- 相關(guān)推薦
我國涉外旅游業(yè)收入的實(shí)證分析
我國涉外旅游業(yè)收入的實(shí)證分析
概況分析
涉外旅游.是指我國旅游部門經(jīng)營對(duì)外招徠并接待外國人、外籍華人和華僑等國際旅行者旅游業(yè)務(wù)的活動(dòng)。涉外旅游業(yè)是一項(xiàng)新型的綜合性的經(jīng)濟(jì)事業(yè)。旅游業(yè)是日益興起的新型“朝陽產(chǎn)業(yè)”,它的發(fā)展無論是對(duì)一國的經(jīng)濟(jì),還是國際間的交流,都有著重要和積極的作用。中國是世界上旅游資源和市場(chǎng)最豐富的國家,把握這一優(yōu)勢(shì),大力發(fā)展旅游業(yè),對(duì)正在深入的改革開放和產(chǎn)業(yè)結(jié)構(gòu)的優(yōu)化,都有著廣泛的促進(jìn)作用。
近年來,我國旅游業(yè)突飛猛進(jìn)。隨著我國對(duì)外開放的逐步深入,涉外旅游業(yè)也獲得了長(zhǎng)足的發(fā)展。它是我國國民經(jīng)濟(jì)和發(fā)展對(duì)外經(jīng)濟(jì)關(guān)系的一個(gè)重要組成部分,是第三產(chǎn)業(yè)的重要部門。
中國旅游市場(chǎng)在21世紀(jì)將進(jìn)一步擴(kuò)大,中國豐富的旅游資源不斷得到開發(fā);旅游產(chǎn)品結(jié)構(gòu)不斷完善;旅游產(chǎn)業(yè)規(guī)模不斷擴(kuò)大;發(fā)展旅游的大環(huán)境逐漸優(yōu)化,這些都為中國旅游市場(chǎng)的擴(kuò)大提供了堅(jiān)實(shí)的保障。我國涉外旅游市場(chǎng)將會(huì)繼續(xù)擴(kuò)大,亞洲是中國的最大客源市場(chǎng),隨著東南亞金融危機(jī)的過去,東南亞、日本的經(jīng)濟(jì)復(fù)蘇,亞洲客源肯定有較大的發(fā)展;歐美遠(yuǎn)程客源國來華人數(shù)都在不斷增長(zhǎng),在中國國際旅游市場(chǎng)上,來自歐美的游客只是一個(gè)全球的平均水平,歐美來華旅游的潛力顯然很大。
二. 模型的建立
我們通過分析我國涉外旅游業(yè)的收入,根據(jù)理論及對(duì)現(xiàn)實(shí)情況的認(rèn)識(shí),建立了一個(gè)單一方程模型:
Y=ß1+ß2X2+ß3X3+ß4X4+ß5X5+U (1.1)
其中:Y——我國涉外旅游業(yè)收入(億元)
X2——涉外飯店數(shù)目(個(gè))
X3——旅游人數(shù)(萬人)
X4——涉外旅游業(yè)職工人數(shù)(人)
X5——涉外旅行社個(gè)數(shù)(個(gè))
U ——隨及擾動(dòng)項(xiàng)
ßi——參數(shù)
模型的估計(jì)和檢驗(yàn)
估計(jì)
設(shè)模型中的隨及誤差項(xiàng)U滿足古典假定,運(yùn)用OLS方法估計(jì)未知參數(shù),利用計(jì)量經(jīng)濟(jì)學(xué)計(jì)算機(jī)軟件Eviews計(jì)算的過程如下:
1.建立文檔,輸入數(shù)據(jù)
首先點(diǎn)擊Eviews圖標(biāo),進(jìn)入Eviews主頁。建立新的Workfile工作框,并輸入數(shù)據(jù),見表一。
表一
obs X2 X3 X4 X5 Y
1991 2130.000 3335.000 38177.00 671.0000 28.40000
1992 2354.000 3811.500 40258.00 852.0000 39.50000
1993 2552.000 4152.700 45431.00 987.0000 46.80000
1994 2995.000 4368.450 57600.00 1110.000 73.23000
1995 3720.000 4638.650 59935.00 1025.000 87.33000
1996 4418.000 5112.750 53093.00 977.0000 102.0000
1997 5201.000 5758.790 48881.00 991.0000 120.7400
1998 5782.000 6347.840 52290.00 1312.000 126.0200
1999 7035.000 7279.560 47153.00 1256.000 140.9900
OLS估計(jì)未知參數(shù)
在主頁上選Group菜單,點(diǎn)擊Estimate Equation項(xiàng),對(duì)數(shù)據(jù)進(jìn)行OLS估計(jì),結(jié)果如表二
表二
Dependent Variable: Y
Method: Least Squares
Date: 05/15/04 Time: 10:29
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X2 0.010831 0.019765 0.548008 0.6128
X3 0.019274 0.031260 0.616577 0.5709
X4 0.001652 0.000559 2.953751 0.0418
X5 -0.046484 0.050411 -0.922104 0.4087
