This study employs an empirical analysis to investigate the impact of AGI development on rural residents’ health. To address potential endogeneity arising from reverse causality and omitted variable problems, we utilize Two-Stage Least Squares–Instrumental Variable (2SLS-IV) method. The baseline regression model is specified as follows:

in Equation 1, Health it denotes the physical and mental health status of individual 𝑖 at time 𝑡, measured across two dimensions: physical health and mental health; AGI it denotes the number of AGIs registered in the county where individual 𝑖 resides at time 𝑡, Controls it refers to a set of control variables; β 0 and β 2 represent the constant term and the regression coefficients of the control variables, respectively; β 1 is the core coefficient of interest in this study; and ε it denotes the random error term.

The individual-level data used in this study are sourced from the China Family Panel Studies (CFPS)³ conducted between 2010 and 2020. This longitudinal survey captures comprehensive micro-level information on China’s social, economic, demographic, educational, and health conditions. The baseline survey covers 25 provinces, municipalities, and autonomous regions, encompassing all members of the sampled households. County-level data are obtained from the China County Statistical Yearbooks for the corresponding years, with missing values imputed using linear interpolation. Data on national AGI development are derived from the China Academy for Rural Development-Qiyan China Agri-research Database (CCAD), which is constructed based on the “Measures for the Administration of Geographical Indications of Agricultural Products”⁴ and data from the Ministry of Agriculture’s registration, review, and evaluation processes. The CCAD includes detailed information on certified products, their regions of origin, and product categories from 2008 to the present.⁵

The data integration process involved three main steps. First, we merged county-level statistics, AGI data, and CFPS datasets (including individual, family member, proxy, and household economic modules), retaining only samples with agricultural household registration (Hukou). Second, micro-level individual data were matched with the AGI data and covariates based on administrative codes. Third, after excluding observations with missing key variables, we obtained a final pooled cross-sectional dataset comprising 21,415 observations.

Descriptive statistics for the full sample are presented in Table 1.

The baseline regression results are presented in Table 2. First, regarding the validity of the instrumental variable, the weak instrument test results reported in Table 3 show an F-statistic of 548.639, substantially exceeding the critical value of 10. This allows us to reject the null hypothesis of weak instruments, confirming the relevance of our instrument.

In terms of health outcomes, the estimated coefficient for self-rated health is 0.050 (p < 0.10), indicating a positive impact on physical health. Conversely, the coefficient for moderate-to-severe depression is 0.024 (p < 0.05), implying an increased risk of mental health issues. These findings reveal that AGI development exerts a “dual health effect” on rural residents: simultaneously generating “development dividends” for physical health while incurring “unintended costs” for mental well-being.

To ensure the robustness of the baseline findings with respect to variable construction and measurement, we employ alternative health indicators with distinct measurement dimensions and properties. First, regarding physical health, the “self-rated health” variable used in the baseline model serves as a subjective composite measure. For robustness, we introduce an objective indicator: “whether hospitalized for illness in the past 12 months”. This proxy captures the severity of physical health conditions and healthcare utilization objectively. Second, for mental health, we use the continuous CES-D depression score instead of the binary depression variable. These alternative indicators capture different dimensions of health from both subjective and objective perspectives, thereby effectively identifying whether the baseline results are driven by measurement biases.

As shown in Table 4, in the instrumental variable estimations, county-level AGI development significantly reduces the probability of hospitalization among rural residents (coefficient is negative and significant), while also showing a significant positive effect on continuous depression scores. This is consistent with the baseline results. This finding not only corroborates the dual health effects of AGI—simultaneously promoting physical health and increasing psychological risk—but also indicates that the core conclusions remain robust across different health measurement approaches, strengthening the credibility and generalizability of the causal inferences. Notably, the decrease in hospitalization rates—an indicator of relatively severe adverse health events—provides objective evidence of the positive role of AGI in improving physical health conditions. Simultaneously, the increase in depression scores reinforces the baseline finding regarding potential risks to mental health. Together, these results support the overall argument of this study regarding the complex health effects of AGI.

