中國經(jīng)濟(jì)增長影響因素實證分析復(fù)習(xí)過程

1、中國經(jīng)濟(jì)增長影響因素實證分析精品文檔中國經(jīng)濟(jì)增長影響因素實證分析摘要：改革開放以來，我國的社會主義經(jīng)濟(jì)取得了突飛猛進(jìn)的發(fā)展，經(jīng)濟(jì)增長速度更是舉世矚目。本文采用經(jīng)濟(jì)增長模型和多元線性回歸分析方法對19932016年中國經(jīng)濟(jì)增長因素進(jìn)行研究，分析了物質(zhì)資本、勞動力、消費對國內(nèi)生產(chǎn)總值的影響，建立計量模型，尋求這些變量與中國國民產(chǎn)出的數(shù)量關(guān)系，進(jìn)行定量分析，對模型進(jìn)行檢驗。關(guān)鍵詞：消費、投資、經(jīng)濟(jì)增長、勞動力一、文獻(xiàn)綜述（一）經(jīng)濟(jì)增長理論經(jīng)濟(jì)增長是指一個國家生產(chǎn)商品和勞務(wù)能力的擴(kuò)大。在實際核算中，常以一國生產(chǎn)的商品和勞務(wù)總量的增加來表示，即以國民生產(chǎn)總值和國內(nèi)生產(chǎn)總值的（GDP）的增長來計算。經(jīng)濟(jì)增

2、長是經(jīng)濟(jì)學(xué)研究的永恒主題。古典經(jīng)濟(jì)增長理論以社會財富的增長為中心，指出生產(chǎn)勞動是財富增長的源泉?，F(xiàn)代經(jīng)濟(jì)增長理論認(rèn)為知識、人力資本、技術(shù)進(jìn)步是經(jīng)濟(jì)增長的主要因素。（二）影響因素的分析從古典增長理論到新增長理論，都重視物質(zhì)資本和勞動的貢獻(xiàn)。物質(zhì)資本是指經(jīng)濟(jì)系統(tǒng)運行中實際投入的資本數(shù)量.然而，由于資本服務(wù)流量難以測度，在這里我們用全社會固定資產(chǎn)投資總額（億元）來衡量物質(zhì)資本。中國擁有全世界近1/4的人口，為經(jīng)濟(jì)增長提供了豐富的勞動力資源。因此本文用總就業(yè)人數(shù)（萬人）來衡量勞動力。居民消費需求也是經(jīng)濟(jì)增長的主導(dǎo)因素。經(jīng)濟(jì)增長問題既受各國政府和居民的關(guān)注，也是經(jīng)濟(jì)學(xué)理論研究的一個重要方面。在19782

3、008年的31中，我國經(jīng)濟(jì)年均增長率高達(dá)9.6%,綜合國力大大增強(qiáng),居民收入水平與生活水平不斷提高，居民的消費需求的數(shù)量和質(zhì)量有了很大的提高。但是，我國目前仍然面臨消費需求不足問題。因此，研究消費需求對經(jīng)濟(jì)增長的影響，并對我國消費需求對經(jīng)濟(jì)增長的影響程度進(jìn)行實證分析，可以更好的理解消費對我國經(jīng)濟(jì)增長的作用。、數(shù)據(jù)收集與模型的建立（一）數(shù)據(jù)收集表1中國經(jīng)濟(jì)增長影響因素模型時間序列表指標(biāo)國內(nèi)生產(chǎn)總值年末就業(yè)人員數(shù)全社會固定資產(chǎn)投資完成額居民消費價格指數(shù)（上年=100）地區(qū)全國全國全國全國頻度年年年年單位億元萬人億元-199335673.26680813072.3114.7199448637.567

4、45517042.1124.1199561339.96806520019.27117.1199671813.66895022974108.31997797156982024941.12102.8199885195.57063728406.1899.2199990564.47139429854.7298.6200010028074100.42001P110863491100.732002121717.47328043499.9199.252003一1374227373655566.61101.172004161840.27426470477.4

5、3103.882005187318.97464788773.61101.812006219438.574978109998.16101.472007270232.375321137323.94104.772008319515.575564172828.4105.862009349081.475828224598.7799.312010P413030.376105251683.771103.322011489300.676420311485.13105.392012P540367.476704374694.741102.652013595244.476977446294.09102.622014

