The relationship between livelihood capital and herdsmen’s usage of digital technology: evidence from Inner Mongolia, China

Digital technology has become a driving force for the development of high-quality agriculture. Existing research mainly focuses on the macro-level factors driving the development of digital technology. In contrast, micro-level investigations considering herdsmen’s resource endowments when analyzing the relationship with digital technology remain relatively scarce. Based on survey data from 356 herdsmen in Inner Mongolia, China, this study uses the entropy weight method to calculate the livelihood capital of herdsmen and then uses IV-probit and moderating effect models to analyze the relationship between livelihood capital and herdsmen’s usage of digital technology. The results show that the livelihood capital positively impacts herdsmen’s use of digital technology at the 1% level of significance and this result is robust. Further analysis of the sub-dimensions reveals that natural, physical, human, and social capital significantly promote herdsmen’s digital technology use. In contrast, the impact of financial capital is not significant. Furthermore, technical training and network quality enhance livelihood capital’s impact on herdsmen’s use of digital technology. Heterogeneity analysis shows that the impact of livelihood capital on the use of digital technology is more pronounced among herdsmen who have a large grassland area, more education, and a greater scale of livestock. Based on our findings, policy recommendations are made for promoting the use of digital technology by enhancing herdsmen’s livelihood capital.

1 Introduction

Under the digital economy, digital technology is emerging as a driver of the modernization and transformation of agriculture and livestock (Wang and Qu, 2024). Developed countries have taken the lead in applying a wide range of digital technologies—such as information systems, big data, and the Internet of Things—to livestock production and management. These technologies have improved production efficiency, enhanced animal welfare, and strengthened resource utilization (Wolfert et al., 2017). In the EU and the US, more than two-thirds of large-scale farms have adopted digital management systems, with significant outcomes observed in livestock monitoring and precision farming. By contrast, China, which is the world’s largest developing country, is still in the early stages of using digital technology for livestock management. According to the “2022 National Evaluation Report on the Development Level of Digital Agriculture and Rural Development,” pastoral areas are significantly below the national average in terms of both digital infrastructure coverage and the extent of technological use. Moreover, more than 70% of herdsmen have yet to systematically integrate digital technologies into their livestock production. To accelerate the digitalization of agriculture, the “Digital Village Development Plan (2022–2025)” prioritizes improvements in digital infrastructure, smart agricultural innovation, and digital governance. The plan aims to accelerate the use of digital technology in rural areas, achieving substantial progress by 2025. However, in agricultural and livestock production, the use of digital technology by farmers and herders remains suboptimal. Currently, the promotion of smart agriculture and digital village initiatives faces multiple barriers, including high costs, mismatched demand, human capital shortage, and underdeveloped commercialization pathways (Huang, 2022). Such challenges limit the widespread adoption and effective use of digital technologies in the process of agricultural modernization. Therefore, to improve technological efficiency and promote broader application in pastoral areas, it is necessary to identify the factors influencing herdsmen’s use of digital technology as well as its underlying mechanisms.

In this context, livelihood capital plays a crucial role in explaining the production decisions and technology adoption behavior of herdsmen, and it is widely regarded as a key concept for analyzing their use of digital technologies (Jansen et al., 2006). Livelihood capital is a core component of the sustainable livelihood framework that refers to the resources and capabilities that herders possess and can mobilize in their production and daily lives (Scoones, 1998). It typically encompasses five dimensions: natural capital, physical capital, human capital, social capital, and financial capital. These types of capital not only determine the production conditions and development potential of herders, but they also substantially shape their cognition, acceptance, and use of new technologies.

Existing research on digital technologies has mainly focused on two areas. The first is the analysis of the factors influencing the use of digital technology by farmers and herdsmen. Studies have indicated that digital technologies have yet to be fully integrated into agricultural production. This has resulted in an imbalance between the supply of digital solutions and the actual needs of agriculture and rural development (Huang, 2024). In addition, talent shortages in the digital sector and inadequate investment in infrastructure pose significant challenges to the effective use of digital technologies for agricultural development (Huang, 2025). There are also imbalances in the application of digital technologies, with more attention paid to crop farming than to livestock production, hindering the role of digital technology in agricultural modernization (Li and Li, 2024). Beyond infrastructure and policy factors, individual cognitive and behavioral characteristics influence the use of digital technology. Studies have found that the perceived benefits of digital technologies among farmers strongly influence their willingness to adopt them, as anticipated improvements in productivity and economic returns increase adoption intention (Bolfe et al., 2020). Other studies suggest that expectations regarding technological performance, perceived complexity, and social influence drive adoption behavior. Furthermore, farmers’ understanding of the basic functionality of digital technologies influences adoption, with higher awareness associated with higher adoption and better utilization (Gabriel and Gandorfer, 2023). In terms of adoption prerequisites, perceived ease of use has been identified as a fundamental condition for adoption (Kabbiri et al., 2018). Furthermore, both perceived ease of use and perceived usefulness are recognized as cognitive mechanisms underpinning farmers’ adoption decisions (Verma and Sinha, 2018). Studies have also indicated that language barriers, poverty, and illiteracy limit smallholders’ access to digital technologies (Nmadu et al., 2013). Similarly, inadequate infrastructure, low literacy rates, insufficient information services, and limited technical capacity hinder adoption, particularly in developing regions, such as Kenya (Odini, 2014). The second research area concerns the role and effectiveness of digital technology. Many studies show that digital technology can improve agricultural production efficiency, increase farmers’ income by optimizing production factor allocation (Hua et al., 2022; Rowe et al., 2019), promote urban–rural integration and nonagricultural employment (Miyajima, 2022), facilitate inclusive financing (Su et al., 2023), increase agricultural mechanization (Peng and Huang, 2024), promote the outflow of rural labor, and promote technical learning among farmers (Van Campenhout et al., 2021). Salim et al. (2015), however, reported that although digital technologies may increase income in the long term, their short-term effect on income is limited. Other studies have shown that digital technologies contribute to agricultural modernization by increasing digital productivity (Yang and Wang, 2023), facilitating industrial transformation (Hou and Hao, 2023), promoting green farming practices (Jiang et al., 2024b; Shikuku, 2019; Wei et al., 2024), increasing the use of organic fertilizers, and supporting the recycling of agricultural film and pesticide packaging (Peng et al., 2024; Xiong et al., 2025).

Despite such work, some deficiencies remain in the research. First, in terms of content, existing studies mainly focus on macro-level factors, such as policy support, infrastructure development, and digital talent cultivation. However, comparatively little attention has been paid to individual-level determinants. In particular, there is limited research on how herdsmen’s livelihood capital influences their use of digital technology. Second, most studies focus on farmers in agricultural areas, whereas relatively few have examined the use of digital technology among herdsmen. Compared with agricultural areas, pastoral areas differ in terms of production methods, geographical conditions, and economic structures. Such differences mean that herdsmen exhibit specific characteristics in their acceptance and use of digital technology. There is a need, then, for research that considers the specific circumstances of pastoral areas. Against this backdrop, this study employs survey data collected from 356 herder households in Xilingol League and Chifeng in Inner Mongolia. Probit and moderating effect models are used to analyze the relationship between livelihood capital and herdsmen’s use of digital technology.

Figure 1

Figure 1. Figure 1 has been replaced (the background color was changed, but the content remains unchanged).

2 Theoretical analysis and research hypotheses

2.1 Direct impact of livelihood capital on herdsmen’s use of digital technology

Livelihood capital encompasses the total resources and assets owned by herdsmen and shapes their digital technology decision-making (Clark et al., 2018). The conceptual mechanism is illustrated in Figure 1. Livelihood capital reflects various factors that influence the vulnerability or intensity of herdsmen’s livelihood strategies (Ahmed et al., 2008; Allison and Ellis, 2001; Carney, 2003). Based on previous studies (Bhandari, 2013; Chen and Zhao, 2024; Huang et al., 2021; Wu et al., 2017) and the distinct characteristics of herding households, this study classifies livelihood capital into five categories: natural, physical, human, financial, and social capital. Natural capital is mainly reflected in herdsmen’s grassland area and quality. Based on field survey data, the average grassland area per household is 9,400 mu, while the average household labor force is only 2.54 people. Thus, each laborer must manage about 3,700 mu of grassland. Such a large grassland scale places high management demands on limited labor, thus forcing herdsmen to choose between hiring external labor and adopting digital technology use in the production process. According to technology adoption theory, herdsmen mainly base their decisions to adopt new technologies on the perception of potential benefits (Davis, 1989). However, with the continual advancement of China’s urbanization policies, the labor supply in pastoral areas has become insufficient, and costs and management uncertainties have increased (Yuan and Chen, 2019), thus leading to a gradual decline in the marginal benefits of hiring labor to manage grassland. Digital technology offers advantages such as reducing labor, improving management efficiency, and lowering risks (Wolfert et al., 2017; Eastwood et al., 2019), thus allowing herdsmen to perceive relatively high benefits and low risks in adopting digital technology for grassland management. Moreover, physical capital is primarily reflected in the livestock scale and machinery assets (Van Campenhout et al., 2021). In one sense, the larger the livestock scale, the more easily the fixed input cost of digital equipment can be spread, thus allowing herdsmen to more clearly perceive the relative advantages of technology in reducing operating costs and improving efficiency (Huang et al., 2022). In another sense, the abundance of machinery assets not only reflects better production conditions and higher technology acceptance but also enhances the compatibility of digital technology with existing production methods, thereby promoting use (Cui and Wang, 2023). At the same time, as the scale increases, the transaction costs and labor burden of herdsmen in monitoring and management increase significantly. Digital technology, as a labor-saving tool, can effectively alleviate this constraint. Therefore, physical capital—through mechanisms such as scale impacts, technological complementarity, and labor substitution—significantly promotes herdsmen’s digital technology use. Human capital is primarily reflected in herdsmen’s education and household labor (Schultz, 1960). That is, herders with a higher education level possess stronger abilities in information acquisition, comprehension, and operation, making it easier for them to perceive the usefulness and ease of use of digital technology during application, thus lowering learning and operational barriers (Davis, 1989). At the same time, better-educated herdsmen can better absorb new knowledge and integrate it into production practices, thus improving the efficiency of technology integration and application. Sufficient household labor enables herdsmen to establish a reasonable division of labor within the family, with some members being dedicated to learning, testing, and maintaining new technologies. This approach ensures adequate training and understanding in the early stages and facilitates later operation and upgrading, thus reducing trial-and-error costs and enhancing the sustainability of technology application. Therefore, human capital effectively promotes herdsmen’s digital technology use by enhancing cognitive ability, learning capacity, and labor support. Furthermore, financial capital mainly includes herdsmen’s income from livestock production and financing ease. As digital technology usually involves high initial investment and ongoing maintenance costs, herdsmen with stronger financial capital can better bear these expenses, reduce funding constraints, and allocate necessary complementary resources such as network infrastructure and training. This approach increases the feasibility of technology investment and increases its application value (Yan et al., 2025). Additionally, higher income and financing capacity strengthen herdsmen’s ability to withstand risks, thus enabling them to resist potential losses during trial and iteration, thereby reducing the risk costs associated with technology failure (Zhou et al., 2022). Therefore—financial capital, through mechanisms such as financial support, risk sharing, and resource integration—effectively enhances herdsmen’s willingness and ability to adopt digital technology use. Social capital functions as a structural social resource (Putnam, 1994). In one sense, rich social networks and active social participation help herdsmen acquire timely knowledge and experience related to digital technology. Through peer demonstration, information exchange, and organized channels such as cooperatives, training programs, and technical exchange activities, they can reduce learning costs, enhance understanding, and improve the trialability and compatibility of digital technology (Liu et al., 2022; Wang and Qu, 2024). In another sense, a high level of social trust allows herdsmen to rely on credible information sources and professional support, thus reducing uncertainty about the reliability and risks of technology (Ren et al., 2022). Thus, social capital—through information acquisition, cognitive enhancement, risk reduction, and organizational support—significantly promotes herdsmen’s digital technology use. Based on this analysis, herdsmen with more livelihood capital are more likely to use and benefit from digital technologies. However, the influence of livelihood capital on the use of digital technology varies across different dimensions. Thus, the following hypothesis is proposed:

