The state's role and position in international trade: A complex network perspective☆
Graphical abstract
Introduction
With globalisation of the world economy, close ties between various countries have combined global trade relationships into an organic whole. Increased international trade is becoming the key in shaping the global economic and political landscape. Mobility of international commodities and factors, technological innovation and development, contribute to economic integration called the “global village”. Economic globalisation and integration lead us to consider the entire countries as a large system. From this perspective, global production and consumption transfers will result in a substantial complex network between countries and their trade partners. A complex network is represented as a set consisting of a number of vertices with links between them. Its representation offers a new level of description that goes beyond the country-specific analyses used in more traditional economic studies of trade.
As theories and the technology of the complex network have improved over the last decade, a significant amount of effort has been devoted to the empirical exploration of the International Trade Network (ITN) from this new perspective. The ITN, also known as the World Trade Web (WTW) and the World Trade Network (WTN), is defined as the network of import/export relationships between world countries in a given year. Because international trade is usually measured using the monetary value of exports and imports between countries, trading relationships are analogous to valued links in a network, which vary from country to country. In research, the ITN has many possible representations: binary or weighted, directed or undirected, and aggregated or disaggregated by commodity across several years. From the perspective of aggregated or disaggregated representations, the ITN can be discussed on a country level, firm level and product level. Based on the starting point that complex network studies focus on the overall, general and most basic universal laws in international trade, most research builds the ITN on a country level.
The earliest contributions studied the simplest type (binary and undirected) of ITN to explore the network structure of international trade relations by measuring trade in commodities from the 2000 dataset. This type means that any two countries can be either connected by a link or not, and link directions (import and export) do not matter. If two countries are connected, we say that they are “partners” or “nearest neighbours”. To formally characterise such types of ITNs, it is sufficient to provide the so-called adjacency matrix. For an undirected ITN and a given year t, the adjacency matrix is a symmetric N × N binary matrix A whose generic entry aij = aji = 1 only if a link between countries i and j exists (and zero otherwise).1 In the early constructions of the network, incomplete data were adopted to build a binary ITN for one particular year (2000 dataset). Building an unweighted ITN allows us access to the simple nature of the network, such as its density and the degree distribution. Studies have shown that density is observed to remain roughly constant, with values of approximately 0.52 for the same period. This network displays a scale-free degree distribution, small-world property, a high clustering coefficient, and degree–degree correlations between different vertices, which are the typical properties of complex networks (Serrano and Boguna, 2003). Further results show that the binary-directed representation of the ITN exhibits a disassortative pattern: countries with many trade partners (i.e., high node degrees) are, on average, connected with countries with fewer partners (i.e., low average nearest-neighbour degrees).2 Further outcomes show significant synchronisation of economic cycles in representative developed countries with the United States (Li et al., 2003). All the properties make the International Trade Network a complex network that is far from being well described through a classical random network description (Serrano and Boguna, 2003).
Using a binary ITN analysis, researchers focus their attention on the problems of the intensity and evolution of international trade by weighted analysis to discuss issues such as the processes of globalisation and regionalisation (Erolay et al., 2011, Tzekina et al., 2008), the role of extensive and intensive margins of trade (Benedictis and Tajoli, 2011, Riccaboni and Schiavo, 2010), the core/periphery structure (Fagiolo et al., 2010, Kali and Reyes, 2007) and the role of WTO (Benedictis and Tajoli, 2011). They adopt weighted-network approaches by defining the elements of the weighted adjacency matrix as , or their combination.3 Due to the results, which indicate that the ITN is a strongly symmetric network (Fagiolo et al., 2007a) where the majority of trade relationships (and their intensities) are reciprocated,4 a good deal of research uses a symmetrical network approach by switching to w through the relation of .5
After finding the weighted adjacency matrix, they either build the ITN using a more complete time-series dataset or develop evolution models to explore the network's structure and feature. Research demonstrates that some important aspects of the International Trade Network have been remarkably stable from 1938 to 2003 (Kastelle et al., 2006), and the topology of the weighted ITN is crucially different from binary methods in a given year. The weighted ITN shows weakly disassortativeness, and, moreover, well-connected countries tend to trade with partners that are strongly connected (Fagiolo et al., 2008). The distribution of the total trade intensity carried by each country (i.e., node strength) is right-skewed, indicating that a few intense trade connections co-exist with the majority of low-intensity ones (Fagiolo, 2010, Fagiolo et al., 2008, Fagiolo et al., 2009). Another universal feature is observed in the power-law growth of the trade strength with the gross domestic product, the exponent being similar for all countries (Bhattacharya et al., 2008). In addition, Tzekina et al. consider the formation of trade “islands” and their evolution to identify community structures and hubs and find mixed evidence for globalisation (Tzekina et al., 2008).6 These studies display that the complex network analysis can be used not only to describe network structures but also to discuss some actual topics of international trade from a new, intuitionistic angle.
Following the network's structure analysis above, scholars (Foti et al., 2011, Kali and Reyes, 2007, Lee et al., 2011, Serrano et al., 2007) try to explore robustness and (crisis) propagation problems of the ITN instead. To investigate robustness (stability of the international trade system), Foti et al.(2011) introduce the notion of extinction analysis, showing that over time, the ITN moves to a “robust yet fragile” configuration where it is robust to random failures but fragile under targeted attack. Scholars believe that a network approach that is capable of incorporating the cascading of interdependent ripples that occur when a shock hits a specific part of the network will provide us with a deeper understanding of economic and financial contagion (Kali and Reyes, 2007). They find that a crisis is amplified if the crisis epicentre country is better integrated into the trade network. Another study develops a general procedure capable to progressively filter out in a consistent and quantitative way the dominant trade channels and provides new quantitative tools for a dynamical approach to the propagation of economic crises (Serrano et al., 2007). A recent study uses a simple toy model of crisis spreading, finding that the GDP of a country cannot fully account for its avalanche size and an individual country's role in crisis spreading is dependent not only on its gross macroeconomic capacities but also on its local and global topological structure in the world economic network (Lee et al., 2011).