C -88.36498 52.76179 -1.674791 0.1693
R-squared 0.988138 Mean dependent var 85.00111
Adjusted R-squared 0.976277 S.D. dependent var 40.71376
S.E. of regression 6.270853 Akaike info criterion 6.809883
Sum squared resid 157.2944 Schwarz criterion 6.919452
Log likelihood -25.64447 F-statistic 83.30613
Durbin-Watson stat 1.852069 Prob(F-statistic) 0.000419
檢驗(yàn)
經(jīng)濟(jì)意義檢驗(yàn)
X5的系數(shù)與其經(jīng)濟(jì)意義不符。我們將通過對(duì)模型的修正看是否能得到更好的結(jié)果。
2. 統(tǒng)計(jì)檢驗(yàn)
對(duì)回歸系數(shù)進(jìn)行整體檢驗(yàn),該檢驗(yàn)是在方差分析的基礎(chǔ)上利用F檢驗(yàn)進(jìn)行的。
由上表數(shù)據(jù),F(xiàn)=248.8175>F0.05(4,4),應(yīng)該拒絕原假設(shè)H0,說明回歸方程顯著。
所以從模型從整體上看,涉外旅游收入與解釋變量之間線形關(guān)系顯著
3. 計(jì)量經(jīng)濟(jì)學(xué)檢驗(yàn)
(1).多重共線性檢驗(yàn)
在Quick菜單中選取項(xiàng)Group Statistics中的Correlation命令,輸入變量名即可得到如下結(jié)果:
表三
X2 X3 X4 X5
X2 1.000000 0.991841 0.297759 0.777257
X3 0.991841 1.000000 0.302919 0.832983
X4 0.297759 0.302919 1.000000 0.573134
X5 0.777257 0.832983 0.573134 1.000000
由表3可以看出,解釋變量之間存在高度線性相關(guān)。同時(shí)由表2得到的可決系數(shù)很大,而且F統(tǒng)計(jì)量值顯著的大于給定顯著性水平下的臨界值,而x2、x3、x5變量的偏回歸系數(shù)t統(tǒng)計(jì)量值并不顯著,且X5的系數(shù)的符號(hào)與經(jīng)濟(jì)意義相悖。盡管整體上線性回歸擬合較好,但模型中解釋變量存在嚴(yán)重的多重共線性。
對(duì)多重共線性的修正
a).運(yùn)用OLS方法逐一求對(duì)各個(gè)解釋變量的回歸。結(jié)合經(jīng)濟(jì)意義和統(tǒng)計(jì)檢驗(yàn)選出擬合效果最好的一元線形回歸方程。
Y對(duì)X2的回歸
表四
Dependent Variable: Y
Method: Least Squares
Date: 05/15/04 Time: 14:56
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X2 0.023098 0.002185 10.57110 0.0000
C -7.868897 9.465154 -0.831354 0.4332
R-squared 0.941052 Mean dependent var 85.00111
Adjusted R-squared 0.932630 S.D. dependent var 40.71376
S.E. of regression 10.56752 Akaike info criterion 7.746577
Sum squared resid 781.7067 Schwarz criterion 7.790404
Log likelihood -32.85959 F-statistic 111.7481
Durbin-Watson stat 0.736732 Prob(F-statistic) 0.000015
將上述回歸結(jié)果整理如下:
Y=--7.868897+0.023098X2 (1.2)
(-0.8313)(10.5711)
R^2=0.9410 ,S.E =10.5675, F=111.7481
Y對(duì)X3的回歸
表五
Dependent Variable: Y
Method: Least Squares
Date: 05/15/04 Time: 15:06
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X3 0.030641 0.003358 9.126027 0.0000
C -67.53941 17.19614 -3.927590 0.0057
R-squared 0.922467 Mean dependent var 85.00111
Adjusted R-squared 0.911391 S.D. dependent var 40.71376
S.E. of regression 12.11937 Akaike info criterion 8.020616
Sum squared resid 1028.153 Schwarz criterion 8.064444