In the mechanism analysis, we adopt a “two-step approach” instead of the traditional stepwise regression model for mediating effects. This decision is driven by methodological rigor: Jiang (2022) argues that within a causal inference framework, if the mediator is an “ex-post variable” influenced by the treatment, traditional stepwise regression is prone to estimation bias due to endogeneity and confounding. Therefore, rather than relying on stepwise testing, we emphasize the identification of core causal chains. Specifically, we first theoretically elucidate the links between mechanisms and health outcomes, and then empirically verify the causal effect of AGI development on each mechanism variable. This method enhances theoretical coherence while avoiding the masking effects potentially arising from interactions among parallel channels.

Although the theoretical framework is primarily grounded in individual-level health capital accumulation, our empirical findings suggest that the health effects of AGI extend beyond the individual, forming a multi-level transmission mechanism characterized by “individual health improvement—family function enhancement—intergenerational health spillover.” Specifically, by generating localized and stable employment opportunities, AGI policies have encouraged the return of many migrant workers—particularly fathers— to their hometowns for work. This fundamentally reshapes the intergenerational health production model within rural families.

Using CFPS data on the duration of co-residence between children and their parents over the past year, we examine this spillover effect. The regression results in Table 11 show that AGI development significantly increases the duration of children’s co-residence with their fathers, while it has no significant effect on co-residence duration with mothers. This asymmetry aligns with our earlier heterogeneity analysis (Section 4.3.1), which suggested that AGI development primarily attracts male migrant workers to return home for local employment.

This structural shift implies a systematic improvement in parental companionship for rural children. The return of fathers transforms parent–child contact from episodic, fragmented “discrete engagement” into stable “integrated co-living.” This not only better addresses children’s emotional needs (Savahl et al., 2019) but also creates a supportive home environment, serving as a significant form of informal health capital (Huang and Hu, 2019). Mechanistically, this process is driven by AGI-induced agglomeration effects, which encourage regional labor concentration and reduce the necessity for outbound migration. The resulting consolidation of family-based living arrangements strengthens household support systems for the current workforce while significantly increasing time and resource investments in child care and education. Ultimately, AGI development extends its positive impact beyond the current generation, laying a stronger foundation for the healthy development and human capital accumulation of the next generation.

This study situates itself within the macro-context of agricultural product branding as a key policy tool for agricultural modernization and rural revitalization. Historically, both policy logic and scholarly discourse surrounding agricultural brand development—represented by AGIs—have primarily focused on “making the pie bigger”. Existing literature confirms that AGIs promote economic growth through brand premiums (Kizos and Vakoufaris, 2011; Van der Ploeg et al., 2000) and enhance ecological sustainability through standardized production (Zhang and Liu, 2025). However, these studies implicitly assume that the resulting social benefits are linear, universal, and positive. They seldom delve into how development gains translate into tangible wellbeing for different groups through actual distribution—specifically—a core dimension of human capital. They seldom delve into how development gains translate into tangible wellbeing for different groups through actual distribution, particularly regarding health—a core dimension of human capital.

Building upon existing research, this study constructs an integrated framework encompassing “economic development–ecological improvement–physical and mental health”. Our findings offer three key theoretical contributions:

First, we elucidate the transmission pathways from AGI development to health outcomes. Beyond confirming economic and ecological benefits, we identify the pivotal bridging role of new agricultural business entities. AGI development significantly promotes the growth of family farms and cooperatives, which leverage advanced production techniques and professional management to achieve “comprehensive outreach” to smallholders. By connecting with these entities, farmers not only share in economic dividends but also gain exposure to safer production methods and scientific health concepts. This implies that health effects are transmitted not just through abstract market mechanisms, but directly through organized, scaled production systems.

Second, we challenge the “optimistic presumption” of AGI’s social neutrality. For the first time, we systematically identify the potential mental health risks AGI may induce—such as anxiety stemming from relative deprivation—and reveal significant heterogeneity across gender, education, and region. These findings warn that AGI development is not inherently inclusive; without careful management, it may reinforce existing socioeconomic inequalities.

Third, we extend the analytical lens to the intergenerational level. By encouraging family reunification, AGI development generates positive spillover effects on the health of rural children. This shifts the benefit assessment from a static cross-sectional perspective to a dynamic intertemporal one, underscoring the profound implications of AGI policies for intergenerational human capital accumulation.

