| Economic fluctuation is a kind of extremely destructive and common phenomenon of economic operation, often manifested as total output fluctuation, total income fluctuation and total employment fluctuation. And its source and diffusion is a hot topic which the academia and decision-making departments have long been concerning about. In recent years, many scholars have found that economic fluctuation not only has macro-level sources, such as aggregate supply, aggregate demand and so on, but also has medium-micro level sources. In particular, the outbreak of the US economic crisis and the formation of the global economic fluctuation, which originated in the US subprime mortgage crisis at the beginning of this century, further pushed the research focus to the meso-level of industry. The output fluctuations of industries caused by various shocks may propagate to the upstream and downstream industries due to the supply and demand relationship among them, causing the fluctuation of the upstream and downstream industries, and spread further among the single-regional economic system and the multi-regional economic system, leading to total output fluctuation of single-regional economic system and the co-movement and interaction of total output fluctuation among multi-regional economic systems.In this paper, the industrial output fluctuation caused by internal and external shocks is defined as industrial fluctuation. The size of the total output fluctuation of a single-regional economic system caused by the diffusion of industrial fluctuation is influenced by the linkage structure of industries in economic system. If there is a kind of industry which provides intermediate inputs for most industries in the economic system, serving as supply hub, and industry which requires most industries in the economic system to provide intermediate inputs for it, serving as demand hub, both of which are referred to hub industries, then hub industries can spread industrial fluctuations to most industries because of local "too linked", and spread orderly lead to total output fluctuation of the economic system. In this paper, "Industrial fluctuations diffuse in a single-regional economic system because of the ubiquitous supply and demand relation among industries, and leads to total output fluctuation of the single-regional economic system." is defined as single-regional industrial fluctuation diffusion effect.In the context of globalization, economic exchanges among countries make the supply and demand linkages between industries cross geographical boundaries. The hub industries of a certain region can not only spread the industrial fluctuation widely to the industries of this regional economic system, leading to total output fluctuation of this region, but also to the industries of other regions, leading to total output fluctuations of other regions. The macroscopic representation is that there exists mutual influence of total output fluctuation among multi regions. In this paper "Because of the inter-regional industrial linkage, industrial fluctuation of a certain region spreads among multi-regional economic systems, resulting in the mutual influence of the total output fluctuation in different regions, and thus making different status and role of a certain region in the multi-regional economic system." is defined as multi-regional industrial fluctuation diffusion effect.In the theoretical circle, there are three kinds of perspectives to study the propagation and diffusion of industrial fluctuation and related meso-sector shock, which are input-output perspective, macroeconomic analysis framework and industrial complex network perspective respectively. Input-output method is the traditional method to study the diffusion of industrial fluctuation and the propagation of sector shock. One way is based on the dynamic input-output model (IO model), by calculating the influence coefficient, reaction coefficient and other indexes to measure the degree of pull and push effect between industries and the impact of the fluctuation of an industry on other industries’ output and total output; and the other way is based on the GVAR model to analyze the propagation path of shock and the impact of shock on prices, output and so on. The analysis of the propagation of shock adopting macroeconomic analysis framework is mainly based on dynamic stochastic general equilibrium model(DSGE model). DSGE Model was first proposed by Kydland and Prescott in 1982 in the paper "Time to Build and Aggregate Fluctuations". The model is used to study how the real shocks (such as technological shocks, monetary policy shocks and preference shocks) spread in the economic system and lead to economic fluctuation under the assumption of perfectly competitive market, within the framework of general equilibrium, and market participants make rational expectations and inter-temporal optimal choice. DSGE model build a communication bridge between macroeconomic theory and microeconomic theory, and has become to an international mainstream macro analysis framework. Although these two kinds of perspectives have studied the propagation way and the impact on macroeconomic of fluctuations or shocks from the sector or industry level, but failed to reveal the mechanism and the fundamental driving force of fluctuations or shocks propagation from the structural level.Existing researches(e.g., Acemglu et al.,2012) have demonstrated the effectiveness of using industrial complex network perspectives to study the diffusion of fluctuations or shocks in economic systems. And this perspectives can explain how the industrial fluctuation or shock can propagate within the economic system and lead to total output fluctuation from the industrial correlation structure level, while related researches on which are limited only to single-regional