Log likelihood -34.09277 F-statistic 83.28438
Durbin-Watson stat 0.775610 Prob(F-statistic) 0.000039
將上述回歸結(jié)果整理如下:
Y= -67.53941+0.030641X3 (1.3)
(-3.9275)(9.1260)
R^2= 0.9224, S.E=12.1193, F=83.2843
Y對(duì)X4的回歸
表六
Dependent Variable: Y
Method: Least Squares
Date: 05/15/04 Time: 15:07
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X4 0.002713 0.001830 1.482325 0.1818
C -48.47780 90.93217 -0.533120 0.6105
R-squared 0.238906 Mean dependent var 85.00111
Adjusted R-squared 0.130178 S.D. dependent var 40.71376
S.E. of regression 37.97137 Akaike info criterion 10.30467
Sum squared resid 10092.77 Schwarz criterion 10.34850
Log likelihood -44.37102 F-statistic 2.197288
Durbin-Watson stat 0.301641 Prob(F-statistic) 0.181813
將上述回歸結(jié)果整理如下:
Y=-48.4778+0.0027X4 (1.4)
(-0.5331) (1.4823)
R^2=0.2389,S.E=37.9713,F(xiàn)=2.1972
Y對(duì)X5的回歸
表七
Dependent Variable: Y
Method: Least Squares
Date: 05/15/04 Time: 15:08
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X5 0.168492 0.046892 3.593223 0.0088
C -86.87950 48.60200 -1.787570 0.1170
R-squared 0.648440 Mean dependent var 85.00111
Adjusted R-squared 0.598217 S.D. dependent var 40.71376
S.E. of regression 6 Akaike info criterion 9.532296
Sum squared resid 4661.995 Schwarz criterion 9.576123
Log likelihood -40.89533 F-statistic 12.91125
Durbin-Watson stat 1.119498 Prob(F-statistic) 0.008819
將上述回歸結(jié)果整理如下:
Y=-86.8795+0.1684X5 (1.5)
(-1.7875)(3.5932)
R^2 =0.6484,S.E=25.8069, F=12.9112
分析:
通過對(duì)多重可決系數(shù)和t統(tǒng)計(jì)量的觀察,X2,X3的可決系數(shù)接近1,且t的絕對(duì)值都遠(yuǎn)大于2,所以模型對(duì)數(shù)據(jù)的擬合程度較好。同時(shí)結(jié)合經(jīng)濟(jì)意義,旅游人數(shù)x3較涉外飯店數(shù)x2對(duì)涉外旅游收入y的影響更大,選出x3,得一元線性回歸方程:
Y=-67.53941+0.030641X3
b. 逐步回歸。引入其余解釋變量,得到以下模型。
表八
Dependent Variable: Y
Method: Least Squares
Date: 05/16/04 Time: 00:02
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X3 -0.003361 0.024767 -0.135705 0.8965
X2 0.025586 0.018485 1.384133 0.2156
C -1.140102 50.62397 -0.022521 0.9828
R-squared 0.941232 Mean dependent var 85.00111
Adjusted R-squared 0.921643 S.D. dependent var 40.71376
S.E. of regression 11.39675 Akaike info criterion 7.965734
Sum squared resid 779.3148 Schwarz criterion 8.031476
Log likelihood -32.84580 F-statistic 48.04823
Durbin-Watson stat 0.742719 Prob(F-statistic) 0.000203
表九
Dependent Variable: Y
Method: Least Squares