H1: Livelihood capital positively influences herdsmen’s use of digital technologies.

H1-1: The impacts of different dimensions of livelihood capital on herdsmen’s digital technology use.

2.2 Indirect impact of livelihood capital on herdsmen’s use of digital technology

Based on technology adoption theory, herdsmen’s perceived ease of use and perceived usefulness are key factors influencing whether they adopt digital technology. However, these are psychological factors, causing herdsmen with different levels of livelihood capital to show heterogeneity in these perceptions. Notably, even herdsmen with similar resource endowments also show heterogeneity in their perceptions of usefulness and ease of use. A possible reason is the differences in herdsmen’s mastery of new technologies or the conditions under which they can use them. Therefore, this study selects technical training and network quality as moderating variables; the mechanism of action is shown in Figure 2.

Figure 2

Figure 2. Mechanisms by which livelihood capital influences herdsmen’s use of digital technology.

2.2.1 Moderating role of technical training

When herdsmen initially encounter new agricultural production technologies, the absence of authoritative information sources increases their information search costs and the perceived risks of adopting these technologies (Jiang et al., 2024a). Technical training is a vital way for herdsmen to obtain production-related information and improve their skills, thereby facilitating the use of digital technology. On the one hand, technical training can offer herdsmen reliable information about digital technology, thus lowering information transmission barriers (Engås et al., 2023; Liu et al., 2023). It enhances their understanding of, acceptance of, and operational proficiency in digital technology, helping them evaluate its feasibility and overall benefits while alleviating their concerns. On the other hand, technical training can offer specialized guidance to trainees (Eastwood et al., 2017; Liu et al., 2022), enhance technical knowledge, and strengthen resource allocation and technical use capabilities (Ammann et al., 2022). Studies have found a positive correlation between farmers’ participation in agricultural training programs and their propensity to adopt agricultural technologies (Song and Qi, 2013).

As a determinant of access to resources for survival and development, livelihood capital comprises multiple dimensions, including human, natural, physical, and social capital. Livelihood capital shapes herdsmen’s ability to receive, assimilate, and apply external technological information. First, herdsmen with higher levels of human capital tend to exhibit stronger cognitive abilities and learning capacities, as well as more proactive attitudes toward adopting new technologies. They are more likely to recognize the benefits of technical training and are better at acquiring the skills needed to operate digital tools, demonstrating an enhanced ability to absorb and internalize knowledge. Second, those with abundant physical capital have more flexibility in terms of time and financial resources, enabling them to cover the costs associated with training and enjoy greater autonomy in accessing such programs. Third, social capital improves access to training information and opportunities. Finally, herdsmen with more natural capital face more pressing demands to improve production efficiency and digital management, thereby displaying higher motivation to engage in training programs. These factors can enable capital-rich herdsmen to leverage existing resources more effectively, thereby enhancing their use of digital technology. Conversely, technical training can help capital-constrained herdsmen compensate for deficiencies in knowledge, skills, and assets, thereby lowering the barriers to digital engagement. Accordingly, technical training is an enabler of the use of digital technology. Thus, the following is proposed:

H2: Technical training moderates the relationship between livelihood capital and herdsmen’s use of digital technology.

2.2.2 Moderating role of network quality

With rapid advancements in internet technologies, economic development across different regions in China is no longer restricted to traditional models (Wang et al., 2023). Yang and Kong (2021) suggested that internet and mobile coverage provide the foundation for herdsmen to participate in digital financial markets. More stable network coverage enhances the efficiency of the use of digital technology, reduces barriers to use, and maximizes the realization of its functions. This enables herdsmen to use digital technology efficiently for precision management, thereby improving economic returns. On the one hand, herdsmen with more natural and physical resources can use digital technology, supported by a reliable network, to optimize pasture monitoring and livestock management. On the other hand, herdsmen with more human capital can apply their technological expertise more effectively in a stable network environment. Furthermore, social capital is strengthened by the efficient dissemination of information within the network, facilitating greater information sharing and technology use. Conversely, in regions with inadequate network coverage or weak signals, digital devices may experience significant functional limitations or become entirely nonoperational (Engås et al., 2023). This will hinder herdsmen’s experience with digital technology. This will not only increase production costs but also affect herdsmen’s expectations of income from using digital technology. Even herders with relatively abundant livelihood capital and a strong desire to use technology lose trust in it, thereby reducing the positive impact of livelihood capital on the use of digital technology. Consequently, the positive impact of livelihood capital on the use of digital technology diminishes in regions with inadequate network infrastructure. Based on the above, the following is proposed:

H3: Network quality moderates the relationship between livelihood capital and herdsmen’s use of digital technology.

3 Data and methods

3.1 Study area and data

The data for this study were obtained from national project surveys conducted by the research team between July and August 2023 and again in January 2024. The survey area covers seven banners (or cities) in Inner Mongolia: Sunite Left Banner, Sunite Right Banner, Abag Banner, Xilinhot City, East Ujimqin Banner, and West Ujimqin Banner in Xilingol League, and Hexigten Banner in Chifeng City. The study was carried out in two cities in the Xilingol League and Chifeng regions, as shown in Figure 3. The selection of survey respondents was guided by the following considerations: First, geographical characteristics. Xilingol League and Chifeng are major pastoral regions in Inner Mongolia. They are characterized by abundant grassland resources and extensive pastoral areas, which makes them representative of pastoral economies. Second, economic diversity. The selection of Xilingol League and Chifeng ensures representation of various levels of economic development in Inner Mongolia. The survey used a stratified random sampling method. In each banner (county), two soums (townships or towns) were randomly selected. Within each soum, two to three gacha (villages) were randomly chosen, and 10–15 households were surveyed in each gacha. A total of 366 households participated in the survey. After excluding 10 incomplete or inconsistent questionnaires, 356 valid questionnaires remained, resulting in an effective response rate of 97%.

Figure 3

Figure 3. Geographical location of the study area.

3.2 Variable selection

3.2.1 Dependent variable

The dependent variable is the herdsmen’s use of digital technology. This behavior is measured based on the survey question, “Has your household adopted any digital devices such as monitoring equipment, smart drinking devices, or electronic ear tags in the course of production and operation?” A 0/1 binary variable is used to represent herdsmen’s use of digital technology, where a user of digital technology = 1 and a nonuser = 0.

3.2.2 Independent variable

The independent variable is the level of herdsmen’s livelihood capital. Based on the DFID Sustainable Livelihoods Analysis Framework (Animato, 1999), livelihood capital is classified into five types: natural, physical, human, financial, and social capital. Drawing on previous research (Chen and Zhao, 2024; Quandt, 2018; Wu et al., 2017; Zhong et al., 2022) and combining the theoretical foundation of the Sustainable Livelihoods Approach with the production and living characteristics of pastoral areas in Inner Mongolia, China, this study selects 13 measurable indicators to comprehensively capture herders’ livelihood capital. For natural capital, grassland serves as the fundamental resource for the survival and development of herders. Grassland area and grassland quality best reflect the status of natural resource endowments; therefore, “total grassland area” and “grassland quality” are selected as the two indicators. In terms of physical capital, herders’ production conditions mainly depend on livestock numbers and the level of mechanization. Accordingly, “total livestock holdings” and “production machinery and equipment” are selected to capture the scale of productive assets. For human capital, the household labor structure, along with education and health conditions, directly determines herders’ ability to learn and adopt new technologies. Thus, “average education level of household members,” “number of family laborers,” and “health status of the household head” are chosen as three key indicators. For financial capital, household income and financing capacity directly influence investment ability and risk-bearing capacity for technology adoption; therefore, “per capita net income” and “financing capacity” are selected as representative indicators. Finally, for social capital, social relationships in pastoral areas are mainly reflected through both online and offline networks, mutual assistance, and organizational participation. Accordingly, “livestock-related WeChat groups,” “household expenditures on personal gifts,” “cooperative participation,” and “trust in relatives and neighbors” are adopted to comprehensively capture social interaction and trust levels among herders. Overall, the 13 selected indicators are theoretically consistent, practically applicable, and data-accessible, thus providing a comprehensive measurement of herders’ livelihood capital. The objective weights of these indicators are determined using the entropy weighting method. It should be noted that the five dimensions of livelihood capital are not entirely independent in empirical measurement, as their formation processes are inherently interrelated. For example, the “number of family laborers” reflects human capital while also influencing the accumulation of physical capital. Such cross-dimensional interconnections are an inherent feature of the Sustainable Livelihoods Framework and reflect the complexity and integrative nature of herders’ livelihood systems.

