neighborhood effects on farmers’ willingness to adopt organic fertilizers should be further investigated.
The above studies have examined the factors affecting farmers’ organic fertilizer application from different perspectives, which laid the analytical foundation for this study. However, most of the existing studies focus on the analysis of influencing factors and less discuss the influence mechanism of agricultural extension services on farmers’ willingness to use organic fertilizer. Therefore, it is unclear how China’s agricultural extension services promote farmers’ organic fertilizer application and how ecological cognition and neighborhood effects play a role in this process. The main contributions of this paper are as follows: we test the mediating role of ecological cognition between agricultural extension services and farmers’ willingness to use organic fertilizer, and further clarify the influence mechanism of extension services on farmers’ willingness to use organic fertilizer with an endogenous conversion model; second, we test the moderating role of neighborhood effect on the above influence mechanism, and verify how the above mechanism is affected by the change of neighborhood effect. The empirical results of this paper can provide a reference basis for understanding farmers’ organic fertilizer application behavior and formulating relevant policy measures.

2. Theoretical Analysis and Research Hypotheses

Agricultural extension bridges agricultural technology research and development and real productivity and is a key link to realizing the transformation of scientific research results into agricultural productivity [22]. Farmers regard agricultural extension services as a major source of agricultural information access [23]. According to agricultural technology diffusion theory, farmers’ technology adoption is a conscious and subjective economic behavior. Due to the heterogeneity of individual receptivity to relevant information and human capital, farmers’ exposure to advanced information is unequal, resulting in differences in their adoption decisions and affecting the diffusion and application of new technologies [24,25]. Effective agricultural extension services can influence farmers’ technology adoption decisions and create favorable conditions for agricultural technology dispersion to accelerate [26,27]. On the one hand, the agricultural extension can serve as a propaganda and teaching tool, increasing farmers’ awareness of new technology and increasing their proclivity to embrace them. On the other hand, agricultural extension services can reduce farmers’ information-seeking costs, provide technical support for their agricultural production, and save farmers’ learning and time costs associated with technology integration, increasing farmers’ willingness to adopt technology [28]. Research shows that farmers pay special attention to external publicity and social services when adopting organic fertilizer technology, so improving the agricultural extension system and developing agricultural social services can effectively enhance their willingness to adopt the technology. [27,29]. Agricultural extension services can assist farmers in overcoming operational challenges associated with organic fertilizer application due to a lack of technical support [30], as well as educate them about the benefits of organic fertilizer application in terms of improving the quality of agricultural products and promoting sustainable land farming, thereby motivating them to use organic fertilizer. In light of this, this research provides Hypothesis 1.
Hypothesis 1.
Acceptance of agricultural extension services can improve farmers’ willingness to use organic fertilizer.
Ecological cognition reflects the extent of an individual’s knowledge of ecology. It is a mental understanding formed by an individual’s continuous accumulation of relevant and ecological environmental knowledge through acquisition, understanding, and learning [31]. According to the theory of planned behavior, under the conditions of uncertainty, behavioral attitudes, subjective norms, and perceived behavioral control jointly influence behavior, while other factors indirectly influence individual behavioral decisions through these three. Therefore, cognitive variables should be included when examining the elements determining an individual’s behavioral intentions [32,33]. While agricultural extension directly affects farmers’ behavioral intentions, it also enhances their ecological cognitive level by acquiring appropriate knowledge and improving foresight, indirectly affecting farmers’ ecological behavioral intentions [34]. Firstly, agricultural extension has a certain guiding and regulating effect on farmers’ production behavior and receiving related extension services will enhance farmers’ understanding of the current pollution situation in agricultural production and their recognition of organic fertilizers as an alternative to chemical fertilizers and prompt them to improve their environmental production behavior. Secondly, agricultural extension services can help raise farmers’ awareness of the environmental pollution caused by excessive application of chemical fertilizers, deepen farmers’ understanding of the role of organic fertilizers in improving the ecological environment and crop quality, and increase their enthusiasm and initiative to adopt organic fertilizers. Given this, this paper proposes Hypothesis 2.
Hypothesis 2.
Agriculturalextension services can enhance farmers’ ecological knowledge and thus promote farmers’ willingness to use organic fertilizers.
According to social cognitive theory, the external environment affects individuals’ cognition, willingness, and behavior [35,36]. Rural China is a vernacular society based on acquaintance networks [37], and the behavioral decisions of households can exhibit significant neighborhood effects [38]. It was found that daily communication among neighbors is an essential channel for farmers to obtain information on pesticide application technologies and can effectively influence behavioral decisions among each other [39]. Xie and Chen (2020) found that an increase in the degree of neighborhood communication can effectively alleviate the negative attitudes of farmers due to ecological farming [40]. Similarly, for farmers’ fertilizer application behavior, a stronger neighborhood effect implies a higher frequency of interactive learning among farmers and a stronger willingness to use organic fertilizer as ecological awareness increases. On the contrary, even if farmers who are weakly influenced by the neighborhood effect are aware of the serious pollution of the agricultural environment caused by the long-term application of chemical fertilizers and that the use of organic fertilizers can improve the ecological environment and soil quality, their willingness to use organic fertilizers will be inhibited by their individual “limited rationality” due to the lack of specific information communication channels and group norms, and they may ignore the negative externalities in the agricultural production process when making production decisions. In view of this, this paper proposes Hypothesis 3.
Hypothesis 3.
The neighborhood effect can strengthen the influence of ecological perception on farmers’ willingness to use organic fertilizer.
This paper integrates agricultural extension, ecological cognition, neighborhood effect, and farmers’ willingness to use organic fertilizer into an analytical framework (Figure 1). Theoretically, it analyzes the mediating effect of ecological cognition on farmers’ organic fertilizer application through agricultural extension and the moderating effect of neighborhood effect on farmers’ organic fertilizer application through ecological cognition.

3. Material and Methods

3.1. Study Area

This paper takes three major mango-growing counties in Hainan Province as the study area (Figure 2). Hainan Province is in the southernmost part of China, a tropical region with a mild climate, abundant rainfall, and fertile land, making it uniquely suited for many tropical crops, such as rubber, coffee, cocoa, coconut, mango and palm [41]. In recent years, the mango industry in Hainan Province has developed rapidly, and mango cultivation has formed a certain scale and achieved considerable economic benefits. However, the extensive use of chemical fertilizers in mango cultivation has triggered a series of ecological and environmental problems, such as soil acidification, soil nutrient imbalance, and surface water eutrophication which has aroused widespread concern. Meanwhile, Hainan is an important ecological province in China, of which the area of nature reserves accounts for 77.22% of the province’s total area. To alleviate the negative impact of excessive fertilizer application on the regional ecological environment, the local government issued the “Hainan Provincal Chemical Fertilizer and Pesticide Application Reduction Promotion and Implementation Program” and relevant policy to promote arable land quality improvement and chemical fertilizer application reduction. Additionally, the government established special subsidies to encourage fertilizer reduction and efficiency enhancement technology, which catalyzes organic fertilizer adoption. Since mango cultivation in Hainan province is mainly distributed in coastal tableland and low hilly areas, we selected three major mango producing counties as research areas, namely Dongfang, Ledong, and Sanya which accounted for 86.52% of the total mango planting area in Hainan, showing certain representativeness to explore the impact of agricultural extension on local farmers’ use of organic fertilizer.

