Farmers' livelihood resilience, as an extension of livelihood resilience theory at the micro level, refers to the comprehensive ability of farmers to maintain, restore, or enhance their livelihood systems when facing multi-dimensional shocks from social, economic, and environmental sources. However, quantifying farmers' livelihood resilience faces numerous challenges. Accordingly, Smith and Frankenberger (2018) established a Farmers' livelihood resilience indicator system based on the mature livelihood sustainability framework, across three dimensions: absorptive capacity, adaptive capacity, and transformative capacity. However, this framework has not yet captured the psychological core of farmers' responses to risk. Cross-disciplinary research in psychology and livelihood studies suggests that adding a psychological dimension is not a mere extension of the indicator set, but a substantive theoretical expansion and deepening of the traditional three-dimensional framework. Integrating psychological capacity into the analysis aligns with the core principle of Sen (1999). Capability Approach, which emphasizes the equal importance of “subjective well-being” and “practical capabilities.” It also addresses a theoretical gap in the sustainable livelihoods framework, namely its insufficient attention to individual psychological empowerment. In doing so, it advances livelihood resilience theory from an analytical perspective focused on “external behavioral adaptation” toward one centered on an “internal–external synergistic drive.” Psychological capacity forms an intrinsic, bidirectional relationship with absorptive, adaptive, and transformative capacities. On the one hand, a positive psychological state serves as a crucial prerequisite for enhancing external livelihood capacities. Farmers often face constraints such as limited education and insufficient skills, compounded by long-term marginalization in the allocation of social resources, which frequently leads to feelings of inferiority and a lack of confidence in development (Fan and Cong, 2024). Psychological distress, such as anxiety and depression triggered by external risk shocks, further diminishes farmers' motivation for self-improvement (Malas and Tolsá, 2024), making it difficult for them to effectively mobilize livelihood capital and adopt adaptive strategies, ultimately constraining the exercise of absorptive and adaptive capacities. On the other hand, strengthening external livelihood capacities can reinforce psychological resilience. When farmers accumulate sufficient livelihood capital and strategic experience by absorbing risks and adapting to change, their expectations for the future and confidence in overcoming adversity increase, creating a virtuous cycle of “psychological empowerment—behavioral optimization—capacity enhancement—psychological reinforcement.”

Based on this, the present study moves beyond the limitations of the traditional three-dimensional framework by constructing a farmer livelihood resilience evaluation indicator system across four dimensions: “absorptive capacity, adaptive capacity, transformative capacity, and psychological capacity.” It also clarifies the theoretical boundaries and explanatory advantages of each dimension, detailed in Table 1, thereby enhancing the theoretical completeness and practical applicability of the farmer livelihood resilience evaluation system. (1) Absorptive capacity refers to the ability of farmers to absorb risk shocks, including the ability to minimize the impact and pressure of shocks before they occur and to recover quickly after they occur. It is the foundation for building livelihood resilience. This study selects natural capital, physical capital, and financial capital as measurement indicators. Natural capital refers to the natural resource endowments that farmers possess, such as land and water resources. In rural areas, it not only provides farmers with basic food and agricultural income but also serves as a safety net. The more arable land farmers own, the more they can achieve income diversification and risk management through scale operations. Physical capital refers to various materials or facilities used by farmers in production or daily life, excluding natural resources, such as housing conditions and household fixed assets. Financial capital primarily refers to cash that farmers can dispose of. The more financial capital farmers accumulate, the more they can convert it into other livelihood capital when facing risks, thereby mitigating adverse impacts. (2) Adaptive capacity refers to the ability of farmers to adapt to new situations and strategies through learning, organizational participation, and the adoption of diversified livelihood strategies. It is a key factor in building livelihood resilience. This study selected social capital, learning ability, livelihood diversification, and human capital as measurement indicators. Social capital refers to the social resources that farmers rely on for survival, such as influence, social networks, group membership, and trust relationships. Learning ability refers to the systematic integration process of farmers' external knowledge elements and technical means, and the ability to transform them into actions to improve livelihood levels. Livelihood diversification is the mainstream strategy for farmers to cope with uncertain environmental changes. Diversified livelihoods show that the stronger the farmer's ability to develop, the better it is to diversify risk. Human capital refers to the quantity and quality of household labor, which is crucial for enhancing adaptability and implementing diverse livelihood strategies. Generally, households with limited human capital have fewer livelihood strategy options, making it difficult for them to participate in high-income activities and resulting in weaker risk adaptation capabilities. (3) Transformative capability refers to the process by which farmers adjust and transform their livelihoods, emphasizing proactive livelihood adjustments and transformations undertaken by households to reduce vulnerability. It is also a key factor in building livelihood resilience. When external risks exceed the system's absorptive and adaptive capacity, transformative capability emphasizes changing the system structure from the root and adopting new coping strategies, thereby recovering from a vulnerable environment. This study selects non-agricultural work experience and potential development opportunities as measurement indicators. (4) Psychological capacity refers to psychological factors such as self-efficacy, hope, and optimism among individual farmers. These factors are crucial for farmers to maintain a positive attitude and motivation when facing adversity, and they are also an important component of livelihood resilience. This study selected farmers' confidence in wealth accumulation, life satisfaction, and future life confidence as measurement indicators.