3.2.3 Moderating variable

The moderating variables in this study are technical training and network quality. Technical training is measured by the number of livestock technical training sessions attended (Groher et al., 2020). This variable reflects herders’ level of professional training and their capacity to acquire and apply new knowledge and technologies. Network quality is measured based on the quality and stability of the network signal (Gozzi et al., 2023). The variable is coded as follows: No network coverage = 1, weak and unstable network signal = 2, weak network signal with average stability = 3, strong and stable network signal = 4, and strong and very stable network signal = 5. This variable captures the level of digital infrastructure available to herders, which influences their ability to access and use digital technologies.

3.2.4 Control variables

The control variables in the model include household head characteristics and regional characteristics. Household head characteristics encompass gender, age, health status, farming experience, and risk preference. The regional characteristic considered is the distance between the household residence and the nearest signal tower. Among these, risk preference is measured through a contextual choice experiment. Specifically, herdsmen are asked to choose between a series of certain payoffs (Option A) and risky payoffs (Option B). Option A is defined as “receiving a fixed amount with certainty,” whereas Option B is “a 50% chance of receiving 500 yuan or 0 yuan.” As the scenarios progress, the fixed amount in Option A is gradually increased to 50, 100, 150, 200, and 250 yuan. By observing herdsmen’s choices across these scenarios, their risk preference tendencies can be effectively characterized. In terms of coding, this study dichotomizes the results: if herdsmen tend to choose certain payoffs (Option A) in most scenarios, the value is set to 0, indicating risk aversion. However, if they tend to choose risky payoffs (Option B) in most scenarios, the value is set to 1, indicating risk preference.

3.2.5 Instrumental variable

This concerns livelihood capital’s impact on herdsmen’s use of digital technology. To address potential endogeneity arising from counterfactual causality and omitted variable bias, two instrumental variables are introduced to correct estimation errors: village average grassland area and village average education. These two variables are derived from the Village average grassland area and the Village average education of other herdsmen within the same survey village to mitigate potential endogeneity bias. Regarding relevance, herdsmen’s livelihood capital is largely influenced by the village’s overall grassland resources and knowledge environment. The more abundant the grassland resources, the more likely individual herdsmen are to acquire larger grassland areas, thus accumulating higher levels of natural capital. However, education not only reflects individual knowledge and skills but also creates externalities at the village level through demonstration impacts, information sharing, and social interaction. When the overall education level of a village is higher, knowledge spillovers among herdsmen are more likely to occur, which enhances individual learning capacity and human capital accumulation, thus strengthening overall livelihood capital. Therefore, the Village average grassland area and the Village average education can significantly influence herdsmen’s livelihood capital, satisfying the relevance requirement of instrumental variables. Regarding exogeneity, even though village-level grassland resources and education levels may influence herdsmen’s natural capital and human capital, they do not directly influence whether individuals adopt digital technology. The digital technology use of herdsmen is more limited by factors such as technical training, network infrastructure, and financial conditions. Thus, the Village average grassland area and the Village average education primarily affect digital technology adoption indirectly through their impact on herdsmen’s livelihood capital, rather than affect it directly by entering the decision-making process of digital technology use. This satisfies the exogeneity requirement of instrumental variables.

3.2.6 Descriptive statistics

Table 1 shows the definitions, assignments, and descriptive statistics of all variables in the model. (1) Herdsmen’s use of digital technology. The descriptive results show that, among the full sample of 356 herdsmen, the average value of the use of digital technology is 0.34. This indicates that approximately 34% of the surveyed herdsmen have used at least one type of digital device in their livestock production activities. (2) In terms of livelihood capital, the mean value of herdsmen’s overall livelihood capital is 0.27, with a standard deviation of 0.19. The mean livelihood capital score is higher for herdsmen using digital technology than for those not using it, indicating that herdsmen with more livelihood capital are more likely to use digital technology. (3) Regarding the subdimensions of capital, the mean values of grassland area and quality for herdsmen using digital technology are 1.19 and 3.93, respectively. Both values are higher than those for nonusers. In terms of physical capital, herdsmen who use digital technology have an average livestock scale that is 174.79 sheep units higher than that of nonusers. Similarly, the mean value of machinery assets is also higher among users of digital technology than among nonusers. In terms of human capital, users of digital technology have higher average levels of education and a larger labor force than nonusers, reflecting stronger human capital among adopters. In terms of financial capital, the average per capita net household income and financing ease values are higher for herdsmen who use digital technologies than for nonusers. Regarding social capital, herdsmen who use digital technology consistently exhibit higher levels than nonusers across all relevant indicators.

Table 1

Table 1. Descriptive statistics of all variables.

We can conclude that the mean value of livelihood capital is higher for herdsmen who use digital technology than it is for nonusers. Therefore, there appears to be a positive correlation between livelihood capital and herdsman’s use of digital technology. However, identifying a causal relationship between the two requires empirical testing through econometric models that control for other factors and address potential endogeneity.

3.3 Methods

3.3.1 Entropy weight method

This study uses the entropy weight method to quantify herdsmen’s livelihood capital. This method is chosen because it is objective in terms of assigning weights according to the degree of variation among indicators, thereby minimizing the potential bias caused by subjective judgment and enhancing the reliability of the results. Furthermore, given the multidimensional nature of livelihood capital, the method facilitates a clearer understanding of disparities between different dimensions of capital, which is particularly beneficial when conducting empirical analyses of the level and structure of livelihood capital. The formulas are given below.

Step 1: Standardization of original data. The raw data are first standardized using the z-score method to eliminate dimensional inconsistency among the indicators, as shown in Equation 1. The standardized value $X_{\mathit{ij}}^{\prime }$ for the i-th herdsmen and the j-th indicator is computed as follows:

$\begin{array}{ll}{X}_{\mathit{ij}}^{\prime }=\frac{X_{\mathit{ij}}-{\overline{X}}_j}{S_j},\text{where}{\overline{X}}_j=\frac{\sum _{i=1}^n{X}_{\mathit{ij}}}{n},S_j^2=\sum _{i=1}^n(X_{\mathit{ij}}^{\prime }-{\overline{X}}_j)2.& (1)\end{array}$

To eliminate the influence of negative values on subsequent logarithmic transformations, a coordinate shift is applied, such that $Z_{\mathit{ij}}=X_{\mathit{ij}}^{\prime }+1.$

Step 2: Calculation of the proportion matrix. To normalize the data across indicators, the proportion of the i-th herdsmen for the j-th indicator is calculated as shown in Equation 2.

$\begin{array}{ll}{P}_{\mathit{ij}}=\frac{X_{\mathit{ij}}^{\prime }}{\sum _{i=1}^n{X}_{\mathit{ij}}^{\prime }}.& (2)\end{array}$

Step 3: Entropy value calculation. The entropy value $e_j$ for the j-th indicator is computed, as shown in Equation 3.

$\begin{array}{ll}{e}_j=-k\sum _{i=1}^n{P}_{\mathit{ij}}\mathit{LN}{P}_{\mathit{ij}},\text{where}k=\frac{1}{\mathrm{LN}n}.& (3)\end{array}$

Step 4: Determination of indicator weights. The indicator weight Wj is calculated, as shown in Equation 4.

$\begin{array}{ll}{W}_j=(1-e_j)/\sum (i=1\text{to}n)(1-e_j)& (4)\end{array}$

Step 5: Calculation of the Composite Livelihood Capital Index. The standardized values are combined with the corresponding indicator weights using weighted averaging to calculate the livelihood capital of each type and the overall livelihood capital for every herdsman (as shown in Equation 5).

$\begin{array}{ll}\mathit{\text{Livelihood capita}}{l}_i=\sum _{j=1}^m{W}_j.Z_{\mathit{ij}}.& (5)\end{array}$

3.3.2 Baseline regression model

This study uses a probit model to examine the relationship between livelihood capital and herdsmen’s usage of digital technology.

Probit model:

$\begin{array}{l}\mathit{\text{Digital technology}}\mathit{us}{e}_i^{\ast }={\alpha }_0+{\alpha }_1\mathit{\text{Livelihood capita}}{l}_i+\\ {\alpha }_2\mathit{\text{Contro}}{l}_i+{\epsilon }_{i1},\\ \mathit{\text{Digital technology}}\mathit{us}{e}_i=1,(\mathit{\text{Digital technology}}\mathit{us}{e}_i^{\ast }>0).\end{array}\text{ (6)}$

In Equation 6, $\mathit{\text{Digital technology}}\mathit{us}{e}_i$ represents the dependent variable in this study, which refers to herdsmen’s use of digital technology. $\mathit{\text{Digital technology}}\mathit{us}{e}_i^{\ast }$ is the latent variable. When $\mathit{\text{Digital technology}}\mathit{us}{e}_{\mathrm{i}}^{\ast }>0$ , $\mathit{\text{Digital technology}}\mathit{us}{e}_i$ takes a value of 1; otherwise, it is 0. $\mathit{\text{Livelihood capita}}{l}_i$ represents the independent variable, which refers to the livelihood capital level. $\mathit{\text{Contro}}{l}_i$ represents control variables, including herdsmen characteristics, regional characteristics, and natural disaster characteristics. ${\alpha }_0$ represents the constant term, ${\alpha }_1$ and ${\alpha }_2$denote the parameters to be estimated, and ${\epsilon }_{i1}$ is the error term.

