影响新疆财政收入增长因素实证分析Word格式.docx

1、39.0710012.9999491.2519795.490145.63102.213.8253497.2619804.030953.24105.120.4652506.3519811.646759.41106.718.3986523.919824.377365.24107.424.7382534.5819835.629778.55109.527.7923545.7519847.221689.75112.233.725558.7519858.4663112.24120.944.4758565.81198610.2008129.04129.748.4658574.68198711.5839148

2、.5139.151.2479584.91198815.4575192.72159.568.4119593.71198919.1521217.2918575.0698599.55199021.7784261.44194.388.7775617.7199126.4691335.91211124.9325638.49199226.0737402.31229.1170.0326646.94199335.1346495.25258.9248.4352655.98199428.6989662.32328.1285.4786657.54199538.2818814.85392.7333.3404676199

3、648.3069900.93433.9387.8472684199757.6171039.85450446.8148715.4199866.56741106.95450.9519.7673680.92199974.44041163.17439.2534.6468694.34200079.07241363.56436.5610.3843672.5200195.09331491.6454705.997685.382002116.47241612.65451.3813.0223701.492003128.22181886.35453.11002.1256721.272004155.7042209.0

4、9465.31161.5245744.492005180.31842604.14468.61352.2757791.622006219.46283045.26474.71567.0521811.752007285.863523.16500.81850.8415830.422008361.06164183.21541.42259.9746847.582009388.78484277.05545.22827.2359866.152010500.57595437.47568.63539.6941894.652011720.43246610.05602.14712.7699953.342012908.

5、96557505.316256258.3831010.44数据来源：新疆统计年鉴2013、新疆辉煌50年。居民价格消费指数以1978年为基期。从表中可以看出新疆财政收入整体趋势是不断增加的，但在1994年有一个明显的低谷，其原因是1994年建立了分税体制，形成了一个符合市场经济基本要求的现代财政税收体制。且在1994年以后，新疆财政收入增长速度明显加快，财政收入额不断增加，说明分税体制改革促进了新疆财政收入的增长。（三）模型参数的估计利用e-views软件，对1978-2012年新疆财政收入、新疆生产总值、居民消费价格指数、固定资产投资、从业人员数资数据，用OLS法回归可得如下结果（表3.1）

6、：表3.1 OLS回归结果Dependent Variable: YMethod: Least SquaresDate: 11/27/13 Time: 14:34Sample: 1978 2012Included observations: 35VariableCoefficientStd. Errort-StatisticProb. C-1.75408437.73723-0.0464820.9632X10.0156630.0081991.9103950.0657X2-0.0888260.033887-2.6212380.0136X30.1313910.00877114.979360.0000

7、X40.0276720.0731390.3783440.7078R-squared0.997621 Mean dependent var133.2512Adjusted R-squared0.997304 S.D. dependent var210.7371S.E. of regression10.94225 Akaike info criterion7.754703Sum squared resid3591.982 Schwarz criterion7.976895Log likelihood-130.7073 F-statistic3145.238Durbin-Watson stat1.9

8、42066 Prob（F-statistic）0.000000得到：= -1.75 + 0.016 X1 - 0.089X2 + 0.131X3 + 0.028 X4 （-0.0465） （-1.9104）（-2.6212）（14.9794）（0.3783）R2 = 0.9976 =0.9973 D.W=1.942 F = 3145.238 （四）模型的检验1经济意义检验按照一般的经济理论，财政收入应和生产总值、固定资产投资、从业人员数呈正相关关系，与物价指数呈负相关关系。所以除2外，其余待定系数应该均为正值。从回归得到模型的结果可以看出各项待定系数的符号与数值，都是符合经济意义的。为了更进一

9、步的检验此模型的真实性，下面将进行统计检验。2.统计检验从上面的回归结果来看, 可决系数R2 = 0.9976 修正的可决系数=0.9973，模型的拟合程度很高。在给定5%的显著性水平下，固定资产投资的P值为0，居民消费价格指数的P值为0.0136，说明这两个解释变量对被解释变量的影响是显著的。但是从P值来看，新疆生产总值和从业人员数对财政收入的影响是不显著的。模型的F值为3145.238，在给定显著水平为0.05的情况下, F（4,30） = 2.53，所以拒绝方程不显著的假设, 回归方程显著。3.计量经济意义检验（1）异方差检验我们使用White检验来对模型进行异方差的检验，检验结果如下表

10、：表4.1 White Heteroskedasticity Test:F-statistic35.27166 ProbabilityObs*R-squared33.637610.002329Test Equation: RESID240-2694.0823739.811-0.7203790.4796-13.440787.212580-1.8635190.0771X12-0.0015910.000558-2.8493150.0099X1*X20.0129110.0056112.3012100.0323X1*X30.0024480.0007983.0662290.0061X1*X40.01532

11、70.0105651.4508260.162316.1721311.352091.4245950.1697X22-0.0248050.006565-3.7783950.0012X2*X3-0.0058800.008306-0.7078960.4872X2*X4-0.0083750.018784-0.4458560.660514.8581510.057601.4773060.1552X32-0.0007110.000317-2.2398130.0366X3*X4-0.0216810.016114-1.3454380.19356.27384314.903940.4209520.6783X42-0.

12、0049880.015243-0.3272530.74690.961075102.62810.933827189.389448.7188110.9075347470.4511.57411-175.88192.902603从上表中可以看出p=0.0020.05，说明模型已不存在一阶序列相关，因而再对模型检验是否存在二阶序列相关，用LM检验，得下表：表4.510.029460.006639549.41962733.386920.2821350.77990.0019560.0083870.2331710.81730.0156690.0308700.5075810.6157-0.0034440.009060-0.3801080.7067-0.0212530.064561-0.3291950.74450.0750760.1888470.3975520.6940RESID（-2）-0.6587550.197789-3.3305950.00240.2865560.1336759.5807687.534249Sum squar