Date: 05/16/04 Time: 00:03
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X3 0.028536 0.002536 11.25191 0.0000
X4 0.001209 0.000441 2.740175 0.0337
C -116.5457 21.75047 -5.358306 0.0017
R-squared 0.965563 Mean dependent var 85.00111
Adjusted R-squared 0.954084 S.D. dependent var 40.71376
S.E. of regression 8.724176 Akaike info criterion 7.431275
Sum squared resid 456.6675 Schwarz criterion 7.497016
Log likelihood -30.44074 F-statistic 84.11512
Durbin-Watson stat 1.174184 Prob(F-statistic) 0.000041
表十
Dependent Variable: Y
Method: Least Squares
Date: 05/16/04 Time: 00:05
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X3 0.030188 0.006551 4.608392 0.0037
X5 0.003566 0.042964 0.083004 0.9365
C -68.92251 24.94512 -2.762966 0.0327
R-squared 0.922556 Mean dependent var 85.00111
Adjusted R-squared 0.896742 S.D. dependent var 40.71376
S.E. of regression 13.08290 Akaike info criterion 8.241691
Sum squared resid 1026.974 Schwarz criterion 8.307432
Log likelihood -34.08761 F-statistic 35.73773
Durbin-Watson stat 0.780006 Prob(F-statistic) 0.000464
Y=-1.140102+0.025586X2+-0.003361X3
(-0.0225) (1.3841) (-0.1357)
Adjusted R^2=0.9216 S.E=11.3967 F=48.0482
Y=-116.5457+0.028536X3+0.001209X4
(-5.3583) (11.2519) (2.7401)
Adjusted R^2=0.9540 S.E= 8.7241 F=84.1151
Y=-68.92251+0.030188X3+0.003566X5
(-2.7629) (4.6083) (0.0830)
Adjusted R^2=0.8967 S.E=13.0829 F=35.7377
可以看出X2和X5的對(duì)模型的擬合優(yōu)度并無改善,同時(shí)對(duì)X3影響較小。
X4在符合經(jīng)濟(jì)意義的前提下,使擬合優(yōu)度提高,每個(gè)參數(shù)統(tǒng)計(jì)檢驗(yàn)顯著,應(yīng)采納該變量。
得到一個(gè)二元回歸方程:
Y=-116.5457+0.028536X3+0.001209X4
表十一
Dependent Variable: Y
Method: Least Squares
Date: 05/16/04 Time: 22:03
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X3 -0.006655 0.013448 -0.494878 0.6417
X4 0.001227 0.000312 3.928074 0.0111
X2 0.026456 0.010020 2.640370 0.0460
C -48.62606 29.98005 -1.621947 0.1657
R-squared 0.985617 Mean dependent var 85.00111
Adjusted R-squared 0.976987 S.D. dependent var 40.71376
S.E. of regression 6.176249 Akaike info criterion 6.780402
Sum squared resid 190.7303 Schwarz criterion 6.868057
Log likelihood -26.51181 F-statistic 114.2115
Durbin-Watson stat 1.743783 Prob(F-statistic) 0.000050
表十二
Dependent Variable: Y
Method: Least Squares
Date: 05/16/04 Time: 22:09
Sample: 1991 1999
Included observations: 9
Variable Coefficient Std. Error t-Statistic Prob.
