Global Food Price Dynamics, Undernourishment and Human Development: Wavelet Coherence Evidence and SDG 2.1 Resilience Scenarios to 2030

This study examines whether international food price dynamics provide a reliable signal of undernourishment and human development outcomes relevant to the attainment of SDG 2 (Zero Hunger) by 2030. We apply wavelet coherence analysis to the FAO Food Price Index and the prevalence of undernourishment (SDG Indicator 2.1.1) over 2001–2023, testing statistical significance against an AR(1) red-noise null hypothesis. Hybrid ARIMA–Random Forest models generate probabilistic price forecasts through 2030. Despite strong raw coherence (R² ≈ 0.77), only 7.8% of time–frequency cells achieve statistical significance, indicating that apparent co-movement largely reflects autocorrelation rather than substantive dependence. Where significant coherence emerges, it concentrates at medium-run horizons (3–6 years), consistent with undernourishment as a habitual dietary adequacy measure linked to sustained affordability pressures affecting health, productivity, and human capital formation. Rolling correlation analysis reveals a regime shift around 2012—from negative to positive correlation—coinciding with a slowdown in progress toward reducing hunger. Price forecasts exhibit rapidly widening confidence intervals (by ±131 index points by 2030), underscoring fundamental limits to predictability. These findings caution against mechanistic inferences from global price indices to hunger and human development outcomes, redirecting policy emphasis toward domestic transmission channels and nutrition-sensitive safety nets.