sustainability

Article
Effects of Social Capital on the Adoption of Green Production
Technologies by Rice Farmers: Moderation Effects Based on
Risk Preferences
Bo Liu 1 , Na Li 1, * and Cairong Liao 1,2, *

1 School of Economics and Management, Jiangxi Agricultural University, Nanchang 330045, China;
1079218971@[Link]
2 Jiangxi Rural Revitalization Strategy Research Institute, Jiangxi Agricultural University,
Nanchang 330045, China
* Correspondence: lina1724@[Link] (N.L.); liaocairong621@[Link] (C.L.)

Abstract: In the context of global climate change and energy and resource scarcity, understanding the
factors that influence the adoption of green production technology by rice farmers and promoting a
transformation in agricultural practices towards low-carbon and environmentally friendly methods
represent a crucial step towards achieving sustainable development. This study employs an ordered
logit model to comprehensively explore the influence mechanism of social capital and risk preference
within the adoption of green production technology by rice farmers. Research indicates that social
capital can significantly and positively affect rice farmers’ uptake of green production technology.
Among various dimensions, it can be seen that the role of social participation emerges as a particularly
salient factor. The analysis of risk preference demonstrates that it exerts a negative moderating
influence. The heterogeneity analysis results indicate that new-generation and part-time rice farmers
prefer adopting new technologies if they possess higher social capital. In light of the findings above,
it is recommended that the organizational support of cooperatives and collective action be reinforced;
the use of digital technologies in agriculture should be increased; the livelihood changes of rice
farmers be considered to meet their heterogeneous needs accurately; and the integration of carbon
Citation: Liu, B.; Li, N.; Liao, C. emission reduction policies and green production technologies be strengthened to promote the
Effects of Social Capital on the widespread adoption of green technologies among rice farmers.
Adoption of Green Production
Technologies by Rice Farmers:
Keywords: social capital; risk preference; technology adoption; professional rice farmer; green
Moderation Effects Based on Risk
energy transformation
Preferences. Sustainability 2024, 16,
8879. [Link]
su16208879

Academic Editors: Minzhe Du and 1. Introduction

Guanglai Zhang
Under the severe challenge of global climate change, agriculture, as a resource-
Received: 11 September 2024 intensive industry, increasingly emphasizes energy consumption and carbon emissions,
Revised: 29 September 2024 and sustainable agricultural development has become a central focus of global farm strate-
Accepted: 10 October 2024 gies [1]. In the data from the latest Statistical Yearbook of World Food and Agriculture 2023
Published: 14 October 2024 released by the Food and Agriculture Organization of the United Nations, it can be found
that global arable land resources currently represent only around 12% of the total land area,
and international agricultural land decreased by 86 million hectares between 2000 and 2021.
This, coupled with the problem of land degradation resulting from the long-term overuse
Copyright: © 2024 by the authors.
of chemical inputs, has severely weakened food production capacity and threatened the
Licensee MDPI, Basel, Switzerland.
livelihoods of more than 3.2 billion people globally, emphasizing the urgency of advancing
This article is an open access article
the sustainability of agricultural production [2]. As the world’s leading rice producer
distributed under the terms and
conditions of the Creative Commons
and consumer, China has significantly contributed to economic and social progress by
Attribution (CC BY) license (https://
developing the rice industry since the reform and opening up. However, this growth
[Link]/licenses/by/ model has frequently been accompanied by high resource consumption and environmental
4.0/). pressures [3]. With the aggravation of resource and environmental constraints, there is an

Sustainability 2024, 16, 8879. [Link] [Link]

Sustainability 2024, 16, 8879 2 of 20

urgent need for China’s rice industry to transition from a primitive development model to a
green, environment-friendly, and resource-saving production model to realize a low-carbon
green transition of industries [4]; it is imperative to enhance the effective utilization of
energy [5]. In this context, serving as a primary body of agricultural production, the success
or failure of the agricultural green transformation cannot be separated from the behavior of
rice farmers in adopting green production technologies based on digital technologies.
The technology of agriculture green production is defined as a collection of innovative
technologies, tools, and rule systems using digital technology and other means, aiming
to reduce environmental pollution, strengthen the quality of agricultural products, and
improve resource efficiency, which contributes a lot to realizing the rural energy transition
and promoting sustainable development of agriculture [6]. However, promoting green
production technologies is not a simple technology diffusion but a complex systemic change
process involving social, economic, and cultural dimensions. He et al. determined that
although green production technologies of agriculture have particular theoretical possibili-
ties to enhance low-carbon efficiency, the adoption rate remains low due to various factors,
including regional differences, farmers’ levels of awareness, and economic conditions in
China [7]. In this change process, the decision-making behavior of farmers represents a
crucial link in implementing technology. Furthermore, the accumulation of social capital
represents an essential factor that cannot be ignored when influencing farmers’ decision-
making. Liu et al. demonstrated that social capital significantly impacts adopting green
production technologies by enhancing farmers’ information acquisition, resource mobiliza-
tion, and digital technology application capabilities [8]. The research conducted by Wen
et al. posits that robust social capital facilitates farmers’ acquisition of technological infor-
mation, thereby bolstering their confidence and proclivity to adopt novel technologies. This
is achieved through establishing trust and demonstrating effects within the social network.
Additionally, social capital plays a pivotal role in resource sharing, effectively addressing
the resource constraints that farmers may encounter during the adoption process [9]. In
addition, Li et al. demonstrated that farmers witnessed a greatly enhanced possibility of
adopting green production technologies due to their participation in cooperatives. This was
predominantly due to the advantages of cooperatives in resource integration, information
sharing, and technical training. These factors have helped farmers accumulate rich social
capital and have promoted green production technologies to be more widely diffused in
the groups of rice farmers [10]. What is more, Liao et al. discovered that digital finance, as
a form of social capital, markedly enhanced the population’s well-being by fostering the
growth of social networks and providing robust social capital support for driving rural ar-
eas to adopt green production technologies [11]. The farmers will be more inclined to adopt
green production technologies by strengthening access to information, promoting resource
integration, providing technical training, and expanding social networks, all facilitated by
social capital.
Based on an analysis of prospect theory, and regarding farmers as ‘rational economic
beings,’ Lin et al. found that an individual’s investment decision-making behavior is not
only significantly and positively influenced by social capital but also that risk preference
contributes a lot to moderating influences of social capital on investment behavior [12].
This highlights the critical role of risk preference in the decision-making of farm households.
Consequently, risk preference represents a pivotal element that must be considered when
rice farmers want to adopt new technologies. To better learn about the impacts of risk pref-
erence on the decision-making of farmers regarding the adoption of technology, Patil et al.
conducted an experimental investigation in eastern India on rice farmers’ risk preferences.
They found that factors including the scale of operation, age, education, and social class
substantially affect farmers’ risk preference decisions and that risk aversion negatively
affects farmers’ willingness to adopt innovative technologies [13]. This finding verifies Mao
et al.’s research results regarding Chinese rice farmers, namely that loss aversion and risk
aversion can damage farmers’ adoption of low-carbon agricultural technologies [14]. Given
the considerable influence that risk preference exerts on farmers’ technology adoption
Sustainability 2024, 16, 8879 3 of 20

behavior, Ghadim et al. conducted further research into how to effectively modify farmers’
risk preferences by enhancing their information acquisition abilities and implementing risk
management measures. This enabled them to alter their decision-making behaviors and
boost the uptake of new technologies [15].
Nonetheless, most studies analyze the influences of policy interventions and market
mechanisms on the behaviors of farmers at the macro level while neglecting to analyze
the psychological, cognitive structure, and behavioral decision-making paths of individual
farmers, particularly the influence of psychological factors, including risk preferences on
technology adoption behavior from the micro perspective. Meanwhile, this kind of detailed
examination at the micro level is crucial for a thorough understanding of the underlying
motivations driving technology adoption among farm households. On the other hand,
social capital as a critical variable influencing farmers’ behavioral decision-making has
been widely acknowledged. However, existing studies are frequently limited to a single
dimension or type of social capital analysis and need a multidimensional, multilevel,
comprehensive framework to thoroughly investigate the intricate mechanism by which
social capital exerts an impact on the technology adoption process among farmers. This
one-sidedness limits our deep comprehension of the full influences of social capital and
does harm to developing more precise and effective policy interventions.
In view of the above, it is essential to focus on China’s professional rice farmers, a
group with distinctive characteristics in the process of agricultural modernization. They are
a group with high economic behavior and market participation, and they are susceptible
to changes in market demand. They are also more susceptible to market fluctuations
due to their vital resource and market integration capacity. In addition, vocational food
and agriculture usually have good record-keeping and management capabilities and can
provide detailed and accurate production and operation data. This lays a solid foundation
for detailed analyses and research and significantly improves the reliability and validity of
the research. This research aims to construct a multidimensional social capital index system
and measure farmers’ risk preferences by utilizing experimental economics under the
background of large-scale operations. This will enable us to quantify farmers’ psychological
characteristics more precisely and thoroughly evaluate the roles of social capital and risk
preferences in adopting green production technologies. The resulting data will provide a
more accurate and scientific basis for formulating relevant policies.

