《2023年计量经济学实验报告新编.docx》由会员分享,可在线阅读,更多相关《2023年计量经济学实验报告新编.docx(26页珍藏版)》请在淘文阁 - 分享文档赚钱的网站上搜索。
1、N限加林XIANGNAN【UNIVERSITY奸量经济学课程实践报告1系 部:经济与管理系专 业:国际经济与贸易任课教师:李祖辉老师年级班级:2 02 3级2班组 员:舒冠、张淑琴、梁湘、冯冬雪税收收入影响因素分析基于Evi ews模型的经济计量分析Inc 1 uded o b s ervat i on s: 55Varia b IeCoefficien tS t d. Error t-Sta t istic Pro b .C-915. 0621932.0 9 9 5 -0. 9 8 17210.330 7X 40 .37 9 6 600 . 0 1 0 4993 6.160170 .0 0
2、00R-squ a re dAdj u st e d R-squa r e0.9 6 1 045M e an dep e nd e n t v a r 16 116.57d0.960310S.D. dependent v a r 2994 2 . 14S . E. of re g r ess iAk a ik e i nfo cr i teron596 5 . 1 4 8ion20.260942 0. 3 33 9Sum squa r ed r e sid1.8 9E+0 9S chwarz c rite r ion4Hanna n Q u inn c riteLo g 1 ikelihood
3、- 5 55.1759r .20. 28917F - s t a tis t ic1 307. 5 58Durbin-Wa t s o n s t at0.287228Prob (F-stat i s t i c )0.000000其中加入X 2的方程修正的R-2最大,其方程为:Y=214. 5864+ 0. 8023 7 1 X2(1. 144910)( 1 7 9.4 8 44)修正的 R2= 0. 9 9 83 2 6 S E=7 9 51 8 46 6 F= 32214.66。所以,以x 2为基础,顺次加入其他的自变量逐步回归:当 a=0 . 0 5 时,t a/2 (n-k-1) =
4、 t a/2 (55-4-1) = t 0. 02 5(50)=2. 009Y= - 3 1 4.805 3 +0. 58 3 6 3 8X2 + 0. 050 5 79X1-2. 1 1 6 0 141 9. 932197. 51 3 0 1 9修正的 R 2=0. 99 9 1 82 S E=3 2 9 86. 0 8应为x 1的引入改善了修正的不2和F值且其他回归参数的t检 查在记录上仍然显著,所以保存X1,方程为:Y= -314. 8 053 +0. 5 8 3 638X2 + 0. 050579X1继续:Y = - 3 588.46 5 + 0. 5 8661 9 X2 + 0. 0
5、4 9 981X1+31. 73905X3-1. 7224972 0 . 27 3 907.51 68 7 11. 575 2 9 8修正的(2= 0. 999205 S E= 363601 8 5 F= 22618. 1 0式中X3不显著,删去。继续:Y = -7 3. 73 2 1 6 + 0. 6 8 3 414X2 + -0. 00 1 2 5 8X1+ 0. 0 61678X 4- 0.7 4070 73 0. 9 82 6 8-0.1703508.6605 9 2修正的 R 八 2二 0. 99 9 662 SE = 1 5 4 326 1 3 F= 53312. 6 8X4虽然显
6、著,但是它的引入影响了其他回归参数的估计值的数值, 使得项和xl的回归参数通但是t检查,所以删去x 4 所以:Y= -31 4. 8 053 +0. 5 836 3 8 X2 + 0. 0505 7 9 XI-2. 116 0 1 41 9. 9 32 1 97.5 1 3 0 1 9修正的 R2=0. 9 9 9182 S E= F= 329 8 6. 0 8最小二乘估计为:Depen d e nt V a riable: YM e t h o d: Le a st S q uar e sDate: 11/0 4/15Date: 11/0 4/15Tim e : 14:42Sample: 1
7、96 0 2023Sample: 196 0 2023Included ob s er vat i on s : 5 5Co e f ficie nVariab 1 et St d . Erro r t-St a t i Stic P r ob.XIX2R-squaredAdju s ted R- s q uaredS.E.o f regr e s s io n-314. 8 0 530. 05 0 5790. 5 8 36381 4 8. 7 72 80. 0 067320. 029 2 810.999212 Mean0.999182 S.D.Akaike85 6 . 3 047rion-2
8、. 1160147.5 130191 9. 9 32 1 9d epende nt vard e pendent vari nf o c r i t e0.0 3 920 .0 0 000.0 0 0 01611 6.5729942. 1416. 39 6 11 6. 5051 6. 505SumSumsquared resid38 1 29403Schwarz cr i t erion62Log1 i kelih o o d-447. 8 936Hannan-Q u inn c r iter.16.4 3 84 7F - s tatistic3298 6 . 08Durb i n-Wa t
