Research On Measurement And Evolution Of World Trade Network

After World War Two, with the acceleration of world economy globalization, more and more countries developed closer mutual trade relationships, countries cooperated and competed at the same time. In this context, any country cannot isolate from the influence of other countries, global economy becomes a wholly organic structure. All the trade relationships among countries form a world trade network, countries are the nodes and trade relationships are the edges. As a node of the network, the change of any country's economy will influence other countries through the trade relationships. This kind of influence usually is not restricted in the local region, but propagate broadly through the world trade network, such as the Asia Financial Crisis in 1990s.Obviously, the healthy development of world trade economy depend on the coordination of countries trade policy and following the self-regulation of world trade system. China has become an important trade country in the world, to plan the correct trade policy and protect our local economy, we must research on the feature and law of world trade network.The dissertation introduces a new perspective on research of world trade system based on some latest research results. World trade system is composed of more than 100 countries with mutual economic linkage, trade is the mutual function of different agents on the micro levels. The dissertation researches on the evolution law of world trade structure from the weighted network perspective, discloses the subjective regulations of mutual economic function between countries.The dissertation measures the structural features and evolution law from space dimension and time dimension. Research on space dimension includes the third, fouth and fifth chapter, time dimension includes the sixth and seventh chapter.Space dimensionSpace dimension means the interception of world trade network snapshot in one year, analyzing the space structure of snapshot based on the data of trade volume. Following three layers:1. Study on the world trade network topological structure. Neglecting the influence of trade flow volume on network structure, constructing the topological adjacent matrix with the 0 or 1 as the existence of trade relationship or not. Research on the characteristics of world trade networks'degree distribution, clustering, degree correlations and reciprocity. The results show that world trade networks are not scale-free, and have the tendency to become more randomly. With more countries participated in the world trade system, their statuses are more orderly, and the coexistence of economic globalization and local integration is strengthened.2 Study on the world trade network weighted structure.Constituting the weighted world trade network by assigning the magnitudes of trade flows to the edges of the network, and study the structural characteristics of weighted network from four perspectives: the relations of node strength and degree, edge weight disparity, clustering and node degree correlations. Comparative analysis with topological structure discloses some new findings, such as the node degree is negatively correlative in topological structure and positively correlative in weighted structure. The results show that the analysis of weighted network could depict more information for economic flow network.3 Study on the world trade network skeleton structure.Apply the concept of spanning tree to decompose the world trade network into the two skeleton structures: maximum BC spanning and maximum flow spanning tree. For the network with great difference in weights, maximum flow spanning tree's structure is more fitted with the original network. The change of out degree of maximum flow spanning tree reflects the historical transfer of countries trade status.Time dimensionResearches on time dimension focus on the characteristics of evolution law of world trade network in the past 51 years. From two perspectives:1. Study on world trade network evolution based on information law. The dissertation borrows the ideas of ecological network to quantitatively depict the evolution characteristics of world trade network: information entropy, redundancy, conditional entropy, mutual information. Empirical results of study on the buffering capability show that it increased after World War Two. In addition, the evolution stability of world trade network is studied by applying the conditional entropy of information theory, results show it is in the middle of ecological network and random network. This means that the evolution of world trade network is more drastic than ecological network, but it is not in the chaos of random network. The results of mutual information means the flexibility of trade flows increased rapidly.2 Study on World trade network flow growth dynamics.In the context of weighted world trade network, the dissertation researches the growth fluctuation of trade volume and trade flow in the time dimension. Finding that both follows the revised laplace distribution, and the power index of relation of variance–size fits with each other. The results prove the proposition of Stanley(1996).The conclusions propose the restricted practical conditions for the building models of world trade system.The innovations include:1. Constructing the weighted world trade network, depicting the relations of trade flow and topological structure of world trade system.World trade network is a typical complex economic network, the magnitudes of trade flows in the network are greatly diverse, so that the simple analysis of topological structure cannot depicts correctly the structural characteristics of world trade network. Constitutes the weighted world trade network by assigning the magnitudes of trade flows to the edges of the network, and studies the structural characteristics of weighted network from four perspectives: the relations of node strength and degree, edge weight disparity, clustering and node degree correlations. Comparative analysis with topological structure discloses some new findings, such as the node degree is negatively correlative in topological structure and positively correlative in weighted structure. The results show that the analysis of weighted network could depict more information for economic flow network.2 Constructing the world trade network skeleton, depicting the relations of center network and original network.The structure of complex network is usually very tremendous and complicated, to depict the main characteristics of main structure, constructing the maximum flow spanning tree of the world trade network to simplify the research. The results means the maximum flow spanning tree reflect the basis characteristics of original network. Because the spanning-tree only occupy the minority of original network, greatly reduce the work of research .This provides a new research method for the flow network research.3 Borrowing the information concept, proposing the quantitative measurement of evolution law of world trade network. The concepts of information law are comprehensively applied in the research of ecological network. Because of the similarity of world trade network and ecological network, the dissertation borrows the information measurement to quantitatively depict the evolution of world trade network. Consider the space restrictions of world trade flow, and apply the criterion of stability to the world trade network. This propose a quantitative measurement of the evolution of economical flow network.4 Depicting the fitting relation of growth rate of trade volume and trade flow in world trade network.Professor Stanley research group find the growth rate of index of agents(firm size,country GDP)follows the power law relations for the complicated economic system.This dissertation proves this regulation in the world trade system ,and further more, disclose the index relations of agent characteristics and mutual function(country GDP and flow volume).This result is the basis of modeling the complicated economic system...