中国科技资源投入产出效率的空间变系数计量分析

A Spatially Varying-coefficient Econometric Analysis on Output Efficiency of S&T Resource Input in China

由于数据在空间上表现出自相关性,使得科技资源投入对产出的作用具有区域差异性。基于前期研究对此关注不够,本研究采用突变级数法、ESDA及地理加权回归模型对1998年、2005年及2013年中国科技资源投入产出的时空差异演变进行了探讨,结果表明:(1)科技资源投入和产出都具有显著的空间集聚态势;(2)科技资源投入对产出具有显著的正面影响,但是其影响有下降的趋势;(3)GWR模型比OLS模型更好地模拟了数据;(4)科技资源投入产出的效果具有明显的空间差异性,自东向西成"梯形"分布,且这种趋势还在不断强化。

Because of the data from the correlation in space, the role of resource input has the regional difference from the output. Based on the previous research which has not enough attention, the space- time difference of the output of S＆T resource input in China are analyzed by the catastrophe progression method, ESDA and geographically weighted regression model. The catastrophe progression method, ESDA and geographically weighted regression model at 1998, 2005 and 2013. The results find that： （1） the output and input of S＆T re- source obviously has both the situation of spatial concentration. （2） Input of S＆T resource has a significant positive impact to input of S＆T resource, but its role has declined. （3） GWR model is better than OLS model to simttlate the data. （4） The influence from S＆T re- source input on S＆T resource output has a significantly spatial variations, presenting a trapezoidal situation from east to west, and this trend continued to strengthen.