Google matrix of the world trade network

TL;DR

This paper constructs and analyzes the Google matrix of the world trade network from 1962 to 2009, revealing stable clusters of rich and poor countries and highlighting the advantages of PageRank-based rankings over traditional methods.

Contribution

It introduces a Google matrix approach to analyze global trade, uncovering stable country groupings and providing a novel perspective independent of economic size.

Findings

Identifies stable rich and poor country domains over decades

Shows PageRank-based rankings differ from traditional export/import rankings

Models trade network properties with a simple random matrix model

Abstract

Using the United Nations Commodity Trade Statistics Database [http://comtrade.un.org/db/] we construct the Google matrix of the world trade network and analyze its properties for various trade commodities for all countries and all available years from 1962 to 2009. The trade flows on this network are classified with the help of PageRank and CheiRank algorithms developed for the World Wide Web and other large scale directed networks. For the world trade this ranking treats all countries on equal democratic grounds independent of country richness. Still this method puts at the top a group of industrially developed countries for trade in {\it all commodities}. Our study establishes the existence of two solid state like domains of rich and poor countries which remain stable in time, while the majority of countries are shown to be in a gas like phase with strong rank fluctuations. A simple random matrix model provides a good description of statistical distribution of countries in two-dimensional rank plane. The comparison with usual ranking by export and import highlights new features and possibilities of our approach.