The digital transformation of agricultural production systems embeds digital technology into the agricultural sector, enabling farmers to carry out related farming activities with the help of digital tools during the agricultural production process. The use of digital tools is constrained by the underlying condition of access to digital technologies (Zhu and Yun, 2025). Existing studies define the spatial heterogeneity of information and communication technology (ICT) infrastructure as the “first-level digital divide” and the gradient of technology application efficiency among agricultural practitioners as the “second-level digital divide” (Li and Ke, 2021). Drawing on this concept, this study defines farmers' use of 4G/5G mobile phones, tablets, laptops, and other devices as “digital technology access.” It describes the information exchange and value realization through technical media in agricultural operations as “digital technology functional use” (Wei et al., 2024). At the same time, drawing on the research of Song et al. (2023), this study reflects the application mechanism from four levels: information retrieval, information communication, digital marketing, and element acquisition. First, information retrieval. In traditional agricultural production, farmers primarily rely on experiential judgment and word-of-mouth from neighbors to obtain information. The lag in information acquisition often leads to production decisions that are out of sync with actual conditions, significantly increasing the risk of decision-making errors. The advent of digital technology has broken this limitation (Musajan et al., 2024). Through agricultural information platforms, government service websites, and other channels, farmers can obtain real-time, accurate information on weather forecasts, pest control guidelines, and the latest government policies. The acquisition of this real-time information enables farmers to adjust their production plans in advance, avoid risks from natural disasters or market fluctuations, thereby improving the scientific basis of their production decisions and consolidating the risk-resilience foundation of their livelihood systems (Ndimbo et al., 2024). Second, information communication. Digital platforms such as WeChat groups and live-streaming classes have broken the spatial limitations between farmers, building cross-regional technical exchange networks. Farmers can learn advanced planting methods in real-time through video live streaming, exchange pest control knowledge in professional communities, and participate in online training courses led by agricultural experts. These interactions not only accelerate the diffusion of green production technologies but also form a “experience-data-practice” feedback loop (Yang et al., 2024). More importantly, when faced with sudden disasters, farmers can quickly share emergency measures through communication tools, enhancing their adaptive capacity to cope with unexpected crises. Third, digital marketing. Before the rise of digital marketing, the distribution process for traditional agricultural products was lengthy and complex. With the widespread application of digital technologies such as e-commerce platforms and live-streaming sales, farmers can communicate directly with consumers through online platforms (Kassie et al., 2015), thereby reducing the intermediate circulation links of agricultural products, lowering transaction costs in the process of agricultural product sales, and consequently increasing economic income (Qiu et al., 2024; Xu et al., 2024). In addition, farmers can adjust their production techniques and product structures in response to consumer needs and feedback, thereby enhancing product value-added and market competitiveness, leading to further income improvement. Increased income not only relieves the economic pressure on rural economically vulnerable groups but also provides support for them to accumulate capital, optimize living environments, purchase agricultural production tools and houses, further increasing the monetary, tangible, and ecological capital of this group, and improving their ability to absorb external sudden shocks (Lin and Xie, 2025). Fourth, element acquisition. In the past, when purchasing agricultural production factors like fertilizers and pesticides from township dealers, farmers often faced problems such as counterfeit fertilizers and pesticides, as well as seasonal price hikes. Nowadays, through e-commerce platforms, farmers can compare the prices and quality of products from multiple suppliers, reduce costs through centralized procurement, and ensure the authenticity of agricultural inputs through methods such as scanning codes (Addis et al., 2021). Additionally, services like soil testing data provided by e-commerce platforms allow farmers to customize fertilizer ratios based on actual soil conditions, further improving production efficiency, increasing farmers' income, and enhancing their ability to absorb unknown risks. Furthermore, when external agricultural input supplies are disrupted, farmers can quickly identify alternative suppliers using platform-stored data, ensuring uninterrupted production, thereby enhancing their transformative capacity against unexpected shocks. Based on the above analysis, the article proposes the following hypothesis:

