Gendered cognition, intra-household preference divergence, and behavioral choice: How gender differences shape GM maize adoption in smallholder families

Agricultural extension research has paid insufficient attention to intra-household heterogeneity, resulting in promotion strategies that often treat farm households as homogeneous decision-making units. Based on paired survey data from 436 smallholder couples in Yunnan Province, China, this study constructs a “gender cognition – preference divergence – behavioral choice” analytical framework to systematically examine how intra-household gender-based cognitive differences trigger disputes over planting decisions through divergent perceptions of safety risks and benefit evaluations, ultimately influencing Genetically Modified (GM) Maize adoption behavior. The study reveals three key findings: (1) Significant gender asymmetry exists in technology risk-benefit assessment, with women demonstrating 3–5 times stronger safety risk perception than men (p < 0.001). Their risk-inhibiting effect (−27.9%) substantially outweighs men's profit-promoting effect (+16.5%), consistent with the loss aversion principle in behavioral economics. (2) Household decisions exhibit gendered bargaining interactions, where cognitive differences create a “female risk veto → risk-priority bargaining (with a benefit-risk compensation threshold of 1.17) → adoption delay” pathway, collectively reducing technology adoption probability by 16.3 percentage points. (3) Household decision-making authority significantly moderates technology diffusion: male-dominant households exhibit a “patriarchal efficiency advantage” (Δ = −24.4), while female-dominant households display heightened risk sensitivity (Δ = −35.5). In contrast, joint-decision households develop a unique preference divergence-driven adoption pathway (+12 %). This study provides a theoretical foundation for designing gender-sensitive policies in agricultural technology extension, recommending optimized diffusion efficiency through differentiated communication strategies and participatory evaluation mechanisms. • Constructs a “gender cognition–preference divergence–behavioral choice” framework using paired data of 436 smallholder couples in Yunnan. • Reveals women's safety risk perception is 3–5x stronger than men's; their risk inhibition outweighs men's profit promotion (p < 0.001). • Identifies a “female risk veto→adoption delay” pathway with a 1.17x benefit-risk threshold, cutting adoption by 16.3%. • Finds household decision-making authority moderates' diffusion: joint-decision households boost adoption by 12%. • Provides basis for gender-sensitive agri-tech policies (e.g., differentiated communication strategies).