松江区创新型企业全要素生产率实证研究-松江统计信息网.pdf

供给侧结构性改革视角下， 松江区创新型企业全要素生产率实证研究 松江区统计局课题组 [内容摘要]： 提高全要素生产率对于地区推进供给侧结构性改革至关重要，本文以松江区三经 普至“十二五”末企业一套表数据为基础资源，以规模以上制造业中创新型企业为研究对象，从 测算、因素挖掘和收敛检验三方面进行实证研究。测算部分，先使用三阶段 DEA 模型测算静态全 要素生产率，再使用 DEA-Malmquist 模型测算动态全要素生产率。因素挖掘部分，先使用 Bagging、 Boosting、随机森林等方法分析全要素生产率各影响因素的重要性，再使用关联规则方法分析全 要素生产率各影响因素的影响方式。收敛检验部分，对全要素生产率分别进行σ收敛、β收敛和 按行业、片区、注册类型、规模分组的俱乐部收敛检验。 [关键词]：全要素生产率；三阶段 DEA；Malmquist；Bagging；Boosting；随机森林；关联规则； 收敛检验 [Abstract]：The improvement of total factor productivity is very important to whether the region economy can achieve sustainable development. This article is based on integrated business enterprise survey database which started from SongJiang’s third economic census to the 12th Five-Year Plan and takes innovative enterprises from above-scale industrial enterprises as research object to carry out empirical study on calculation, factor mining and convergence test. On calculation, we first use three-stage DEA model to calculate the static total factor productivity and then use DEA-Malmquist model to calculate dynamic total factor productivity. On factor mining, we first use methods like Bagging, Boosting, Random Forest to analyze the importance of effect factors of total factor productivity and then use Association Rule method to analyze effect ways of effect factors of total factor productivity. On convergence test, we carry out σ convergence, β convergence on total factor productivity and club convergence test on total factor productivity based on groups which are grouped by industry, area, registration type, and scale. [Key Words]:total factor productivity；three-stage DEA；Malmquist；bagging；boosting； random forest；association rule；convergence test 1 一、推进供给侧结构性改革需要提高全要素生产率 1.供给侧结构性改革的概念 供给侧结构性改革是指从提高供给质量出发，用改革的办法推进结构调整，矫 正要素配置扭曲，扩大有效供给，提高供给结构对需求变化的适应性和灵活性，提 高全要素生产率，更好满足广大人民群众的需要，促进经济社会持续健康发展。