Abstract:
Composite indicators have gained popularity in various research areas. However, the determination of an appropriate weighting method is challenging. Subjective weighting methods are criticised for their potential bias that may reduce stakeholders’ trust in the results of a composite index. By contrast, objective weighting processes are perceived to provide unbiased results that may overcome trust issues. The Global Food Security Index (GFSI) is a composite indicator that measures the comparative level of food insecurity for 113 countries. The initial components of the GFSI included the affordability, availability and quality and safety components. In 2017, the GFSI added a fourth component for natural resources and resilience (NRR) as a risk to food security. The Economist Intelligence Unit’s (EIU) panel of experts uses a subjective weighting of indicators in the GFSI model. This study set out to assess whether an objective weighting of the NRR component of the GFSI significantly changed the country scores and ranks compared to the subjective weighting process. The GFSI data was analysed using a principal component analysis (PCA) to derive objectively weighted NRR scores and ranks. The objectively and subjectively weighted NRR ranks were strongly correlated (rho = 0.831), implying that the GFSI model was not strongly statistically biased. The study concluded that subjective weighting of the NRR component of the GFSI may still provide relatively fair country scores and ranks. However, an objective weighting of the NRR component could improve the reliability of the NRR component of the GFSI and build greater trust.