Hypothesis 1: Digital technology use has a positive impact on farmers' livelihood resilience.

Hypothesis 1.1: Access to digital technology has a positive impact on farmers' livelihood resilience.

Hypothesis 1.2: The functional use of digital technology has a positive impact on farmers' livelihood resilience.

The theory of innovation diffusion suggests that the adoption of new technologies essentially represents farmers' re-optimization of production factors and livelihood strategies based on expected returns (Rogers, 2003). As a factor-augmenting technology, agricultural green technology enhances farmers' livelihood resilience primarily through two pathways. First, adopting agricultural green technologies can reduce production costs and resource consumption, increase farmers' income, and improve their ability to absorb unknown risks. For instance, soil testing and formulated fertilization technology, as a precision fertilization technique, can prioritize measuring soil nutrient content and then apply fertilizers precisely. This helps mitigate agricultural pollution, prevents soil contamination from indiscriminate fertilization, conserves resources, improves fertilizer use efficiency, while increasing agricultural yields and boosting farmers' income (Wen and Ma, 2024). Second, the adoption of agricultural green technology can reduce labor time in agriculture and provide farmers with more opportunities to work outside. For example, unmanned plant protection technology and integrated water and fertilizer technology. Unmanned plant protection technology is highly efficient, using large drones for intelligent spraying, which significantly increases application speed and improves operational efficiency. Furthermore, integrated water and fertilizer technology combines modern information technology to change traditional production models. It is a new type of irrigation method that, as a form of water-saving irrigation, delivers synthetic fertilizers to plant roots through pipelines, with fully mechanical and intelligent operation throughout the process. This green, environmentally friendly, and efficient production model significantly reduces labor time (Yang et al., 2024). With the reduction in labor time, farmers can allocate their time more flexibly, engage in off-farm work, and thereby enhance their ability to adapt to unforeseen shocks. Based on this, this study proposes the following hypothesis:

Hypothesis 2: Adoption of agricultural green technology has a positive impact on farmers' livelihood resilience.

Agricultural green production technology refers to farming techniques and management methods with environmental protection and sustainability as core principles. The core goal is to achieve holistic optimization of ecological, social, and economic values, specifically manifested through minimizing environmental pollution and hazards, maximizing resource utilization efficiency, and optimally satisfying human material and spiritual needs (Zhu et al., 2025). From the perspective of technological complementarity theory, the large-scale deployment and effective application of agricultural green technologies depend on precise data support and intelligent management. Digital technologies provide irreplaceable advantages in core domains such as real-time data collection, multidimensional information analysis, and dynamic decision support. This creates a highly complementary “demand–support” relationship between the two in functional terms (Markard and Hoffmann, 2016), thereby forming the fundamental premise for technological synergy. This technological complementarity is manifested not only at the level of functional compatibility but also aligns with the core logic of General-Purpose Technology (GPT) theory. Specifically, as a typical GPT, digital technology requires concrete industrial contexts and specific demands to realize its enabling value. The promotion and application of agricultural green technologies provide precisely such tangible implementation scenarios and well-defined demand orientations (Lipsey et al., 2005), driving the transformation of digital technology from a “general-purpose tool” into an “agriculture-specific empowering instrument,” thereby enhancing the relevance and efficiency of technological empowerment and realizing its productive value (Li et al., 2024). The theory of technology-institution coevolution further indicates that the reason digital technology and agricultural green technology can form efficient synergistic effects in the Chinese rural context lies in the dynamic adaptation between technological innovation and rural grassroots institutions, as well as the institutional environment (North, 1990). Institutional improvements and organizational optimization in rural China—in domains such as digital infrastructure, agricultural technology extension systems, and digital financial services —have provided crucial institutional support and service guarantees for the co-implementation of these technologies. This effectively addresses the “last-mile” problem in rural technology adoption, thereby facilitating the full release of their synergistic potential in rural settings. This study defines the “synergistic effect” as the dynamic interaction between digital and agricultural green technologies. Rooted in functional complementarity and technological integration, this synergy evolves through continuous adaptation to rural institutional environments and organizational systems. Its core characteristics are manifested in the synergistic amplification of technology-enabled outcomes and the dual enhancement of agricultural productivity and resilience. This conceptualization distinguishes it from the independent effects of single-technology applications and avoids the overgeneralized use of the “synergy” concept in academic research.

