Systemic Risk in the Multilayer Network of Global Production and Trade of Food

Abstract The recent Russian invasion into Ukraine highlighted the vulnerability of the network of global food supply. In an interconnected world the local loss of a basic staple crop can lead to shortages around the globe that are not limited to the crop itself but include all food products that rely on its availability as an input for their production. Therefore, a thorough risk assessment has to include the trade networks of multiple food products as well as their interconnections by production processes. The importance of losses caused by a lack of different inputs to produce a given food product and the resulting multidimensionality of systemic risk have not been widely acknowledged yet. Here we construct a multilayer-network model of global trade and production of food, based on the supply- and use-tables of a physical multi-regional input-output model for 191 countries and 125 products. We use the propagation of supply shocks on this network to compute the relative losses (RL) of a given product that a country suffers if another product is no longer produced in a different country. This total RL can exceed the losses attributed to direct trade of the former product by more than an order of magnitude, as we exemplify by studying the effects of a lost agricultural production in Ukraine on countries around the globe. Computing the RL for all combinations of shock-source and shock-target countries and food products, enables us to quantify systemic risk as a multidimensional object. We advocate to take into account this multidimensionality in the future design of more resilient supply chains to enhance global food security.