A Bayesian insight into improving national food security

Abstract

The disruptions in food systems caused by extreme events have repeatedly challenged food security at multiple levels. Recently, the COVID-19 pandemic has exacerbated the vulnerabilities of existing global food systems and has resulted in food stress for an additional 145 million people. This paper addresses the critical need for enacting and strengthening policies targeted at securing food systems to achieve the Sustainable Development Goal (SDG) of Zero Hunger by 2030. We propose a novel systematic approach through the Bayesian network modeling framework to enhance national food security and build resilient food systems by effectively prioritizing areas where interventions are most critical and will have the greatest positive impact on investment. Our analysis utilizes annual data from the Global Food Security Index (GFSI) for Thailand from 2012 to 2020, which includes 59 indicators across four dimensions of food security. The GFSI data is sourced from international organizations including the FAO, WHO, World Bank, and others. Our results, supported by literature, showcase the Bayesian approach as an efficient and convenient decision-support tool that provides concrete and actionable recommendations for policymakers with clearly defined constraints and uncertainties. Further research could explore applying this approach to specific regional contexts, incorporating additional data sources to refine the prioritization of interventions.