3.2. Data

Research data were obtained from a field survey of mango farmers conducted in Hainan Province in July 2020. The survey was conducted by members of our research group who have extensive field research experience through face-to-face interviews. Questionnaires were distributed through multistage stratified sampling and random sampling. In the first stage, three counties were purposively selected with comprehensive consideration of representativeness, economic development, and agricultural production; In the second stage, three towns were randomly selected from each county, and five villages were randomly selected in each town. Finally, farmers who grow mango were randomly selected from each village. The survey questionnaire mainly included the basic characteristics of the family, agricultural production and operation status, the willingness to use organic fertilizers and the status of agricultural extension services. In order to ensure the quality of the survey and data, the questionnaire was optimized through a literature review, a pilot survey, and in-depth discussions with experts in related fields and local government departments. Unified training was also conducted for all interviewers in our research team. A total of 518 questionnaires were given out, and 492 were valid after inspection of invalid ones with an effective rate of 94.98%.
Table 1 presents the socio-demographic characteristics of the sample households and household heads. Male farmers account for 93.90% of the sample, as men head the majority of Chinese families. The sample farmers are relatively middle-aged, with an average age of 51.2, and 61.99% of farmers are over 50 years old. The sample farmers were generally low-educated, accounting for 84.5% of junior high school graduates, but some new professional farmers received higher education. Each sample household has a certain mango planting scale. The average planting area of sample households is 35.07 acres, most of which are from 16 acres to 45 acres, accounting for 43.1%, and the householders with a planting scale of more than 15 acres account for 67% of all the samples. Accordingly, the household income of the samples was mainly 100,000 to 500,000 yuan, which accounts for 47.6%. Additionally, 41.67 percent of respondents indicated that they had never received AES.
Generally, 73.58% of the sample farmers were strongly willing to use organic fertilizer shown in Table 1. Farmers were further divided into groups according to whether they had received AES, and their willingness to use organic fertilizer in different groups was statistically analyzed. The results are shown in Table 2. Among the farmers who had received AES, 80.49% had a strong willingness to use organic fertilizer. Among the farmers who had not received AES, 63.90% had a strong willingness to use organic fertilizer. AES play a certain role in promoting farmers’ willingness to use organic fertilizer.

3.3. Econometric Model

3.3.1. Using Endogenous Switching Regression (ESR) Model to Analyze the Impacts of AES on Farmers’ Willingness to Use Organic Fertilizer

In some cases, the acceptance of AES by farmers does not necessarily happen randomly but may be the result of other unobservable factors. Therefore, considering the selection bias caused by observable and unobservable factors, this paper refers to the study of Lokshin and Sajaia [42] and adopts the ESR model. Constructing a counterfactual analysis framework based on the benchmark regression results and then estimating the treatment effect of accepting AES on WTU to more accurately obtain the impact of AES on farmers’ willingness to use organic fertilizer.
In the first stage, we estimate the determinants of farmers’ decisions to accept AES, if AES is accepted, T i = 1 , otherwise T i = 0 . Z i represents the factors that influence farmers’ acceptance of AES. γ is the parameter to be estimated and μ is the random error term. The outcome equations are estimated in the second stage to analyze factors influencing farmers’ willingness to use organic fertilizer. Whether farmers accept AES can be expressed as:
T i = Z i γ + μ , T i = { 1 , T i > 0 0 , T i 0
Secondly, the result equation of farmers’ willingness to use organic fertilizer in different situations can be further defined:
Y i a = X i a β a + σ i a λ i a + ε i a , T i = 1
Y i a = X i a β a + σ i a λ i a + ε i a , T i = 1
Y i a and Y i n denote WTU in the state of receiving AES and non-receiving AES, respectively, and X i a and X i n are the factors affecting WTU, ε i a and ε i n are random error terms. To address the sample selection bias caused by unobservable factors, λ i a , λ i n and covariances σ i a and σ u n were introduced for joint estimation using the maximum likelihood method.
When there is sample selection bias which may cause heterogeneity issues, the ESR model can estimate the average treatment effect of WTU with accepting AES or not. The average treatment effect on the treated (ATT) and the average treatment effect on the untreated (ATU) can be calculated to measure how AES acceptance affects the WTU of farmers who accepts AES and who does not.
E ( Y i a | A i = 1 ) = X i a β a + σ i a λ i a
E ( Y in | A i = 0 ) = X i n β n + σ u n λ i n
E ( Y i n | A i = 1 ) = X i a β n + σ u n λ i a
E ( Y i a | A i = 0 ) = X i n β a + σ i a λ i n
Equations (4) and (6) are the calculation equations for the average WTU who accept AES. The average treatment effect (ATT) of WTU who have accepted AES is:
E ( Y i a | A i = 1 ) - E ( Y i n | A i = 1 ) = X i a ( β a - β n ) + ( σ i a σ u n ) λ i a
The average treatment effect (ATU) of WTU for farmers who have not received AES is:
E ( Y i a | A i = 0 ) - E ( Y in | A i = 0 ) = X i n ( β a - β n ) + ( σ i a σ u n ) λ i n

3.3.2. The Mediating Effect of Ecological Cognition

With the help of the mediation effect model, the influence path and mechanism of AES on WTU can be revealed. The meaning of the mediation effect is that in the influence of the explanatory variable X on the explained variable Y, if the variable X will influence the explained variable Y through the variable M, then the variable M is called the mediating variable. The influence mechanism X-M-Y is the mediating effect model [43]. The commonly used mediation effect test methods are mainly the causal steps and bootstrap approaches. The causal steps approach is the most popular and easy to understand, and simple to operate. Since the bootstrap approach cannot effectively test the case where the mediating variable is binary [44], this paper adopts the causal stepwise regression approach proposed by Baron and Kenny [45] to test the mediating effect of ecological cognition. The principle is as follows:
ln p r o b ( y i ) 1 p r o b ( y i ) = a 1 + a 2 E x t i + a 3 X i + ε 1 i
ln p r o b ( C og i ) 1 p r o b ( C og i ) = b 1 + b 2 E x t i + b 3 X i + ε 2 i
ln p r o b ( y i ) 1 p r o b ( y i ) = c 1 + c 2 E x t i + c 3 C o g i + c 4 X i + ε 3 i
where C o g i is the mediating variable, representing the ecological cognition of the farmers; E x t i reflects whether the farmers accept AES X i refers to control variables. Equation (10) represents the basic regression model of the first step to test whether AES as an impact on WTU. Equation (11) tests the effect of AES on the mediating variable (farmers’ ecological cognition). Equation (12) tests the influence of ecological cognition on WTU based on controlling the E x t i .
In addition, when calculating the proportion of the mediating effect of ecological cognition, since the coefficients a 2 and b 2 obtained by the regression cannot be directly compared, this paper refers to the research of MacKinnon et al. (2007) [46], and the original equation coefficients are treated as follows:
First, the expressions of the original Equations (10)–(12) are reset.
ln p r o b ( y i ) 1 p r o b ( y i ) = a 1 + a 2 E x t i + a 3 X i + ε 1 i
ln p r o b ( C o g i ) 1 p r o b ( C o g i ) = b 1 + b 2 E x t i + b 3 X i + ε 2 i
ln p r o b ( y i ) 1 p r o b ( y i ) = c 1 + c 2 E x t i + c 3 C o g i + c 4 X i + ε 3 i
To make the coefficients to be estimated comparable, the coefficients to be estimated were processed. The comparable coefficients are obtained by multiplying each variable coefficient by its standard deviation and then dividing it by the standard deviation of the dependent variable of the equation in which it is located. This is done as follows.
a 2 s t d = a 2 S D ( Ext ) S D ( y )
b 2 s t d = b 2 S D ( Ext ) S D ( Cog )
c 2 s t d = c 2 S D ( Ext ) S D ( y )
c 3 s t d = c 2 S D ( Cog ) S D ( y )
Next, the variance and standard deviation of the standard logistic distribution are calculated.
V a r ( y ) = a 2 2 · V a r ( Ext ) + π 2 / 3
V a r ( Cog ) = b 2 2 · V a r ( Ext ) + π 2 / 3
V a r ( y ) = c 2 2 · V a r ( Ext ) + c 3 2 · V a r ( Cog ) + 2 c 2 c 3 · C o v ( Ext , Cog ) + π 2 / 3
Finally, the standard errors of the standardized regression coefficients are calculated.
S E ( a 2 s t d ) = S E ( a 2 ) S D ( Ext ) S D ( y )
S E ( b 2 s t d ) = S E ( b 2 ) S D ( Ext ) S D ( Cog )
S E ( c 2 s t d ) = S E ( c 2 ) S D ( Ext ) S D ( y )
S E ( c 3 s t d ) = S E ( c 3 ) S D ( Cog ) S D ( y )