The deep integration of these two technologies generates synergies that go beyond simple juxtaposition. By optimizing compatibility and coordination within the technological system, it reshapes agricultural risk prevention and control mechanisms, thereby substantially enhancing farmers' capacity to withstand both natural and market risks. Furthermore, adopting green agricultural technologies gives farmers advantages in resource allocation. With the help of new tools such as digital finance, farmers can access more financial support and use these funds to purchase efficient production tools and green technologies, rather than being limited to traditional production materials (Gao et al., 2022). This optimized resource allocation not only improves the input-output ratio of digital technology but also amplifies its positive impact on livelihood resilience. Meanwhile, the synergistic use of green technology and digital technology significantly shortens agricultural labor time, providing farmers with more off-farm employment opportunities. Through automated and intelligent production methods, farmers can be freed from traditional labor, allowing them to allocate more time to non-agricultural employment or other high-yield activities, thereby enhancing their ability to cope with unforeseen shocks (Wang et al., 2024). Based on this, the study proposes the following hypotheses:

Hypothesis 3: The synergistic effect between digital technology use and agricultural green technology adoption has a positive impact on farmers' livelihood resilience.

Income diversification is the central mediating mechanism connecting the synergistic use of digital and green technologies to farmers' livelihood resilience. It systematically converts the economic potential arising from this technological synergy into concrete enhancements across multiple dimensions of resilience in farmers' livelihood systems. This transmission mechanism corresponds to the central tenet of technological complementarity theory, which posits that synergies between technologies enhance production efficiency, thereby improving livelihood quality (Bresnahan and Trajtenberg, 1995). It is also consistent with the logic of technology–institution coevolution theory. This theory contends that technology's external empowerment must be channeled through suitable mediating carriers to activate the endogenous dynamism of livelihood systems, facilitating the transition from technology adoption to resilience building (Geels, 2002). First, by taking advantage of the low interdependence among different income sources to diversify risk, farmers enhance their absorptive capacity, allowing them to better withstand short-term shocks such as natural disasters and market fluctuations. Second, by capitalizing on the resource slack and strategic flexibility created by diversified income streams, farmers enhance their adaptive capacity to cope with medium- and long-term trends, including climate change and policy adjustments. Third, by promoting the accumulation of physical, human, and social capital through income growth and expanding livelihood networks, farmers are empowered to break path dependence and enhance their transformative capacity for achieving modal transition. Fourth, a stable and diversified income structure strengthens farmers' sense of economic security and control over their livelihoods, thereby enhancing their psychological capacity to cope with uncertainty. Based on this, the study proposes the following hypotheses:

Hypothesis 4: Income diversification mediates the positive relationship between the synergistic interaction of digital technology use and agricultural green technology adoption and farmers' livelihood resilience.

Based on the above theoretical analysis, Figure 1 presents the analytical framework used in this study.

The micro-level data used in this study were obtained from the China Rural Revitalization Survey (CRRS) conducted by the Institute of Rural Development, Chinese Academy of Social Sciences, during August-September 2020. The survey collected data from 3,833 sample households covering 50 counties (cities, districts), 156 townships, and 300 administrative villages across 10 provinces in China. To ensure sample representativeness, the survey first randomly selected 10 provinces from the eastern, central, and western regions. These 10 provinces are Guangdong, Zhejiang, Shandong, Anhui, Henan, Heilongjiang, Guizhou, Sichuan, Shaanxi, and Ningxia Hui Autonomous Region. Then, county-level units within each province were stratified into five tiers based on economic indicators, with 5counties (cities, districts) randomly selected from each tier following geographic distribution balance principles. Next, three townships were randomly selected from each county (cities, districts) according to economic development gradients. Within each selected township, one economically developed village and one less-developed village were chosen. Finally, 12–14 farm households were sampled per village using equal-distance sampling. The survey data were divided into three levels: individual, household, and village. Based on research objectives and following principles of data completeness and accessibility, the data required for this study were selected. The data cleansing process methodically eliminated questionnaires with missing key information, abnormal data distribution, logical inconsistencies, and non-agricultural business tendencies. At the same time, considering insufficient valid sample sizes from Zhejiang and Guangdong provinces, where economic structures have entered a highly modernized stage, with agriculture accounting for a low proportion of total social output (Sun et al., 2024). Based on these factors, samples from these two provinces were excluded during sample selection, yielding a final sample of 615 farming households.