X3 0.036303 0.003155 11.50697 0.0001
X4 0.001866 0.000371 5.036388 0.0040
X5 -0.070169 0.024064 -2.915919 0.0332
C -115.9801 14.50022 -7.998507 0.0005
R-squared 0.987248 Mean dependent var 85.00111
Adjusted R-squared 0.979597 S.D. dependent var 40.71376
S.E. of regression 5.815561 Akaike info criterion 6.660054
Sum squared resid 169.1038 Schwarz criterion 6.747709
Log likelihood -25.97024 F-statistic 129.0310
Durbin-Watson stat 1.949370 Prob(F-statistic) 0.000037
分析:
由表11、表12知,分別引入x2或x5后,他們對(duì)y的影響并不顯著,故將x2和x5刪除,
此后統(tǒng)計(jì)檢驗(yàn)效果均有較大改善。
綜上所述,選擇此模型為修正后的模型:
Y=-116.5457+0.028536X3+0.001209X4
(2)自相關(guān)檢驗(yàn)
D-W檢驗(yàn):
根據(jù)表9估計(jì)的結(jié)果,由DW=1.174184,給定顯著性水平a=0.05,查Durbin-Watson表,n=9,k=2,得下限臨界值dl=0.629,du=1.699,因?yàn)镈W統(tǒng)計(jì)量為1.174184,在(0.629,1.699)中,所以根據(jù)判定區(qū)域值,這時(shí)隨機(jī)誤差項(xiàng)不能確定是否存在自相關(guān)
圖示法:
用上述OLS估計(jì),可以得到殘差resid,運(yùn)用GENR生成序列E1和E2(E2=E1^2).
在Quick菜單中選Graph項(xiàng),在對(duì)話框中鍵入E1 E1(-1),得到下圖:
考慮用Cochrane-Orcutt迭代法檢驗(yàn)
Dependent Variable: Y
Method: Least Squares
Date: 05/21/04 Time: 11:41
Sample(adjusted): 1992 1999
Included observations: 8 after adjusting endpoints
Convergence not achieved after 100 iterations
Variable Coefficient Std. Error t-Statistic Prob.
C 1416.631 14361.16 0.098643 0.9262
X3 -0.002176 0.024544 -0.088676 0.9336
X4 9.28E-05 0.000734 0.126480 0.9055
AR(1) 0.988777 0.132381 7.469197 0.0017
R-squared 0.969215 Mean dependent var 92.07625
Adjusted R-squared 0.946126 S.D. dependent var 37.14207
S.E. of regression 8.620954 Akaike info criterion 7.453121
Sum squared resid 297.2834 Schwarz criterion 7.492842
Log likelihood -25.81249 F-statistic 41.97768
Durbin-Watson stat 2.766962 Prob(F-statistic) 0.001759
Inverted AR Roots .99
分析:從上圖可以看出殘差et并未呈線形回歸,表明隨機(jī)誤差ut不存在自相關(guān)。
2.異方差的檢驗(yàn)
ARCH檢驗(yàn)
表十三
ARCH Test:
F-statistic 0.237244 Probability 0.865536
Obs*R-squared 1.574786 Probability 0.665120
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 05/18/04 Time: 10:14
Sample(adjusted): 1994 1999
Included observations: 6 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 97.03369 115.0346 0.843517 0.4877