3.3.3. Moderating Effect Test: The Moderating Effect of Neighborhood Effect

This paper draws on the research results of Hayes [47] to test whether the neighborhood effect can be used as a moderating variable to change the effect of ecological perception on WTU. Although the neighborhood effect variable is multi-categorical, it is assumed that the stronger the neighborhood effect is, the larger the moderating effect is. Therefore the variable can be treated as a continuous variable. Equation (28) is constructed to test the hypotheses to verify the moderating effect of the neighborhood effect. The specific form of the two-stage equation is as follows.
ln p r o b ( y i ) 1 p r o b ( y i ) = d 1 + d 2 C o g i + d 3 N e i i + d 4 X i + ε 4 i
ln p r o b ( y i ) 1 p r o b ( y i ) = e 1 + e 2 C o g i + e 3 N e i i + e 4 C o g i · N e i i + e 5 X i + ε 5 i
In Equations (27) and (28), N e i i and C o g i represent the neighborhood effect and ecological cognition of farmer, respectively, X i refers to the control variables, d 1 , d 2 , d 3 , d 4 , e 1 , e 2 , e 3 , e 4 , and e 5 are the coefficients to be estimated, and ε 4 i and ε 5 i are the residual items. If R 2 of Equation (28) is significantly higher than R 2 of Equation (27), or C o g i · N e i i has a significant effect, it will prove that the neighborhood effect has a moderating effect.

3.4. Variables

3.4.1. Explanatory Variables

The explanatory variable in this paper is “farmers’ willingness to use organic fertilizer”. Considering the availability of organic fertilizer to farmers, organic fertilizer in this paper covers both home-produced organic fertilizer and commercial organic fertilizer purchased by farmers in the market. In the questionnaire, we set the question “Are you willing to use organic fertilizer?” to reflect the willingness of farmers to use organic fertilizer. Respondents rated their willingness and assigned a value to their answers from 1 to 5, which stand for very reluctant to very willing.

3.4.2. Core Explanatory Variables

The core explanatory variables of this paper are “agricultural extension services”, “ecological perception,” and “neighborhood effect.” Firstly, for “agricultural extension services “, the dummy variable was introduced into the model by asking the household head whether they have received any extension services related to mango cultivation, and the dummy variable was “0” if the household said they had not received any agricultural extension services, and “1” if the opposite. Second, for ecological cognition, based on the existing literature [48], ecological cognition refers to the degree of personal knowledge about the environment and the capacity to acquire relevant ecological knowledge, which indicates whether farmers know how to reduce damage to the ecological environment in production and life or how to protect the ecological environment. In view of this, we questioned farmers whether excessive fertilizer application would cause environmental pollution. If the farmer answered “no,” the value was assigned as “0”, and vice versa as “1”. Finally, the neighborhood effect refers to the mutual influence of interpersonal behavior and because the information transfer and communication between neighbors play an important role in the process of neighborhood effect affecting farmers’ behavior [49], this paper examines whether farmers can obtain helpful information from their neighbors. As a result, we question farmers in this research whether they receive beneficial information from their neighbors as a major variable of the neighborhood effect and assign values to various levels of access to information: “rarely = 1, less = 2, average = 3, more = 4, very much = 5”.

3.4.3. Control Variables

To avoid the interference of other factors that may affect farmers’ ecological perceptions and WTU, the respondents’ characteristics, family characteristics, and production characteristics were set as control variables in the model based on existing research results. Given that the endogenous conversion model requires the inclusion of exclusion variables, this paper selects “Importance of extension Information” as an instrumental variable by asking respondents, “Is the information obtained from extension agents important?”. When farmers believe that the information obtained from extension services is helpful for agricultural production decisions, they will rely heavily on agricultural extension services as a primary source of information for agricultural production decisions. Farmers’ subjective evaluation of AES information will affect their decision to accept AES or not but will not directly affect WTU We also validated the plausibility of this instrumental variable using econometric analysis. The coefficient of the instrumental variable was −0.382, with a p-value of 0.206 when we regress WTU on influencing factors, indicating that the importance of extension information does not directly affect WTU. But when we regress farmers’ acceptance of agricultural extension services on influencing factors, the coefficient of the instrumental variable was 1.602 with a p-value of 0.000, indicating instrumental variable “the importance of extension information” directly affects farmers’ acceptance of AES. The correlation requirement was satisfied.

4. Results

4.1. Comparison of Characteristics of Sample Farmers

Table 3 shows the descriptive statistical analysis of the variables and t-test results of the difference in the mean between the farmer who received AES and who did not. WTU of the group who received AES is significantly higher than those who have not received AES at a significant level of 5%. According to the statistical results, there is a significant difference in ecological cognition, neighborhood effect, political appearance, planting experience, WeChat usage, village cadres, and importance of AES information between the two groups. Since the difference in WTU probably does not stem from the direct impact of AES, so it is necessary to use the ESR model to measure whether AES affects WTU.

