中国生产性服务业发展的影响因素分析——基于贝叶斯模型平均(BMA)方法的实证研究  

Influencing Factors of the Development of Producer Services in China——Empirical Studies Based on Bayesian Averaging Model( BMA)

作　　者：张超 郑长娟[1] 

机构地区：[1]宁波大红鹰学院金融贸易学院,浙江宁波315175

出　　处：《山东财经大学学报》2018年第2期43-51,59,共10页Journal of Shandong University of Finance and Economics

基　　金：浙江省哲学社会科学规划项目"浙江县域金融服务业的集聚特征和效应分析--基于产业共同集聚的视角"(17NDJC105YB);浙江省人力资源和社会保障厅科研项目"供给侧改革背景下浙江产业结构调整的收入分配效应研究"(2017095);慈溪市社会科学研究项目"加快生产性服务业发展推动慈溪产业转型升级对策研究"(2017SKB001)

摘　　要：基于模型不确定性视角,选取2004-2015年省际面板数据,尝试利用贝叶斯模型平均方法对影响中国生产性服务业发展的重要因素予以识别和检验。研究结论表明:在事先选取的16个解释变量中,服务效率、工业化程度、经济发展水平、劳动力投入以及人口老龄化等5个解释变量是影响中国生产性服务业发展的主导因素,此外,专业化程度、人力资本投入以及基础设施水平等3个解释变量对生产性服务业发展也具有较好的解释能力。同时,证实了BMA方法在处理模型不确定性问题上优于单一模型。最后,提出了经济新常态下加快生产性服务业发展的政策建议。Based on the model uncertainty,using Bayesian model averaging method and provincial panel data from 2004 to 2015,this paper discerns and analyzes the dominant factors of the producer services in China. The result shows that in the selection of sixteen explanatory variables in advance,there are five explanatory variables including service efficiency,degree of industrialization,level of economic development,labor input,aging of the population,which are the dominant factors for the producer services in China. Three other explanatory factors including degree of specialization,human capital investment,infrastructure level also have certain explanatory abilities. In the meantime,we prove that the BMA method is superior to single model in dealing with model uncertainty. Finally,the paper puts forward some suggestions.

关 键 词：生产性服务业 影响因素 模型不确定性 贝叶斯模型平均方法 

分 类 号：F127[经济管理—世界经济]