The above research provides a good snapshot of the structure and features of ITNs, enabling us to understand international trade as a whole. In this paper, we provide another picture of the mid-level structure and even micro-level elements of ITNs using the latest complex network theories. We first explore the cascading influence of the interdependent ripples that occur when trading relationships change in the ITN, which cannot be measured by traditional economic methods.
In this paper, after presenting the main topology and structure of the ITN (Section 2), we introduce the coarse graining method and the weighted extremal optimisation algorithm (WEO) to divide the countries into communities, which help us discuss the trade patterns on a mid-level structure of the ITN (Section 3). With this context of mid-level structures, six centrality indicators are raised to measure vertices' importance in the ITN, and issues of the roles of countries and the EU will be discussed. The results provide countries' ranking order and some conclusions about trade imbalance, globalisation and regionalisation. Moreover, by using a simple bootstrap percolation process, we discuss the topic of cascading failure caused by disconnecting vertices' linked in the ITN, which is the key objective of this paper (Section 5).
Section snippets
2010 International Trade Network
As a supplementary and typical example of an ITN study, we build a 2010 International Trade Network, and make empirical analyses using the latest achievements in complex network theory.
We employ international trade data that are available from the Direction of Trade Statistics, provided by the IMF e-library, which contains data on the value of trades between countries and their trading partners. The IMF provides both import and export data. Because trade flows are uneven, the adjacency matrices
Community structure in International Trade Network
The first issue we address concerns the study of the trade meso-structure in international trade. Exploring the community structure can assist us in understanding patterns well. According to Newman (Newman, 2003), the community structure means the network can be divided into several groups. The connections between the inner nodes are denser, and the connections of nodes among groups are relatively sparse. Based on above definition, we introduce the weighted extremal optimisation algorithm (WEO)
Importance of vertices in International Trade Network
Because there is strong heterogeneity between countries and because countries play very different roles in the network structure, which is evidence that is difficult to reconcile with traditional trade models (Benedictis and Tajoli, 2011), we then shift our attention to the analysis of states' importance in the ITN.12
Bootstrap percolation process
Studies have shown that the global trading system is highly complex and can be best viewed as a complex network of interacting macroeconomic agents. The world is still made up of nation states and a global marketplace but is moving further toward economic “globalisation”. Based on historical experiments, one important factor that will dictate the future direction of globalisation is sovereign governments, which should not be overlooked. They still have the power to erect significant obstacles
Conclusions
With the multi-polarisation of today's world and the depth of development of economic globalisation, countries are more interdependent. The world is undergoing a huge, profound and complicated change. This paper applies an analysis of complex networks to empirically research international trade and countries' trading relations. The structure of global trade is quantitatively described and analysed.
On the basis of mid-level structures, according to the theory of a network's community, we use the
References (47)
- et al.
Eigenvector-like measures of centrality for asymmetric relations
Soc. Networks
(2001) On variants of shortest-path betweenness centrality and their generic computation
Soc. Networks
(2008)- et al.
The physics of living neural networks
Phys. Rep.
(2007) - et al.
On the topological properties of the world trade web: a weighted network analysis
Physica A
(2008) - et al.
Structure and evolution of the world trade network
Phys. A
(2005) - et al.
Complexity and synchronization of the World trade Web
Physica A
(2003) - et al.
Numerical study on the kinetic Ising model for glass transition
Phys. Lett. A
(1986) - et al.
Transmission of order in some unusual dilute systems
Phys. Rev. Lett.
(1987) - et al.
Bootstrap percolation on the hypercube
Electron. J. Probab.
(2006) - et al.
Bootstrap percolation on random regular graph
Random Struct. Algoritm.
(2007)
Bootstrap percolation on bipartite networks
Acta Phys. Sin.
Bootstrap percolation on complex networks
Phys. Rev. E
The world trade network
World Econ.
The international trade network: weighted network analysis and modelling
J. Stat. Mech: Theory Exp.
Centrality and AIDS
Connections
'Bootstrap percolation on a Bethe lattice
J. Phys. C: Solid State Phys.
Community detection in complex networks using extremal optimization
Phys. Rev. E
Trade synchronization in the World Trade Web
Int. J. Complex Syst. Sci.
Constrained diffusion dynamics in the hard — square lattice gas at high density
J. Chem. Phys.
The international-trade network: gravity equations and topological properties
J. Econ. Interac. Coord.
The Evolution of the World Trade Web
International Trade and Financial Integration: a Weighted Network Analysis
World trade web: topological properties dynamics and evolution
Phys. Rev. E
Cited by (119)
The interconnectedness of European Banking and Shadow Banking for sustainable development goals: Insights from a network GVAR model
2024, Research in International Business and FinanceIntroducing the GVAR-GARCH model: Evidence from financial markets
2024, Journal of International Financial Markets, Institutions and MoneyRegional economic integration and machine learning: Policy insights from the review of literature
2023, Journal of Policy Modeling
- ☆
This work was supported by the National Basic Research Development Program of China (grant no. 2011CB952001), National Science Foundation of China (grant no. 41271542 and 61174150); Program for New Century Excellent Talents in University (grant no. NCET-09-0227, NCET-09-0228), and the Fundamental Research Funds for the Central Universities.