The impacts of digital technology use and agricultural green technology adoption on farmers' livelihood resilience are displayed in Table 4 columns (1) and (2), respectively. The results show that the impacts of digital technology use and the adoption of agriculture green technology on farmers' livelihood resilience are significantly positive at the 1% and 5% statistical levels, respectively. Column (3) simultaneously considers the impact of both factors on farmers' livelihood resilience. The regression coefficient, which is still significantly positive and demonstrates the positive impacts of digital technology use (β = 0.123, P < 0.01) and green agricultural technology adoption (β = 0.011, P < 0.05) on farmers' livelihood resilience, verifies hypotheses 1 and 2. The possible reasons for this result can be explained by the four traits that are used to measure farmers' livelihood resilience. First, absorptive capacity. Digital technology has constructed a “physical-economic-social” three-dimensional buffer system from the three dimensions of disaster warning, economic capital accumulation, and social networks, enhancing the risk-bearing threshold of the agricultural production and operation system and improving its risk absorption capacity. Second, adaptive capacity. The use of digital technology helps break down information barriers between urban and rural areas, enabling farmers to obtain key information such as market prices, policies and regulations, and technical standards in real time. This not only enhances farmers' ability to seize investment opportunities and avoid market risks, but also strengthens the foresight of their economic decision-making. Therefore, farmers have stronger adaptive adjustment capabilities when facing unknown shocks. Third, transformative capability. The industry convergence trend driven by digital technologies has broken the boundaries of traditional agriculture, promoting deep integration between agricultural production and modern service industries such as e-commerce and leisure tourism, creating a large number of job opportunities, enriching the income-generating channels for rural families, and thereby strengthening farmers' transformative capacity to cope with unknown risks. Fourth, psychological capability. Relying on a multi-dimensional information perception network, farmers can timely obtain market dynamics and disaster warning information, thus significantly improving their risk identification ability and perception accuracy. This significantly reduces the anxiety index of farmers when facing uncertainties. Meanwhile, the cognitive loop of “risk analysis—strategy evaluation—action selection” constructed by digital technology encourages farmers to shift from passively enduring risks to actively building defense systems, providing them with a solid psychological guarantee for responding to various risks.

Table 5, columns (4)–(8), shows the estimation results of the adoption of agricultural green technology when combined with digital technology access and use of digital technology functions. The results show that both access to and the functional use of digital technology significantly improve farmers' livelihood resilience, verifying hypotheses 1.1 and 1.2. It is important to note that while farmers' livelihood resilience is significantly impacted by both access to (β = 0.019, P < 0.01) and use of digital technology, digital marketing (β = 0.050, P < 0.01) has the most obvious facilitating effect among these uses. In contrast, information retrieval (β = 0.029, P < 0.01) has a relatively smaller effect. This could be because digital marketing, with its efficient transmission and acquisition of information about agricultural products and offers favorable market matching mechanisms, can effectively expand sales channels for agricultural products, lower distribution costs, and improve the market competitiveness of agricultural products—all of which significantly increase farmers' livelihood resilience. The information farmers retrieve in the current information-explosion period might not be in line with their actual demands for productivity and life. Additionally, some information may be inaccurate, timely, or of limited use, which would make it hard to support farmers in making decisions. The enhanced benefit of information retrieval on farmers' livelihood resilience has not yet fully materialized because it is challenging to translate this information into actionable livelihood enhancement strategies.

Combining digital technology use, agricultural green technology adoption, and their synergistic effects into a single model results in a significantly positive synergistic effect on farmers' livelihood resilience at the 1% level, verifying hypothesis 3. This is demonstrated in Table 6, column (9).