9、so n stat1 . 0 89983Pr o b(F-st a t istic)0.000000(2)、自相关(序列相关)(2)、自相关(序列相关)令 e= r esidD e pen dent V a ri a ble: EMet hod: Leas t S quar e sDate: 11 / 04/15 Time: 14:49S ample (adjusted): 1 9 6 1 2 0 23Included observa t i o ns: 5 4 af t e r ad justm e n t sVar i ableCo e ffici e ntS t d . Er r ort
10、-Statis t icProb.C3 27. 0 3 7 22 02.74 1 81. 6 1 3 0 730. 1 132X10.0 3 0 1700.0120402 . 5058240. 0156X2-0.1111 9 30.04 4 899-2. 4 765080.016 8T1A2-0.9950640.417083-2. 3 85 7 720. 0 210E(-1)0. 4 376820. 1 284333.40 7 8 570.0013-1.7788R-squa red0. 2 93 9 4 6 M e an depend e nt var9 5Adju s ted R- s qu
11、a r ed 0 . 2 3 6 30 S .D. d e pendent v a r 8 4 8.0834S. E. o f regression 7 4 1 . 13492 6 9 1 476Sum s q ua r ed resi dLog likel i hood-43 0. 8411F-s t at i Stic5.099957Akaike info crit e r ion 16.142 2 6S c h wa r z c riterion 1 6 .32 6 4316.21 3 2Han n a n-Quin n criter.9Du r b in-Wa t s o n sta1
12、.52 1 476Pro b(Fsta t ist i c)0.0 016 23匕乂检查二5 4*不2= 15. 8 730 8 4 X2(1)=3. 84o 然后查看 e (一1)的t检查=3. 407857t (50)=2. 00 9 ,所以,存在一阶自相关。继续:Dependent V a r ia b 1 e: EMe t h od: L east SquaresDa t e : 11/0 4/1 5 Tim e : 15:08Sample (a d juste d ) : 196 2 2 023Included o b servations: 5 3 a ft e r a d j u
13、 stmentsVa r iab 1 eC o e f f i cien tStd. E r r ort-Statisti cProb.C3 5 2. 3 572208. 2 82 31.6 9 17290. 0 9 73XI0. 0 2572 00.0 123302.0859130.042 4X 2-0.0915240.0463 0 8-1.9 764030.0540T1 A2-0.9718 9 00 .422 1 9 9-2 .30 1 9680.0258E(-l )0 .52 5 3 890.1 36 7553.8418180. 0 004E(-2)-0.2 5 6 42 10.1464
14、05-1.75 1 4540 . 0864R- s q u ared0 .33712 1Mean de p e n d ent v a r-5. 6 6124885 5 . 7 1 4Adjusted R-square d0.266602S.D. d e p endent v a r6S. E. of r e gression732.82 2 3Akaike in f oc riterion16. 1 37951 6 .36 1 0Sum sq u ar e d resid25240341S c hwar z cr i ter i on11 6 .223 7Lo g like 1 ihood-
15、4 2 1.6558Hannan-Quin n c r iter.3Fstatis t ic4.78 0 57 5Du r b i n-Wat son stat1.530441Prob(F-sta t i s tic)0. 0 0 129 8aLM检查=5 4*R-2= 17.867413X2(2)=5. 9 9。然后查看 e(-1)e(-2)的t检查,经查表可知e( 2) t值不满足条件,所以,不存在二阶自相关,即存在一阶自相关。自相关的修正:使用迭代法修正得到:D e pe n d e nt Variable: YMe t hod: L e ast S q u a r e sDa t e:
16、 11/04/1 5Da t e: 11/04/1 5Time: 1 5:14Samp 1 e (adjusted): 1961 2023In c lu d ed o bse r v ati o ns: 54 after a d j u stm e ntsConvergence achie v ed a f ter 11 iterationst-S t at i s tiVari a bleCoe fficie n t Std. Err o rProb.X1T1A2AR(1 )R-s q u ared25 4 4.36 622 7 3.4 391. 1 1 91710.2 6 850.1 8
