基于随机前沿引力模型的省际技术转移效率及影响因素研究

Research on Technology Transfer Efficiency and Influencing Fac⁃tors Among Provinces Based on Stochastic Frontier Gravity Model

省际技术转移日益成为缩小地区间技术鸿沟的措施之一。借鉴相关文献研究成果并结合技术转移特征,梳理了影响省际技术转移效率的14个因素,利用新近发展的Lasso(Least absolute shrinkage and selection operator)回归模型筛选出重要变量,然后采用SFA(随机前沿分析)结合引力模型、随机前沿非效率模型考察了省际技术转移效率的主要影响因素及非效率项,并对各省(区、市)的技术转移效率进行了测度。结果表明:①技术输出和输入地区的有效专利数、两者间的技术距离显著影响省际技术转移效率,而省际地理距离对该效率无显著影响;②技术输入地区的技术市场化及知识产权保护度等非效率项是影响技术转移非效率的原因,提高技术市场化程度及知识产权保护力度均能够缓解技术转移的政策约束,从而对技术转移非效率有一定的抑制作用;③我国各省(区、市)的技术转移效率均值在0.4~0.5水平上波动,东部地区的技术转移效率明显高于中部、西部及东北地区。

Inter-provincial technology transfer has increasingly become one of the measures to narrow the technological gap between regions.This paper draws on relevant literature research results and combines the characteristics of technology transfer to sort out 14 factors affecting the efficiency of inter-provincial technology transfer.Due to these factors with the wide and complex sources,by using the recently developed method of Lasso(Least absolute shrinkage and Selection operator)regression model to screen out six important variables that can significantly influence technology transfer:technology absorption capability of the technology introduction area,quality of technological innovation,geographical distance,technological distance,protection of intellectual property rights and marketization of technologies.Due to the bias of Lasso regression,then,it uses the method of SFA(Stochastic Frontier Analysis)and combined with gravity model and stochastic frontier inefficiency model to investigate the main influencing factors and inefficiency terms of inter-provincial technology transfer efficiency,and measures the technology transfer efficiency of each province.The empirical results show that:①The applicability of the four models is verified:Traditional linear regression model(model 1),the time-varying stochastic frontier model(model 2),the time-varying stochastic frontier model item(model 3)and the efficiency is time-varying and"technology marketization"and"intellectual property protection"variables of the linear combination model(model 4),the result shows model 2,model 3 and model 4 are significantly better than that of model 1,that is,the non-efficiency term of technology transfer exists;and model 4 is significantly better than model 3,model 2 and model 1,it means the stochastic frontier model has a significant rationality to extract the inefficiency factor from the random factor.Compared with Model 3,the results of model 3 are better than those of model 2,and the real-time variable stochastic frontier model is better than the non-timevarying stochastic frontier model.②The number of effective patents(quality of technological innovation and technology absorption capability of the technology introduction area)and the the efficiency of inter-provincial technology transfer is significantly affected by the technological distance between the two regions,while the geographical distance between provinces has no significant effect on it.③The inefficiency of technology transfer is caused by technology marketization and intellectual property protection in the technology import areas.Improving the degree of technology marketization and intellectual property protection can alleviate the policy constraints of technology transfer,thus inhibiting the inefficiency of technology transfer to a certain extent.④The average technology transfer efficiency of Beijing and the provinces fluctuated from 0.4 to 0.6,and the technology transfer efficiency of the eastern region was significantly higher than that of the central,western and northeastern regions.⑤The efficiency of inter-provincial technology transfer is still at a low level and has great potential for development.