2. Theoretical Analysis
2.1. Social Capital and Rice Farmers’ Green Production Technology Adoption Behavior
Social capital, a crucial resource that farmers depend on for agricultural production
and daily
Sustainability 2024, 16, x FOR PEER REVIEW
life, is developed within individual or group social structures through dimensions
5 of 21
such as social networks, social participation, and social trust, and impacts farmers’ behav-
ioral decisions [16], as shown in Figure 1. Therefore, the hypothesis below is proposed:

Figure 1. Theoretical analysis framework.

Figure 1. Theoretical analysis framework.

2.2. Moderating Effects of Risk Preferences in Social Capital and Farmers’ Green Production
Technology Adoption
Risk preference is defined as an individual’s attitude and tendency toward risk when
uncertain. It reflects the decision maker’s tolerance for risk, cognitive depth, and personal
preferences, and it can vary with changes in uncertainty factors [26]. In promoting and
Sustainability 2024, 16, 8879 4 of 20

Hypothesis H1. Rice farmers accumulate social capital through access to information, trust, and
resources, enhancing their willingness and ability to adopt green production technologies.

As a core element of social capital, the height and breadth of the social network signifi-
cantly shape the channels and efficiency of farmers’ access to information and influence
their decision-making process of perceiving, evaluating, and ultimately adopting new
technologies [17]. The extent of a social network indicates the hierarchy and diversity of
its members regarding knowledge, experience, and social status. The increase in social
network height for rice farmers means they have access to more network members with pro-
fessional knowledge, rich practical experience, and higher social status. Moreover, Zhang
et al. highlighted that farmers with higher social network height have the advantages of
higher quality access to information, more pronounced technology demonstration effects,
and a stronger foundation of trust, which in turn promotes their adoption of more advanced
and authoritative production technologies [18]. On the contrary, the breadth of a social
network refers to the number, diversity, and distribution range of network members. Ma
et al. argued that a broader social network could offer rice farmers a more comprehensive
range of information and resources, reduce information asymmetry, and facilitate resource
sharing [19]. Jia et al. argued that farmers’ extensive connectivity and high interactiv-
ity through social networks enable the rapid and accurate dissemination of information
concerning new technologies, resulting in more informed decision-making [20].

Hypothesis H1a. The social networks of rice farmers facilitate access to more sophisticated
technological information and resources, which expedites the dissemination and implementation of
environmentally sustainable production technologies.

Social trust is a stable relationship of trust between individuals or groups based

on mutual understanding, shared values, and historical interaction experiences. Being
the indispensable cornerstone of social capital, social trust is conducive to driving social
cooperation and improving the overall efficiency of society. As Bonan et al. put forward,
farmers tend to be conservative due to unknown risks and uncertainties in the face of new
technologies. However, they gradually build up their confidence in the new technology
by observing and imitating the behavior of the object of their trust. This process facilitates
the social diffusion of the technology and enhances farmers’ identification with the new
technology [21]. Tao et al. demonstrated that social trust decreased the perceived risk
and uncertainty in farmers’ adoption of green production technologies, which acted as a
risk-buffering mechanism and enabled farmers to be more willing and to actively adopt
new technologies [22]. Social trust is key to driving rural areas to widely adopt green
production technologies.

Hypothesis H1b. Social trust among rice farmers promotes mutual assistance and cooper-
ation in adopting new technologies, which is conducive to the widespread adoption of green
production technologies.

Social participation is the intensity of social capital. Luo pointed out in his research that
members of society actively participate in a series of social activities, including production,
consumption, politics, social interaction, etc., and this behavior has a clear purpose and
organization, as well as promotes the development process of the society at a deeper
level [23]. For rice farmers, engaging with and joining rice professional cooperatives via
the Internet has become a crucial means of integrating into society, accessing information,
and exchanging experiences. Under the wave of digital economy, the popularity of Internet
technology provides unprecedented opportunities for rice farmers. A study by Zhu et al.
pointed out that the online platform constructed a free and diversified public space for
farmers, allowing them to interact face-to-face with agricultural experts and enhancing
their technical expertise and operational skills [24]. Weng et al. believe that cooperatives
act as offline communication platforms for rice farmers; rice farmers were able to access
Sustainability 2024, 16, 8879 5 of 20

high-quality and cost-effective agricultural inputs, professional technical guidance, and

services, all of which assisted in expanding their understanding of green production
technologies, which in turn substantially influenced their green production technology
adoption behavior [25].

Hypothesis H1c. Social engagement positively influences rice farmers’ adoption of green produc-
tion technology.

2.2. Moderating Effects of Risk Preferences in Social Capital and Farmers’ Green Production
Technology Adoption
Risk preference is defined as an individual’s attitude and tendency toward risk when
uncertain. It reflects the decision maker’s tolerance for risk, cognitive depth, and personal
preferences, and it can vary with changes in uncertainty factors [26]. In promoting and
applying green production technologies, farmers, as ‘rational economic beings’, primarily
consider profit maximization in their production decisions [27]. Rice farmers who take
risks use social network data, technical training, and new technology quickly, hoping for a
significant return. Risk-averse farmers are more circumspect. Suri’s study demonstrates
that farmers consider stability, maturity, and potential risks and require more informa-
tion [28]. They often hesitate between new and old technologies. Risk preferences affect
the impacts exerted by social capital on the extent to which rice farmers adopt green
production technologies.
The research thus makes the hypothesis as follows:

Hypothesis H2. Risk preference moderates the effect of social capital on rice farmers’ green
production technology adoption.

Hypothesis H2a. Risk preference plays a moderating role in the influence of social trust, participa-
tion, and networks on the adoption behavior of rice farmers’ green agricultural production technology.

3. Materials and Methods

3.1. Data
3.1.1. Study Area
As a significant rice-growing area in China, Jiangxi Province not only has a long
history of rice cultivation but is also known for its extensive planting area, high yields, and
high-quality rice. Since different Jiangxi Province regions have differences in climate and
soil, a diverse demand exists for rice technology. This makes Jiangxi Province an ideal place
to study the adaptability of different technologies and the match between farmers’ needs
and technology extension strategies. The government of Jiangxi Province has provided
robust support for agricultural technology extension and has actively worked to create
a policy environment conducive to this extension. This support has had a significant
demonstration effect and a broad influence on disseminating agricultural technology.
Therefore, selecting Jiangxi Province’s rice farmers as the research sample helps uncover
the underlying motivations behind farmers’ behavior and provides valuable insights for
enhancing the production efficiency of farmers and optimizing the promotion strategy of
green production technology. The research area is shown in Figure 2.

3.1.2. Data Sources

Selecting rice farmers in Jiangxi Province as the research sample not only helps uncover
the underlying motivations behind farmers’ behavior but also provides valuable insights for
optimizing the promotion strategy of green production technology and enhancing farmers’
production efficiency. A pre-survey was conducted in June 2023 prior to the formal launch
of the questionnaire survey. The research team, in conjunction with the subject team of the
School of Economics and Management of South China Agricultural University, recruited a
 Sustainability 2024, 16, 8879 6 of 20

total of 15 current master’s and doctoral students to research Guangdong Province. We

employed a combination of stratified sampling and typical sampling methods to select rep-
resentative regions in northern, western, and southern Guangdong. Specifically, we chose
Sustainability 2024, 16, x FOR PEER REVIEW
four cities—Shaoguan, Qingyuan, Jiangkoumen, and Yangjiang—to conduct a one-week 6 of 21
random sampling pre-survey. Based on the results of the 70 pre-survey questionnaires
collected, the questionnaires were further revised and enhanced using scientific methods,
and the content of the official questionnaires was finalized.

Study area.
[Link]
Figure area.