economic system. The studies on the propagation, co-movement and mutual influence of inter-regional economic fluctuation caused by inter-regional diffusion of industrial fluctuation mostly are ■ correlation analysis of economic fluctuations between two countries (regions) based on the impact factors or macroeconomic variables. But such studies can’t quantify the intensity of mutual influence of inter-regional economic fluctuations and the significance and role of the stability of a certain region on the stability and development of the other regional economic. Therefore, this paper uses industrial complex network perspective to study the single-region and multi-regional industrial fluctuation diffusion effect, and focuses on the study of the multi-regional industrial fluctuation diffusion effect. Thus we can reveal the formation of the single-regional total output fluctuation and its key influencing factors from the structural level, and quantify the intensity of the interaction between the multi-regional total output fluctuations and the status and role of each region in the multi-regional economic system.Industries in the economic system have various categories, and they are also different in nature, status and role. Due to the impact of technical, economic and social factors, there exist close and complex relationship between industries, and the inter-industrial relationship is also undergoing profound and complex changes, thus forming complex industrial network. In this paper, the so-called complex network perspective refers to the industrial complex network perspective, and is called industrial network for short. Different nodes in industrial network are of different importance, usually measured by various types of centrality, such as degree centrality, eigenvector centrality, closeness centrality and betweenness centrality etc.. Hub industries correspond to industries with larger degree centrality, the existence of which leads to the asymmetry of industrial network structure, reflecting the local correlation characteristics of the industrial network. In this paper, the industrial network structure described by the degree distribution is defined as industrial network basic structure. Eigenvector centrality, closeness centrality and betweenness centrality are different indexes reflecting the node importance formed based on node degree value. Eigenvector centrality also describes the local association characteristics of the nodes, while closeness centrality and betweenness centrality reflect the broader association of the nodes, and its k - shell structure describes the global association characteristics of industrial network. The node’s k - shell entropy, which is calculated based on the k - shell structure, reflects the extent to which the industrial nodes are linked diversifiedly to each shell, and the larger the value is, the stronger the ability of industry spread fluctuation.Therefore, based on the total output fluctuation analysis framework of Acemoglu et al., considering the local association characteristics of nodes, and based on the industrial network basic structure, chapter 3 analyzes the formation mechanism of signal-regional industrial fluctuation diffusion effect, constructs single-regional industrial fluctuation diffusion effect basic network structure model, and uses China’s industrial network time series data to carry out case analysis. Furthermore, chapter 3 combining the industrial network method with GRA method, proposes a new method- NSPG method to study the impact of industrial network basic structure on industrial fluctuation diffusion effect, and carries out a case study with the cross section data of Chinese industrial network. The study on single-regional industrial fluctuation diffusion effect based on industrial network basic structure demonstrates that the asymmetry of industrial network basic structure affects the size of single-regional industrial fluctuation diffusion effect, and the greater of the asymmetry degree of the industrial network basic structure is, the greater of the industrial fluctuation diffusion effect of the single regional economic system is. The study of the single-regional industrial fluctuation diffusion effect can explain the outbreak of economic crisis of a certain region, such as the outbreak of the US economic crisis at the beginning of this century. The research conclusion is of great significance for a country (region) to maintain economic stability and promote economic development through industrial policy making.Inter-regional industrial linkages form interregional industrial networks, and the existence of supply hubs leads to the asymmetry of inter-regional industrial network basic structure. Asymmetric inter-regional industrial network basic structure influence the diffusion of industrial fluctuation among multi-regional economic systems. Chapter 4 based on the study of signal-regional industrial fluctuation diffusion effect, analyzes the mechanism of multi-regional industrial fluctuation diffusion effect from the local association level, based on the influence of inter-regional industrial network basic structure on the multi-regional diffusion of industrial fluctuation. And according to the industry sum of other regions which have the first-order direct linkage with supply hubs in each region, multi-regional industrial fluctuation diffusion effect basic network model which takes regions as nodes reflecting the influence of inter-regional total output fluctuations is constructed, named regional weighted influence basic network. Then indexes of degree linkage structure, basic linkage structure and core structure are selected to quantify the intensity of inter-regional total output fluctuations and the status and role of each region in the multi-regional economic system from the