3.3.3 Moderating effect model

To estimate the moderating effects of technical training and network quality on the relationship between livelihood capital and herdsmen’s use of digital technology, an extended regression model incorporating interaction terms between livelihood capital, technical training, and network quality is developed based on Equation 6 (Wen et al., 2005). The model is specified as follows:

$\begin{array}{l}\mathit{\text{Digital technology}}\mathit{us}{e}_i={\alpha }_0+{\alpha }_1\mathit{\text{Livelihood capita}}{l}_i+{\alpha }_2{M}_i+\\ {\alpha }_3\mathit{\text{Livelihood capita}}{l}_i\times \\ M_i+{\alpha }_4\mathit{\text{Contro}}{l}_i+{\epsilon }_{i1}.\end{array}\text{ (7)}$

In Equation 7, $\mathit{\text{Digital technology}}\mathit{us}{e}_i$represents herdsmen’s use of digital technology, and $M_i$ denotes the moderating variable, which includes technical training and network quality. The meanings of the other variables are the same as in the probit model.

4 Results

4.1 Entropy weight calculation results

Table 2 presents the descriptive statistics of herdsmen’s livelihood capital and its five constituent dimensions. In terms of overall livelihood capital, herdsmen’s average score is 0.294, with a standard deviation of 0.177. The significant difference between the maximum and minimum values highlights the substantial disparity in capital endowments. This suggests that while some herdsmen have relatively sufficient resources, many still face significant constraints. In terms of individual capital dimensions, herdsmen’s livelihood capital shows an overall pattern of scarcity, with relatively low scores across all dimensions. Unlike the idealized pentagonal structure of balanced capital endowment, the distribution of capital is uneven. The ranking from highest to lowest mean value is as follows: social capital > financial capital > human capital > natural capital > physical capital. This pattern indicates that herdsmen are relatively well endowed with social capital, reflecting stronger interpersonal networks, information exchange, and community participation, while physical capital remains the most deficient (Zhang et al., 2017).

Table 2

Table 2. Estimation results for herdsmen’s livelihood capital.

From the comparison of mean values across different types of livelihood capital, herdsmen who use digital technology often show significantly higher levels of all types of capital than nonusers, with differences significant at the 1% level. Among these, the differences in financial, social, and physical capital are the most pronounced, while those in natural capital and human capital are less overt. This result indicates that digital technology use largely depends on the support of financial, social, and physical conditions. Specifically, herdsmen with higher financial capital levels can more easily afford the costs of equipment acquisition and maintenance while also expanding financing channels through digital financial tools. Thus, they can further enhance their financial capital accumulation. In terms of social capital, digital technology use significantly improves the efficiency of herdsmen’s external connections, thereby facilitating online communication and resource sharing, which allows them to better embed in social networks and foster reciprocal cooperation. Consequently, they possess clear advantages in social capital. Concerning physical capital, herdsmen with better production facilities and larger-scale livestock operations are more motivated and capable of adopting digital technology. The application of digital management, in turn, enhances production efficiency and resource utilization, further strengthening their physical capital. In contrast, natural capital mainly depends on natural endowments such as grassland area and policy allocation; it is less directly influenced by digital technology use. Human capital, while somewhat related to technology adoption, exhibits no significant difference, likely because overall education levels and training opportunities in pastoral areas are relatively limited.

4.2 Baseline regression model results

Table 3 presents the baseline regression results for livelihood capital’s impact on herdsmen’s use of digital technology. Columns (1) and (2) show the baseline regression results of overall livelihood capital and its five sub-dimensions on herdsmen’s digital technology use. Columns (3) and (4) show the results from the IV-probit model. The baseline regression results show that livelihood capital has a significant positive impact on herdsmen’s digital technology use at the 1% statistical level, confirming Hypothesis 1. Further analysis of the sub-dimensions shows that natural, physical, human, and social capital all significantly promote herdsmen’s digital technology use. Among these, the coefficients from largest to smallest are physical, natural, human, and social capital, suggesting that material conditions and natural resource endowments play the most prominent roles in driving digital technology use. By contrast, the impact of financial capital is statistically insignificant, indicating that at the current stage, financial capital has not yet offered impactive support for herdsmen’s digital technology use. Field survey results show that the price of a single monitoring device is generally between 200 and 1,000 yuan, which is relatively low. This implies that the financial investment required for herdsmen to purchase and use such equipment is limited, and that the influence of financial capital on digital technology use is not significant. From the regression results of the IV-probit model, first, endogeneity is addressed through the use of instrumental variables. To mitigate potential endogeneity in livelihood capital, two village-level instrumental variables are employed: village grassland area and village education level. First, the overidentification test results support the exogeneity of the instruments. The J-statistic is 0.09, corresponding to a p-value of 0.762. Furthermore, the Amemiya–Lee–Newey test yields a similar conclusion, suggesting that the model does not suffer from overidentification. Second, the weak instrument diagnostics further confirm the relevance of the instruments. Both the conditional likelihood ratio (CLR) (Moreira, 2003) and K-statistic are statistically significant at the 1% level. This shows that the selected instrumental variables do not suffer from the problem of weak instrumental variables. Third, the first-stage regression results (column 3 of Table 3) demonstrate that both village-level grassland area and village education level have statistically significant and positive impacts on herdsmen’s livelihood capital at the 1% significance levels. These findings confirm that the selected instruments are strongly correlated with the endogenous explanatory variable, thereby satisfying the relevance condition required for valid instrumental variable estimation. Second, livelihood capital has a significant and positive impact on herdsmen’s use of digital technology, This result further confirms H1. This indicates that herdsmen with more livelihood capital are substantially more likely to adopt digital technologies such as monitoring equipment, smart drinking devices, and electronic ear tags.

Table 3

Table 3. Baseline regression results for livelihood capital’s impact on herdsmen’s use of digital technology.

4.3 Indirect impact of livelihood capital on herdsmen’s use of digital technology

4.3.1 Moderating effect analysis of livelihood capital on herders’ digital technology use

Table 4 shows the results of the moderating effect test for livelihood capital’s impact on herdsmen’s use of digital technology. Columns (1) and Columns (2) report the moderating role of technical training, while Columns (3) and Columns (4) report that of network quality. Regarding the moderating effect of technical training, the interaction term between herdsmen’s livelihood capital and technical training is statistically significant at the 5% level. This result indicates that herdsmen with higher levels of livelihood capital derive more benefits from the use of digital technology when they receive relevant training. Such training reduces information asymmetry and skill barriers (Rogers et al., 2014), enhances their knowledge absorption and operational competence, and further strengthens their confidence in technology adoption. Regarding the moderating effect of network quality, the interaction term between herdsmen’s livelihood capital and network quality is significant at the 5% level. This finding suggests that improved network quality reinforces the positive relationship between herdsmen’s livelihood capital and their use of digital technology. Specifically, stable and accessible internet infrastructure not only ensures the smooth functioning of digital tools, but it also reduces the transaction costs associated with information acquisition (Aker, 2011), thereby enabling herdsmen to more effectively leverage their capital endowments. Moreover, better network quality facilitates information sharing and collaboration within social networks, amplifying the role of social capital in technology diffusion (Zhang et al., 2025).

Table 4

Table 4. Regression results of the moderating effect test.

This finding suggests that improved network quality reinforces the positive relationship between herdsmen’s livelihood capital and their use of digital technology. Specifically, stable and accessible internet infrastructure enables herdsmen to better leverage their capital endowments in the application of digital technologies, thereby amplifying the practical impact of their livelihood capital. These results support H2 and H3.

4.3.2 Moderating effect analysis of different livelihood capitals on herders’ digital technology use

Tables 5, 6 list the moderating effects of technical training and network quality under different types of livelihood capital. In these tables, the odd-numbered columns show the main impacts of each type of capital, and the even-numbered columns add the interaction terms to test whether technical training and network quality significantly alter the influence of different types of livelihood capital on herdsmen’s digital technology use. According to the regression results in Table 5, the interaction term between physical capital and technical training is significantly positive at the 5% level. This indicates that technical training can effectively bolster equipment operation and management ability in households with large-scale livestock production and better equipment conditions, thus amplifying the positive impact of physical capital. Additionally, the interaction term between social capital and technical training is also significantly positive at the 5% level, suggesting that technical training further strengthens the role of social capital. Specifically, technical training not only improves individual knowledge and technical ability but also promotes information sharing and collaborative diffusion through integration with social networks. This enhances the impact of social capital in promoting digital technology use (see Columns 4 and 10). In contrast, the interaction terms of natural, human, and financial capital with technical training are statistically insignificant (see Columns 2, 6, and 8). Regarding natural capital, field research found that most herdsmen mainly apply monitoring equipment to the internal management of cattle sheds and sheep pens. Thus, the impact pathway of technical training on natural capital is limited. For human capital, descriptive statistics reveal that the gap between digital technology users and nonusers in terms of human capital is insubstantial. Moreover, given that digital technology applications in pastoral areas are still at an early stage and devices such as monitoring equipment and smart drinking devices are relatively simple to operate, the marginal impact of training remains insufficient. In terms of financial capital, as the current prices of digital equipment are relatively low, the financial burden of acquisition is limited, so training has not significantly amplified the role of financial capital in promoting digital technology use.

Table 5

Table 5. Moderating effect of technical training under different types of livelihood capital.

Table 6

Table 6. Moderating effect of network quality under different types of livelihood capital.

