虚拟消去后向关联研究

虚拟消去后向关联包括Celia、Clements、Diezenbacher ＆ vanderlinden、Miller ＆ Lahr产业关联。以中国1987、1990、1992、1995、1997、2000、2002年投入产出表为基础，测算并比较了四种后向关联。就数值而言，大小次序为：D＆V、M＆L法、Clements、Cella法。四种后向关联中，Clements后向关联要明显大于Cella后向关联，Clements、M＆L、D＆V三种后向关联差别不大。利用最终需求消除了规模因素的四种相对后向关联与Rasmussen后向关联的相关系数较高，尤其是Cella后向关联。随着部门分类细化，Clements、M＆L、D＆V三种后向关联之间的差异减小。

Research on Backward Linkages of Hypothetical Extraction

backward linkages of hypothetical extraction include Cella,Clements,Diezenbacher & van der linden,Miller & Lahr’s sector linkages.Based on the China’s input-output tables of years of 1987,1990,1992,1995,1997,2000,2002,four backward linkages are calculated and compared.In terms of values,order of size is D&V’s linkage、M&L’s、Clements’s、Cella’s.In the four kinds of backward linkages,Clements backward linkage is significantly greater than Cella backward linkage,Clements’s,M & L’s,D & V’s have little difference.Relative backward linkage which the scale factor is Eliminated using final demand has higher correlation with Rasmussen’s backward linkage,especially Cella’s method.With the detailed sector,the differences between Clements’,M & L’s,D & V’s become smaller.