Considering that different digital technology functions may have distinct functional characteristics. This study, based on a comprehensive examination of whether the synergistic effects between digital technology use and green agricultural technology adoption have exerted positive impacts on farmers' livelihood resilience, further analyzes the synergistic effects of adopting agricultural green technology and using four digital technology functions—information retrieval, information communication, digital marketing, and element acquisition—on the resilience of farmers' livelihoods. The synergistic impacts of information communication and digital marketing with the adoption of agricultural green technologies are considerably positive at the 5% and 1% statistical levels, respectively, according to the regression results in Table 6 columns (11) and (12). This suggests that agricultural green technology adoption and the synergistic effects of these two digital technology functions (information communication and digital marketing) significantly improve farmers' livelihood resilience. Some possible explanations for this include: information communication “lowers technical thresholds and stabilizes production expectations” to resolve the “implementation difficulty” issue of green technology, and digital marketing “transmits value signals and expands market boundaries” to address the “monetization difficulty” issue of green technology. Additionally, from the “production end” and “sales end,” the two create a closed-loop synergy with green technology, directly influencing the three main facets of farmers' livelihood resilience (stability, profitability, and risk response).

This study employs two methods to further test the robustness of the research findings: the entropy-weighted TOPSIS and the alternative models method. First, the method used to calculate the total score of farmers' livelihood resilience was altered. To eliminate estimation biases caused by different calculation methods, this study follows Çam and Kagizman (2023) methodology by recalculating farmers' livelihood resilience levels using the entropy-weighted TOPSIS method and then performing regression analysis on the updated data. The results obtained are consistent with the regression results previously mentioned, as shown in Table 11 columns (14)–(16). Second, this study follows Zhang et al. (2024) methodology and employs the Tobit model to examine the effects of “digital-green” dual technology use and its synergistic effects on farmers' livelihood resilience, given that the dependent variable's scores range from 0 to 1. The use of “digital-green” dual technologies and their synergies continue to have a significant positive influence on farmers' livelihood resilience, as shown by columns (17)–(19). The robustness of the regression results is verified by systematically combining the findings of the two robustness analysis methods mentioned above.

Building on the above analysis, this study employs income diversification as a mediating variable to investigate the mechanism by which the synergistic effect of digital technology use and agricultural green technology adoption influences farmers' livelihood resilience. The results are presented in Table 12. The results in Column (1) indicate that the synergy between digital technology use and agricultural green technology adoption exerts a significantly positive effect on income diversification (β = 0.044, p < 0.1). The results of column (2) show that the synergistic effect of digital technology use and agricultural green technology adoption is still statistically significant and beneficial for farmers' livelihood resilience even after adjusting for the mediating variable of income diversification (β = 0.065, p < 0.05). Furthermore, income diversification exerts a significant positive influence on farmers' livelihood resilience (β = 0.114, p < 0.01). Based on the preceding analysis, it can be concluded that the synergistic effect of digital technology use and agricultural green technology adoption indirectly enhances farmers' livelihood resilience by increasing income diversification. Thus, hypothesis 4 is supported.

Drawing on previous studies, this study discusses heterogeneity from the viewpoints of the age and degree of part-time employment of the farmers surveyed to better understand the heterogeneity of the impact of various farmer types' use of digital technology, adoption of agricultural green technology, and the synergistic effects of the two on their livelihood resilience.

This study first compares and analyzes the impacts of digital technology use, adoption of agricultural green technology, and the synergistic effects of both on the livelihood resilience of farmers in various age groups, taking into account the age of the farmers who were polled. Based on Telles et al. (2022) age grouping approach and the World Health Organization's definition of the elderly, this study divides farmers into two groups: younger and middle-aged (age ≤ 55) and elderly (age > 55). The cutoff age for this category is 55. As indicated in Table 13 columns (1) and (2), the employment of digital technology can significantly enhance the livelihood resilience of the middle-aged, young (β = 0.113, P < 0.01), and elderly groups (β = 0.140, P < 0.01), with the elderly group enhancing the most. Some of the possible explanations are: The role that elderly agricultural workers play in agricultural productivity is changing significantly as their computer literacy levels continue to improve. For instance, the use of smart agricultural equipment, environmental monitoring systems, and other technological tools has significantly reduced the demand for physically demanding work in agricultural production. This has allowed elderly workers to continue working despite their physical limitations. Second, elderly agricultural workers can now respond more composedly to uncertainties like market fluctuations and natural risks, which enhances their adaptability and stability in agricultural production. This is made possible by the deep integration of technological tools and production knowledge, which has significantly improved the efficacy and precision of production decisions.