Per the regression results in Table 6, the interaction term between physical capital and network quality is significantly positive at the 1% level, indicating that improvements in network conditions can significantly bolster the role of physical capital in promoting digital technology use. Moreover, a favorable network environment enables herdsmen with large-scale livestock operations and better equipment conditions to use digital devices more efficiently, thus amplifying the marginal benefits of physical capital. Additionally, the interaction term between social capital and network quality is also significantly positive at the 10% level, suggesting that network quality further strengthens the impact of social capital. Better network conditions improve information sharing and collaboration efficiency among herdsmen, enabling mutual trust and cooperation within social relationships to be more smoothly transformed into the motivation for technology adoption (see Columns 4 and 10). Finally, the interaction term between human capital and network quality is significantly negative at the 10% level. A possible explanation is that herdsmen with higher education levels or stronger skills can often independently solve problems in digital technology use through self-learning and accumulated experience, making them relatively less reliant on network conditions. However, herdsmen with lower education levels are more likely to take advantage of online resources under favorable network conditions to access information and technical support, thus creating a reverse differential impact and a negative interaction between human capital and network quality. In comparison, the interaction terms of natural capital and financial capital with network quality are statistically insignificant (see Columns 2 and 8). Field research further revealed that about 65% of herdsmen mainly use monitoring equipment within livestock pens, while its application in grasslands is limited. As most herdsmen’s grasslands are far from their residences (commonly 30 km, 50 km, or even over 100 km), installing network lines and monitoring equipment on grasslands would represent a considerable expense for small-scale households. Therefore, even if herdsmen have sufficient capital, they will have difficulty independently building and maintaining network infrastructure. Consequently, the roles of natural and financial capital in improving network conditions are severely constrained and cannot be effectively transformed into drivers of digital technology use.

4.4 Robustness test

Table 7 shows the results of the robustness tests for livelihood capital’s impact on herdsmen’s use of digital technology. To ensure the robustness of the findings, two approaches are used to test livelihood capital’s impact on herdsmen’s use of digital technology. First, the dependent variable is replaced for robustness testing. To further assess the robustness of livelihood capital’s impact on herdsmen’s use of digital technology, the original binary dependent variable—“whether digital technology is used”—is replaced with a continuous variable representing the number of digital devices used by herdsmen. Specifically, this variable measures the total count of digital tools employed in livestock production, including monitoring equipment, smart drinking devices, and electronic ear tags. Further, the model is reestimated using the two-stage least-squares (2SLS) approach. Second, to mitigate the potential influence of extreme values on the results, two-sided 5% winsorization is applied to the core explanatory variable (i.e., livelihood capital). Specifically, observations below the 5th and above the 95th percentile of the distribution are replaced with the corresponding percentile values. After this adjustment, the IV-probit model is reestimated to examine the robustness of the findings. The results obtained from both robustness tests are consistent with the baseline regression results in Table 3, thus confirming again the robustness of the main findings.

Table 7

Table 7. Robustness tests.

4.5 Heterogeneity analysis

Table 8 shows the heterogeneity analysis results for livelihood capital’s impact on herdsmen’s use of digital technology. In terms of grassland area, livelihood capital has a positive impact on the use of digital technology by herders with grassland areas greater than 7,500 mu, with a coefficient that is significant at the 1% level. This suggests that herders will demand greater efficiency in production management as the size of the grassland area increases, thus amplifying the promotional impact of livelihood capital on their digital technology use. In terms of the scale of livestock, the livelihood capital of herdsmen has a positive impact on their use of digital technology if they have more than 449 sheep at the 1% significance level. This suggests that herdsmen with more livestock are more likely to use digital technology. Larger-scale operations tend to involve more complex production and management processes, thereby increasing the need for digital tools to enhance operational efficiency. Regarding the educational level, herdsmen whose families have an average educational level of more than seven years have a positive impact on their use of digital technology, with a coefficient that is significant at the 1% level. This suggests that herdsmen with more formal education have greater capabilities in terms of understanding and using digital tools.

Table 8

Table 8. Heterogeneity analysis results.

Figure 4 further visualizes the marginal effects of livelihood capital on herders’ digital technology adoption rate. The horizontal axis represents the grouping of livelihood capital, which is divided into 20 equal-interval groups according to the distribution of herders’ comprehensive livelihood capital scores, with each group representing a specific range of livelihood capital levels. This approach, while maintaining the continuity of the variable, allows for a smooth depiction of the gradient from low to high livelihood capital. The vertical axis indicates the mean rate of digital technology use within each group, reflecting the extent of technology utilization among herders with different livelihood capital levels. The three panels, respectively, categorize herders by grassland area, livestock scale, and education level, illustrating differences in the marginal effects across heterogeneous groups. The results indicate that as livelihood capital increases, the overall adoption rate of digital technology tends to rise. Further, herders with larger grassland areas, greater livestock scales, and higher education levels exhibit stronger responsiveness to changes in livelihood capital, thus leading to faster growth in adoption rates. This pattern is consistent with the regression results presented in Table 8, providing further evidence that the influence of livelihood capital on digital technology adoption is heterogeneous across groups.

Figure 4

Figure 4. Marginal effects of livelihood capital on digital technology adoption.

5 Discussion

5.1 The role of livelihood capital

The empirical results of this study show that livelihood capital is an important factor in promoting herdsmen’s digital technology use. Overall, the higher the level of livelihood capital, the more likely herdsmen are to use digital technology. This conclusion aligns with the view of Feder et al. (1985) that resource endowment is a prerequisite for farmers to adopt new technologies, and it also aligns with the conclusion of Tey and Brindal (2012) that farmers’ technology adoption is jointly influenced by capital constraints and capacity differences. However, differently from previous studies that mainly emphasized “financial capital constraints” (Abdulai and Huffman, 2014), this study finds that under circumstances where equipment costs are relatively manageable, nonfinancial capital plays a more central role in driving technology adoption. This suggests that in the early stage of digital technology promotion in China’s pastoral areas, the constraints differ from those in the context of traditional agricultural technology adoption (see Table 3).

In the sub-dimensional analysis, the impacts of natural capital and physical capital are the most significant. This study finds that grassland resources and production assets not only determine herdsmen’s operating scale but also govern their capacity and willingness to adopt digital technology. This result aligns with that of Li et al. (2024), who examined the relationship between large-scale operations and the adoption of smart agriculture. Further, resource endowments still play a prerequisite role in the digitalization process of pastoral areas. Human capital also significantly promotes digital technology use, which aligns with the logic of Rogers et al. (2014) diffusion of innovations theory, emphasizing that a knowledge base is a prerequisite for innovation adoption. Empirical studies also support this point; for example, Larson et al. (2008) found that human capital factors such as age and education are the main determinants influencing farmers’ technology adoption decisions. From the view of pastoral areas, this study further demonstrates that education and skill reserves help herdsmen better understand and master digital technology, thus reducing uncertainty and complexity. The facilitating role of social capital is also confirmed, which aligns with Bandiera and Rasul (2006) findings on “peer impacts and network diffusion” Herdsmen exchange information and share experiences through kinship, cooperatives, and neighborhood ties, thus reducing adoption risks to some extent. This indicates that digital technology adoption is not only the outcome of individual rational choice but an aspect of the unique social networks of pastoral areas. By contrast, the role of financial capital is insignificant. Differently from studies such as Bai et al. (2025), which emphasize the importance of credit constraints in the adoption of agricultural technology, this study finds that in the early stage of digitalization in pastoral areas, the relatively low cost of equipment means that financial capital is not the primary bottleneck. This result is similar to Aker (2011) findings on the diffusion of mobile phones in Africa, where the importance of financial capital diminishes when technology costs are relatively low, while information and social capital become more explanatory. Considering the characteristics of pastoral areas further clarifies this difference: China’s pastoral regions are vast, with a low population density, underdeveloped networks and public service infrastructure, and relatively low levels of income and education among herdsmen. Against this backdrop, herdsmen depend more on natural capital, physical conditions, and social networks to overcome environmental constraints, rather than relying solely on financial capital. Similar phenomena have been seen in some developing countries, where studies in India and Africa show that land resources, social networks, and education levels explain farmers’ adoption of mobile agricultural services more effectively than financial capital (Aker, 2011; Goyal, 2010).

This study not only confirms the theoretical expectation of resource endowments in technology adoption but also reveals the marginal weakening impact of financial capital under conditions of low-cost digital equipment. This result enriches the existing research perspective on farmers’ technology adoption and highlights the fundamental role of natural, physical, human, and social capital in promoting herdsmen’s digital technology use. Therefore, the core conclusion can be drawn that in the early stage of digital technology promotion, nonfinancial capital serves as the primary driving force of herdsmen’s adoption behavior, while the role of financial capital is relatively limited. This conclusion not only provides new empirical evidence for understanding the digital transformation of pastoral areas but also offers a valuable reference for policymakers in terms of resource allocation and priority-setting. It suggests that greater investment should be directed toward strengthening the use of natural capital, improving physical conditions, enhancing human capital, and cultivating social capital, aiming to more effectively promote the digitalization process in pastoral regions.

In addition to the core explanatory variable, some control variables also show significant effects. Specifically, the coefficient of risk preference is positive and significant, indicating that herders with higher risk preferences are more likely to adopt digital technologies. This suggests that herders who are more open to exploration and innovation tend to experiment with new technologies to enhance production efficiency and potential returns. From a policy perspective, promoting pilot programs, improving information dissemination and technical training, and developing risk-compensation or insurance mechanisms could help mitigate the perceived risks of innovation and encourage risk-averse herders to participate in digital technology adoption.

5.2 Moderating effects of technical training and network quality

The empirical results of this study show that technical training and network quality produce significant moderating effects between herdsmen’s livelihood capital and digital technology use, although different types of livelihood capital yield heterogeneous responses in this process. Overall, both technical training and network quality bolster the positive impact of livelihood capital endowments on digital technology adoption, but such an enhancement is selective and conditional, not universal.

In the sub-dimensional analysis, technical training significantly strengthens the impacts of physical and social capital on digital technology use. Specifically, the interaction term between physical capital and technical training is significantly positive at the 5% level, indicating that for households with larger livestock scales and better equipment conditions, technical training can further improve equipment operation and management capacity, thus enhancing the efficiency of digital technology use. This result is consistent with Liu and Yan (2025), who emphasize the critical role of capability accumulation in technology diffusion, and it also confirms the logic in Rogers’ (2014) diffusion of innovations theory that a knowledge base reduces technological uncertainty. The interaction term between social capital and technical training is also significantly positive, suggesting that once herdsmen receive training, then information sharing, experience exchange, and cooperation within their social networks further promote digital technology adoption. This result agrees with Bandiera and Rasul (2006)‘s on peer effects and social learning, showing that training not only improves individual technical ability but also amplifies the value of social capital through network impacts. In contrast, the interaction effects of natural capital, human capital, and financial capital are insignificant, suggesting that in the early stage of digitalization, technical training mainly strengthens existing advantages, rather than compensating for weak capital endowments.