17、89350.0 1 52001 2 .430 0 90. 0 0 000.0 9 310 80.05 4 43 51.7 1 0 4350 .0935-5. 7 4 07271. 7 008 8 3-3.37 5 1440.0 0150.91 4 2 670. 0 6474614. 1207 30. 0 0000.999710Mean depe n d ent var164 1 0. 7 53 0 1 42. 9Adj u sted Rsq u aredAdj u sted Rsq u ared0.999687S. D. d e penden t v arAka i k e i nfo c r
18、 i terS.E.o f r e g r ession5 3 3. 5 896on15.4 851 5139511 7SumSums qu a red residSch war z criter i on1 5. 669321 5.5 5 61LogLog1 i k e lih o o d-413. 0 9 9 1Hannan-Q u i n n criter.F-statis t ic4227 1 .44Durbin Wat son stat1. 1 4607 00.0 0 0 0 0P rob(F-sta t istic)In v er t e d AR R o otsDW= 1 . 1
19、460 7 0,查表得出上限为1. 724,下限为DW= 1.1 4 6 0 701.41 4,所以存在正相关。再次修正自相关使用迭代法修正得到:Depen d e nt V a riab 1 e: YMethod: L e a st Squ a r e sDa t e: 1 1/04/15 T i me: 1 5:1 9Sample (adjusted): 1 9 62 2 023Included obs e r v a t ion s : 5 3 a f ter a d justmentsquaredquared0. 9 99 7 63S. D. depe n d e n t v a rC
20、onvergenc e ach i eve d after 1 0 iteration sV a riabl eCoeff i cie n t S t d. E r ro rt-S t at i sti cP r ob.C1 5 21.2 789 3 4.58 8 71 . 6277510.1 1 03X10.2 1 0 8320.0 1 1 9 141 7.6 963 30.0 0 00X2-0.0058670.0475 5 9-0. 1 23 3 7 00. 9 0 23-5.3 0 340T1A26 1. 080 3 40-4. 909 0 150.0000AR (1)1.44277 2
21、0.1283201 1 .2 4 35 60.0 0 00AR(2)-0. 576584 0. 1 2 9 935-4.4 3 7 4930.0 0 01R- s qu a red0.9 9 9 7 8 6 Mean d e p endent var167 17.4030346.2A d j u s te d R- sS. E. o f r egressonon466. 90 3 9Akaike i n fo cri t e r i o n15.23 63 91 5.4 5S um sq u ared residS um sq u ared resid102459 66S c hw a rz
22、cr i te r io n9455.322Log lik e lihoodLog lik e lihood-3 97. 7 64 5Han nan-Quin n c r iter.F-st a t i s tic43 9 2 3.39Dur b i n-Wa t s o n stat1.8833 6 0Prob (F- s tat i Stic)0.000000.7 2 -.24iInverted A R Roo t s.72+.24 i,DW= 1 . 883 3 6 0 ,查表得出上限为1.7 6 8,下限为1.3 74,由于1. 7 68 DW=1.8 8 3360 x 2 (5) =
23、 11. 07,所以拒绝原假设,表白 模型中的误差存在异方差。修正异方差:使用W二l/abs( r esid)作为权数,得出:De p en d ent Variable: YMeth o d : Least Sq u a r esDate: 11 / 0 4/15 T i me: 1 5:28Sample: 1960 2 0 2 3I ncluded ob s erv a t ions: 55Weighti n g seri e s: 1/ABS(RESID)C o effici e nV a r i abl et Std. Err o r t-St a tistic Prob.C-3 1
24、2 .89 7 0 6 . 0 832 7 2 -5 1 .43 5 630.0000X10. 0 51 3 79 0. 001 7 5629.2 61060.0000X 20.579503 0.0 0 8 8 0 2 6 5. 8 40 1 10 .0 0 00Weighted Stati s tic sMean depend ent vaR-square dR-square d0 . 9997543107.38 5Adj us t ed R-squar e d0.9 9 9744S.D. d e p ende n t var5 593.49 0Akai k e i n fo c r i t
