Space Impact Factors Of The Factor Productivity Analysis

In recent years,research of Chinese economic spatial distribution is in the ascendant. This situation not only derives from the breakthrough of the relative theory, but also from the obviously unbalanced development across the areas. Previous theoretical and empirical studies have pointed out that the phenomenon of agglomeration economies will promote the improvement of productivity. However, the formation mechanisms of agglomeration effects are so varied: within an industry, between industries, within a region, and between regions, etc.. So it is only through empirical research to fully understand and clarify the many types of spatial factor's impact on regional productivity that we can propose the right policy recommendations.Basing on Davis & Weinstein (2003), this paper combines Urban Economics and New Economic Geography perspectives, and testifies Chinese Market Potential (MP), Forward/Backward Linkages, Div(diversification) and Spe(specialization) these five spatial factors'influence on factor productivity for the regions. Compared to Davis & Weinstein(2003), innovation of this paper is: (1)To testify spatial factors'influence on factor productivity on Chinese data. (2) To obtaine factor productivity by regression productivity function. (3) To adjust the definition of Cost linkage. (4) To check the effect of delay method in solving the endogenous problem between factor productivity and agglomeration.Based on the data from "1997 inter-regional input-output table of China", "2004 Industrial micro-enterprise database" and "2005 China Statistical Yearbook",in the way of OLS, Robustness Test and Seemingly Unrelated Regression, this paper shows that:(1)MP, forward/backward linkages and Div have significant effects on capital factor productivity, and forward linkages show most stable. (2) MP, forward/backward linkages and Div have significant impact on labor factor productivity, but forward linkages most stable with low educational labor productivity and backward linkages most stable with high educational labor productivity. (3) Seemingly Unrelated Regression shows most stable of forward linkages. (4) The delay method can solve endogenous relationship between agglomeration economy and factor productivity.This paper shows that our industries have quite the scale of agglomeration economies. This clustering phenomenon is mostly controled by the cost of intermediate inputs. Therefore, to enhance inter-enterprise logistics and transport capacity, to reduce the resistance between the product flow are for the government to further enhance the development the most effective means.