The improvement of network quality also demonstrates differentiated impacts across capital types. The interaction term between physical capital and network quality is significantly positive at the 1% level, indicating that favorable network conditions enable large-scale households to use digital technology more efficiently, thus amplifying the marginal benefits of physical capital. This is consistent with the view of Guo (2025) that infrastructure improvement is a necessary condition for technology adoption. The interaction between social capital and network quality is also significantly positive at the 10% level, showing that stable network conditions strengthen communication, cooperation, and information flows among herdsmen, thus allowing social capital to be more effectively transformed into the motivation for technology adoption. Notably, the interaction between human capital and network quality is significantly negative at the 10% level. A possible explanation is that herdsmen with human capital levels can solve technical problems independently through their skills and experience, making them less reliant on network conditions. In contrast, herdsmen with lower levels of human capital rely more on favorable network conditions to access external information and technical support, thus creating a differentiated impact. This is similar to Aker (2011) findings on the adoption of mobile agricultural services in Africa, where low human capital groups are more likely to gain “latecomer advantages” under favorable information conditions.

Overall, technical training and network quality play important amplifying roles in the digital transformation of pastoral areas, but their impacts vary by capital type. Technical training is more effective when combined with physical capital and social capital, thus promoting digital technology adoption through capacity building and social diffusion. Thus, improvements in network quality not only strengthen the impacts of physical and social capital but also provide additional support for groups with low human capital levels. However, the improvement of network quality in pastoral regions is still constrained by high infrastructure and maintenance costs, which may limit the pace of digital transformation. Therefore, ensuring cost-effective and fiscally sustainable network expansion remains an important policy challenge that requires coordinated planning and phased implementation.

5.3 Limitations and future research

(1) The sample data are relatively limited, which restricts the generalizability of the findings to some extent. China is a vast country, and there are significant differences between pastoral areas in terms of geographical environment, climatic conditions, grassland ecosystems, and ethnic backgrounds. These factors influence herdsmen’s livelihood capital structure and their use of digital technology. (2) With the continuous development of digital technology, herdsmen’s application behaviors are showing dynamic evolutionary characteristics, but the mechanisms through which livelihood capital exerts its impacts may differ across stages of development. At the same time, different types of digital technology differ in terms of cost, complexity, applicability, and expected returns, meaning that the capital constraints faced by herdsmen when adopting monitoring equipment, smart drinking devices, or information service platforms are not the same. Therefore, future research can employ panel data or follow-up surveys to reveal the long-term dynamic impacts of livelihood capital from a temporal perspective while also analyzing its mechanisms in diversified digital contexts from the perspective of technological differences. (3) Although this study examines the moderating effects of technical training and network quality, it does not further combine case studies to analyze the underlying logic and mechanisms. Thus, future research should incorporate typical cases to more deeply investigate the indirect chains and mechanisms through which training and network conditions operate in the capital–behavior relationship, thus providing more precise references for policy practice.

6 Conclusions and policy implications

6.1 Conclusion

Based on survey data from 356 herdsmen across seven banners in Inner Mongolia, this study uses the entropy weight method to evaluate the livelihood capital of herdsmen. It then uses IV-probit and moderating effect models to analyze the relationship between livelihood capital and herdsmen’s usage of digital technology. The main findings are as follows: (1) In terms of overall livelihood capital, the mean score of herdsmen’s livelihood capital is 0.294, with a standard deviation of 0.177, indicating generally low capital levels and considerable variation across households. In terms of individual capital dimensions, herdsmen’s livelihood capital shows an overall pattern of scarcity, with relatively low scores across all dimensions. Unlike the idealized pentagonal structure of balanced capital endowment, the distribution of capital is uneven. The ranking from highest to lowest mean value is as follows: social capital > financial capital > human capital > natural capital > physical capital. (2) Livelihood capital positively affects herdsmen’s use of digital technology at the 1% significance level; this result is further validated by robustness tests. This suggests that enhancing herdsmen’s livelihood capital can promote their use of digital technologies. (3) The moderation analysis reveals that both technical training and network quality significantly strengthen the positive impact of livelihood capital on herdsmen’s use of digital technology. Specifically, the interaction between livelihood capital and technical training is significant at the 5% level, while the interaction with network quality is significant at the 1% level. These findings suggest that when herdsmen have higher levels of livelihood capital, supported by training related to farming techniques and stable network infrastructure, it increases their use of digital technologies. (4) Heterogeneity analysis shows that the impact of livelihood capital on the use of digital technology is more pronounced among herdsmen who have a large grassland area, more education and a greater scale of livestock.

6.2 Policy implications

Based on this study’s findings, the following three policy recommendations are made:

1. Implement targeted policies to enhance livelihood capital and strengthen herdsmen’s capacity for digital technology adoption. The central government should take the lead in advancing top-level policy design and resource allocation, while local pastoral governments and grassland management authorities should be responsible for implementing measures related to natural capital, such as promoting rangeland tenure registration, optimizing the grass–livestock balance system, improving ecological compensation mechanisms, and exploring rangeland resource-sharing arrangements. Fiscal authorities, financial regulators, and agricultural and rural departments should jointly introduce subsidy programs and preferential loan policies, guiding financial institutions and agricultural insurers to strengthen inclusive financial systems and insurance mechanisms to ease the financial burdens of herdsmen during the phases of digital equipment upgrading and large-scale application. Education and human resource departments should focus on enhancing basic education and vocational skills training in pastoral areas, encourage young herders to return home and engage in entrepreneurial activities, and thereby strengthen human capital. Meanwhile, cooperatives and herders’ associations should assume primary responsibility for fostering social capital by building networks for information sharing, technical collaboration, and demonstration-based diffusion, ultimately promoting collective learning and imitation.

2. Improve the technical training system to enhance herdsmen’s capacity for digital technology use. Agricultural and rural authorities should collaborate with research institutes and universities to establish a multi-tiered digital technology training system, thereby institutionalizing a model of “centralized instruction—targeted assistance—remote services.” Local governments are expected to mobilize extension agencies and township-level technicians to deliver regular on-site training in pastoral areas, focusing on equipment operation, troubleshooting, and software application. Digital equipment manufacturers and third-party technology service providers should assume responsibility for socialized services by forming specialized technical teams to provide “household-based, one-to-one” guidance. In addition, they should build “online Q&A and operational demonstration” platforms via WeChat, short-video applications, and other digital channels to ensure the feasibility of remote and continuous technical support. Through such multi-stakeholder collaboration, not only can the cognitive and operational barriers to adoption be reduced, but targeted support for resource-constrained households can also be effectively strengthened.

3. Optimize network infrastructure strategies and prioritize accessibility in key areas. Information technology regulators and telecommunications operators should serve as the primary actors in network infrastructure development, adopting a co-construction and sharing model that combines government financial support with market-based enterprise operations. Priority should be given to ensuring network stability in herder settlements, cooperative offices, and centralized grazing areas. For remote rangeland regions, local governments are encouraged to promote the application of emerging technologies—such as low-orbit satellite internet, portable base stations, and solar-powered signal boosters—to guarantee supplementary network coverage. Digital equipment manufacturers, in turn, should integrate adaptive features into product design, including “offline functionality, disconnection buffering, and local data recording,” to mitigate the impacts of short-term connectivity interruptions. Given the relatively high cost of satellite internet and related infrastructure, a phased implementation strategy supported by government subsidies and public–private partnerships could improve fiscal feasibility and cost-effectiveness. Pilot programs should be prioritized in areas with higher herder density and stronger demand for digital services, ensuring that resources are allocated efficiently and affordability is maintained. Through multi-stakeholder collaboration, it is possible to enhance both the service capacity and the long-term sustainability of digital infrastructure in pastoral areas while balancing cost considerations with coverage needs.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

Ethics statement

Ethical review and approval were not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent from the patients/participants or patients/participants’ legal guardian/next of kin was not required to participate in this study, in accordance with the national legislation and the institutional requirements.

Author contributions

MY: Writing – original draft, Methodology, Data curation, Software, Conceptualization, Writing – review & editing, Investigation. YW: Methodology, Writing – review & editing, Funding acquisition.

Funding

The author(s) declare that financial support was received for the research and/or publication of this article. This research was funded by the National Natural Science Foundation of China, grant number 72163024, and the Inner Mongolia Autonomous Region “Integrated Rural and Pastoral Area Development Innovation Team,” grant number NMGIRT2223.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The authors declare that no Gen AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Abdulai, A., and Huffman, W. (2014). The adoption and impact of soil and water conservation technology: an endogenous switching regression application. Land Econ. 90, 26–43. doi: 10.3368/le.90.1.26

Crossref Full Text | Google Scholar

Ahmed, N., Allison, E. H., and Muir, J. F. (2008). Using the sustainable livelihoods framework to identify constraints and opportunities to the development of freshwater prawn farming in Southwest Bangladesh. J. World Aquacult. Soc. 39, 598–611. doi: 10.1111/j.1749-7345.2008.00198.x

Crossref Full Text | Google Scholar

Aker, J. C. (2011). Dial “a” for agriculture: a review of information and communication technologies for agricultural extension in developing countries. Agric. Econ. 42, 631–647. doi: 10.1111/j.1574-0862.2011.00545.x

Crossref Full Text | Google Scholar

Allison, E. H., and Ellis, F. (2001). The livelihoods approach and management of small-scale fisheries. Mar. Policy 25, 377–388. doi: 10.1016/S0308-597X(01)00023-9

Crossref Full Text | Google Scholar

Ammann, J., Walter, A., and El Benni, N. (2022). Adoption and perception of farm management information systems by future Swiss farm managers–an online study. J. Rural. Stud. 89, 298–305. doi: 10.1016/j.jrurstud.2021.12.008

Crossref Full Text | Google Scholar

Animato, C. (1999). Sustainable livelihoods: Lessons from early experience. London: Department for International Development.