RESID^2(-1) -0.265903 0.819938 -0.324296 0.7765
RESID^2(-2) -0.465400 0.619188 -0.751629 0.5307
RESID^2(-3) 0.201437 1.401556 0.143724 0.8989
R-squared 0.262464 Mean dependent var 56.66340
Adjusted R-squared -0.843839 S.D. dependent var 71.89270
S.E. of regression 97.62169 Akaike info criterion 12.23480
Sum squared resid 19059.99 Schwarz criterion 12.09597
Log likelihood -32.70439 F-statistic 0.237244
Durbin-Watson stat 1.972026 Prob(F-statistic) 0.865536
由表13輸出結(jié)果知,obs*R2=1.574786,查X2分布表,給定a=0.05,自由度為P=3,得臨界值為7.81,因?yàn)?.574786<7.81,所以接受原假設(shè)H 0。表明模型中隨機(jī)誤差項(xiàng)不存在異方差。
圖示法
分析:從上圖中可以看出,模型中隨機(jī)誤差項(xiàng)不存在異方差。
我們進(jìn)行了一系列檢驗(yàn)和修正后該模型的最終結(jié)果如下:
Y=-116.5457+0.028536X3+0.001209X4
(-5.3583) (11.2519) (2.7401)
Adjusted R^2=0.9540 S.E= 8.7241 F=84.1151
從模型中可看出:
X3,X4均符合經(jīng)濟(jì)意義的檢驗(yàn)。從經(jīng)濟(jì)意義上看,涉外旅游業(yè)的收入隨著旅游人數(shù)和涉外旅游職工人數(shù)的增加而增加。
模型表明:涉外旅游業(yè)的收入與旅游人數(shù)和涉外旅游職工人數(shù)有明顯的相關(guān)關(guān)系。實(shí)際上,這個(gè)結(jié)論也是很合理的。
模型的修正可決系數(shù)及F值比較理想,模型的擬合優(yōu)度好。
由上述分析可知,我們的模型還是成功的。
結(jié)論部分:
通過以上的實(shí)證分析可以看出涉外旅游業(yè)的收入與旅游人數(shù)、涉外旅游業(yè)職工人數(shù)有著密切的關(guān)系。
2000年我國接待入境過夜旅游人數(shù)達(dá)3122.88萬人次,旅游外匯收入則達(dá)到162.24億美元,國際排名分別為第5和第7。在未來旅游業(yè)的發(fā)展方面,我國的自然風(fēng)光資源和社會(huì)歷史文化資源還遠(yuǎn)遠(yuǎn)沒有得到開發(fā),尤其是我國的中西部,眾多的自然資源還沒有向游客揭開神秘的面紗。隨著東部旅游資源的深度開發(fā),中西部旅游資源的相繼開發(fā)和旅游條件的改善,我國對(duì)世界的吸引力將會(huì)越來越大。據(jù)世界旅游組織預(yù)測(cè),2020年,我國將成為世界最大的旅游目的地國家,接待旅游人次達(dá)1.37億,同時(shí)也成為世界十大旅游客源國之一,出游人次達(dá)1億。另外,旅游收入在國民生產(chǎn)總值中的份額也呈明顯的增長(zhǎng)趨勢(shì),旅游業(yè)與國民經(jīng)濟(jì)的相關(guān)程度越來越高。
主要參考文獻(xiàn):
中華人民共和國國家旅游局,《中國旅游統(tǒng)計(jì)年鑒》,中國旅游出版社,2000
林南枝、李天元,《旅游市場(chǎng)學(xué)》,南開大學(xué)出版社,1996
羅明義,《旅游經(jīng)濟(jì)學(xué)》,高等教育出版社,1998
【我國涉外旅游業(yè)收入的實(shí)證分析】相關(guān)文章:
我國中小企業(yè)成長(zhǎng)性的實(shí)證研究06-08
我國涉外產(chǎn)品責(zé)任法律適用立法之完善06-10
科技進(jìn)步貢獻(xiàn)率測(cè)算及預(yù)測(cè)實(shí)證分析05-29
我國督促程序的現(xiàn)狀及原因分析06-12
我國物權(quán)法中預(yù)告登記的范圍分析07-19
我國國內(nèi)旅游業(yè)發(fā)展現(xiàn)狀和趨勢(shì)研究10-29
我國區(qū)域經(jīng)濟(jì)發(fā)展形勢(shì)分析06-04
分析我國競(jìng)爭(zhēng)力低下的原因及對(duì)策08-27