基于SBM-Tobit-GWR模型的交通结构效率

在国家节约资源和生态环境保护战略的引导下，为了提高交通运输效率，改善交通运输结构，实现交通行业的绿色可持续发展，从交通投入与系统产出两个层面分析影响交通结构效率的因素，引入非期望产出的超效率SBM （slack based measure）模型，并考虑交通的环境效益，系统分析了我国30个省市的区域交通运输效率. 同时，应用Tobit回归方法与地理加权模型（geographically weighted regression，GWR）解析交通结构效率差异成因、空间分异特征，并提出相应的结构调整策略. 研究结果表明:交通运输综合效率存在地域分布差异，效率值排名前5的省份和效率值分别为上海（1.567）、广东（1.366）、云南（1.292）、江西（1.181）、安徽（1.160）；全国范围内，第二产业产值占GDP比重、人口密度、人均地区生产总值的回归系数分别为0.9513、0.7659、0.5691，三者对于交通结构效率的影响最为显著；分变量系数空间分布图显示不同地区各社会经济要素对于交通结构效率的影响程度存在空间异质性；现阶段我国整体交通效率的提升需要将交通基础建设的大规模、粗犷式发展转变为交通布局规划的精细化设计，优化资源配置，促进公共交通发展. 

Efficiency of Traffic Structure Based on SBM-Tobit-GWR Model

Under the guidance of resource conservation and ecological environmental protection strategies, in order to improve transportation efficiency, optimize transportation structure, and achieve environmental friendly and sustainable developments of transportation industry, the factors affecting the transportation structure efficiency are analyzed from the two aspects of traffic input and system output. By introducing the super-SBM (slack based measure) undesirable model, which considers the environmental benefits of the transportation structure system, the transportation structure efficiency of 30 provinces in China is systematically analyzed. Then, the Tobit regression and geographically weighted regression method are used to analyze the causes of transportation structure efficiency differences, the spatial differentiation of factors, and accordingly propose the adjustment strategies for the transportation structure. The results show that the comprehensive efficiency of transportation has obvious regional differences. The top five provinces are Shanghai (1.567), Guangdong (1.366), Yunnan (1.292), Jiangxi (1.181) and Anhui (1.160). The regression coefficients in terms of the proportion of secondary industry output in GDP, population density, and regional per capita GDP are 0.9513, 0.7659 and 0.5691 respectively, which have the most significant impact on the transportation structure efficiency. The spatial distribution of sub-variable coefficients shows that there is spatial heterogeneity for different regions in the influence level of socioeconomic factors on the transportation structure efficiency. To improve the overall transportation efficiency in China, it is necessary to transform the large-scale and rough development of transportation infrastructure into the refined design and planning of transportation layout, optimize resource allocation, and promote the development of public transportation.