Formal research was conducted between July and September 2023 by a panel of 32 MSc
3.1.2.
and Data Sources The stratified and typical sampling methods were both employed
PhD students.
Selecting
to choose rice farmers
77 townships in Jiangxi
within Province
17 counties acrossas10the research
district sample
cities not Province,
in Jiangxi only helps un-
including Fuzhou, Xinyu, Ganzhou, Nanchang, Ji’an, Yichun,
cover the underlying motivations behind farmers’ behavior but also providesJiujiang, Shangrao, Yingtan,
valuable
and Jingdezhen. After coordinating with local township government
insights for optimizing the promotion strategy of green production technology and en-staff, we randomly
selectedfarmers’
hancing rice farmers with planting
production sizes exceeding
efficiency. A pre-survey 2 hectares from the provided
was conducted roster.
in June 2023 prior to
To ensure the objectivity and fairness of the survey, we obtained only the names and
the formal launch of the questionnaire survey. The research team, in conjunction with the
home addresses of the farmers during the questioning period to avoid the interference of
subject team of the During
other information. Schooltheof survey,
Economics and Management
the investigator conducted of interviews
South China Agricultural
with each
University,
sample farmer for over one hour to gather information on six key aspects, includingto
recruited a total of 15 current master’s and doctoral students research
farm
Guangdong
endowmentProvince.
and income,Werice
employed
businessacharacteristics,
combination of stratifiedagricultural
household sampling and typical sam-
machinery
pling methods
services, to select representative
rice production decision-makingregions
behavior inand
northern, western,
sustainability, riskand southern
perception andGuang-
order production,
dong. Specifically, andwegreen production
chose four characteristics
cities—Shaoguan, and non-cognitive
Qingyuan,abilities. A total and
Jiangkoumen,
of 661 questionnaires
Yangjiang—to conduct were obtained for
a one-week this research.
random sampling pre-survey. Based on the results of
the 70With rice farmers’
pre-survey adoption behavior
questionnaires collected,of the
green production technology
questionnaires were furtheras therevised
core and
research theme, the collected sample data were meticulously screened and compiled. Firstly,
enhanced using scientific methods, and the content of the official questionnaires was fi-
we excluded 60 sample questionnaires that were not filled out completely or were not filled
nalized.
out about privacy issues. Second, given that the head of the rice farmer family can more
Formal research
comprehensively learnwas
aboutconducted between
the inputs and outputs July and September
of household 2023 production,
agricultural by a panel of 32
MSc and PhD students. The stratified and typical sampling methods
we opted to use the head of the household as the study population to achieve were bothmore
employed
toaccurate
choose and77 townships
scientificallywithin 17 counties
valid empirical across
results. 10 on
Based district cities
this, we in Jiangxi
deleted Province,
93 samples of in-
questionnaires
cluding Fuzhou, completed by non-heads
Xinyu, Ganzhou, of households.
Nanchang, Ji’an, Ultimately, we obtained
Yichun, Jiujiang, 508 valid
Shangrao, Yingtan,
questionnaires,
and [Link]
Afterprovided a solid foundation
coordinating with local for the subsequent
township researchstaff,
government and analysis,
we randomly
as shown in Table 1.
selected rice farmers with planting sizes exceeding 2 hectares from the provided roster.
To ensure the objectivity and fairness of the survey, we obtained only the names and home
addresses of the farmers during the questioning period to avoid the interference of other
information. During the survey, the investigator conducted interviews with each sample
farmer for over one hour to gather information on six key aspects, including farm endow-
ment and income, rice business characteristics, household agricultural machinery ser-
vices, rice production decision-making behavior and sustainability, risk perception and
order production, and green production characteristics and non-cognitive abilities. A total
Sustainability 2024, 16, 8879 7 of 20

Table 1. Data sources.

City Valid Samples City Valid Samples

Fu zhou 93 Jiu jiang 58
Gan Zhou 79 Nan chang 48
Ji an 37 Shang rao 32
Jing dezhen 23 Xin yu 39
Ying tan 23 Yi chun 77

3.1.3. Descriptive Statistics

In this study, the distinct dimensions of the sample structure were analyzed in depth
through descriptive statistical analysis, and the relevant results are indicated in Figure 3.
The analyzed data demonstrated that the intensity of green production technologies adop-
tion illustrated an apparent non-uniform distribution among farmers. There is the highest
proportion of farmers having adopted just one technology, at 51.18%, highlighting the
‘threshold’ phenomenon in technology adoption [29], and the ratio of farmers having
adopted four technologies merely accounts for 2.36%, which further highlights that technol-
ogy adoption depth in various farmer groups has great differences. This study was based
on household heads, and the proportion of male heads of households was as high as 98 per-
cent. In addition, the formation of this phenomenon is not accidental; it represents a finding
of intertwined influences of several factors. Probing from the ideological and cultural
level, Confucianism profoundly influences Chinese society, underscoring traditional family
structures and gender roles. In this cultural context, men are typically regarded as the
family’s economic pillars and decision-making core. As a result, they have more of a chance
to assume the role of head of household within the family. Regarding age composition,
over half of the respondents were 50 or older, indicating the aging trend within China’s
agricultural labor force. Regarding educational attainment, 57.50 percent of the respondents
had a junior high school education or below. This proportion suggests that the widespread
adoption of green production technologies cannot be separated from the overall education
of farmers. Regarding health status, this sample of farmers were generally in good physical
condition, providing a solid foundation for successful technology adoption. The percentage
of farmers participating in professional rice cooperatives reached 42.72 percent, which
implies that cooperatives play an active role in technology promotion.

3.2. Models
The adoption behavior of green production technologies plays an essential role in envi-
ronmental protection and sustainable agricultural development, and the adoption behavior
of rice farmers towards green production technologies is frequently not a simple binary
choice but a multilevel decision-making process. Due to that, the green technology adop-
tion level serves as the dependent variable; its values are categorized into different ordered
groups in accordance with how many green production technology adoption behaviors
are shown by the rice farmers. This approach aligns with the continuity and hierarchy
observed in rice farmers’ technology adoption behaviors in agricultural production.
The ordered logit regression model is based on the assumption of one latent variable.
Let this latent variable be Y*, demonstrating the intensity of rice farmers’ adoption behavior
towards green production technologies. Y* indicates a linear function of social capital X
and other control variables (e.g., age, gender, literacy, etc., denoted here as Z), i.e.,:

Y ∗ = β 0 + β 1 X1 + β 2 X2 + · · · + β i Z + α (1)

where β0 demonstrates the intercept term, β1 , β2 , . . . , βi indicate the coefficient vectors

corresponding to social capital X1 , social network X2 , social participation X3 , social trust
X4 , and control variable Z, which are the direction and magnitude of the impact that
independent variables have on rice farmers’ technology adoption behavior, and α illustrates
the error term, assuming to abide by the standard normal distribution.
Sustainability 2024, 16,
Sustainability 2024, 16, 8879
x FOR PEER REVIEW 88of 21
of 20

Figure 3. Descriptive statistics.

Figure 3. Descriptive statistics.
3.2. Models
Since Y* serves as a latent variable, we cannot directly observe it. Nonetheless, we can
Thethe
classify adoption
intensitybehavior of green
of the green production
production technologies
technology adoptionplays an essential
behavior role in
of rice farmers
environmental protection and sustainabl e agricultural development, and the adoption
into five ordered categories based on the value of Y* (Y = 1, 2, . . . , 5). This is achieved
behavior of rice farmers towards green production technologies is frequently not a sim-
ple binary choice but a multilevel decision-making process. Due to that, the green tech-
nology adoption level serves as the dependent variable; its values are categorized into
Sustainability 2024, 16, 8879 9 of 20

through a series of thresholds µ1 , µ2 , µ3 , µ4 where µ1 < µ2 < µ3 < µ4 , specifically if the

following occurs:
Y* ≤ µ1 , then Y = 1 (not adopted at all)
µ1 < Y* ≤ µ2 , then Y = 2 (adopt item 1)
µ2 < Y* ≤ µ3 , then Y = 3 (adopt item 2)
µ3 < Y* ≤ µ4 , then Y = 4 (adopt item 3)
Y* > µ4 , then Y = 5 (adopt item 4)
Specifically, the probability that a rice farmer is on Y = j (j = 1, 2, . . . , 5) can be expressed
as follows:
P(Y = j | X, Z) = P(µ j−1 < Y∗ ≤ µj | X, Z) (2)
This can be further translated as follows:
exp(−µ j−1 + β 0 + β 1 X1 + β 2 X2 + · · · β i Z ) exp(−µ j + β 0 + β 1 X1 + β 2 X2 + · · · β i Z )
P(Y = j | X, Z ) = 4
− (3)
1 + ∑k=1 exp(−µk + β 0 + β 1 X1 + β 2 X2 + · · · β i Z ) 1 + ∑4k=1 exp(−µk + β 0 + β 1 X1 + β 2 X2 + · · · β i Z )

When j = 1, the above equation simplifies to the following:

1
P(Y = 1 | X, Z ) = (4)
1 + ∑4k=1 exp(−µk + β 1 X1 + · · · β i Z )

3.3. Variables
3.3.1. Explained Variables
The explained variable was selected as the adoption behavior of rice farmers’ green
production technology. As proposed by Cui et al. [30], four of the eight green production
technology types listed in the questionnaire—namely soil formulation and fertilizer ap-
plication technology, biopesticide technology, efficient mechanical application technology,
and collaborative promotion and recommendation technology—were chosen as the focus
of the study. First, a number of green production technologies adopted by the sample
farmers were summarized. If rice farmers did not adopt all four technologies, they were
assigned the value of ‘1’, one the value of ‘2’, two the value of ‘3’, and three of these
technologies a value of ‘1’, adopting three items was assigned as ‘4’, and adopting four
items was assigned as ‘5’. For testing the scientific validity and robustness of the sample
data, ‘whether rice farmers adopt green production technology behavior’ was selected as a
replacement variable, with ‘0’ assigned to the non-adoption of green production technology
and ‘1’ assigned to the adoption of green production technology.