Our findings suggest that while continuing to rely on AGI branding to “grow the pie,” policy practice must attach equal importance to “dividing the pie well.” To ensure that brand-based dividends are translated into comprehensive, balanced, and sustainable well-being, we propose the following recommendations:

First, policymakers should embed equity-oriented distribution mechanisms in brand support. It is necessary to move beyond simple production subsidies to establishing fair benefit-sharing mechanisms. Models such as “guaranteed procurement plus premium dividends” and “enterprise + cooperative + farmer” linkages should be promoted to ensure smallholders share in value-added gains. Furthermore, local governments should explore secondary distribution mechanisms for brand-generated revenues, such as village-level development funds, to support vulnerable households and mitigate the mental health risks associated with relative deprivation.

Second, policy efforts should leverage new agricultural business entities to build an inclusive industrial system. Strong support should be directed toward family farms and cooperatives, reinforcing their social responsibility to “link and empower farmers.” Specifically, when creating local employment opportunities, policies should encourage “family-friendly” and “gender-friendly” practices—such as flexible working hours and community-based childcare support—to alleviate the specific pressures faced by female farmers. This ensures that the modernization of agriculture does not come at the cost of marginalized groups.

Finally, it is essential to foster long-term capacity building for intergenerational health. Policies should transcend short-term economic metrics to focus on long-term human capital cultivation. We recommend integrating systematic health literacy promotion and production skills training into AGI industrial chains. Additionally, a portion of brand revenues could be dedicated to improving rural health services and supporting parent–child care programs. Such measures would amplify the social benefits of AGI in fostering family reunification, thereby transforming economic gains into lasting drivers of intergenerational health and sustainable rural development.

This study systematically reveals the multidimensional and complex impacts of AGI development on the health of rural residents, with key findings summarized in four main aspects:

First, AGI development has a dual health effect characterized by “physical benefits alongside psychological stress.” The study finds that AGI development significantly improves the self-rated health of rural residents, reflecting a “development dividend” for physical well-being; however, it also significantly increases the risk of moderate or severe depression, revealing a “hidden cost” to mental health that may accompany such development.

Second, this health effect demonstrates significant group heterogeneity. Specifically, men, farmers with lower education levels, those engaged in farming, and residents in central China show more pronounced benefits in terms of physical health. In contrast, women, higher-educated groups, and residents in western China face more prominent negative effects on mental health. These disparities are closely related to differences in employment patterns, market competitiveness, and regional resource endowments across groups.

Third, mechanism analysis indicates that AGI development improves physical health through pathways such as income growth, local employment and resource agglomeration, promotion of health behaviors, and development of new agricultural business entities. At the same time, however, it may negatively affect mental health by exacerbating intra-rural income inequality and inducing relative deprivation.

Fourth, the study identifies an intergenerational health spillover effect. By creating local employment opportunities that encourage migrant fathers to return to their hometowns, AGI development significantly increases the duration of co-residence between children and their fathers. This not only improves family structure and strengthens parent–child companionship but also creates favorable conditions for the health and human capital accumulation of the next generation.

In summary, the health impacts of AGI development are multidimensional, heterogeneous, and intergenerationally extended. This suggests that in advancing rural industrial branding, it is essential to balance between economic growth and mental health, efficiency gains and social equity, and short-term benefits and the long-term sustainability of human capital.

This study also has several limitations. First, regarding sample coverage, the sample distribution across counties and villages is not fully uniform, and not all AGI-producing regions are covered. Furthermore, the analysis primarily focuses on residents who retain agricultural household registration, with limited inclusion of groups who have completely exited agriculture or changed their household registration status due to AGI development, potentially affecting the comprehensiveness of the conclusions.

Second, regarding temporal dynamics, while our analysis covers the 2010–2020 period, it primarily captures average effects and may not fully trace the dynamic evolution of AGI’s health impacts. For instance, does its influence strengthen or weaken over time? Are there threshold effects or burnout effects? These questions warrant further exploration through longer-term longitudinal data or case studies.

Finally, regarding granularity, the study does not conduct more detailed typological analysis of AGIs, leaving unexplored questions such as “which types of AGI, implemented in what ways, are more conducive to health promotion or risk intensification.” This limitation somewhat reduces the specificity and targeted relevance of the policy implications derived from the findings.