Google Scholar

Bai, X., Peng, G., and Hu, H. (2025). Digital capabilities and farmers’ smart agriculture technology adoption behavior. J. Agro-For. Econ. Manage. 24, 514–524. doi: 10.16195/j.cnki.cn36-1328/f.2025.04.54

Crossref Full Text | Google Scholar

Bandiera, O., and Rasul, I. (2006). Social networks and technology adoption in northern Mozambique. Econ. J. 116, 869–902. doi: 10.1111/j.1468-0297.2006.01115.x,

PubMed Abstract | Crossref Full Text | Google Scholar

Bhandari, P. B. (2013). Rural livelihood change? Household capital, community resources and livelihood transition. J. Rural. Stud. 32, 126–136. doi: 10.1016/j.jrurstud.2013.05.001,

PubMed Abstract | Crossref Full Text | Google Scholar

Bolfe, É. L., Jorge, L. C., Sanches, I. D. A., Luchiari Júnior, A., and Ramirez, A. R. (2020). Precision and digital agriculture: adoption of technologies and perception of Brazilian farmers. Agriculture 10:653. doi: 10.3390/agriculture10120653

Crossref Full Text | Google Scholar

Carney, D. (2003). Sustainable livelihoods approaches: Progress and possibilities for change. London: Department for International Development.

Google Scholar

Chen, Y., and Zhao, M. (2024). Influence and mechanism of capital endowment on farmers’ adoption of organic fertilizer substitution technology. J. Northwest A&F Univ. 24, 116–127. doi: 10.13968/j.cnki.1009-9107.2024.03.13

Crossref Full Text | Google Scholar

Clark, B., Jones, G., Kendall, H., Taylor, J., Cao, Y., and Chen, L. (2018). A proposed framework for accelerating technology trajectories in agriculture: a case study in China. Front. Agric. Sci. Eng. 5, 485–898. doi: 10.15302/J-FASE-2018244

Crossref Full Text | Google Scholar

Cui, L., and Wang, W. (2023). Factors affecting the adoption of digital technology by farmers in China: a systematic literature review. Sustainability 15:14824. doi: 10.3390/su152014824

Crossref Full Text | Google Scholar

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q. 13, 319–340. doi: 10.2307/249008

Crossref Full Text | Google Scholar

Eastwood, C., Klerkx, L., Ayre, M., and Dela Rue, B. (2019). Managing socio-ethical challenges in the development of smart farming: from a fragmented to a comprehensive approach for responsible research and innovation. J. Agric. Environ. Ethics 32, 741–768. doi: 10.1007/s10806-017-9704-5

Crossref Full Text | Google Scholar

Eastwood, C., Klerkx, L., and Nettle, R. (2017). Dynamics and distribution of public and private research and extension roles for technological innovation and diffusion: case studies of the implementation and adaptation of precision farming technologies. J. Rural. Stud. 49, 1–12. doi: 10.1016/j.jrurstud.2016.11.008

Crossref Full Text | Google Scholar

Engås, K. G., Raja, J. Z., and Neufang, I. F. (2023). Decoding technological frames: an exploratory study of access to and meaningful engagement with digital technologies in agriculture. Technol. Forecast. Soc. Change 190:122405. doi: 10.1016/j.techfore.2023.122405

Crossref Full Text | Google Scholar

Feder, G., Just, R. E., and Zilberman, D. (1985). Adoption of agricultural innovations in developing countries: a survey. Econ. Dev. Cult. Change 33, 255–298. doi: 10.1086/451461

Crossref Full Text | Google Scholar

Gabriel, A., and Gandorfer, M. (2023). Adoption of digital technologies in agriculture—an inventory in a European small-scale farming region. Precis. Agric. 24, 68–91. doi: 10.1007/s11119-022-09931-1

Crossref Full Text | Google Scholar

Goyal, A. (2010). Information, direct access to farmers, and rural market performance in Central India. Am. Econ. J. Appl. Econ. 2, 22–45. doi: 10.1257/app.2.3.22

Crossref Full Text | Google Scholar

Gozzi, N., Comini, N., and Perra, N. (2023). The adoption of non-pharmaceutical interventions and the role of digital infrastructure during the COVID-19 pandemic in Colombia, Ecuador, and El Salvador. EPJ Data Sci. 12:18. doi: 10.48550/arXiv.2202.12088

Crossref Full Text | Google Scholar

Groher, T., Heitkämper, K., and Umstätter, C. (2020). Digital technology adoption in livestock production with a special focus on ruminant farming. Animal 14, 2404–2413. doi: 10.1017/S1751731120001391,

PubMed Abstract | Crossref Full Text | Google Scholar

Guo, Y. (2025). Technology empowering high-quality agricultural development in China: internal mechanisms and implementation pathways. Agric. Econ. 9, 6–8. Available online at: https://kns.cnki.net/kcms2/article/abstract?v=ewTrO3D4JaPWnciQ_-gU8A5T8kQZqW4JxA-yTOK_DFiwUL-okGgxza7JvsldQihq_vnKqnBDHGx12uhVJRLDsBtOoD9DXRIKRgxMN38n0UK7RqDEBLNG0LyGPRqIHoLR69xa5R0X-W3xupYN4iT1ugfBHgOlqGIeC73fD3GEvNUgTalJxPyfcg==&uniplatform=NZKPT&language=CHS

Google Scholar

Hou, M., and Hao, X. (2023). How does digital technology promote high-quality agricultural development — the mediation effect based on factor flow and the regulating effect of industrial structure transformation. J. Henan Norm. Univ. (Philos. Soc. Sci.) 50, 21–28. doi: 10.16366/j.cnki.1000-2359.2023.06.04

Crossref Full Text | Google Scholar

Hua, Z., Lin, W., and Xu, N. (2022). Digital gap or digital dividends?———the impact of the use of digital technology on the income of rural low-income households. J. China Agric. Univ. 39, 133–154. doi: 10.13240/j.cnki.caujsse.2022.05.012

Crossref Full Text | Google Scholar

Huang, J. (2022). Facilitating farmer's income growth and common prosperity through accelerating rural economic transformation. Issues Agric. Econ. 43, 4–15. doi: 10.13246/j.cnki.iae.2022.07.009

Crossref Full Text | Google Scholar

Huang, J. K. (2024). New quality productive forces in agriculture: connotation and denotation,potential and challenges, and development ideas. China Rural Surv. 45, 19–34. doi: 10.20074/j.cnki.11-3586/f.2024.05.002

Crossref Full Text | Google Scholar

Huang, C. (2025). The internal logic,practical dilemma and breakthrough path of digital technology empowering rural revitalization. Reform 37, 55–64. Available online at: https://link.cnki.net/urlid/50.1012.f.20240806.1417.002

Google Scholar

Huang, Z., Li, J., and Wang, J. (2021). Optimization analysis of the combination effect of livelihood capitals on family income in poverty areas———based on 778 rural households’survey data in Shaanxi province, China. J Agrotech Econ 40, 79–91. doi: 10.13246/j.cnki.jae.2021.07.006

Crossref Full Text | Google Scholar

Huang, Z., Zhang, Y., Huang, Y., Xu, G., and Shang, S. (2022). Sales scale, non-pastoral employment and herders’ technology adoption: evidence from pastoral China. Land 11:1011. doi: 10.3390/land11071011

Crossref Full Text | Google Scholar

Jansen, H. G., Pender, J., Damon, A., Wielemaker, W., and Schipper, R. (2006). Policies for sustainable development in the hillside areas of Honduras: a quantitative livelihoods approach. Agric. Econ. 34, 141–153. doi: 10.1111/j.1574-0864.2006.00114.x

Crossref Full Text | Google Scholar

Jiang, I., Huang, H., and He, K. (2024a). The impacts of technical training and economic subsidies on farmers’diffusion behaviors of bio-pesticide application technology. China Rural Surv. 45, 163–184. doi: 10.20074/j.cnki.11-3586/f.2024.04.009

Crossref Full Text | Google Scholar

Jiang, L., Zhao, W., and Zhang, H. (2024b). Digital technology use, green technology adoption and agricultural productivity. Sci. Decis. Mak. 31, 1–14. doi: 10.3773/j.issn.1006-4885.2024.09.001

Crossref Full Text | Google Scholar

Kabbiri, R., Dora, M., Kumar, V., Elepu, G., and Gellynck, X. (2018). Mobile phone adoption in Agri-food sector: are farmers in sub-Saharan Africa connected? Technol. Forecast. Soc. Change 131, 253–261. doi: 10.1016/j.techfore.2017.12.010

Crossref Full Text | Google Scholar

Larson, J. A., Roberts, R. K., English, B. C., Larkin, S. L., Marra, M. C., Martin, S. W., et al. (2008). Factors affecting farmer adoption of remotely sensed imagery for precision management in cotton production. Precis. Agric. 9, 195–208. doi: 10.1007/s11119-008-9065-1

Crossref Full Text | Google Scholar

Li, C., Hong, Z., Xinye, L., and Xue, L. (2024). Research on the impact of digital divide on carbon emissions and its spatial effects. J Agrotech Econ 9, 22–36. doi: 10.13246/j.cnki.jae.2024.09.003

Crossref Full Text | Google Scholar

Li, N., and Li, Y. (2024). Digitally empowered agricultural modernization: intrinsic mecha⁃nism, effectiveness review, and practical direction. J. Huazhong Agric. Univ. 44, 32–43. doi: 10.13300/j.cnki.hnwkxb.2024.05.004

Crossref Full Text | Google Scholar

Liu, Y., Li, H., Lin, B., and Fu, X. (2022). Impact of rice cultivation technology training on fertilizer application amount of farmers: empirical analysis of control function approach based on 70 counties. J Agrotech Econ 41, 114–131. doi: 10.13246/j.cnki.jae.2022.10.003

Crossref Full Text | Google Scholar

Liu, L., Sun, W., Liu, L., and Jiang, Q. (2023). Study on the influence of heterogeneous agricultural training subjects on farmers' green agricultural technology adoption behavior. Chin. J. Agric. Resour. Reg. Plan. 44, 162–171. doi: 10.7621/cjarrp.1005-9121.20230416