3.3.2. Explanatory Variables

Based on existing research, social capital was chosen as the core independent variable.
Its multidimensional composition effectively captures rice farmers’ resource acquisition
ability, participation level, and trust base within their social networks. This approach offers
a comprehensive explanatory perspective for understanding the adoption behavior of green
production technologies. To be specific, Zhou et al. demonstrated that farmers rapidly
disseminated and exchanged agricultural information and technical knowledge through the
dual mechanism of strong and weak connections in social networks, representing a dynamic
process that significantly accelerated the rate of new technology adoption [31]. Ahaya et al.
indicated that farmers who were members of social cooperatives demonstrated a greater
propensity to accept and implement new technologies, thereby underscoring the beneficial
influence of organizational participation on technology adoption [32]. The digital economy
has witnessed notable surges in farmers’ utilization of digital technology, facilitating their
integration into social interactions and access to information resources. This has paved
the way for a novel avenue of technology adoption. Granovetter’s theoretical contribution
emphasizes the pivotal role of interpersonal trust in bolstering the sense of identity among
farmers and fostering their inclination to disseminate information. Establishing such a
trusting ambiance emerges as a pivotal factor in advancing the potential for technology
Sustainability 2024, 16, 8879 10 of 20

adoption [33]. In view of the abstract nature of the social network, trust, and participation
concepts, which can be hard to observe directly, this paper uses indirect indicators to
measure the two observed variables of ‘Number of relatives’ households with whom there
is regular interaction’ and ‘Total amount spent on interpersonal exchanges in 2022’, which
is selected to quantify the breadth and height of the social network [34,35]. ‘How frequently
does your search for desired agricultural information online?’ and ‘Are you a member of a
specialized rice production cooperative?’ were selected to measure social participation [36].
The social trust dimension was measured by ‘Are you frequently entrusted with challenging
tasks by those around you?’ and ‘Do you enjoy interacting with the elderly and children in
the village?’, which are indirectly reflected by two observed variables [37]. Social capital
and all dimensions are objectively assigned by the entropy value method, which determines
the weights through the principle of information entropy to objectively evaluate farmers’
social capital [38].

3.3.3. Moderating Variable

As proposed by Tang et al. [39], risk preference is chosen as a moderating variable.
Assign values to the sample farmers by choosing the experimental measure: If you have
CNY 100,000: (1) At fixed interest, CNY 2000 income per year; (2) Purchase low-risk financial
products yielding CNY 6000 annually, which may not be as profitable as a fixed-interest
income or could result in a loss of CNY 2000; (3) Purchase high-risk financial products,
which may be as high as CNY 20,000 income per year, but perhaps a loss of CNY 5000.
Risk-averse people are assigned a value of ‘1’, risk-neutral people ’2’, and risk lovers ’3’.

3.3.4. Control Variables

During the design process, to comprehensively and accurately analyze factors that
affect green production technologies adoption by rice farmers, nine control variables
spanning three dimensions—individual characteristics of the household head, agricultural
characteristics and household characteristics, and agricultural operations—were selected
in accordance with the farmers’ behaviors theory and through referring to Wang et al. [40]
and Qu et al. [41]. Four variables in total were chosen for the individual characteristics of
household heads, including age, gender, education, and education level; two variables were
selected for household characteristics, including labor force employment in agricultural
production and how many laborers are in your household at the beginning of 2022; and
three variables were selected for agricultural operation characteristics, including the scale
of operation, the user-friendliness of agricultural machinery, and ease of irrigation. As
shown in Table 2.

Table 2. Variable descriptions and descriptive statistics.

Variables Variable Name Variable Description and Assignment (Unit) Average Value Standard Deviation
Explained Variables
Number of green production technologies adopted
Adoption intensity 2.010 0.902
Green Production (items): 0 = 1; 1 = 2; 2 = 3; 3 = 4; 4 = 5.
Technologies Whether to adopt green production technologies?
Adoption behavior 0.715 0.452
No = 0; Yes = 1.
Explanatory Variables
Social Capital Entropy value method results 1.421 0.476
Total amount spent on interpersonal exchanges in
High Level of 2022: CNY 2500 or less = 1; CNY 2501–5000 = 2; CNY 3.874 1.825
Social Network 5001–7500 = 3; CNY 7501–10,000 = 4; CNY
10,001–12,500 = 5; Over CNY 12,501 = 6.
Social Network
Number of relatives’ households with whom there is
Breadth of Social Network regular interaction: 9 households or less = 1; 2.211 0.677
10–29 households = 2; 30 households or more = 3.
Sustainability 2024, 16, 8879 11 of 20

Table 2. Cont.

Variables Variable Name Variable Description and Assignment (Unit) Average Value Standard Deviation
How frequently does your family search for desired
agricultural information online? Completely
Network Participation 2.732 1.489
disagree = 1; Somewhat disagree = 2; Neutral = 3;
Social Participation Somewhat agree = 4; Completely agree.

Participation of Cooperatives Are you a member of a specialized rice production 0.427 0.495
cooperative? No = 0; Yes = 1.
Are you frequently entrusted with challenging tasks
by those around you? Completely disagree = 1;
Trust in Family and Friends 3.415 1.186
Somewhat disagree = 2; Neutral = 3; Somewhat
agree = 4; Completely agree = 5.
Social Trust
Do you enjoy interacting with elderly and children
in the village? Completely disagree = 1; Somewhat
Trust in the Neighborhood 4.203 0.948
disagree = 2; Neutral = 3; Somewhat agree = 4;
Completely agree = 5.
Moderating Variable
Measured through choice experiment:
Risk Preference 1.673 0.789
Risk Averse = 1; Risk Neutral = 2; Risk Lover = 3.
Control Variable
Gender Gender: Female = 1; Male = 2. 1.980 0.139
Age Age: head of household’s age in years. 52.179 7.385
Education: Not Attended School = 1; Elementary
Individual Characteristics Education School = 2; Junior High School = 3; High School = 4; 3.299 0.953
College Degree or Above = 5.
Health state: Very poor = 1; Poor = 2; Average = 3;
Health 4.274 0.832
Good = 4; Excellent = 5.
Labor Employment Whether to employ labor or not? No = 0; Yes = 1. 0.906 0.293
Household Number of household laborers: actual number of
Characteristics Number of Household laborers, excluding students, enlisted individuals, 2.833 1.177
Laborers
and full-time homemakers.

Scale of Operation How much land is there? 28.788 54.628

Actual rice paddy area in hectares.
Is agricultural machinery easy to use?
Agricultural Operation Convenience of Agricultural
Unable to Use = 1; Inconvenient = 2; Average = 3; 4.041 0.908
Characteristics Machinery Use
Somewhat Convenient = 4; Very Convenient = 5.
Is agricultural irrigation equipment convenient?
Convenience of Irrigation Very Poor = 1; Poor = 2; Average = 3; Good = 4; 3.083 1.096
Very Good = 5.

4. Regression Results and Analysis

4.1. Impact of Social Capital on Rice Farmers’ Green Production Technology Adoption Behavior
Before baseline regression analyses, covariance diagnostics were initially conducted to
verify both the validity and accuracy of model estimates. Diagnostic findings revealed that
in each of the independent variables, the variance inflation factor (VIF) ranged between
1.020 and 1.282 with a tolerance close to 1, indicating no significant issues with covariance
among the independent variables.
The ordered logit model was selected as the primary effect regression model, and the
robustness test used the ordered probit regression model. These approaches were adopted
to explore the influences of social capital on green production technologies adopted by rice
farmers from different perspectives to enhance reliability.
Table 3 indicates in detail the results of empirical calculations on the influences of
social capital on the adoption behavior of green production technologies by rice farmers.
The findings show that social capital exhibits a significant positive effect at the 0.001 level
in all the models, consistent with our Hypothesis H1. As an informal institution, social
capital substantially enhances rice farmers’ decision-making propensity to adopt green
production technologies by providing information, resources, and support.
Sustainability 2024, 16, 8879 12 of 20

Table 3. Results of empirical analysis.