4.2. Effect of Agricultural Extension Services on Farmers’ Willingness to Use Organic Fertilizer

In this paper, Stata (Version 16.0, created by StataCorp LLC in Texas, USA) software was used for model estimation. The variables were tested for multicollinearity before conducting the regression analysis to circumvent multicollinearity. The results showed that the variables’ variance inflation factors (VIF) were all less than 5, indicating no problem with multicollinearity. Considering that sample selection bias may lead to endogeneity problems, which may make the traditional model misleading in estimating the effect of AES on WTU, this paper adopts an Endogenous Switching Model (ESR) for further estimation and selects “Does the information obtained from extension workers matter?” as the exclusive variable. The model estimation results are shown in Table 4.
As can be seen from Table 4, the Wald chi-square value is significant at the 1% level, indicating a good overall fit of the model, and the error correlation coefficient rho1 is significant at the 5% statistical level, indicating that there are unobservable factors that affect both farmers’ willingness to accept the AES and organic fertilizer application, and the baseline regression model may have selective bias, which suggests that the adoption of the endogenous transformation model is necessary. The positive estimate of rho1 indicates that farmers who received AES are more willing to use organic fertilizer. The negative estimate of rho0 indicates that farmers who did not receive agricultural extension have a lower WTU than the average farmers in the sample.
The regression model results of farmers’ decision to accept AES are shown in column (1) in Table 4. First, farmers’ subjective determination of the importance of AES has a significant positive effect on their acceptance of AES at the 1% level. Farmers will actively accept extension services when they believe that AES may help their production decisions. Second, farmers’ years of cultivation had a significant positive effect on their acceptance of AES at the 5% level, which may be due to the fact that the longer farmers have been engaged in the mango cultivation, the more they can appreciate the benefits of AES for their cultivation. Moreover, WeChat usage significantly affected farmers’ acceptance of AES at the 1% level. As a new channel for farmers’ daily communication and information transfer, WeChat can strengthen the information exchange among village people and transmit village-related information through the “group effect,” thus keeping villagers informed of AES activities. Finally, village cadres in the household had a significant positive effect on the acceptance of AES at the 5% level, probably because village cadres were the first to be exposed to information about AES activities in the village and actively participated in agricultural extension-related activities.
The model estimation results of the WTU among farmers receiving AES are shown in columns (2) to (3) in Table 3. From regression (2), it can be seen that years of education significantly affect WTU of farmers who received AES at the 5% level. At the same time, there is no significant effect on WTU of farmers who did not receive AES. This may be due to the fact that after receiving AES, farmers with higher education are more likely to combine their existing knowledge to improve their resource allocation, environmental protection, and innovation abilities, thus better promoting their WTU. Ethnicity has a significant negative effect on WTU for farmers who did not receive AES at the 5% level, indicating that ethnic minority farmers are more willing to use organic fertilizer. Field research found that the early generations of Hainan ethnic minorities lived in a more primitive ecological environment and formed a harmonious coexistence between human and natural environment customs and behaviors. They were more inclined to adopt environmentally friendly production methods than the Han Chinese. From regression (3), it can be seen that after receiving AES, the behavioral differences between ethnic groups may disappear due to equal access to information. The number of phone contacts had a significant positive effect on organic fertilizer use willingness among farmers who received extension services at the 1% level, indirectly verifying the positive role of social networks in promoting green technology adoption behavior, in line with the study of Ogunleye (2021) [50]. Annual household income has a significant positive impact on the willingness to use organic fertilizer of farmers who did not participate in the extension services at the level of 1%, which may be due to the fact that as economic capital rises, farmers’ attention will gradually shift from output and income to comprehensive benefits such as the quality of agricultural products and ecological benefits, so they are more willing to use organic fertilizer. However, after receiving AES, lower-income farmers can improve their WTU, and the difference caused by income disappears. In addition, the effect of cropping year on WTU was significant at the 1% and 5% levels for the two types of farmers, confirming whether farmers received AES cropping years had a significant effect on the WTU. Since organic fertilizers tend to take more time to perform their functions, such as improving the physical and chemical properties of soil and enhancing the quality of agricultural products compared to traditional chemical fertilizers. The older the farmers who have been cultivating for a longer time, the more deeply they can appreciate the benefits of organic fertilizer application for mango cultivation, which is in line with the conclusion of Xu et al. [51] that the technology adoption effect of farmers strengthens their willingness to adopt.
This paper further measured the average treatment effect of AES on WTU using Equations (8) and (9). The estimated results are shown in Table 5. The average treatment effect (ATT) for farmers who had received AES in the counterfactual analysis framework (assuming no acceptance) was 1.300, indicating accepting AES could increase WTU to the extent of 1.300. For farmers who had not received AES in the counterfactual analysis framework (assuming acceptance), the average treatment effect (ATU) was 0.287, which means receiving AES could increase farmers’ willingness to the extent of 0.287. The results proved that acceptance of AES could effectively increase farmers’ WTU. With the continuous activity of AES in rural areas, the content of AES has shifted from focusing only on yield and technical application to more dimensions such as agricultural product quality and ecological protection. The benefits of applying organic fertilizer have been recognized by farmers and stimulated their enthusiasm for using organic fertilizer. Hypothesis 1 was tested.

4.3. Testing the Mediating Effect of Ecological Cognition

This paper tests the mediating effect of ecological cognition in the process of AES’ influence on farmers’ WTU with the help of the Ordered Logit model. The regression results are shown in Table 6. According to regressions (4) and (5), AES has a significant positive effect on farmers’ ecological perceptions and WTU, both at the 1% level. From regression (7), it is clear that AES still had a significant effect on farmers’ WTU when AES and ecological perceptions were put into the same model for regression, but the regression coefficient was reduced. The criteria for judging the mediating effect indicated that in the process of AES’ influence on farmers’ WTU, farmers’ ecological cognitive level has a partial mediating effect. That is, AES has a direct effect on farmers’ WTU and has an indirect effect on farmers’ WTU by affecting farmers’ ecological cognitive level. When farmers receive AES, they are guided by agronomists to perform reasonable practices to increase their willingness to apply organic fertilizer. They are also aware of the benefits of organic fertilizer application for environmental protection. Therefore, while agricultural extension service promotes farmers’ WTU, they also increase their willingness by improving ecological cognition. Hypothesis 2 was tested.
In order to clarify the weight of the mediating effect and the direct effect in the total effect, the original coefficients were processed, and comparable coefficients were obtained using the method summarized by Mackinnon (2008) [52]. The results are shown in Table 7. The proportion of the mediating effect to the total effect was calculated as, that is, the proportion of the mediating effect of farmers’ willingness to accept AES on organic fertilizer was 17.843%. This indicates that the effect of AES on farmers’ WTU is mainly a direct effect. At the same time, the AES will positively influence farmers’ WTU indirectly by raising their ecological awareness level.

4.4. Tesing the Moderating Effect of Neighborhood Effect

Based on the previous analysis, this paper tests the moderating effect of neighborhood effects on the relationship between the level of farmers’ ecological perceptions and their WTU to understand the boundary conditions under which farmers’ ecological perceptions play a mediating role. Drawing on the findings of Wen et al. (2005) [53], a two-stage regression analysis was used, and the Ecological cognition and Neighborhood effect were mean-centered, respectively. The regression results are shown in Table 8. As shown in Table 8, the interaction term (neighborhood effect, ecological perception) positively affected farmers’ WTU at the 10% level. The regression (9) was significantly larger than that in regression (8), which indicated that there was a significant moderating effect of the neighborhood effect, i.e., the more neighbors had access to more helpful information, the stronger the influence of farmers’ ecological perception on their WTU. Hypothesis 3 was tested.

4.5. Robustness Test

4.5.1. Robustness Test 1: The Effect of Agricultural Extension Services on Farmers’ Willingness to Use Organic Fertilizer

In order to verify the robustness of the effect of agricultural extension on farmers’ WTU, this paper uses the propensity score matching method (PSM) to analyze the net effect of farmers’ acceptance of agricultural extension on their WTU (ATT). The specific idea is as follows: farmers who received AES are set as the experimental group, and those who did not receive agricultural extension are set as the control group accordingly; then, each farmer’s score is measured by the logit model, and three methods, K proximity matching (K value is set to 4), radius matching (radius is set to 0.050) and kernel matching (bandwidth is set to 0.050), are selected to match the experimental group with the control group; finally, the differences in farmers’ willingness to adopt organic fertilizer application between the experimental and control groups were analyzed, and the average treatment effect (ATT) was calculated. As shown in Table 9, after propensity score matching, the acceptance of AES significantly increased the probability of farmers’ willingness to adopt organic fertilizer application, consistent with the previous ESP model estimation results. It indicates that the findings of the previous analysis are robust. The difference is that the ATT obtained from the ESP model estimation is larger than that obtained from the PSM model, which is due to the fact that the PSM model ignores the selectivity bias caused by unobservable variables in the estimation, and the results obtained may be biased.

4.5.2. Robustness Test II: The Mediating Effect of Ecological Cognition

This paper tested the mediating effect of ecological cognition by randomly selecting 85% of the samples to form 418 new samples and replacing the original model with an Ordered Probit model. As shown in Table 10, the sign and significance of the coefficients of the core explanatory variables are consistent with the previous results, further confirming that the study results have good robustness.

4.5.3. Robustness Test III: Moderating Effect of Neighborhood Effect

To further test the robustness of the moderating effect, this paper used subsamples to analyze the variability of ecological cognition as a mediating variable on farmers’ willingness to use organic fertilizer. According to the farmers’ scores on the extent of obtaining helpful information from people around them, “never = 1, less = 2” was classified as low, and “more = 4, a lot = 5” was classified as high. For the option of “average = 3”, this paper draws on Jiang et al.’s [48] method, which suggests that the choice of “average” indicates that the respondents disapprove of the opinion, so it is categorized as low. The regression results are shown in Table 11.
As shown in Table 11, in the subsample with a “low degree” of neighborhood effect, the level of ecological awareness had no significant effect on farmers’ WTU; in the subsample with a “high degree” of neighborhood effect, the level of ecological awareness had a positive effect on farmers’ WTU at the 1% level. In the sub-sample with a “high degree of neighborhood effect”, the level of ecological awareness positively affected farmers’ intention to fertilize organically at the 1% level. Thus, the positive effect of ecological perception on farmers’ WTU was gradually strengthened with the increasing degree of neighborhood effect, which verified the neighborhood effect’s moderating effect.