local association level.The industrial network basic structure reflects the local association characteristics of the industrial network, and the industrial network k-shell structure reflects the global association characteristics of the industrial network. The nodes with larger k - shell entropy can link to more shells, so that the output fluctuation of the industry with higher k - shell entropy, named dominant industry can propagate to more shells, causing output fluctuations of more industries. Chapter 5 analyzes the mechanism of multi-regional industrial fluctuation diffusion effect from the global association level, based on the influence of inter-regional industrial network k-shell structure on the multi-regional diffusion of industrial fluctuation. And according to the k - shell sum of other regions which have the first-order direct linkage with dominant industries in each region, multi-regional industrial fluctuation diffusion effect k - hell network model which takes regions as nodes reflecting the influence of inter-regional total output fluctuations is constructed, named regional weighted influence k - shell network. Then indexes of degree linkage structure, basic linkage structure and core structure are selected to quantify the intensity of inter-regional total output fluctuations and the difference of the status and the role of each region in the multi-regional economic system from the global association level to study the multi-regional industrial fluctuation k - shell structure diffusion effect.Furthermore, chapter 6 considers the impact of local correlation and global correlation on the diffusion of industrial fluctuation among multi-regional economic systems, defines strong diffusion industry according to the integrated value of degree and k - shell entropy, and construct multi-regional industrial fluctuation diffusion effect integrated network model which takes regions as nodes reflecting the influence of inter-regional total output fluctuations according to the k - shell sum of other regions which have the first-order direct linkage with strong diffusion industries in each region, named regional weighted influence integrated network. Then indexes of degree linkage structure, basic linkage structure and core structure are selected to quantify the intensity of inter-regional total output fluctuations and the difference of the status and the role of each region in the multi-regional economic system comprehensively and objectively from integrated level of local association and global association to study the multi-regional industrial fluctuation integrated structure diffusion effect.Chapters 4,5 and 6 all carry out case study to analyze multi-regional industrial fluctuation basic structure diffusion effect, multi-regional industrial fluctuation k-shell structure diffusion effect and multi-regional industrial fluctuation integrated structure diffusion effect adopting the latest Asian International Input-Output Table (trade table) published by official website of the Institute of Economics Japan External Trade Organization (IDE-JETRO) in 2013 as the common data. Calculation results prove that although regional weighted influence basic network, regional weighted influence k-shell network and regional weighted influence integrated network are constructed based on different industrial fluctuation diffusion mechanism, there are tiny difference of the intensity of the interaction between the total output fluctuation of countries (regions) quantified by three models, and the status and role of each country (region) in the ten countries (regions) economic system are not various significantly, thus proving the validity-of the study on multi-regional industrial fluctuation diffusion effect based on complex network perspective. And the research conclusions are helpful for countries to find ways to reduce economic fluctuation, maintain economic stability and promote economic development, to form regional value chain or global value chain, and to improve international competitiveness and comparative advantage from the aspect of industrial policy and trade policy formulation.In summary, based on the industrial complex network perspective, this paper studies the single-regional industrial fluctuation diffusion effect and multi-regional industrial fluctuation diffusion effect considering the influence of different structures of industrial complex network on industrial fluctuation diffusion effect, reveals the source and diffusion of total output fluctuation from the structural level, emphasizes the influence of the industrial network structure on the total output fluctuation of single-regional economic system, and quantifies the intensity of mutual influence of inter-regional total output fluctuations and the different status and role of each region in a multi-regional economic system. Based on the two kinds of structure of industrial network, the mechanism of industrial fluctuation propagate among different regions is analyzed. According to the interregional industrial network, the paper construct three kinds of regional weighted influence network, realize the coupling of industrial network and regional network, and embody the principle and idea of the "industry-region" two-layer super-network. Therefore, this paper has high innovation in the mechanism analysis, the model construction aspect, the model thought aspect, index selection aspect and so on. The related conclusions of the study are of great theoretical and practical significance to analyze the source and diffusion of economic fluctuation and the interaction of interregional economic fluctuation, formulate policies to promote economic stability and development, and build regional co-creating value chains, enhance international competitiveness and comparative advantages. |
PDF Full Text Request