Key points are not available for this paper at this time.
Introduction Most global food security assessments rely on subjective expert weights, which limits temporal comparability and risks underestimating governance as a structural determinant. This study aims to address these limitations by developing an objective, data-driven Food Security Index that prioritizes empirical weighting over qualitative assumptions. Methods The study applies Principal Component Analysis (PCA) to eight indicators across the four FAO pillars (Availability, Access, Utilization, and Stability) for 122 countries from 2000 to 2021. Weights were derived empirically from the data rather than expert opinion. Furthermore, countries were grouped into four distinct security tiers using K-means clustering, and convergent validity was tested against the Global Food Security Index (GFSI) using Spearman's rank correlation. Results The first principal component (PC ₁) explains 60. 36% of the total variance, identifying political stability and protein supply as the strongest contributors to the index. Sub-Saharan Africa and South Asia are concentrated in low-security tiers regardless of income variation, while several upper-middle-income countries underperform due to inequality and institutional fragility. The index demonstrates strong convergent validity with the GFSI (ρ = 0. 81). Discussion The findings suggest that income alone is neither necessary nor sufficient for food security; governance quality and dietary adequacy (e. g. , protein supply) are more significant “anchoring” factors than simple calorie availability. These results carry direct implications for SDG 2, suggesting that achieving “Zero Hunger” requires a shift from income-centric strategies toward governance reforms and targeted nutritional quality investments. Policy priorities should be tiered: focusing on foundational infrastructure in low-security contexts, shock-resilience in middle tiers, and subnational interventions in higher-performing nations where internal deprivation is often masked by national aggregates.
Oyebamiji et al. (Thu,) studied this question.