5. Discussion

Increased use of organic fertilizers by farmers will not only serve to cultivate the land, but also effectively alleviate the problem of surface pollution caused by excessive application of chemical fertilizers, thereby protecting the ecological environment [54,55]. Therefore, a lot of publicity and guidance work has been done around organic fertilizers in China’s agricultural extension services, which are mainly aimed at improving farmers’ awareness level of organic fertilizer application for ecological protection, as well as their technical capacity in organic fertilizer application, etc. In general, the efforts made by the Chinese government in agricultural extension services have achieved certain effect, the amount of chemical fertilizer applied per unit area has started to trend downward, and the proportion of organic fertilizer use has been increased to some extent. However, the role of agricultural extension services in improving farmers’ fertilizer use is still relatively limited, and the structure of fertilizer use in Chinese agricultural production is still highly irrational [3,8]. The above is one of the important reasons we want to explore the impact of agricultural extension services on farmers’ willingness to use organic fertilizer. We also believe our findings can help the government improve the existing agricultural extension services. In addition, there are several aspects worthy of further discussion for the current work.
This paper focused on the influence mechanism of agricultural extension services on farmers’ willingness to use organic fertilizer and introduced the mediating effect of ecological cognition and the moderating effect of neighborhood effect, which helped us to have a deeper understanding of the influence mechanism. Moreover, according to the existing literature, we found that some scholars focused on the effects of both social capital and ecological cognition on farmers’ eco-friendly production behavior [56,57], and they took ecological cognition as a mediating variable in the influence mechanism of social capital on farmers’ eco-friendly production behavior, which is significantly different from our study. In our view, agricultural extension services play a more significant role in enhancing farmers’ ecological cognition, which is also one of the critical objectives of agricultural extension services, so we consider ecological cognition as a mediating variable between agricultural extension services and farmers’ ecologically friendly willingness. Meanwhile, the neighborhood effect may have a moderating effect on the above influence mechanism, because there may be some discussion and communication between neighbors after receiving agricultural extension services.
All these above are the improvements in this study compared with other studies, but there is also room for further study. For example, the influence mechanism of agricultural extension services on farmers’ willingness to use organic fertilizers is discussed only from the perspective of ecological cognition, but whether other mediating variables can be included? This remains to be further verified. Furthermore, the skill level of farmers is also one of the important targets of agricultural extension services concern, and it is likely to be one of the ideal mediating variables. However, due to the omission of our preliminary questionnaire design, we did not include relevant questions for measuring farmers’ skill levels in organic fertilizer application, which prevented us from verifying this mediating mechanism. In view of this, a more complete and targeted questionnaire and theoretical model could be designed to explore this issue in more depth in future studies.
In terms of policy application, according to our main findings, we tentatively put forward the following possible supplementary measures or improvement directions for China’s agricultural extension policy. Firstly, agricultural extension departments should further strengthen agricultural technology promotion and publicity, and improve the construction of the extension system. Secondly, more technical training and educational campaigns should be actively carried out to make farmers aware of the importance of agricultural product safety and ecological environmental protection, green production knowledge should be popularized to improve farmers’ awareness. Thirdly, we encourage neighborly exchanges among rural residents, and cultivate local elite farmers to promote other farmers’ environmentally friendly behavior.

6. Conclusions

In recent years, the problem of agricultural non-point source pollution caused by excessive chemical fertilizer application has sparked widespread public concern. Many regions have implemented policies to expedite the transition to organic fertilizers and promote green and sustainable agricultural development through increased agricultural extension services efforts. This paper uses the survey data of 492 mango farmers in Hainan to analyze the effect of agricultural extension on farmers’ willingness to use organic fertilizer with an ESR model, and explore the mediating effect of ecological cognition and the moderating effect of neighborhood effect with a moderated mediating effect model. The main conclusions are as follows: (1) agricultural extension services have a significant positive effect on farmers’ willingness to use organic fertilizer and can enhance farmers’ willingness to use organic fertilizer through agricultural promotion; (2) agricultural extension services have an indirect effect on farmers’ willingness to use organic fertilizer, which shows that it can enhance farmers’ willingness to use organic fertilizer by improving their ecological cognition, and ecological cognition has a partial mediating effect, and the mediating effect accounts for 17.84% of the total effect; (3) in the influence mechanism of ecological cognition on farmers’ willingness to use organic fertilizer, the neighborhood effect has a positive moderating effect.

Author Contributions

Conceptualization, D.Q., T.X. and N.L.; methodology, N.L.; software, N.L.; validation, T.X. and D.Z.; formal analysis, D.Q. and N.L.; investigation, D.Q. and Y.Z.; resources, D.Q. and D.Z.; data curation, N.L.; writing—original draft preparation, D.Q.; writing—review and editing, D.Q., N.L. and L.C. visualization, N.L.; supervision, D.Z. and T.X.; project administration, D.Q.; funding acquisition, T.X. and D.Q. All authors have read and agreed to the published version of the manuscript.

Funding

The research was supported by the National Natural Science Foundation of China (No. 72003054, No. 72103052, No. 71773134, No. 72073135), The Youth Project of Humanities and Social Science Foundation of the Ministry of Education (No. 19XJC790011), Hainan Provincial Natural Science Foundation of China (No. 719QN197, No. 720RC576).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to personal privacy and non-open access to the research program.

Conflicts of Interest

We declare that there is no conflict of interest.