25、erS.E. of reg r ession97.82912on1 2.0 5 732Sum squar e d r e sid497667. 9S c h warz c r it e rion1 2 . 16681Hann a n Qu i nn c rLog 1 ik e li h oo d-328. 5 76 4 iter.1 2.0996610 5 6F-statis t icF-statis t ic37. 8 Durbin-W a tson stat0 . 656664P r ob(F-s t atist i c )0.000 0 00Un we i g h t e d Sta t
26、 i sti c sR-squa red0.9 99211Mean dependentvar1 611 6 .57Adjust e d R-squ a redS.D.dependent var29942. 1 4S .E. o f re g ression8 5 6.813 3Sum sq u ared resi d3 8 174711Du r b i n-Wat s on s t at 1 .07933 9估计的结果为:+ 0.5795 0 3x 2y =-3 12 .8 9 70 +0.051379x151.435651.4356(29. 2 6 1 0 6 )(6 5.84 0 1 1R
27、 2=0. 9 9 9754R 2=0. 9 9 9754D W= 0 . 656664F= 1 05 6 37. 8再次做怀特检查:Depende n t Vari a bl e : EMetho d : Lea s t S q uar e sD ate: 11/04/1 5 Tim e : 15: 34Sampl e : 19 6 0 2023I nclude d ob s e rvations: 55Wei g hting series: 1/ABS (RESID)VariableCoefficie n t Std. E r ror t -StatisticProb.X11 1 9 .8
28、7460. 0 2 144 23 1 .1 54373. 8 4 7 76 30.00030.0039285. 45824 90.0 0 00-0 .1 7 82 7X2X20.0 3 9257-4.54 1 2850.000 05.0 7 E 0X1A2X1A21 . 87E-07- 2.7 096590. 00 9 3-1.4 1E-0 53.74E- 0 6-3.7746330.00045.63E-061 .65E-063.4 1 24310 .0013Weighted Stat i sties2 6.712Rsquared0.3 9 4 1 56 Mean de p ende n t
29、v a r24Adj us ted R s q uard e pendent var9 1.44191Akaik ei nfo c r i tS.E.o f regress i o n77.76 0 88erion11.647 8 2S u m sq u a r e d r e s0 S ch warz c ri t erion11.86 680Log li k e 1 iho o d-3 1 4. 3 1 5 1 Han nan- Qui n n cr i ter.1 1 .7325 1Fst a t i SticProb( F -sta t is t ic)R-sq u ar e dAdj
30、 u sted R-S q u a r e dS .E. o f r e gr e s s io nDurb i n-Watson stat0 . 0 00 1 246 .375776 Durbi n W a tson stat 0. 7 0 27 7 0Un weigh t ed Statist i c s-1.41 E-128 40.2976236700880.3 7 921 7 Mean d epe n dent var0.31 5 8 72 S. D. dep e nd e nt va r695.0273 S um sq u ared resi d0 .96467 1从表中可以看出,n
31、K2=5 5 * 0.3 9 4 1 56=21. 6 7 85 8 ,由 wh i te 检查知,在a=0. 0 5下,查表得知X2 (5)=1 1 . 0705,由于 nlT 2=21.67858 X2 (5)= 11. 0 7 05,所以拒绝原假设,表白模型 中的误差存在异方差。继续修正:使用w= 1/(resid厂2作为权数,得出:D e pendent Variab 1 e: YMethod: L e ast S q u a resDa t e: 11/04/1 5 Time: 15:39Sample: 1960 2 023I nc 1 u d ed o b s erv a tio
32、ns: 5 5Weig h t i n g se r ies: 1 / (RE S ID) A 2Co e ffic i eV a r i abl ent Std. E r ror tSta t istic Prob.C-291.44 3 50 . 512662 -5 6 8.4 9 020.0000X10.0 5 6905 0. 000 1 573 61.723 20.0000X20.537500 0. 0 0 0 9 01596. 47750.000 0Weight e d StatisticsR-squar e d1.0000 0 0 Mean d e pe n dent var 2 0
