Abstract: This paper applies the economic data of two provinces and one city in the national regional typical planning-Yangtze River Delta regional planning to build a model of fixed effect and variable intercept regression in different time window; contrasted by the changes of regression coefficient in different time window, it obtains policy effect of national regional planning. The innovate and character is that: First ,it obtains the increment changes size of effect variable which results in the unit policy by the slope coefficient changes of variable intercept regression model in different time window of all three provinces. Second, contrasted by the regression equation intercept changes of planning pre and post , all three provinces and different time window, it gains the increased size of effect variable amount. The innovate changes the status that current research has gained the evaluation result only by the comparison of different region transverse data; it has no comparability because of the differences of objective base and condition in different region in fact. The third, in different provinces, compared to the sorting size changes of the pre-and-post planning for effect regression function’s intercept , it discriminates the changes of development balance level. The main conclusions are that: Firstly, the variation of the unit input of central fixed assets investment to urban and rural residents increases evidently in planning region; It is no more notable because of the increased economic development level and the decreased variation of the unit input of central subsidy income to the disposable income of urban and rural residents. Secondly, the effect function value of disposable income per capitals for urban and rural residents in planning region increase evidently. Thirdly, the gap of spontaneous income of disposable income per capitais among every provinces for Yangtze River Delta regional planning level is reduced.

Key words: technique economy and management, policy effect evaluation, the fixed effects regression, regional planning

摘要： 应用国家重大区域规划中的典型规划“长江三角洲地区区域规划”中两省一市的经济数据,建立了不同时间窗口的固定效应变截距回归模型。通过对比不同时间窗口模型回归系数的变化,得到国家重大区域规划的政策效果。本研究的创新与特色:一是通过全部三个省份不同时间窗口变截距回归模型斜率系数的变化,得到单位政策变量导致效果变量增量变化的大小。二是通过规划前后全部三个省份不同时间窗口回归方程截距变化的对比,得到效果变量总额增加自发部分的大小。本文的创新点一和二改变了现有研究通过不同区域的横向数据对比获取评价效果的现状;事实上,不同区域由于客观基础和条件的差异是不具备可比性的。三是通过对比规划前后不同省份效果回归函数截距大小排序的变化,判别不同省份之间发展的均衡程度的变化。主要结论:一是规划区域中央固定资产投资的单位投入对城乡居民人均可支配收入的改变量有显著增加;由于经济发展水平的提高中央补助收入的单位投入对城乡居民可支配收入的变化量减少,其影响不再显著。二是规划区城乡居民人均可支配收入自发水平有显著增加。三是长三角规划区各省间人均可支配收入的自发性收入水平的差距缩小。

关键词: 技术经济及管理, 政策效果评价, 变截距回归, 区域规划