References

  1. Chen, X.; Zeng, D.; Xu, Y.; Fan, X. Perceptions, Risk Attitude and Organic Fertilizer Investment: Evidence from Rice and Banana Farmers in Guangxi, China. Sustainability 2018, 10, 3715. [Google Scholar] [CrossRef]
  2. Qin, G.; Niu, Z.; Yu, J.; Li, Z.; Ma, J.; Xiang, P. Soil heavy metal pollution and food safety in China: Effects, sources and removing technology. Chemosphere 2021, 267, 129205. [Google Scholar] [CrossRef] [PubMed]
  3. Wu, H.; Hao, H.; Lei, H.; Ge, Y.; Shi, H.; Song, Y. Farm Size, Risk Aversion and Overuse of Fertilizer: The Heterogeneity of Large-Scale and Small-Scale Wheat Farmers in Northern China. Land 2021, 10, 111. [Google Scholar] [CrossRef]
  4. Tabaxi, I.; Kakabouki, I.A.K.C.; Folina, A.; Karydogianni, S.; Kalivas, A.; Bilalis, D.J. Effect of organic fertilization on soil characteristics, yield and quality of Virginia Tobacco in Mediterranean area. Emir. J. Food Agric. 2020, 32, 610–616. [Google Scholar] [CrossRef]
  5. Samuel, A.D.; Bungau, S.G.; Fodor, I.K.; Tit, D.M.; Melinte, C.E. Effects of Liming and Fertilization on the Dehydrogenase and Catalase Activities. Rev. Chim. 2019, 70, 3464–3468. [Google Scholar] [CrossRef]
  6. Samuel, A.D.; Bungau, S.G.; Tit, D.M.; Frunzulica, C.; Badea, G.E. Effects of Long Term Application of Organic and Mineral Fertilizers on Soil Enzymes. Rev. Chim. 2018, 69, 2608–2612. [Google Scholar] [CrossRef]
  7. Du, Y.; Cui, B.; Zhang, Q.; Wang, Z.; Sun, J.; Niu, W. Effects of manure fertilizer on crop yield and soil properties in China: A meta-analysis. Catena 2020, 193, 104617. [Google Scholar] [CrossRef]
  8. Li, B.; Shen, Y. Effects of land transfer quality on the application of organic fertilizer by large-scale farmers in China. Land Use Policy 2021, 100, 105124. [Google Scholar] [CrossRef]
  9. Yang, Y.; He, Y.; Li, Z.W. Social capital and the use of organic fertilizer: An empirical analysis of Hubei Province in China. Environ. Sci. Pollut. Res. 2020, 27, 15211–15222. [Google Scholar] [CrossRef]
  10. Fang, P.; Abler, D.; Lin, G.; Sher, A.; Quan, Q. Substituting Organic Fertilizer for Chemical Fertilizer: Evidence from Apple Growers in China. Land 2021, 10, 858. [Google Scholar] [CrossRef]
  11. Abebe, G.; Debebe, S. Factors affecting use of organic fertilizer among smallholder farmers in Sekela district of Amhara region, Northwestern Ethiopia. Cogent Food Agric. 2019, 5, 1669398. [Google Scholar] [CrossRef]
  12. Yi, X.; Yu, L.; Chang, S.; Yin, C.; Wang, H.; Zhang, Z. The effects of China’s Organic-Substitute-Chemical-Fertilizer (OSCF) policy on greenhouse vegetable farmers. J. Clean. Prod. 2021, 297, 126677. [Google Scholar] [CrossRef]
  13. Guo, L.; Li, H.; Cao, X.; Cao, A.; Huang, M. Effect of agricultural subsidies on the use of chemical fertilizer. J. Environ. Manag. 2021, 299, 113621. [Google Scholar] [CrossRef] [PubMed]
  14. Martey, E. Welfare effect of organic fertilizer use in Ghana. Heliyon 2018, 4, e00844. [Google Scholar] [CrossRef] [PubMed]
  15. Li, J.; Liu, P.; Li, Z. Optimal design and techno-economic analysis of a solar-wind-biomass off-grid hybrid power system for remote rural electrification: A case study of west China. Energy 2020, 208, 118387. [Google Scholar] [CrossRef]
  16. Brown, S.A. Household technology adoption, use, and impacts: Past, present, and future. Inf. Syst. Front. 2008, 10, 397–402. [Google Scholar] [CrossRef]
  17. Bernstein, H. Food sovereignty via the ‘peasant way’: A sceptical view. J. Peasant Stud. 2014, 41, 1031–1063. [Google Scholar] [CrossRef]
  18. Castillo, G.; Engler, A.; Wollni, M. Planned behavior and social capital: Understanding farmer? behavior toward pressurized irrigation technologies. Agric. Water Manag. 2021, 243, 106524. [Google Scholar] [CrossRef]
  19. Ji, C.; Jin, S.; Wang, H.; Ye, C. Estimating effects of cooperative membership on farmers’ safe production behaviors: Evidence from pig sector in China. Food Policy 2019, 83, 231–245. [Google Scholar] [CrossRef]
  20. Daxini, A.; Ryan, M.; O’Donoghue, C.; Barnes, A.P. Understanding farmer? intentions to follow a nutrient management plan using the theory of planned behaviour. Land Use Policy 2019, 85, 428–437. [Google Scholar] [CrossRef]
  21. Junaidi, J. Social Status of the Fish-farmers of Floating-net-cages in Lake Maninjau, Indonesia. J. Aquac. Res. Dev. 2015, 7, 1. [Google Scholar] [CrossRef]
  22. Cook, B.R.; Satizábal, P.; Curnow, J. Humanising agricultural extension: A review. World Dev. 2021, 140, 105337. [Google Scholar] [CrossRef]
  23. Qiao, D.; Lu, Q.; Xu, T. Social network, extension service and farmers water-saving irrigation technology adoption in Minqin County. Resour. Sci. 2017, 3, 441–450. [Google Scholar] [CrossRef]
  24. Faure, G.; Desjeux, Y.; Gasselin, P. New Challenges in Agricultural Advisory Services from a Research Perspective: A Literature Review, Synthesis and Research Agenda. J. Agric. Educ. Ext. 2012, 18, 461–492. [Google Scholar] [CrossRef]
  25. Cai, J.; Jia, Y.; Hu, R.; Zhang, C. Four Decades of China’s Agricultural Extension Reform and its Impact on Agent? Time Allocation. Aust. J. Agric. Resour. Econ. 2019, 64, 104–125. [Google Scholar] [CrossRef]
  26. Anderson, J.R.; Feder, G.S.R. Rural Extension Services; Agriculture and Rural Development Department World Bank: Washington, DC, USA, 2003. [Google Scholar] [CrossRef]
  27. Anderson, J.R.; Feder, G. Chapter 44 Agricultural Extension. In Handbook of Agricultural Economics; Evenson, R., Pingali, P., Eds.; Elsevier: Amsterdam, The Netherlands, 2007; Volume 3, pp. 2343–2378. [Google Scholar]
  28. Wossen, T.; Abdoulaye, T.; Alene, A.; Haile, M.G.; Feleke, S.; Olanrewaju, A.; Manyong, V. Impacts of extension access and cooperative membership on technology adoption and household welfare. J. Rural Stud. 2017, 54, 223–233. [Google Scholar] [CrossRef]
  29. Wossen, T.; Berger, T.; Di Falco, S. Social capital, risk preference and adoption of improved farm land management practices in Ethiopia. Agric. Econ-Blackwell 2015, 46, 81–97. [Google Scholar] [CrossRef]
  30. Ghosh, S. in Innovations in public sector-led agricultural extension. Sci. Res. Essays 2012, 7, 4170–4175. [Google Scholar]
  31. Muhammed, S.H.; Waktola, M.; Adunea, D. Determinants of adoption of agricultural extension package technologies by smallholder households on sorghum production: Case of Gemechis and Mieso districts of West Hararghe Zone, Oromia Regional State, Ethiopia. J. Agric. Ext. Rural Dev. 2020, 12, 62–75. [Google Scholar] [CrossRef]
  32. Ajzen, I. The theory of planned behavior. Organ. Behav. Hum. Decis. 1991, 50, 179–211. [Google Scholar] [CrossRef]
  33. Kotchen, M.J.; Reiling, S.D. Environmental attitudes, motivations, and contingent valuation of nonuse values: A case study involving endangered species. Ecol. Econ. 2000, 32, 93–107. [Google Scholar] [CrossRef]
  34. Rezaei, R.; Safa, L.; Ganjkhanloo, M.M. Understanding farmers’ ecological conservation behavior regarding the use of integrated pest management- an application of the technology acceptance model. Glob. Ecol. Conserv. 2020, 22, e941. [Google Scholar] [CrossRef]