Model 1 Model 2 Model 3 Model 4

Item
(Ordered Logit) (Ordered Logit) (Ordered Probit) (Ordered Probit)
1.240 *** 0.724 ***
Social Capital
(5.894) (6.073)
0.159 * 0.092 *
Social Network
(2.371) (2.391)
0.225 * 0.127 *
Social Participation
(2.315) (2.277)
0.942 *** 0.554 ***
Social Trust
(4.976) (5.157)
−0.036 −0.063 −0.015 −0.031
Gender
(−0.057) (−0.100) (−0.040) (−0.085)
−0.004 −0.003 −0.001 −0.002
Age
(−0.277) (−0.260) (−0.194) (−0.224)
0.088 0.080 0.060 0.056
Education
(0.894) (0.804) (1.058) (0.986)
0.141 0.132 0.079 0.075
Health
(1.300) (1.216) (1.269) (1.190)
1.199 *** 1.226 *** 0.639 *** 0.647 ***
Labor Employment
(3.800) (3.879) (3.497) (3.537)
−0.062 −0.069 −0.047 −0.051
Number of Household Laborers
(−0.798) (−0.891) (−1.054) (−1.139)
0.006 *** 0.007 *** 0.004 *** 0.004 ***
Scale of Operation
(3.741) (3.849) (3.687) (3.761)
Convenience of Agricultural 0.234 * 0.238 * 0.132 * 0.134 *
Machinery Use (2.345) (2.389) (2.299) (2.337)
0.031 0.028 0.017 0.016
Convenience of Irrigation
(0.388) (0.344) (0.368) (0.351)
x2 (10) = 117.566, x2 (12) = 119.315, x2 (10) = 117.223, x2 (12) = 118.819,
Likelihood Ratio Test
p = 0.000 p = 0.000 p = 0.000 p = 0.000
McFadden R2 0.097 0.098 0.096 0.098
Note: * p < 0.05, *** p < 0.001, and the values in brackets are standard errors.

In Model 2, the regression coefficient value of a social network is 0.159 and demon-
strates significance at the 0.05 level (z = 2.371, p = 0.018 < 0.05). This indicates that social
networks can greatly and positively affect adoption behavior by rice farmers on green
production technologies, i.e., the technical information, experience sharing, as well as
demonstration effect acquired by rice farmers through the social network positively drive
green production technologies adoption, thus validating Hypothesis H1a. The regression
coefficient value of social trust was 0.225. It revealed significance at the 0.05 level (z = 2.315,
p = 0.021 < 0.05), which implies that social trust likewise exerts a significant favorable
influence on rice farmers’ green production technology adoption behavior, validating
Hypothesis H1b. Nonetheless, its direct effect is relatively constrained, which may be
attributed to the complexity of the trust mechanism. Trust may affect technology adoption
indirectly, including influencing information access and willingness to cooperate, rather
than directly determining technology adoption behavior. The regression coefficient value
of social participation was 0.942 and illustrated significance at the 0.01 level (z = 4.976,
p = 0.000 < 0.01). This indicates that social participation has a significant positive effect on
rice farmers’ adoption of green production technology. Rice farmers who actively engage
in social activities have a greater likelihood of understanding new technologies and becom-
ing better aware and accepting such technologies through communication and learning,
thus promoting the widespread adoption of green production technologies, validating
Hypothesis H1c.
Regarding the control variables, household employed labor and scale of operation
illustrated a significant positive impact relationship at the 0.001 level, which implied
that rice farmers with better resource endowments are more predisposed to adopt new
Sustainability 2024, 16, 8879 13 of 20

technologies. The ease of use of agricultural machinery substantially and positively affects
rice farmers’ adoption behavior of green production technologies at the 0.05 level, implying
that the increase in agricultural mechanization can drive green production technologies to
be more widely adopted and applied.

4.2. Moderating Effects of Risk Preferences

Models 5–8 determine if risk preference moderates the influence of social capital on
the uptake of rice farmers’ green technologies. This can be achieved through the method
of stepwise regression. Since there are certain space limitations, no regression analysis is
offered for each step. Therefore, the results are shown.
According to Table 4 in Model 5, the interaction term of risk preference and social capital
has a substantial adverse effect on the technology adoption behavior of rice farmers (t = −2.397,
p = 0.017 < 0.05). This result suggests that risk preference negatively moderates the effect of
social capital on technology adoption behavior. Specifically, the positive impact of social capital
may be more pronounced among risk-averse rice farmers, as risk-averse people are more
inclined to rely on information and support within the social network, which can help mitigate
the uncertainty and risk associated with technology adoption, assuming that H2 holds.

Table 4. Results of moderation effect analysis.

Item Model 6 Model 7 Model 8 Model 9

Constant 1.050 1.013 0.875 1.014

(1.598) (1.523) (1.302) (1.488)
0.497 ***
Social Capital (5.896)
0.050
Risk Preference (1.069)
Social Capital * Risk Preference
−0.233 *
(−2.397)
0.060 *
Social Network (2.123)
0.064
Risk Preference (1.329)
Social Network * Risk Preference
−0.127 ***
(−3.526)
0.409 ***
Social Participation (5.279)
0.044
Risk Preference (0.932)
Social Participation * Risk Preference
−0.138
(−1.474)
0.096 *
Social Trust (2.317)
0.056
Risk Preference (1.161)
0.046
Social Trust * Risk Preference (0.865)
Control Variable The variable has been controlled
Sample Size 508 508 508 508
R2 0.219 0.181 0.201 0.162
Adjusted R2 0.200 0.161 0.182 0.142
F-value F(12,495) = 11.539, F(12,495) = 9.126, F(12,495) = 10.404, F(12,495) = 7.984,
p = 0.000 p = 0.000 p = 0.000 p = 0.000
∆R2 0.009 0.021 0.004 0.001
Note: * p < 0.05, *** p < 0.001, and the values in brackets are standard errors.

In Model 6, the interaction term between social network and risk preference was
found to be significantly negative (t = −3.526, p = 0.000 < 0.05), which indicates that risk
preference significantly and negatively moderates the impact of social networks on rice
farmers’ adoption of green production technologies. However, in Model 7 and Model
8, the interaction terms did not show significance, rendering Hypotheses H2a partially
valid. Possible reasons include that social participation and social trust, as critical elements
of social capital, directly affect new technology acceptance and adoption by rice farmers,
and the moderating effect of risk preference may be relatively weak and insufficient to
Sustainability 2024, 16, 8879 14 of 20

statistically substantially change the effects of social participation and social trust on
technology adoption behavior.

4.3. Robustness Tests

The findings were tested for robustness through three methods to guarantee their
generalizability and reliability (refer to Table 5). Firstly, the Tobit regression model (Model
9) was employed as a baseline test, and the results illustrated that social capital had a
significant positive effect (β = 0.511, p < 0.001) on the adoption behavior of rice farmers
towards green production technologies. This finding aligns with expectations and under-
scores the crucial role of social capital in advancing the adoption of green agricultural
practices. Secondly, further by replacing the dependent variable in a binary logit regres-
sion approach (Model 10), social capital still demonstrated a significant facilitating effect
(β = 0.924, p < 0.001), even though the specific values were changed, which demonstrated
that our findings remained consistent across distinct dependent variable settings, enhancing
the robustness of the findings.

Table 5. Robustness test results.

Model 10 Model 11 Model 12

Item
(Tobit Regression Analysis) (Replacing the Dependent Variable) (Binary Probit)
0.511 *** 0.924 *** 0.542 ***
Social Capital
(6.115) (3.352) (3.426)
Control Variable The variable has been controlled
0.277 −4.566 * −2.696 *
Intercept
(0.420) (−2.131) (−2.209)
McFadden R2 0.089 0.174 0.168
Likelihood Ratio Test x2 (10) = 118.645, p = 0.000 x2 (10) = 105.745, p = 0.000 x2 (10) = 102.181, p = 0.000
Note: * p < 0.05, *** p < 0.001, and the values in brackets are standard errors.