6、64397477253512020.65101.992015689052.177451561999.83101.442016744127.277603606465.66102資料來源：中經(jīng)網(wǎng)統(tǒng)計數(shù)據(jù)庫（二）模型設(shè)計為了具體分析各要素對我國經(jīng)濟(jì)增長影響的大小，我們可以用國內(nèi)生產(chǎn)總值（y）作為對經(jīng)濟(jì)發(fā)展的衡量，代表經(jīng)濟(jì)發(fā)展；用總就業(yè)人員數(shù)（x2）衡量勞動力；用固定資產(chǎn)投資總額（x3）衡量資本投入：用價格指數(shù)（x4）去代表消費需求。運用這些數(shù)據(jù)進(jìn)行回歸分析。采用的模型如下：Y?i2X23X34X4i其中，y代表國內(nèi)生產(chǎn)總值，x2代表社會就業(yè)人數(shù)，x3代表固定資產(chǎn)投資，x4代表消費價格指數(shù)，i代表隨機(jī)

7、擾動項。我們通過對該模型的回歸分析，得出各個變量與我國經(jīng)濟(jì)增長的變動關(guān)系。三、模型估計和檢驗（一）模型初始估計表2模型初始估計結(jié)果DependentVariable:YMethod:LeastSquaresDate:12/02/17Time:21:55Sample:19932016Includedobservations:24VariableCoefficientStd.Errort-StatisticProb.C-1275271.219020.5-5.8226090.0000X215.836642.1838457.2517240.0000X30.9780160.03143931.108490

8、.0000X42050.542769.38612.6651660.0149R-squared0.995694Meandependentvar273572.7AdjustedR-squared0.995048S.D.dependentvar228431.4S.E.ofregression16074.20Akaikeinfocriterion22.35883Sumsquaredresid5.17E+09Schwarzcriterion22.55517Loglikelihood-264.3060Hannan-Quinncriter.22.41092F-statistic1541.649Durbin-

9、Watsonstat0.587012Prob(F-statistic)0.000000Y?127537115.83664X20.978016X32050.542X4t=(-5.822609)(7.251724)(31.10849)(2.665166)1541.649, DW 0. 5870122R20.99569,R20.99505,F（二）多重共線性檢驗表3相關(guān)系數(shù)矩陣被解釋變量X2X3X4R20.7119540.9822820.088145VIF3.471756.43981.0967X2X21X30.7829237374353251X4-0.6097869809492669X30.7829

10、2373743532511-0.2549475574563468X4-0.6097869809492669-0.25494755745634681表4輔助回歸的R2值根據(jù)多重共線性檢驗，解釋變量之間存在著線性相關(guān)。通過采用剔除變量法，多重共線性的修正結(jié)果如下：剔除X4表5修正多重共線性后的模型DependentVariable:YMethod:LeastSquaresDate:12/02/17Time:22:02Sample:19932016Includedobservations:24VariableCoefficientStd.Errort-StatisticProb.C-776314.9

11、129136.0-6.0116050.0000X211.867971.8147766.5396370.0000X31.0158210.03187431.869690.0000R-squared0.994165Meandependentvar273572.7AdjustedR-squared0.993609S.D.dependentvar228431.4S.E.ofregression18261.19Akaikeinfocriterion22.57941Sumsquaredresid7.00E+09Schwarzcriterion22.72667Loglikelihood-267.9530Han

12、nan-Quinncriter.22.61848F-statistic1788.997Durbin-Watsonstat0.554622Prob(F-statistic)0.000000Y?776314.911.86797X21.015821X3it=(129136.0)(1.813776)22R20.994165,R20.993609,F1788.997,DW0.554622（三）自相關(guān)性檢驗1、殘差圖法20,00015,000,10,000.*5,000.0.*E-5,000_*-10,000_,.-15,000.-20,000_*-25,000!,-30,000-20,000-10,00

13、0010,00020,000E(-1)ResidualActual-Fitted2、回歸檢驗法表6回歸結(jié)果DependentVariable:EMethod:LeastSquaresDate:12/02/17Time:23:30Sample(adjusted):1424Includedobservations:11afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.E(-1)0.0353470.3166090.1116420.9133R-squared-0.006486Meandependentvar-853.3129Adju