  35. Bandura, A. Social foundations of thought and action: A social cognitive theory. In Social Foundations of Thought and Action: A Social Cognitive Theory; Prentice-Hall, Inc.: Englewood Cliffs, NJ, USA, 1986; p. 617. [Google Scholar]
  36. Shahangian, S.A.; Tabesh, M.; Yazdanpanah, M. Psychosocial determinants of household adoption of water-efficiency behaviors in Tehran capital, Iran: Application of the social cognitive theory. Urban Clim. 2021, 39, 100935. [Google Scholar] [CrossRef]
  37. Fei, X.T. Rural China; Shanghai People’s Publishing House: Shanghai, China, 2008; Volume 1, pp. 1–9. [Google Scholar]
  38. Wydick, B.; Hayes, H.K.; Kempf, S.H. Social Networks, Neighborhood Effects, and Credit Access: Evidence from Rural Guatemala. World Dev. 2011, 39, 974–982. [Google Scholar] [CrossRef]
  39. Tsusaka, T.W.; Kajisa, K.; Pede, V.O.; Aoyagi, K. Neighborhood effects and social behavior: The case of irrigated and rainfed farmers in Bohol, the Philippines. J. Econ. Behav. Organ. 2015, 118, 227–246. [Google Scholar] [CrossRef]
  40. Xie, X.X.; Chen, M.Q. Years of Farming, Neighborhood Communication and Farmers’ Ecological Farming Adoption: Based on Data Validation in Jiangxi Province. Resources and Environment in the Yangtze Basin. China/Asia Demand 2020, 4, 1016–1026. [Google Scholar] [CrossRef]
  41. Qiao, D.; Li, W.; Zhang, D.; Yan, Y.; Xu, T. How do You Want to restore?—Assessing the Public Preferences and Social Benefits of Ecological Restoration for Natural Rubber Plantation in China. Front. Environ. Sci. 2022, 10, 92. [Google Scholar] [CrossRef]
  42. Lokshin, M.; Sajaia, Z. Impact of interventions on discrete outcomes: Maximum likelihood estimation of the binary choice models with binary endogenous regressors. Stata J. 2011, 11, 368–385. [Google Scholar] [CrossRef]
  43. Wen, Z.L.; Ye, B.J. Analyses of Mediating Effects: The Development of Methods and Models. Adv. Psychol. Sci. 2014, 5, 731–745. [Google Scholar] [CrossRef]
  44. Iacobucci, D. Mediation analysis and categorical variables: The final frontier. J. Consum. Psychol. 2012, 22, 582–594. [Google Scholar] [CrossRef]
  45. Baron, R.M.; Kenny, D.A. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J. Personal. Soc. Psychol. 1986, 51, 1173–1182. [Google Scholar] [CrossRef]
  46. Mackinnon, D.P.; Fairchild, A.J.; Fritz, M. Mediation analysis. Annu. Rev. Psychol. 2007, 58, 593–614. [Google Scholar] [CrossRef] [PubMed]
  47. Hayes, A. Introduction to mediation, moderation, and conditional process analysis. J. Educ. Meas. 2013, 51, 335–337. [Google Scholar] [CrossRef]
  48. Jiang, W.J.; Yan, T.W.; Zhang, J.B. Can Internet Use Promote Farmers to Adopt Straw Returning Technology?—An Empirical Analysis Based on Endogenous Switching Probit Mode. J. Agrotech. Eco. 2021, 3, 50–62. [Google Scholar] [CrossRef]
  49. Timothy, G.C.; Christopher, R.U. Learning about a New Technology: Pineapple in Ghana. Am. Econ. Rev. 2010, 100, 35–69. [Google Scholar] [CrossRef]
  50. Ogunleye, A.; Kehinde, A.; Mishra, A.; Ogundeji, A. Impacts of farmers’ participation in social capital networks on climate change adaptation strategies adoption in Nigeria. Heliyon 2021, 7, e8624. [Google Scholar] [CrossRef] [PubMed]
  51. Xu, T.; Zhao, M.J.; Li, E.H. The impact of technology perception and subsidy policy on different phases of farmers’ water-saving irrigation technology adoption. Res. Sci. 2018, 4, 809–817. [Google Scholar] [CrossRef]
  52. Mackinnon, D.P. Introduction to Statistical Mediation Analysis; Routledge: Abingdon, UK, 2008. [Google Scholar]
  53. Wen, Z.L.; Hou, J.T.; Zhang, L. Comparison and application of moderating effects and mediating effects. J. Psycho. 2005, 2, 268–274. [Google Scholar]
  54. Bungau, S.; Behl, T.; Aleya, L.; Bourgeade, P.; Aloui-Sossé, B.; Purza, A.L.; Abid, A.; Samuel, A.D. Expatiating the impact of anthropogenic aspects and climatic factors on long-term soil monitoring and management. Environ. Sci. Pollut. Res. 2021, 28, 30528–30550. [Google Scholar] [CrossRef]
  55. Behl, T.; Kaur, I.; Sehgal, A.; Singh, S.; Sharma, N.; Bhatia, S.; Al-Harrasi, A.; Bungau, S. The dichotomy of nanotechnology as the cutting edge of agriculture: Nano-farming as an asset versus nanotoxicity. Chemosphere 2022, 288, 132533. [Google Scholar] [CrossRef]
  56. Xiao, Y.; Qi, Z.H.; Yang, C.T.; Liu, Z. Social Capital, Ecological Cognition and Rarional Fertilization Behavior of Farmers: Empirical Analysis Based on Structural Equation Model. J. China Agric. Univ. 2021, 26, 249–262. [Google Scholar]
  57. Jiang, W.J.; Yan, T.W.; Jiang, X.; Zhang, J.B. Influence of Social Network and Ecological Cognition on farmer’s willingness of straw returning. J. China Agric. Univ. 2019, 24, 203–216. [Google Scholar]
Figure 1. The theory model.
Figure 1. The theory model.
Sustainability 14 07166 g001
Figure 2. Study area.
Figure 2. Study area.
Sustainability 14 07166 g002
Table 1. Descriptive statistics of sample farmers’ characteristics.
Table 1. Descriptive statistics of sample farmers’ characteristics.
ItemsLevelsObservationsFrequency (%)ItemsLevelsObservationsFrequency (%)
WTULow13026.42Received AES or notYes28758.33
High36273.58No20541.67
GenderMale46293.90EthnicityHan25852.44
Female306.10Minority23447.56
Age (years)≤2971.42Education (years)≤610821.90
30~395210.577~930562.60
40~4912826.0210~126813.80
≥5030561.99≥13183.70
Planting scale (acres)≤1516533.0Annual household income(one thousand yuan)≤1102.0
16~4521243.11~1021543.7
46~909719.710~5023447.6
≥91367.3≥50336.7
Table 2. Statistical results of farmers’ willingness to use organic fertilizers.
Table 2. Statistical results of farmers’ willingness to use organic fertilizers.
WTUReceived AESNot Received AES
ObservationsPercentage (%)ObservationsPercentage (%)
Low5619.51%7436.10%
High23180.49%13163.90%
Total287100205100
Table 3. Variable Definition and Description Statistics.
Table 3. Variable Definition and Description Statistics.
ItemsVariablesDefinitionMean of the Observations Received AESMean of the Observations Not Received AESDifference
Core explanatory variablesWTUAre respondents willing to use organic fertilizers? 1 = very unwilling, 2 = unwilling, 3 = average, 4 = more willing, 5 = very willing3.9903.556−0.433 ***
AESHave the respondents received AES related to mango cultivation? 0 = no, 1 = yes
Ecological cognitionDo respondents think that excessive application of fertilizers will pollute the environment? 0 = no, 1 = yes0.5440.410−0.134 ***
Neighborhood EffectIs the respondent able to obtain useful information from the surrounding neighbors? 1 = rare, 2 = few, 3 = fair, 4 = more, 5 = very much3.8683.605−0.263 ***
Control variablesGenderMale = 1, Female = 00.9410.937−0.004
AgeRespondent’s age (years)51.36950.951−0.418
EthnicityHan = 1, Minority = 00.5300.517−0.013
Political appearanceParty member = 1, non-party member = 00.2090.1220.087 **
EducationNumber of years of schooling of respondents7.9657.3240.641 **
Planting experienceNumber of years respondents have been engaged in mango farming (years)18.36216.576−1.787 ***