A binary probit model (Model 11) was adopted as another robustness test. These
findings reconfirmed that social capital contributed positively to the behavior of rice farmers
in adopting green production technologies (β = 0.542, p < 0.001), which not only validates
the accuracy of the previous two models but also enhances our understanding of how
social capital operates across different statistical models.
Whether under different model settings or after substituting the dependent variable
and controlling for potential confounders, the validation results consistently demonstrate
that social capital positively impacts the adoption of green production technologies. The
critical effects of social capital on rice farmers’ adoption of green production technologies
are again demonstrated.

4.4. Heterogeneity Analysis

According to generational theory, the development and utilization of social capital
in farm households are very different [42]. As an essential demographic variable, age is
closely related to farmers’ risk preference, information access, social network structure, and
technology acceptance. Consequently, age grouping can reveal the distinctions in green tech-
nology adoption behaviors under the influence of social capital among farmers of distinct
generations and provide a basis for developing generation-targeted agricultural policies.
With the development of the rural economy and the advancement of agricultural mod-
ernization, more and more professional rice farmers are attempting to combine agriculture
with tourism, e-commerce, and other emerging businesses to diversify their operations and
increase their incomes [43]. This is employed as a grouping variable to reveal the role of
the occupational structure of farming households in the relationship between social capital
and technology adoption and further enhance the understanding of farmers’ behavior.
The empirical results indicate (refer to Tables 6 and 7) that social capital contributes
much to promoting the adoption of green production technologies, and its effect remains
Sustainability 2024, 16, 8879 15 of 20

consistent across various age groups and occupational backgrounds, all of which are
positive and statistically significant. In particular, social capital plays a more significant
role in advancing green technology adoption among the new generation of rice farmers
and those in part-time farming groups, which may be related to the fact that they are more
receptive to new technologies and maintain a higher level of community participation.

Table 6. The result of age heterogeneity regression.

Model 12
Item New Generation Older Generation New Generation Older Generation
Social 0.541 *** 0.406 ***
Capital (3.617) (3.899)
0.441 ** 0.338 ***
Social Participation (3.043) (3.682)
Social 0.009 0.123 *
Trust (0.125) (2.544)
0.076 0.023
Social Network (1.426) (0.706)
Control Variable The variable has been controlled
Constant −0.604 0.490 −0.527 0.337
(−0.526) (0.744) (−0.452) (0.505)
Sample Size 172 336 172 336
R2 0.233 0.195 0.235 0.206
Adjusted R2 0.191 0.173 0.182 0.179
Note: * p < 0.05, ** p < 0.01, *** p < 0.001, and the values in brackets are standard errors.

Table 7. The result of employment heterogeneity regression.

Model 13
Item Pure Agricultural Type Part-time Agricultural Type Pure Agricultural Type Part-time Agricultural Type
Social 0.451 *** 0.703 ***
Capital (4.632) (4.201)
0.383 *** 0.556 ***
Social Participation (4.378) (3.484)
Social 0.082 0.084
Trust (1.700) (1.109)
0.030 0.090
Social Network (0.899) (1.763)
Control Variable The variable has been controlled
Constant 0.924 −2.082 0.927 −2.225
(1.229) (−1.180) (1.201) (−1.226)
Sample Size 361 147 361 147
R2 0.189 0.341 0.194 0.341
Adjusted R2 0.166 0.292 0.166 0.282
Note: *** p < 0.001, and the values in brackets are standard errors.

5. Discussion
Investigating the driving mechanisms of farmers’ adoption of green production tech-
nologies represents a crucial area of research in the agricultural low-carbon green transition.
While the existing literature identifies the potential impacts that social capital exerted on
farmers’ decision-making regarding technology adoption, more comprehensive empirical
analyses are needed, particularly concerning professional rice farmers in China. Accord-
ingly, this study seeks to contribute to this discourse by analyzing in-depth research data
from 508 households of professional rice grain farmers in China. The objective is to explore
social capital’s influence on rice farmers’ adoption of green production technologies while
examining the moderating role of risk preferences. A multidimensional social capital indi-
cator system was established and used with experimental economics to measure farmers’
risk preferences. Rigorous empirical analyses were conducted using ordered logit models
and moderated effects models. This approach contributes to the existing literature by
elucidating the intricate interrelationships between social capital, risk preferences, and
technology adoption among Chinese rice farmers.
The results of the study indicated that social capital contributed greatly to rice farmers’
adoption of green production technologies. In addition, it is noted that social participa-
Sustainability 2024, 16, 8879 16 of 20

tion, social networks, and social trust also influenced the adoption of technology, with
social participation having the most prominent effect, which agrees with the findings of
Katungi [44].
Heterogeneity analyses indicate that traditional social networks may have less im-
pact on technology adoption as occupational structures diversify and digital technologies
become more prevalent [45]. The study revealed that the social capital of the younger gen-
eration of rice farmers exerted a more pronounced influence on the adoption of green pro-
duction technologies than older generations. This is evidenced by the fact that the younger
generation was more inclined to acquire knowledge and adopt new technologies through
collective activities and modern information technology. Furthermore, the adoption of
technologies by part-time farmers was found to be more significantly affected by social
capital than that of purely agriculturalists. This is attributed to diversifying their economic
sources and, more significantly, risk tolerance. Notably, social participation demonstrated a
significant positive impact across the various heterogeneous grouping results. The find-
ings revealed that rice farmers with higher social participation preferred adopting green
production technologies. Generational differences and differences in livelihood strategies
influenced this. During this research, several Ganzhou Tenkai Rice Professional Coopera-
tive members were consulted. It was established that the Cooperative had established an
expert technical guidance team. This team guides members in disseminating and apply-
ing green production technologies. Furthermore, it was revealed that the Cooperative’s
platform provides members greater access to online agricultural information. This, in turn,
promoted the frequency of raised green production technologies in agriculture. This cannot
be separated from the fact that as professional rice cooperative members, rice farmers prefer
to engage in collective actions, such as attending technical training sessions together and
purchasing new equipment collectively. This facilitates the adoption of new technologies
by lowering the threshold for individual adoption and inspiring more farmers to adopt
new production technologies via demonstration effects shown by successful cases.
The results indicated that risk preference significantly moderated the connection between
social capital and green production technology adoption behavior. This finding aligns with
the conclusions that Yu et al. drew in their study on the impact of green production technology
adoption behavior among vegetable farmers [46]. Further analysis revealed that the social
network dimension is particularly salient. Even when rice farmers have access to information
about relevant technologies through their social networks, they may be reluctant to adopt
the technology due to excessive concerns about risk. This, in turn, affects the adoption of the
technology. This finding is consistent with the tenets of prospect theory. It was determined
that risk preferences did not substantially moderate the effects of social participation and
social trust on technology adoption. This may be attributed to the fact that in an environment
where social participation and social trust foster robust social capital, rice farmers are more
likely to share information, exchange experiences, and engage in collective actions. This
accelerates the diffusion of green production technologies and mitigates the risks and costs
associated with adopting new technologies for individual rice farmers. While risk preference,
as an individual-level psychological factor, affects the decision-making process of individuals
in the face of new technologies, this individual-level risk preference may be diluted by the
group effect, i.e., on the condition that rice farmers are in a highly participatory and trusting
social environment, they may be more influenced by collective decisions and actions rather
than individual risk preferences. While distinct rice farmers may have different levels of risk
preferences, these differences may be compensated for by intra-group communication and
mutual support in an environment of high social participation and trust.
The following points represent the marginal contributions of this paper: In terms
of research methodology, this study is distinguished from most previous studies in that
it employs a more comprehensive approach to understanding farmers’ decision-making
behaviors under different risk conditions. The design of specific choice experiments allows
for a more accurate representation of farmers’ decision-making processes, enhancing the
reliability and explanatory power of the results. Furthermore, combining the ordered
Sustainability 2024, 16, 8879 17 of 20

logit model and the ordered probit model provides a detailed analysis of the effects of
social capital and risk preference on adopting green production technologies. Regarding
sample selection, this study focuses on a distinctive group of professional rice farmers
in China with a higher degree of specialization and more intensive production activities.
This is a departure from the broadly selected sample of farmers or region-specific farmers
that have been the focus of previous studies. Atsriku [47] and Yengoh et al. [48] selected
a larger sample of farmers, providing a comprehensive sector overview. However, this
approach may not fully capture the specific challenges professional food and agriculture
farmers face in the context of increasing agricultural intensification. Accordingly, this study
aims to elucidate the role and dynamics of this particular group in the context of modern
agricultural development. This study found that social capital encourages rice farmers
to adopt green production technologies. This conclusion aligns with existing studies but
differs in dimensions and influence. The study by Katungi et al. showed that farmers’
social capital affects the likelihood of adopting green technologies [44]. This study further
refines the effects of different social capital dimensions on technology adoption and reveals
the complex role of risk preferences.
Despite this study’s many valuable insights, several limitations and areas for further
exploration remain. First, since this research is mainly based on exploring cross-sectional
data of vocational food and agriculture in Jiangxi Province, it cannot indicate the chrono-
logical relationship between social capital and technology adoption. Future studies may
consider methods, including longitudinal data or experimental design, to more precisely
reveal the causal relationship between the two. Second, although we revealed differences in
technology adoption among distinct groups in our subgroup analyses, the deeper reasons
behind these differences were insufficiently explored. Future research could explore the
social, economic, and psychological factors underlying these differences through in-depth
qualitative studies or more refined variable designs.

