【正文】
而文教支出在財(cái)政支出中也占著很大的比例,因此必然存在著長(zhǎng)期均衡的關(guān)系。進(jìn)一步做誤差修正模型如下,結(jié)果顯示非常良好。Dependent Variable: D_WJMethod: Least SquaresDate: 01/06/10 Time: 22:24Sample (adjusted): 1954 2006Included observations: 53 after adjustmentsVariableCoefficientStd. ErrortStatisticProb.D_GDPD_CZZCECM(1)CRsquaredMean dependent varAdjusted Rsquared. dependent var. of regressionAkaike info criterionSum squared residSchwarz criterionLog likelihoodHannanQuinn criter.FstatisticDurbinWatson statProb(Fstatistic)然后為了拓展我們的研究,在不區(qū)分解釋變量與被解釋變量的前提下嘗試進(jìn)行基于VAR模型的Johanson檢驗(yàn)。得到結(jié)果如下:表12:Johanson檢驗(yàn)結(jié)果在EVIEWS中的表達(dá):Date: 01/06/10 Time: 22:03Sample (adjusted): 1955 2006Included observations: 52 after adjustmentsTrend assumption: No deterministic trendSeries: WJ GDP CZZCLags interval (in first differences): 1 to 1Unrestricted Cointegration Rank Test (Trace)HypothesizedTraceNo. of CE(s)EigenvalueStatisticCritical ValueProb.**None *At most 1 *At most 2Trace test indicates 2 cointegrating eqn(s) at the level* denotes rejection of the hypothesis at the level**MacKinnonHaugMichelis (1999) pvaluesUnrestricted Cointegration Rank Test (Maximum Eigenvalue)HypothesizedMaxEigenNo. of CE(s)EigenvalueStatisticCritical ValueProb.**None *At most 1 *At most 2Maxeigenvalue test indicates 2 cointegrating eqn(s) at the level* denotes rejection of the hypothesis at the level**MacKinnonHaugMichelis (1999) pvalues根據(jù)判別原則,跡統(tǒng)計(jì)量和最大特征根值均大于任何顯著性水平下的臨界值,因此,我們拒絕協(xié)整關(guān)系個(gè)數(shù)為0的原假設(shè),即協(xié)整向量的個(gè)數(shù)至少為1,或者說(shuō)國(guó)內(nèi)生產(chǎn)總值增量、文教支持增量以及財(cái)政支出增量三者之間至少存在一個(gè)協(xié)整關(guān)系。我們可以初步推斷出文教支出必然會(huì)隨著國(guó)內(nèi)生產(chǎn)總值的增加而增加。結(jié)果與現(xiàn)實(shí)經(jīng)濟(jì)相吻合。基于以上研究,我們進(jìn)一步做出三個(gè)變量的Granger因果檢驗(yàn)結(jié)果如下:表13:原變量Granger因果關(guān)系檢驗(yàn)在EVIEWS中的表達(dá):Pairwise Granger Causality TestsDate: 01/06/10 Time: 20:31Sample: 1953 2006Lags: 2Null Hypothesis:ObsFStatisticProb.WJ does not Granger Cause GDP52GDP does not Granger Cause WJCZZC does not Granger Cause GDP52GDP does not Granger Cause CZZCCZZC does not Granger Cause WJ52WJ does not Granger Cause CZZC可見(jiàn),均拒絕了原假設(shè),也就是說(shuō)這三個(gè)變量?jī)蓛芍g都不能互相作為Granger原因,即不能夠互相解釋。我們對(duì)未差分的原變量進(jìn)行Granger因果檢驗(yàn),結(jié)果未改變。九、總結(jié)本文主要使用計(jì)量經(jīng)濟(jì)學(xué)的分析方法,分析了我國(guó)文教支出的影響因素,得出以下幾個(gè)結(jié)論: 我國(guó)文教支出增長(zhǎng)率受GDP增長(zhǎng)率和財(cái)政支出增長(zhǎng)率的影響。這兩個(gè)因素對(duì)文教支出增長(zhǎng)率的作用都是正向的。也就是說(shuō),隨著GDP增長(zhǎng)加速和財(cái)政支出增長(zhǎng)加速,文教支出的增長(zhǎng)也將加速。而且,它們的影響大小基本相等。 改革開(kāi)放前后,我國(guó)文教支出有明顯的不同。 我國(guó)文教支出和GDP、財(cái)政支出之間存在一個(gè)協(xié)整關(guān)系,即GDP和財(cái)政支出對(duì)文教支出有長(zhǎng)期均衡存在。31 / 31