Frequent contactsNumber of relatives and friends with whom the respondent often interacts15.48815.9900.502
WeChat usageDoes the respondent use WeChat? 0 = no, 1 = yes0.8920.717−0.175 ***
Phone contactsNumber of contacts in the phone address book117.693107.351−10.342
Agricultural laborNumber of family farm laborers2.1992.2630.065
Village cadresWhether there are village cadres among the family members. 0 = no, 1 = yes0.1850.078−0.107 ***
Annual incomeThe total annual income of respondents’ households (ten thousand yuan)17.06218.3341.172
Farm incomePercentage of agricultural income in respondents’ total household income (%)0.6220.6400.018
Land fragmentationNumber of parcels of land owned by respondents2.4692.224−0.244
Planting scaleArea of land operated by respondents (mu)33.95236.6542.702
Instrumental variableImportance of AES InformationIs the information obtained from the promoter important? important = 1, not important = 00.1430.034−0.109 ***
Note: (1) ***, ** indicate significance at the 1%, 5% levels, respectively; (2) “−” stands for no estimated results, same for the following tables.
Table 4. Regression results of ESR model.
Table 4. Regression results of ESR model.
VariablesAES
Regression (1)
WTU
Received AES
Regression (2)
Not received AES
Regression (3)
Importance of AES Information0.996 *** (0.233)
Gender0.118 (0.263)−0.393 (0.252)−0.425 (0.340)
Age0.008 (0.007)−0.008 (0.007)−0.009 (0.008)
Ethnicity0.140 (0.137)0.064 (0.121)−0.369 ** (0.184)
Political appearance0.179 (0.186)0.134 (0.160)−0.227 (0.258)
Education0.013 (0.020)0.038 ** (0.019)0.017 (0.025)
Planting experience0.018 ** (0.009)0.027 *** (0.008)0.022 ** (0.011)
Frequent contacts−0.003 (0.004)0.001 (0.004)−0.002 (0.005)
WeChat usage0.759 *** (0.169)−0.277 (0.201)0.287 (0.248)
Phone contacts0.001 (0.001)0.001 *** (0.001)0.001 (0.001)
Agricultural labor−0.029 (0.060)0.014 (0.056)0.028 (0.076)
Village cadres0.424 ** (0.200)0.048 (0.167)−0.097 (0.310)
Annual income−0.002 (0.002)0.005 (0.004)0.015 *** (0.006)
Farm income−0.161 (0.317)0.038 (0.297)0.450 (0.393)
Land fragmentation0.037 (0.033)0.001 (0.024)−0.058 (0.049)
Planting scale−0.002 (0.002)−0.002 (0.002)0.001 (0.003)
Constant−1.386 *** (0.550)3.630 *** (0.618)2.925 *** (0.690)
rho1 0.443 ** (0.174)
rho0 −0.470 (0.274)
log-likelihood−967.91
Wald test35.63 ***
Note: (1) ***, ** indicate significance at the 1%, 5% levels, respectively; (2) Standard errors are in parentheses.
Table 5. Average treatment effect of AES on farmers’ willingness to use organic fertilizer.
Table 5. Average treatment effect of AES on farmers’ willingness to use organic fertilizer.
ObservationsReceived AESNot Received AESATTATU
Received AES3.991
(0.019)
2.691
(0.028)
1.300 ***
(0.034)
Not received AES3.556
(0.036)
3.270
(0.026)
0.287 ***
(0.044)
Note: (1) *** indicates significance at the 1% level; (2) Standard errors are in parentheses.
Table 6. Effects of AES and ecological cognition on farmers’ willingness to use organic fertilizer.
Table 6. Effects of AES and ecological cognition on farmers’ willingness to use organic fertilizer.
VariablesWTU
Regression (4)
Ecological Cognition Regression (5)WTU
Regression (6)
Coef.Std. Err.Coef.Std. Err.Coef.Std. Err.
AES0.678 ***0.1800.485 ***0.1920.648 ***0.181
Ecological cognition0.294 *0.175
Control variablesControlledControlledControlled
LR chi275.27024.4178.11
p-value0.0000.0580.000
Pseudo R20.0590.0350.060
Note: *** and * indicate significance at the 1% and 10% levels, respectively.
Table 7. Comparable processing results of regression coefficients.
Table 7. Comparable processing results of regression coefficients.
VariablesEcological CognitionWTUWTU
Initial Coef.Comparable Coef.Initial Coef.Comparable Coef.Initial Coef.Comparable Coef.
Ecological cognition0.294 *0.077 *
(0.175)(0.046)
AES0.484 ***0.577 ***0.677 ***0.249 ***0.648 ***0.203 ***
(0.192)(0.229)(0.180)(0.066)(0.181)(0.057)
Note: (1) *** and * indicate significance at the 1% and 10% levels, respectively; (2) Standard errors are in brackets.
Table 8. Moderating effect of neighborhood effect.
Table 8. Moderating effect of neighborhood effect.
VariablesRegression (7)Regression (8)
Coef.Std. Err.Coef.Std. Err.
Ecological cognition0.392 **0.1750.405 **0.176
Neighborhood effect0.456 ***0.0810.460 ***0.081
Interactive term0.263 *0.158
Control variablesControlledControlled
LR chi296.7199.49
p-value0.0000.000
Pseudo R20.0750.078
Note: (1) ***, ** and * indicate significance at the 1%, 5% and 10% levels, respectively; (2) to avoid the multicollinearity between ecological cognition, neighborhood effect, and the interaction term, we mean-center ecological cognition and neighborhood effect, and the obtained coefficient is the centralization coefficient.
Table 9. PSM Model estimates.
Table 9. PSM Model estimates.
VariablesApproachAverage Treatment Effectt-Test
AESNearest neighbor matching0.417 ***3.63
(K = 4)(0.115)
Radius matching0.398 ***3.69
(radius = 0.050)(0.108)
Kernel matching0.394 ***3.63
(bandwidth = 0.050)(0.109)
Note: (1) *** indicates significance at the 1% (2) Standard errors are in parentheses.
Table 10. The results of the mediation effect estimation based on the ordered Probit model.
Table 10. The results of the mediation effect estimation based on the ordered Probit model.
VariablesEcological Cognition
Regression (9)
WTU
Regression (10)
WTU
Regression (11)
Coef.Std. Err.Coef.Std. Err.Coef.Std. Err.
Ecological cognition0.221 **0.110
AES0.310 **0.1290.434 ***0.1110.411 ***0.112
Control variablesControlledControlledControlled
LR chi221.1967.3873.46
p-value0.0970.0000.000
Pseudo R20.0360.0610.065
Note: ***, ** indicate significance at the 1%, 5% levels, respectively.
Table 11. The effect of ecological cognition under subgroup regression.
Table 11. The effect of ecological cognition under subgroup regression.
VariablesLow Neighborhood Effect
Regression (12)
High Neighborhood Effect
Regression (13)
Coef.Std. Err.Coef.Std. Err.
Ecological cognition−0.1240.3100.689 ***0.224
Control variablesControlledControlled
Observations174318
LR chi230.5162.86
p Value0.0100.000
Pseudo R20.0630.081
Note: *** indicates significance at the 1% level.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Qiao, D.; Li, N.; Cao, L.; Zhang, D.; Zheng, Y.; Xu, T. How Agricultural Extension Services Improve Farmers’ Organic Fertilizer Use in China? The Perspective of Neighborhood Effect and Ecological Cognition. Sustainability 2022, 14, 7166. https://doi.org/10.3390/su14127166

AMA Style

Qiao D, Li N, Cao L, Zhang D, Zheng Y, Xu T. How Agricultural Extension Services Improve Farmers’ Organic Fertilizer Use in China? The Perspective of Neighborhood Effect and Ecological Cognition. Sustainability. 2022; 14(12):7166. https://doi.org/10.3390/su14127166

Chicago/Turabian Style

Qiao, Dan, Ningjie Li, Li Cao, Desheng Zhang, Yanan Zheng, and Tao Xu. 2022. "How Agricultural Extension Services Improve Farmers’ Organic Fertilizer Use in China? The Perspective of Neighborhood Effect and Ecological Cognition" Sustainability 14, no. 12: 7166. https://doi.org/10.3390/su14127166

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop