河南省經(jīng)濟(jì)增長(zhǎng)影響因素分析

1、資源與環(huán)境學(xué)院計(jì)量地理學(xué)課程論文河南省經(jīng)濟(jì)增長(zhǎng)影響因素分析班 級(jí)姓 名學(xué) 號(hào)專 業(yè)地理科學(xué)專業(yè)河南省經(jīng)濟(jì)增長(zhǎng)影響因素分析摘要：改革開放以來(lái)，河南省的經(jīng)濟(jì)一直在以極快的速度增長(zhǎng)，本文采用經(jīng)濟(jì)增長(zhǎng)模型和多元線性回歸分析方法對(duì)20012014年河南省經(jīng)濟(jì)增長(zhǎng)因素進(jìn)行研究，分析了物質(zhì)資本、消費(fèi)、財(cái)政支出對(duì)河南省生產(chǎn)總值的影響，建立計(jì)量模型，尋求這些變量與河南省國(guó)民產(chǎn)出的數(shù)量關(guān)系，進(jìn)行定量分析，對(duì)模型進(jìn)行檢驗(yàn)。 關(guān)鍵詞：消費(fèi)、投資、經(jīng)濟(jì)增長(zhǎng)、財(cái)政支出 一、前言  （一）經(jīng)濟(jì)增長(zhǎng)理論 經(jīng)濟(jì)增長(zhǎng)是指一個(gè)國(guó)家或地區(qū)的生產(chǎn)商品和勞務(wù)能力的擴(kuò)大。在實(shí)際核算中，常以生產(chǎn)的商品和勞務(wù)總量的增加來(lái)表示，即以國(guó)

2、民生產(chǎn)總值和地區(qū)生產(chǎn)總值的（GDP）的增長(zhǎng)來(lái)計(jì)算。經(jīng)濟(jì)增長(zhǎng)是經(jīng)濟(jì)學(xué)研究的永恒主題。 古典經(jīng)濟(jì)增長(zhǎng)理論以社會(huì)財(cái)富的增長(zhǎng)為中心，指出生產(chǎn)勞動(dòng)是財(cái)富增長(zhǎng)的源泉?，F(xiàn)代經(jīng)濟(jì)增長(zhǎng)理論認(rèn)為知識(shí)、人力資本、技術(shù)進(jìn)步是經(jīng)濟(jì)增長(zhǎng)的主要因素。（二）影響因素的分析 從古典增長(zhǎng)理論到新增長(zhǎng)理論，都重視物質(zhì)資本和勞動(dòng)的貢獻(xiàn)。物質(zhì)資本是指經(jīng)濟(jì)系統(tǒng)運(yùn)行中實(shí)際投入的資本數(shù)量.然而，由于資本服務(wù)流量難以測(cè)度，在這里我們用全社會(huì)固定資產(chǎn)投資總額（億元）來(lái)衡量物質(zhì)資本。居民消費(fèi)需求和政府投資也是經(jīng)濟(jì)增長(zhǎng)的主導(dǎo)因素。經(jīng)濟(jì)增長(zhǎng)問(wèn)題既受各國(guó)政府和居民的關(guān)注,也是經(jīng)濟(jì)學(xué)理論研究的一個(gè)重要方面。 在20012014年的14中,我省經(jīng)濟(jì)年均增長(zhǎng)

3、率高達(dá)11.5%,綜合實(shí)力大大增強(qiáng),居民收入水平與生活水平不斷提高,居民的消費(fèi)需求的數(shù)量和質(zhì)量有了很大的提高。但是,我省目前仍然面臨消費(fèi)需求不足問(wèn)題。因此,研究消費(fèi)需求對(duì)經(jīng)濟(jì)增長(zhǎng)的影響,并對(duì)我省消費(fèi)需求對(duì)經(jīng)濟(jì)增長(zhǎng)的影響程度進(jìn)行實(shí)證分析，可以更好的理解消費(fèi)對(duì)我省經(jīng)濟(jì)增長(zhǎng)的作用。二、數(shù)據(jù)收集與模型的建立（一）數(shù)據(jù)收集本文采用了2001-2014年的河南省生產(chǎn)總值等數(shù)據(jù)，來(lái)源于河南省統(tǒng)計(jì)年鑒，具體數(shù)據(jù)表如下：年份生產(chǎn)總值全社會(huì)固定資產(chǎn)投資總額(億元)居民消費(fèi)價(jià)格指數(shù)（上年為100）財(cái)政支出（億元）20015533.011627.99106.9508.5820026035.481820.45108.6

4、629.1820036867.702310.54108.6716.6020048553.793099.38109.5879.96200510587.424378.69107.71116.04200612362.795907.74112.31440.09200715012.468010.11109.11870.61200818018.5310490.65114.32281.61200919480.4613704.65112.42905.76201023092.3616585.85114.134160317770.51112.04248.82201229599.31214

5、49.99110.45006.40201332191.3026087.45109.95582.31201434938.2430782.17108.66028.69（二）模型設(shè)計(jì)為了具體分析各要素對(duì)河南省經(jīng)濟(jì)增長(zhǎng)影響的大小，我們可以用河南省生產(chǎn)總值（y）作為對(duì)經(jīng)濟(jì)發(fā)展的衡量，代表經(jīng)濟(jì)發(fā)展；用固定資產(chǎn)投資總額(x1)衡量資本投入；用價(jià)格指數(shù)（x2）去代表消費(fèi)需求；用財(cái)政支出（x3）代表政府投資。運(yùn)用這些數(shù)據(jù)進(jìn)行回歸分析。采用的模型如下： 其中，y為河南省生產(chǎn)總值，x1為固定資產(chǎn)投資總額，x2為消費(fèi)價(jià)格指數(shù)，x3為財(cái)政支出，ui代表隨機(jī)擾動(dòng)項(xiàng)。我們通過(guò)對(duì)該模型的回歸分析，得出各個(gè)變量與我省經(jīng)濟(jì)增長(zhǎng)的

6、變動(dòng)關(guān)系。三、模型估計(jì)和檢驗(yàn)（一）模型初始估計(jì)在Evivw中利用最小二乘法進(jìn)行初步回歸分析得到如下的分析結(jié)果：Dependent Variable: YMethod: Least SquaresDate: 01/02/17 Time: 13:32Sample: 2001 2014Included observations: 14VariableCoefficientStd. Errort-StatisticProb.  C-33005.4911023.17-2.9941910.0135X10.0821930.2129260.3860190.7076X2340.6070100

7、.73083.3813580.0070X34.6890971.0548284.4453640.0012R-squared0.995022    Mean dependent var17800.28Adjusted R-squared0.993529    S.D. dependent var10143.41S.E. of regression815.9620    Akaike info criterion16.48157Sum squared resid6657939.&#

8、160;   Schwarz criterion16.66416Log likelihood-111.3710    Hannan-Quinn criter.16.46467F-statistic666.3206    Durbin-Watson stat1.630732Prob(F-statistic)0.000000可以看出，經(jīng)濟(jì)檢驗(yàn)合理，沒(méi)有出現(xiàn)數(shù)字和符號(hào)的錯(cuò)誤。并且可決系數(shù)R2 =0.995022，修正的可決系數(shù)為0.993529。可以看出，擬和效果十分的好。因此，該

9、模型的設(shè)定是合理的 ，將表中的數(shù)字帶入模型得：   （二）多重共線性檢驗(yàn) 計(jì)算解釋變量的簡(jiǎn)單相關(guān)系數(shù)矩陣YX1X2X3Y 1.000000 0.989035 0.341552 0.994639X1 0.989035 1.000000 0.263767 0.993818X2 0.341552 0.263767 1.000000 0.270700X3 0.994639 0.993818 0.270700

10、0;1.000000由相關(guān)系數(shù)矩陣可以看出，x1和x3相互之間的相關(guān)系數(shù)比較高，證實(shí)確實(shí)存在多重共線性。采用逐步回歸的辦法，去檢查和解釋多重共線性問(wèn)題。分別做Y對(duì)x1、x2、x3的一元回歸，結(jié)果如下：Dependent Variable: YMethod: Least SquaresDate: 01/02/17 Time: 14:25Sample: 2001 2014Included observations: 14VariableCoefficientStd. Errort-StatisticProb.  C5537.514673.08148.2271090.0000X1

11、1.0466540.04511623.199420.0000R-squared0.978190    Mean dependent var17800.28Adjusted R-squared0.976373    S.D. dependent var10143.41S.E. of regression1559.159    Akaike info criterion17.67324Sum squared resid29171707    

12、;Schwarz criterion17.76454Log likelihood-121.7127    Hannan-Quinn criter.17.66479F-statistic538.2130    Durbin-Watson stat0.814233Prob(F-statistic)0.000000Dependent Variable: YMethod: Least SquaresDate: 01/02/17 Time: 14:27Sample: 2001 2014Included observation