Young and middle-aged farmers benefit greatly (β = 0.108, P < 0.01) from the synergistic effect of digital technology use and agricultural green technology adoption, while elderly farmers suffer significantly (β = −0.248, P < 0.05). Due to their greater capacity for learning, middle-aged and younger farmers may be able to use digital technology more effectively and realize the synergy of “digital technology reducing information costs of green technology + green technology increasing agricultural product value-added.” for instance, Young and middle-aged farmers can use live streaming e-commerce to promote “green planting” labels, sell agricultural products at premium rates, and turn them into a direct source of livelihood income. However, elderly farmers are less adept at using green and digital technologies, and may have “burdens outweighing income” due to operational errors or cost accumulation when combining them. For example, elderly farmers may not be able to access online information on green technology, which could lead to careless investment that depletes their livelihood resources instead of yielding the desired results.

Second, taking the part-time employment level of surveyed farmers as a grouping variable, we further examine the heterogeneous impacts of digital technology use, green technology adoption, and their synergistic effect on farmers' livelihood resilience across different part-time employment levels. Drawing on Andersson et al. (2003) classification of farmers' part-time employment levels, this study employs the share of non-agricultural income in total household income as a metric: farmers with non-agricultural income accounting for more than 50% of total household income are categorized into the high part-time group, while those with a share of 50% or less are grouped into the low part-time group. Columns (3) and (4) of Table 13 reveal that digital technology use significantly enhances livelihood resilience for both high (β = 0.137, P < 0.01) and low (β = 0.113, P < 0.01) part-time groups, with a more prominent impact observed for the high part-time group. Plausible explanations include the following: High part-time farmers have greater access to non-agricultural employment opportunities, thereby diversifying their income sources and reducing reliance on a single agricultural production activity. This diversified employment structure strengthens their capacity to withstand risks associated with a single industry. Meanwhile, high part-time farmers are more adept at leveraging digital technology to optimize time allocation—they use digital tools to efficiently manage both agricultural production and non-agricultural work, achieving a dynamic balance between the two. This time management capability not only improves production efficiency but also enhances their flexibility in responding to sudden risks. Furthermore, high part-time farmers possess wider social networks; through these extensive personal connections, they can quickly access information on the application prospects and promotion status of new technologies, and take the lead in integrating these technologies into agricultural production or non-agricultural employment. This proactive adoption enhances their competitiveness and strengthens their adaptability and risk resistance in the market.

Digital technology use, agricultural green technology adoption, and the synergistic effect of these two factors have significantly enhanced the livelihood resilience of farmers in the high part-time group at the statistical levels of 1% and 5%, respectively. The potential reasons are as follows: high part-time farmers have stable non-agricultural income and stronger risk-bearing capacity, allowing them to cover the upfront investments needed to adopt green technologies. At the same time, digital technology helps them link green agricultural products to high-end markets, realizing a synergistic effect of “green technology enhancing quality + digital technology expanding sales channels. ”For example, high part-time farmers can utilize funds accumulated from non-agricultural work to purchase water-saving equipment, then promote the concept of “water-saving agriculture” via short video platforms to secure premium orders.

This study is based on a sample of 615 rural households from the 2020 CRRS. By constructing a four-dimensional livelihood resilience indicator system and using Ordinary Least Squares (OLS), it reveals the mechanisms of digital technology use, agricultural green technology adoption, and their synergistic effects on farmers' livelihood resilience, and further explores their impact heterogeneity. On this basis, the PSM method is used to alleviate the self-selection bias in farmers' behaviors regarding digital technology use and agricultural green technology adoption, and a counterfactual framework is established to further empirically analyze the impact effects of digital technology use and agricultural green technology on farmers' livelihood resilience leading to the following research conclusions.

Based on empirical research findings, to fully leverage the use of digital technology, the adoption of agricultural green technology, and their synergistic effects to enhance farmers' livelihood resilience, this study proposes the following policy recommendations.