AI-Driven Price Elasticity Analysis for Health Foods in Consumer Markets

Abstract

Although the price elasticity of demand plays an essential role in shaping consumers' behavior in terms of healthy consumption, existing econometric estimates rely on assumptions of linearity and homogeneity in consumers' responses to changing prices, which are far from reality. This paper introduces and evaluates an AI-driven approach that combines a hybrid gradient boosting and deep learning representation of the price elasticity of healthy foods with a PLS-SEM nomological network. Three reflective constructs (AI-based Fiscal Policy and Healthy Food Consumption (HFC), Behavioral Price Sensitivity (BPS), and Hybrid AI Elasticity Estimation (AIE)) were tested on a sample of 250 consumers. The measurement model demonstrated that all indicators satisfied reliability and validity criteria (composite reliability = 0.803-0.877; HTMT < 0.75). The structural model showed that AIE significantly explained BPS (β = 0.568, f² = 0.477, R² = 0.323) and HFC (β = 0.470, f² = 0.283, R² = 0.220), with statistical power ≈ 1.000 at the 1% level. These results support all three hypotheses: the hybrid AI representation model has high explanatory power, behavioral covariates strongly affect price sensitivity among low-income populations, and AI signals for subsidies or taxation are positively correlated with healthy consumption among poorer groups. This framework provides an evidence-based basis for policymaking when considering food pricing.