6. Conclusions and Suggestions

6.1. Conclusions
This study empirically analyzes the impacts that social capital exerted on the behavior of
rice farmers in Jiangxi Province, China, in adopting green production technology. Employing
an ordered logit model, we found that social capital contributed greatly and positively to
technology adoption, with social participation playing the most significant role, emphasizing
the crucial role of collective action and cooperative networks in advancing green practices.
Social trust and networks also positively influence technology adoption, but with a weaker
relative role. Risk preference was found to negatively moderate the effect of social networks,
reducing their influence on technology adoption. Further heterogeneity analysis revealed
that social capital positively and greatly affected the green production technology adoption
behavior of rice farmers within different grouped regression models; however, among the
dimensions, only social participation had a significant positive effect on rice farmers’ adoption
behavior at the 0.001 level in the grouped regression model. In contrast, social trust positively
affected the adoption behavior of the older rice farmers only at the 0.05 level, and the effects
of the other dimensions were insignificant, highlighting the need for differentiated policies.
Overall, this study provides empirical support for further optimizing the promotion strategies
of green production technologies for rice vocational food farmers and exhibits that social
capital and risk preference contribute a lot to green agriculture.

6.2. Suggestions
First, strengthen the organization and support of cooperatives and collective action. It
is recommended that the government provide policy and financial support to cooperatives,
integrate digital platforms with traditional extension channels, and establish an information-
sharing mechanism within cooperatives to ensure that farmers have timely access to new
technologies. The digital platform facilitates collective interaction through online and
offline organizations, disseminating training previews in advance and providing online
Sustainability 2024, 16, 8879 18 of 20

learning resources. In-field training is oriented towards practical operation, addressing the
diverse needs of farmers. Collective learning and feedback activities are organized for rice
farmers, fostering trust and enabling timely adjustments and optimizations in strategies,
thus ensuring the sustainable and stable development of cooperatives and collective action.
Second, the use of digital technologies in agriculture should be increased. The gov-
ernment should encourage the creation of a unified digital agriculture service platform,
explore the innovative organizational model of ‘Internet + cooperatives,’ and advance the
use of intelligent equipment and technical training through these digital platforms. This
will enhance farmers’ ability to access and effectively use green production technologies,
which improves their efficiency and uses digital tools to enhance networking and trust rela-
tionships in farmers, reduces the adverse impact of risk appetite on technology adoption,
and increases the green technologies’ adoption rate.
Third, it is recommended that the monitoring and evaluation of rice farmers’ liveli-
hoods be conducted to gain insights into changes in heterogeneous needs. Those responsi-
ble for formulating policy should make effective use of modern information technology
to gain insight into the diverse characteristics of rice farmers and their evolving trends
and comprehend the needs and perspectives of the various categories of rice farmers.
In particular, new-generation farmers require training in digital technology and support
for entrepreneurship. Part-time farmers can adopt green production technologies more
readily by providing flexible technological services and implementing flexible subsidy
and incentive policies. For those engaged in part-time farming, the provision of flexible
technical services and the implementation of flexible subsidies and incentives are more
likely to assist them in overcoming the barriers to adopting green production technologies.
Fourth, the interface between carbon emission reduction policies and green production
technologies must be strengthened. Through policy guidance and market mechanisms,
explore incorporating agricultural carbon emissions into the carbon trading market system
and incentivize farmers to adopt green production techniques to reduce carbon emissions
through financial subsidies and market mechanisms. It has also established a scientific
and reasonable accounting and trading mechanism for agricultural carbon emissions to
encourage farmers to adopt green production techniques to reduce energy consumption
and carbon emissions.

Author Contributions: B.L.: Conceptualization, Methodology, Software, Investigations, Data curation,

Formal analysis, and Writing—original draft. N.L.: Writing—review and editing, Supervision, and
Funding acquisition. C.L.: Supervision, Validation, Investigations, Writing—review and editing, and
Funding acquisition. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Natural Science Foundation of China (Project
No. 72264013), the National Natural Science Foundation of China (Project No. 71964016), and the
National Natural Science Foundation of China (Project No. 42467067).
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 privacy restrictions.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Saravanan, R.; Suchiradipta, B. Agricultural Innovation Systems: Fostering Convergence for Extension; MANAGE: Hyderabad, India, 2017.
2. Ekka, P.; Patra, S.; Upreti, M.; Kumar, G.; Kumar, A.; Saikia, P. Land Degradation and its impacts on Biodiversity and Ecosystem
services. In Land and Environmental Management through Forestry; Raj, A., Jhariya, M.K., Banerjee, A., Nema, S., Bargali, K., Eds.;
Wiley: Chichester, UK, 2023; pp. 77–101.
3. Zou, L.; Liu, Y.; Wang, Y.; Hu, X. Assessment and analysis of agricultural non-point source pollution loads in China: 1978–2017.
J. Environ. Manag. 2020, 263, 110400. [CrossRef] [PubMed]
Sustainability 2024, 16, 8879 19 of 20

