北京市产业网络结构特征与调节效应

Beijing industrial network structural features and moderating effect

经济增长的影响因素分析是一个非常有吸引力的科学议题。资源、劳动力、资本、技术或知识、制度以及空间因素等诸多因素逐步纳入分析的视野。投入要素的规模与产业规模均与产业网络的结构特征有关,需要将产业网络的结构特征引入到产业规模的解释框架之中。利用2005年北京市42部门的投入产出表数据,建立了产业交易链接网络,分析了该网络的基本特征;基于调节变量回归分析方法,讨论了产业网络在产业发展规模中的效能。结果发现:①不同产业间特有的经济技术联系使得产业间交易链接稀疏,集聚系数不高;②不同产业的供给市场数和需求市场数组合差异悬殊;③产业网络结构参量对产业规模的影响是双重的,其一表现为解释变量,如介数中心性;其二是半调节变量,强化了劳动力投入、固定资产投资与产业规模的因果关系,同时,对产业规模也有一定的解释作用。

The analysis of influencing factors on economic growth is an attractive scientific issue.Capital,labor power,technology or knowledge,principle and space have gradually become the analytic factors.These factors are causally related with economic output.The sectors in national economy construct the regional economic system.Each sector takes the other outputs as production factors,and supplies its own output for other sectors.The complicated demand-supply relationship integrated all sectors in national economy into an interdependent organization through the technical economy relation.Under the restriction of industrial network,various economic activities are accelerating the regional economic growth.If the sectors are regarded as nodes,and transaction activities as links,these factors could turn the demand-supply system into a complicated network.That is a new angle to research regional economic growth from industrial network.This paper maintains that both of the scale of input factors and industry are related to the structural characteristics of industrial networks,and emphasizes that it is necessary to bring these structural characteristics into the explanation framework of industrial output change.In order to do research on the essential features of industrial networks,we use the input-output tables of 42 industrial sectors in Beijing（2005） to set up a network based on industrial exchange link.And then,we discuss the efficiency of industrial networks in industrial development based on regression via moderator.It is found that： first,the specific connections in economy and technology between different industries make less industrial transaction links,and lead to the low cluster coefficient.Second,different industries have different numbers of supply-demand market.There are relatively few industrial sectors that have many supply and demand markets in Beijing＇s industrial network in 2005.Third,the structure parameters of the industrial network can be divided into two categories according to the influences on the industrial scale.One is the explanatory variable such as betweenness centrality.The other is the half moderator,which put emphasis on the causality between labor input,fixed assets investments and industrial scale.Meanwhile,the half moderators also have explanation function on industrial scale.