14、stedR-squared-0.006486S.D.dependentvar10172.61S.E.ofregression10205.54Akaikeinfocriterion21.38576Sumsquaredresid1.04E+09Schwarzcriterion21.42193Loglikelihood-116.6217Hannan-Quinncriter.21.36296Durbin-Watsonstat1.7616643、D慚驗法取顯著性水平為0.05,本例中k=3,n=24查DW表得臨界值，di1.7616640.5546221.2070,du2.3163，而模型估計的DW值是

15、0.554622,DW值落入?yún)^(qū)域0DWdi,認(rèn)為隨機(jī)擾動項存在正的自相關(guān)。4、相關(guān)圖和QB計量Series:R因口WorlcfileUNTrrLEDr:Untitled_BX|ProcObjectPropertieaPrintNameFreezeSampleGenrSheet0叩tiStatsICorrelogramofRESIDDate:12f03/17Time:00:01Sample:124Includedobservations;11AutccorrelationPartialCorrelationACPAOQ-StatProb1111J11。0490.0490C3410B5311111

16、12-Q39S-0402255720.27C1匚11匚130219-0.2043.40530.3331匚1I二14-0.12S-0.3443.74040.W1n115024000705.11660.40211匚160106-0.171543790.4SS1匚11匚17-0.205-0220594Q9口4351111300970.113739220.495111匚190000-0U37.39220.59C1111110004?-00307.C5730662殘差相關(guān)系數(shù)圖Breusch-Godfrey檢驗法表7BG檢驗結(jié)果Breusch-GodfreySerialCorrelationLMTest

17、:F-statistic0.849638 Prob. F(2,7)0.4674Obs*R-squared2.344024 Prob. Chi-Square(2)0.3097TestEquation:DependentVariable:RESIDMethod:LeastSquaresDate:12/02/17Time:23:22Sample:1324Includedobservations:12Presamplemissingvaluelaggedresidualssettozero.VariableCoefficientStd.Errort-StatisticProb.C155027.8146

18、0100.0.1061760.9184X2-2.08333219.58771-0.1063590.9183X30.0121730.1067170.1140710.9124RESID(-1)0.0950140.3607980.2633430.7999RESID(-2)-0.4843870.372710-1.2996350.2349R-squared0.195335Meandependentvar2.44E-09AdjustedR-squared-0.264473S.D.dependentvar10139.63S.E.ofregression11401.90Akaikeinfocriterion2

19、1.81528Sumsquaredresid9.10E+08Schwarzcriterion22.01733Loglikelihood-125.8917Hannan-Quinncriter.21.74048F-statistic0.424819Durbin-Watsonstat2.078170Prob(F-statistic)0.786883（四）自相關(guān)問題的處理CO迭代法估計參數(shù)表8CO迭代法結(jié)果DependentVariable:YMethod:LeastSquaresDate:12/03/17Time:00:08Sample(adjusted):224Includedobservatio

20、ns:23afteradjustmentsConvergenceachievedafter12iterationsVariableCoefficientStd.Errort-StatisticProb.C-1124646.775903.1-1.4494670.1635X216.6882110.560181.5802950.1305X30.9479700.0966069.8127070.0000AR(1)0.7588740.1904243.9851850.0008R-squared0.997169Meandependentvar283916.1AdjustedR-squared0.996721S

21、.D.dependentvar227746.3S.E.ofregression13040.42Akaikeinfocriterion21.94627Sumsquaredresid3.23E+09Schwarzcriterion22.14374Loglikelihood-248.3821Hannan-Quinncriter.21.99593F-statistic2230.435Durbin-Watsonstat1.603193Prob(F-statistic)0.000000InvertedARRoots.76上圖表明，經(jīng)過調(diào)整后的樣本共包含23個觀測值，迭代12次后收斂，AR(1)前面的系數(shù)為0.758874,即為p的估計值。從以上回歸結(jié)果可知原模型的估計結(jié)果為：y?1124616.68821x20.9480x3t=(-1.4495)(1.5803)(9.8127)修正后的DW=1.603193,進(jìn)行自相關(guān)檢驗，Q統(tǒng)計量如下圖通過上圖可以看