4. Zhu, D.; Ye, L. Green resilience of Chinese agriculture: Level measurement and spatio-temporal evolution. Stat. Decis. 2024,
40, 118–123.
5. Du, M.; Feng, R.; Chen, Z. Blue sky defense in low-carbon pilot cities: A spatial spillover perspective of carbon emission efficiency.
Sci. Total Environ. 2022, 846, 157509. [CrossRef]
6. Wu, F.; He, J.; Cai, L.; Du, M.; Huang, M. Accurate multi-objective prediction of CO2 emission performance indexes and industrial
structure optimization using multihead attention-based convolutional neural network. J. Environ. Manag. 2023, 337, 117759.
[CrossRef]
7. He, P.; Zhang, J.; Li, W. The role of agricultural green production technologies in improving low-carbon efficiency in China:
Necessary but not effective. J. Environ. Manag. 2021, 293, 112837. [CrossRef]
8. Liu, Y.; Liu, Z.; Liu, J.; Qiu, L.; Wang, Y.; Fu, X. Research on the impact of members’ social capital within agricultural cooperatives
on their adoption of IPM in China. Int. J. Environ. Res. Public Health 2022, 19, 11538. [CrossRef]
9. Wen, S.; Chen, Z.; Xia, X.; Zhang, J.; Xie, S. Impact of digital agricultural technology promotion services on farmers’ adoption of
green production technologies. J. Arid. Land Resour. Environ. 2024, 38, 43–52.
10. Li, D.; Dang, H.; Yu, J. Can Participation in cooperatives promote the adoption of green production techniques by Chinese apple
growers: Counterfactual estimation based on propensity score matching. Front. Environ. Sci. 2022, 10, 1026273. [CrossRef]
11. Liao, L.; Du, M. How digital finance shapes residents’ health: Evidence from China. China Econ. Rev. 2024, 87, 102246. [CrossRef]
12. Lin, W.Y.; Hu, Y.R. Social Capital, Risk Preference, and Household Risk Financial Asset Allocation. Product. Res. 2024, 5, 141–146.
13. Patil, V.; Veettil, P.C. Experimental evidence of risk attitude of farmers from risk-preference elicitation in India. In Proceedings of
the International Association of Agricultural Economists (IAAE), Vancouver, BC, Canada, 28 July–2 August 2018.
14. Mao, H.; Quan, Y.-R.; Fu, Y. Risk preferences and the low-carbon agricultural technology adoption: Evidence from rice production
in China. J. Integr. Agric. 2023, 22, 2577–2590. [CrossRef]
15. Ghadim, A.K.A.; Pannell, D.J.; Burton, M.P. Risk, uncertainty, and learning in adoption of a crop innovation. Agric. Econ. 2005,
33, 1–9. [CrossRef]
16. Zhu, Q.Y.; Chen, Y.R.; Hu, W.Y.; Mei, Y.; Yuan, K.H. A study on the relationship between social capital, cultivated land value
cognition and farmers’ willingness to pay for cultivated land protection: An empirical analysis based on a moderated mediator
model. China Popul. Resour. Environ. 2019, 29, 120–131.
17. Zhang, W.E.; Luo, Y.; Zhao, M.J. Social Networks, Information Acquisition and Households’ Behavior of Agricultural Film
Recycling: A Sample of Farm Households in Yellow River Basin. J. Agro-For. Econ. Manag. 2022, 21, 40–48.
18. Zhang, T.C.; Yan, T.W.; Wang, Z. Social Networks, Income Uncertainty and the Self-Employed Women’s Adoption Behavior of
Conservation Tillage Technology. J. Agrotech. Econ. 2020, 8, 101–116.
19. Ma, X.D.; Shao, L.Q.; Huo, X. Is Pattern of Difference Sequence the Cause of Technology Lock-in in Farmers’ Production?—Based
on the Perspective of Technology Network Embedding. J. Huazhong Agric. Univ. (Soc. Sci. Ed.) 2018, 6, 20–28+151.
20. Jia, J.; Jiang, D.; Wei, L. Application of Agricultural Science and Technology Transformation; Science Press: Beijing, China, 2013; Volume 65.
21. Bonan, J.; Battiston, P.; Bleck, J.; LeMay-Boucher, P.; Pareglio, S.; Sarr, B.; Tavoni, M. Social Interaction and Technology Adoption:
Experimental Evidence from Improved Cookstoves in Mali. World Dev. 2020, 144, 105467. [CrossRef]
22. Tao, Y.; Qiu, X.Y.; Zhou, Y.X.; Hu, J.L. The Conflict between Risk Perception, Social Trust and Farmers’ Organic Fertilizer
Substitution Behavior. J. Agrotech. Econ. 2022, 5, 49–64.
23. Luo, B.L. Organizational Characteristics, Reputation Mechanism and Its Associated Game in the Condition of Village Environment.
Reform 2009, 2, 72–80.
24. Zhu, S.B.; Xiong, F.X.; Zhu, J. Impact of Internet Use on Rural Households’ Income: An Analysis of the Mediating Effect Based on
Social Capital. J. Agro-For. Econ. Manag. 2022, 21, 518–526.
25. Weng, Z.L.; Lan, D.W.; Tang, J.; Liao, Y.; College of Economics and Management/Research Institute of Rural Revitalization
Strategy; Jiangxi Agricultural University. Risk Preference, Cooperative Participation, and Adoption Behavior of Green Rice
Production Technology—Based on 520 Surveys of Rice Growers in Jiangxi Province. Agric. Econ. Manag. 2023, 5, 13–22.
26. Shou, Y.; Olney, J. Attitudes toward risk and uncertainty: The role of subjective knowledge and affect. J. Behav. Decis. Mak. 2021,
34, 393–404. [CrossRef]
27. Popkin, S. The rational peasant: The political economy of peasant society. Theory Soc. 1980, 9, 411–471. [CrossRef]
28. Suri, T. Selection and comparative advantage in technology adoption. Econometrica. 2011, 79, 159–209.
29. Zhang, W.; Wang, D.; Lu, Q. Impact of agricultural productive services on green production behavior of farmers. J. Arid Land
Resour. Environ. 2024, 38, 21–30.
30. Cui, S.; Li, Y.; Jiao, X.; Zhang, D. Hierarchical linkage between the basic characteristics of smallholders and technology awareness
determines Small-Holders’ willingness to adopt green production technology. Agriculture 2022, 12, 1275. [CrossRef]
31. Zhou, R.; YU, D.K.; TU, G.P. Farmers’ Knowledge Behaviors in the Agriculture Technology Diffusion Process of “Corporation &
Farmer” Mode. J. Anhui Agric. Sci. 2015, 43, 387–392.
32. Ahaya, I.; Zereyesus, Y.A.; Nakelse, T.; Haruna, B. Complementarity of technology adoption and social capital participation: The
case of systems of rice intensification in Ghana. J. Int. Dev. 2019, 31, 601–616. [CrossRef]
33. Granovetter, M.S. Economic action and social structure: The problem of embeddedness. Am. J. Sociol. 1985, 91, 481–510. [CrossRef]
Sustainability 2024, 16, 8879 20 of 20

34. Zhang, W.W.; Yang, C.; Yao, K.W.; Wu, Z.-X. Research on Livelihood Strategy Selection of Resettlement Based on DEMATEL-ISM
Model. China Rural Water Hydropower 2024, 1–15. Available online: [Link]
CAPJ&dbname=CAPJLAST&filename=ZNSD20240612007&uniplatform=OVERSEA&v=M1XYbiiBuh2gfWHHOn9JTaR4V3
pEWzLG_2pqr3R1PLF2NEeeptHakBHEsghrSMa_ (accessed on 9 October 2024).
35. Liu, Z.; Wei, Y.; Li, Q.; Lan, J. The mediating role of social capital in digital information technology poverty reduction: An
empirical study in urban and rural China. Land 2021, 10, 634. [CrossRef]
36. Wang, Y.Q.; Zhang, J.N.; Wang, H.Y. The Impact of Social Capital and Knowledge Stosk on The Adoption of Fertilizer and
Drug Reductions by Grain Growers Configuration Path Research. Chin. J. Agric. Resour. Reg. Plan. 2024, 1–12. Avail-
able online: [Link]
06&uniplatform=OVERSEA&v=vcbNuYayeRAs6ins5fVa1fjp1QXY5874hoSMgaX-kvnAvlnKEvxT3xziHd274ed4 (accessed on
9 October 2024).
37. Han, M.; Liu, R.; Ma, H.; Zhong, K.; Wang, J.; Xu, Y. The impact of social capital on farmers’ willingness to adopt new agricultural
technologies: Empirical evidence from China. Agriculture 2022, 12, 1368. [CrossRef]
38. Wan, J.; Deng, W.; Song, X.; Liu, Y.; Zhang, S.; Su, Y.; Lu, Y. Spatio-temporal impact of rural livelihood capital on labor migration
in Panxi, southwestern mountainous region of China. Chin. Geogr. Sci. 2018, 28, 153–166. [CrossRef]
39. Tang, W.S.; Weng, Z.L.; Yan, C.H. Information acquisition, risk preference and the socialized services of technology-intensive
farm machinery: A study based on rice scale management households in Jiangxi Province. J. China Agric. Univ. 2022, 27, 270–280.
40. Wang, L.; Hu, Y.; Kong, R. The Impact of Bancassurance Interaction on the Adoption Behaviour of Green Production Technology
in Family Farms: Evidence from China. Land 2023, 12, 941. [CrossRef]
41. Qu, X.; Zhou, W.; He, J.; Xu, D. Land Certification, Adjustment Experience, and Green Production Technology Acceptance of
Farmers: Evidence from Sichuan Province, China. Land 2023, 12, 848. [CrossRef]
42. Ren, Z.; Guo, Y. The effect of environmental regulation and social capital on farmers adoption behavior of low-carbon agricultural
technology. J. Nat. Resour. 2023, 38, 2872–2888. [CrossRef]
43. Luo, X.; Li, L.; Zhang, L.; Ning, C.; Liu, X. The Impact of Forestland Tenure Security on Rural Household Income: Analysis of
Mediating Effects Based on Labor Migration. Forests 2024, 15, 1336. [CrossRef]
44. Katungi, E.M. Social Capital and Technology Adoption on Small Farms: The Case of Banana Production Technology in Uganda.
Ph.D. Thesis, University of Pretoria, Pretoria, South Africa, 2007.
45. Skaalsveen, K.; Ingram, J.; Urquhart, J. The role of farmers’ social networks in the implementation of no-till farming practices.
Agric. Syst. 2020, 181, 102824. [CrossRef]
46. Yu, L.; Chen, C.; Niu, Z.; Gao, Y.; Yang, H.; Xue, Z. Risk aversion, cooperative membership and the adoption of green control
techniques: Evidence from China. J. Clean. Prod. 2021, 279, 123288. [CrossRef]
47. Atsriku, G.E. The Adoption of Agriculture Technology in Small-Scale Farming in the Adumasa Community in Ghana. Agricultural,
Food and Environmental Studies. Master’s Thesis, International Institute of Social Studies, Den Haag, The Netherlands, 2020.
48. Yengoh, G.; Frederick, A.; Svensson, M.G.E. Technology adoption in small-scale agriculture: The case of Cameroon and Ghana.
Sci. Technol. Innov. Stud. 2009, 5, 111–131.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.