Power of Countries in the Context of Food Security

Food security has been a topical question for the last decades. A huge number of organizations were founded during this period and their main purposes are to evaluate the level of food security in each country and to decrease the level of undernourishment in the world. However, the evaluation of the level of food security is usually based on statistical indicators of each country that are frequently received through household surveys. The main disadvantage of these measures is that they do not consider relations between countries. We suggest to combine both statistical indicators of countries and their bilateral trade relations through the social network analysis in order to detect the most influential traders. In order to assess the most influential actors, classical centrality measures can be applied. Unfortunately, they do not consider such essential factors for the influence estimation as individual attributes of nodes, the group influence as well as long-range connections. In this work several new centrality measures that take into account all these factors are proposed. The approach helps us to assess node-to-node influence in a network, which further may be aggregated into the index of the total influence or the dependency index. We apply classical centralities and proposed measures to the trade network based on World Integrated Trade Solution database. Our analysis is focused on three main products for food security (wheat, rice, poultry meat) that are studied separately. Direct and indirect connections were analyzed in order to explain why some countries are import dependent. We also provide time series analysis of influence indices considering 1996 – 2016 period. Finally, cluster analysis is applied to initial networks of trade and the networks of direct influences. The results show that the network of direct influence is more preferable for the community detection.