13、s: 14VariableCoefficientStd. Errort-StatisticProb.  C-145762.7129954.9-1.1216410.2840X21482.7001177.7971.2588760.2320R-squared0.116658    Mean dependent var17800.28Adjusted R-squared0.043046    S.D. dependent var10143.41S.E. of regression9922.695

14、0;   Akaike info criterion21.37460Sum squared resid1.18E+09    Schwarz criterion21.46589Log likelihood-147.6222    Hannan-Quinn criter.21.36615F-statistic1.584768    Durbin-Watson stat0.216216Prob(F-statistic)0.232014Dependen

15、t Variable: YMethod: Least SquaresDate: 01/02/17 Time: 14:27Sample: 2001 2014Included observations: 14VariableCoefficientStd. Errort-StatisticProb.  C4183.866502.12908.3322530.0000X35.2040850.15618533.320020.0000R-squared0.989307    Mean dependent var17800.28Adjusted R-

16、squared0.988416    S.D. dependent var10143.41S.E. of regression1091.732    Akaike info criterion16.96048Sum squared resid14302546    Schwarz criterion17.05178Log likelihood-116.7234    Hannan-Quinn criter.16.95203F-stati

17、stic1110.224    Durbin-Watson stat0.611681Prob(F-statistic)0.000000經(jīng)過(guò)比較得，X3與Y的t檢驗(yàn)和擬和效果最好 ，因此把X3作為基準(zhǔn)變量引入，然后在逐步的引如其他的解釋變量。Dependent Variable: YMethod: Least SquaresDate: 01/02/17 Time: 14:29Sample: 2001 2014Included observations: 14VariableCoefficientStd. Errort-StatisticProb.

18、  C4237.181623.67676.7938740.0000X34.9743001.4677093.3891590.0060X10.0467660.2968610.1575340.8777R-squared0.989331    Mean dependent var17800.28Adjusted R-squared0.987391    S.D. dependent var10143.41S.E. of regression1138.993    

19、Akaike info criterion17.10109Sum squared resid14270351    Schwarz criterion17.23803Log likelihood-116.7076    Hannan-Quinn criter.17.08841F-statistic510.0129    Durbin-Watson stat0.599772Prob(F-statistic)0.000000Dependent Variable: YMethod:

20、 Least SquaresDate: 01/02/17 Time: 14:29Sample: 2001 2014Included observations: 14VariableCoefficientStd. Errort-StatisticProb.  C-32889.7510584.28-3.1074150.0100X35.0935810.11647543.731000.0000X2338.693796.638793.5047380.0049R-squared0.994948    Mean dependent var17800

21、.28Adjusted R-squared0.994030    S.D. dependent var10143.41S.E. of regression783.7642    Akaike info criterion16.35350Sum squared resid6757150.    Schwarz criterion16.49044Log likelihood-111.4745    Hannan-Quinn criter.1

22、6.34083F-statistic1083.206    Durbin-Watson stat1.608830Prob(F-statistic)0.000000從所得的結(jié)果中可以看出，x2的調(diào)整后可決系數(shù)最大，當(dāng)去除x1后多重共線性消失，得到的檢驗(yàn)結(jié)果如上。從上面修正的回歸結(jié)果可以看出，R2=0.994948，并且它的修正的可決系數(shù)值也達(dá)到了0.994030，顯然，它的擬和效果十分的好，并且t檢驗(yàn)值顯著的大于它的臨界值，即t值檢驗(yàn)十分的顯著，因此多重共線性消失，得到修正后的模型為：  （三）異方差檢驗(yàn)White檢驗(yàn)：Heteroskedast

23、icity Test: WhiteF-statistic3.114913    Prob. F(5,8)0.0746Obs*R-squared9.249117    Prob. Chi-Square(5)0.0995Scaled explained SS4.944554    Prob. Chi-Square(5)0.4227Test Equation:Dependent Variable: RESID2Method: Least SquaresDate: 01/02/17

24、Time: 14:53Sample: 2001 2014Included observations: 14VariableCoefficientStd. Errort-StatisticProb.  C2.89E+084.12E+080.7012960.5030X313759.847463.8951.8435210.1025X32-0.3173790.101175-3.1369190.0139X3*X2-107.303565.05182-1.6495070.1377X2-5077554.7471862.-0.6795570.5160X2222162.6833832.380.6550730.5308R-squared0.660651    Mean dependent var482653.5Adjusted R-squared0.448558    S.D. dependent var659161.2S.E. of regression489487.3    Akaike info crite