33、 231.12A dj u s t ed R-squared 1.0 0 0 000 S.D. depen d e nt var146874.0S. E. of r e gre s s i o n 2.48 1 680 Akaike info cri t e r ion4. 7 0 8一、意义1960年以来,中国的经济基本处在高速增长之中。经济增长的 高速发展,势必会影响国家财政政策和国家福利水平。而税收作为国 家财政收入中最重要的部分对这些政策的实行也会有很大的影响。近 些年来,国家的税收也受到多种因素的影响。经济发展水平决定税收收入水平,税收同时也反作用于经济。要 实现经济的连续增长,必须
34、规定与经济紧密关联的税收符合其发展的 规定,即政府筹集的税收收入应尽也许的满足其实现职能的需求,同 时又不至于损害经济的发展。影响未来的需求,我们需要研究影响中 国税收收入的重要因素,分析中央和地方税收收入增长的数量规律, 从结构上对税收收入的影响做一个很好的了解,对于预测中国税收未 来的增长趋势具有重要的作用,对于我国的社会主义现代化建设具有 重要意义。二、研究综述影响税收收入的因素有很多,如经济发展水平、税收制度的设 计、政府职能范围等。李卫刚认为,影响税收增长的因素是多元的, 重要有经济增长、税制结构、税收征管水平和价格因素;孙玉栋认为, 影响税收收入增长的因素重要有经济增长、物价、税收
35、政策调整和税 收征管等几个方面;安体富认为,税收收入重要受价格、经济结构的变 动、经济效应的变动、税收政策、财税制度、税收征管和税款虚收的 影响。以上学者都说明税收收入的影响因素是多方面的,同时都认为 经济因素的重要性,但他们多集中于对税收收入影响因素的全面分析,750Sum s q u ar e d re s3 2 0. 2 543 S c hw a rz criterion4.818 24 1Hanna n-Quinn c4. 75 1 09Log li k e 1 ih o odF-stat i s t icP ro b (F -stat i Stic)-12 6.49 0 6riter
36、.Dur binWa t s on s t2.56E+ 09 at0. 0000 0 01.3070 1 6Unw e i g h t ed Statist i csR-s q u aredR-s q u aredAdj us t ed Rs q u ared0.9 9 8 50 6S .E. of r e gression 11 5 7. 354Mean dependent var 161 1 6.57S . D. d epen dent v ar 299 4 2 .14 Sum s qua red resid 69652 3 9 1Durbin-Watso n statDurbin-Wat
37、so n stat0. 5597 0 4估计的结果为:537 5 OOx 2(59y =-29 1 . 4 4 3 5 +0. 0 5 6 905x1 + 0 .(-56 8 . 4 902)(36 1 . 7232)6.4775)R 2=1. 0 0 0000 DW= 1 . 307016 F=2. 56E+ 0 9从以上结果可看出存在自相关。同上自相关的修正,最终得D epen d ent Vari a bl e : EMethod: Least Squar e sDa t e : 1 1 / 04/15 T i me: 1 5: 5 1Samp 1 e (adj usted): 1962
38、 2 023I nclud e d obse r v a t ion s : 5 3 aft e r a dju s tmentsC onver g e n c e ach i e ve d after 9 ite r ationsVa r i able Coef f icient St d . Error t-Statis t i c Prob.c1836. 078934. 574 81.9 6 461 30. 05 5 4XI0.16 0 2530 .011 9 1 41 3 .4 5 0 8 60 .0 0 00X2-0. 5895050. 04 7 559-1 2 .395 3 60.
39、 0 000T1A2AR(1)AR (2)T1A2AR(1)AR (2)-5.3 0 3 3941.0 8 032 6-4.9 0 90 670. 0 0001. 44277 40.12 8 32011.243 5 90.0000-0.5765 850.1 2 9935 - 4.4375 0 6 0.0 0 0 1R-squa red0. 7309 14Mean d e pendent var-5. 6 6 1248Ad ju s ted R- s qua r ed0 .70 2 2 8 7S.E. o f reg r e ssion 4 66. 9 039Sum s q u a red res i d 10 2 459 6 6S .D. d e p ende nt v ar 855.71 4 6 Akaike info c rit e r i on 15.2 3 6 39S c h war z criterio n 15.45945 Hanna n-Qui n n c r i tL o g lik e lihoodF-s t atist i c-3 97. 764 5 e r .25.5330 2 Du r bin-W a t son stat1 5.322171.883 3 63 7 2 -.24iPro b (F-s t ati s ti
限制150内