Crossref Full Text | Google Scholar

Liu, B., and Yan, Z. (2025). Research on the impact of digital divide on carbon emissions and its spatial effects. World Surv. Res. 9, 17–27. doi: 10.13778/j.cnki.11-3705/c.2025.09.002

Crossref Full Text | Google Scholar

Miyajima, K. (2022). Mobile phone ownership and welfare: evidence from South Africa’s household survey. World Dev. 154:105863. doi: 10.1016/j.worlddev.2022.105863

Crossref Full Text | Google Scholar

Moreira, M. J. (2003). A conditional likelihood ratio test for structural models. Econometrica 71, 1027–1048. doi: 10.1111/1468-0262.00438

Crossref Full Text | Google Scholar

Nmadu, J., Aiyelitsoya, F., and Sallawu, H. (2013). Use of ICT in securing marketing information among small scale farmers in Niger state. Int. J. Trade Econ. Financ. 4, 66–72. doi: 10.7763/IJTEF.2013.V4.262

Crossref Full Text | Google Scholar

Odini, S. (2014). Access to and use of agricultural information by small scale women farmers in support of efforts to attain food security in Vihiga County, Kenya. J. Emerg. Trends Econ. Manag. Sci. 5, 80–86. doi: 10.10520/EJC152937

Crossref Full Text | Google Scholar

Peng, J., and Huang, S. (2024). Mechanisms of digital technology’s impact on grain yield from perspective of food security. J. South China Agric. Univ. (Soc. Sci.) 23, 77–92. doi: 10.7671/j.issn.1672-0202.2024.04.007

Crossref Full Text | Google Scholar

Peng, X., Yan, X., and Wang, H. (2024). Study on the effect of digital technology adoption and farmers’ cognition on fertilizer reduction and efficiency improvement behavior. Agriculture 14:973. doi: 10.3390/agriculture14070973

Crossref Full Text | Google Scholar

Putnam, R. D. (1994). Making democracy work: Civic traditions in modern Italy. Princeton: Princeton University Press.

Google Scholar

Quandt, A. (2018). Measuring livelihood resilience: the household livelihood resilience approach (HLRA). World Dev. 107, 253–263. doi: 10.1016/j.worlddev.2018.02.024

Crossref Full Text | Google Scholar

Ren, Z., Fu, Z., and Zhong, K. (2022). The influence of social capital on farmers’ green control technology adoption behavior. Front. Psychol. 13:1001442. doi: 10.3389/fpsyg.2022.1001442,

PubMed Abstract | Crossref Full Text | Google Scholar

Rogers, E. M., Singhal, A., and Quinlan, M. M. (2014). Diffusion of innovations. New York: Routledge.

Google Scholar

Rowe, E., Dawkins, M. S., and Gebhardt-Henrich, S. G. (2019). A systematic review of precision livestock farming in the poultry sector: is technology focussed on improving bird welfare? Animals 9:614. doi: 10.3390/ani9090614,

PubMed Abstract | Crossref Full Text | Google Scholar

Salim, R., Mamun, S. A. K., and Hassan, K. (2015). Role of communication technologies in broadacre agriculture in Australia: an empirical analysis using panel data. Aust. J. Agric. Resour. Econ. 60, 243–264. doi: 10.1111/1467-8489.12114

Crossref Full Text | Google Scholar

Schultz, T. W. (1960). Capital formation by education. J. Polit. Economy. 68, 571–583. doi: 10.1086/258393

Crossref Full Text | Google Scholar

Scoones, I. (1998). Sustainable rural livelihoods: a framework for analysis. IDS Work. Pap. 72, 1–22. doi: 10.1111/j.2040-0209.1998.tb00758.x

Crossref Full Text | Google Scholar

Shikuku, K. M. (2019). Information exchange links, knowledge exposure, and adoption of agricultural technologies in northern Uganda. World Dev. 115, 94–106. doi: 10.1016/j.worlddev.2018.11.012

Crossref Full Text | Google Scholar

Song, J., and Qi, C. (2013). Factors influencing farmers' agricultural technology adoption: a contractual perspective. China Rural Surv. 34, 52–94. doi: 10.20074/j.cnki.11-3586/f.2013.06.006

Crossref Full Text | Google Scholar

Su, Q., Xing, H., and Liu, C. (2023). Digital technology empowers farmers to increase income: action mechanism and influential effect. J. Shanxi Univ. Finance Econ. 45, 72–86. doi: 10.13781/j.cnki.1007-9556.2023.12.006

Crossref Full Text | Google Scholar

Tey, Y. S., and Brindal, M. (2012). Factors influencing the adoption of precision agricultural technologies: a review for policy implications. Precis. Agric. 13, 713–730. doi: 10.1007/s11119-012-9273-6

Crossref Full Text | Google Scholar

Van Campenhout, B., Spielman, D. J., and Lecoutere, E. (2021). Information and communication technologies to provide agricultural advice to smallholder farmers: experimental evidence from Uganda. Am. J. Agric. Econ. 103, 317–337. doi: 10.1002/ajae.12089

Crossref Full Text | Google Scholar

Verma, P., and Sinha, N. (2018). Integrating perceived economic wellbeing to technology acceptance model: the case of mobile based agricultural extension service. Technol. Forecast. Soc. Change 126, 207–216. doi: 10.1016/j.techfore.2017.08.013

Crossref Full Text | Google Scholar

Wang, H., and Qu, L. (2024). Development process, application scenarios, and optimization pathways of digital agriculture: a discussion based on digital agriculture application scenarios. J. Nanjing Agric. Univ. (Soc. Sci.) 24, 146–158. doi: 10.19714/j.cnki.1671-7465.2024.0081

Crossref Full Text | Google Scholar

Wang, W., Wang, M., Zhang, Y., and Wang, F. (2023). Research progress of internet and digital rural in China. World Reg. Stud. 32, 70–81. doi: 10.3969/j.issn.1004-9479.2023.02.2021147

Crossref Full Text | Google Scholar

Wei, J., Zheng, S., and Zhang, Q. (2024). Digital technology use, green cognition and adoption of green control technologies by farmers—taking pear farmers in Shaanxi and Hebei provinces as an example. World Agric. 46, 99–112. doi: 10.13856/j.cn11-1097/s.2024.03.009

Crossref Full Text | Google Scholar

Wen, Z., Hou, J., and Zhang, L. (2005). A comparison of moderator and mediator and their applications. Acta Psychol. Sin. 50, 268–274.

Google Scholar

Wolfert, S., Ge, L., Verdouw, C., and Bogaardt, M.-J. (2017). Big data in smart farming–a review. Agric. Syst. 153, 69–80. doi: 10.1016/j.agsy.2017.01.023

Crossref Full Text | Google Scholar

Wu, Y., Su, R., Xu, L., Yang, Z., and Wang, M. (2017). The relationship between herders' livelihood capital and livelihood strategies: a case study of Xilinhot city and west Ujimqin banner, Inner Mongolia. J Agrotech Econ 36, 71–77. doi: 10.13246/j.cnki.jae.2017.07.007

Crossref Full Text | Google Scholar

Xiong, F., You, C., and Zhu, S. (2025). Effect of digital technology application on grain grower’s behavior of green production technology adoption. Chin. J. Agric. Resour. Reg. Plan. 46, 1–12. doi: 10.7621/cjarrp.1005-9121.20250106

Crossref Full Text | Google Scholar

Yan, Y., Chen, L., Zhou, Z., and Wei, Y. (2025). Digital financial inclusion and agricultural modernization development in China—a study based on the perspective of agricultural mechanization services. Humanit. Soc. Sci. Commun. 12, 1–11. doi: 10.1057/s41599-025-04821-z

Crossref Full Text | Google Scholar

Yang, S., and Kong, R. (2021). Does participation in digital financial markets improve rural households’ income? J. Huazhong Agric. Univ. (05), 180–190+200. doi: 10.13300/j.cnki.hnwkxb.2021.05.019

Crossref Full Text | Google Scholar

Yang, J., and Wang, Q. (2023). Digital technology and high-quality agricultural development———from the perspective of digital productivity. J. Shanxi Univ. Finance Econ. 45, 47–63. doi: 10.13781/j.cnki.1007-9556.2023.04.004

Crossref Full Text | Google Scholar

Yuan, B., and Chen, C. (2019). Evolution of a development model for fruit industry against background of rising labor cost: intensive or extensive adjustment? Sustainability 11:3864. doi: 10.3390/su11143864

Crossref Full Text | Google Scholar

Zhang, B., Luo, L., and Fu, X. (2025). Research on the mechanism and effect of digital literacy empowering farmers' common prosperity. Rural Econ. 10, 124–132. Available online at: https://kns.cnki.net/kcms2/article/abstract?v=ewTrO3D4JaOapgl8IzFzV2SSftZRLc87DCYh3QlV8LdD3ksW9lfvI1_TRMw-vdJdxdcGt9opYV0EjULWg1pmrj-7M9t4JglpCAjrMuiAllWkMhdyHnZ42n2qbrQilsCsvJ0_jpUdAmQKvW6ilVGvqx0HMUGJPvdwmEhGdGSvVqBYUFyvmHi0hQ==&uniplatform=NZKPT&language=CHS

Google Scholar

Zhang, T., Yan, T., He, K., and Zhang, J. (2017). Impact of capital endowment on peasants’ willingness to invest in green production: taking crop straw returning to the field as an example. China Popul. Resour. Environ. 27, 78–89. doi: 10.12062/cpre.20170422

Crossref Full Text | Google Scholar

Zhong, F., Ying, C., and Fan, D. (2022). Public service delivery and the livelihood adaptive capacity of farmers and herders: the mediating effect of livelihood capital. Land 11:1467. doi: 10.3390/land11091467

Crossref Full Text | Google Scholar

Zhou, Z., Zhang, Y., and Yan, Z. (2022). Will digital financial inclusion increase Chinese farmers’ willingness to adopt agricultural technology? Agriculture 12:1514. doi: 10.3390/agriculture12101514

Crossref Full Text | Google Scholar