我国国内生产总值的实证分析

文档格式：DOCX| 30 页|大小 147.88KB|积分 20|2022-09-23 发布|文档ID：155369596

下载文档

下载文档到电脑，查找使用更方便还剩页未读，继续阅读>>

侵权申诉举报

1 / 30

此文档下载收益归作者所有下载文档

版权提示

文本预览

常见问题

计量经济学课程设计——我国国内生产总值的实证分析目录Ⅰ. 摘摘要…………………………………………………………………………………… 2 关关键词 …………………………………………………………………………………… 2Ⅱ. 正正文……………………………………………………………………………………2 11. 序序言言……………………………………………………………………………………2 22. 模模型设定定……………………………………………………………………………… 33. 参参数估计计 ………………………………………………………………………………… 44. 检检验修正正 ………………………………………………………………………………… 经济意意义检验验 ……………………………………………………………………… 统计意意义检验验 ……………………………………………………………………… 计量经经济学检检验 …………………………………………………………………… 多多重共线线性检验验 ……………………………………………………………… 相关关系数检检验 ……………………………………………………………… 逐步步回归修修正 ……………………………………………………………… 异异方差性性检验 ………………………………………………………………… 异方方差检验验 ………………………………………………………………… 模型型修正 …………………………………………………………………… 序序列相关关性检验验 ……………………………………………………………… GBB 检验 ……………………………………………………………………… 模型型修正 …………………………………………………………………… 模型预预测检验验 ……………………………………………………………………… 模模型确认认……………………………………………………………………………… 55. 模模型评价价 …………………………………………………………………………………… 66. 政政策建议议 …………………………………………………………………………………… 77. 参参考文献献 ……………………………………………………………………………………我国国内内生产总总值的实实证分析析【摘要】：：本文主主要是从从宏观经经济的角角度，对对影响我我国自119900年至220099年的国国内生产产总值的的主要因因素进行行实证分分析。

结合我我国特定定国情选选取了六六个影响响我国国国内生产产总值的的主要因因素，并并对其时时间序列列分析，建建立多元元线性模模型，利利用OLLS方法法进行参参数估计计并进行行计量经经济学模模型的四四大检验验经济济意义检检验中，发发现储蓄蓄总额前前参数不不符合经经济理论论常识，并并在后面面的工作作中得到到了修正正；计量量经济学学检验中中，发现现初建模模型具有有多重共共线性，采采用逐步步回归法法进行修修正，消消除了多多重共线线性；在在异方差差性检验验中，发发现模型型具有异异方差性性，采用用对数变变换法进进行修正正，消除除了异方方差性；；利用GGB检验验法发现现模型随随机干扰扰项存在在2阶序序列自相相关性，采采用广义义差分变变换法修修正模型型，消消除了模模型序列列相关性性；利用用20110年数数据，模模型通过过了经济济预测检检验，并并确定了了最终模模型，得得出结论论：进出出口额、职职工工资资总额和和上期国国内生产产总值对对国内生生产总值值有很大大影响最后，进行了模型评价并结合模型及我国国情给出了相应的可供参考的政策建议关键词词】：国内生生产总值值进出出口额职职工工资资总额经经济意义义检验计计量经济济学检验验时间序序列多元线线性回归归 OOLS方方法逐步步回归法法多重共共线性异异方差性性对数数变换法法 GBB检验法法序列自自相关性性广义差差分变换换法经济济预测检检验序言自19885年国国家统计计局建立立起相应应的核算算制度以以来，国国内生产产总值核核算已经经成为我我国宏观观经济管管理部门门了解经经济运行行状况的的重要手手段，制制定经济济发展战战略、中中长期规规划、年年度计划划和各种种宏观经经济政策策的重要要依据。

因因此研究究国内生生产总值值的影响响因素对对我国的的经济发发展有重重大意义义20010年年国内生生产总值值39779833亿元，按按可比价价格计算算，比上上年增长长10..3%，增增速比上上年加快快1.11个百分分点总总量跃居居世界第第二本本文主要要运用计计量经济济学和统统计经济济学研究究一些经经济指标标对国内内生产总总值的影影响和相相关关系系GDDP == C++ C11*LNNX1 + CC2*LNNX3 + CC3*LNNX5一、模型型的设定定选国内生生产总值值GDPP为被解解释变量量，而影响国国内生产产总值的的因素有有很多，但但普遍看看来，进进出口额额、财政政支出总总额、职职工工资资总额、税税收总额额、上期期国内生生产总值值和储蓄总总额这六个因素素对国内内生产总总值影响响较大，因因此，我我们搜集集了这六六个因素素的时间间序列数数据作为为解释变变量，希希望建立立一个合合适的经经济模型型来从理理论上探探讨影响响国内生生产总值值的因素素，进而而提出相应应的建议议把上述述六个因因素分别别设定为为X、X、XX、X、XX、X6设定模模型为：GDP==+++U经查资料料得国内内生产总总值样本本观测数数据（单位/亿元）：：年份GDP进出口额额财政支出出职工工资资总额税收收入入上期GDDP储蓄余额额19900186667.8855600.130833.59929511.128211.866169992.3312100.219911217881.5572255.833866.62233233.929900.177186667.881610019922269223.5591199.637422.239399.232966.911217881.5523122.319933353333.9911277146422.349166.242555.3269223.5530955.219944481997.99203881.9957922.62266566.451266.888353333.9946800.119955607993.77234999.9968233.7228100060388.044481997.9958844.119966711776.66241333.8879377.5559080069099.822607993.7776477.619977789773269667.2292333.56694055.382344.044711776.66100553.1119988844002.33268449.77107998.11892966.592622.8789773116115.9919999896777.11298996.22131887.66798755.5106882.558844002.33146666.7720000992114.66392773.22158886.55106556.22125881.551896777.11181990.77200111096655..2421883.66189002.558118330.99153001.338992114.66223227.662001203332..7513778.22220553.115131661.11176336.4451096655..2281221.77200331358822..8704883.55246449.995147443.55200117.3311203332..7351119200441598878..3955339.11284886.889169000.22241665.6681358822..8414116.55200551849937..41169921..8339330.228197889.99287778.5541598878..3487887.55200662163314..41409971..5404222.773232665.99348004.3351849937..4585775.99200772658810..31667740..2497881.335282444456221.9972163314..4675999.77200883140045..41799921..5625992.66633711454222.37792658810..3785885.22200993405506..91506648..1762999.993402888.116 595221.5593140045..41005541..3——数据据来自中中国统计计年鉴二、模型型的参数数估计对设定模模型用OOLS法法进行参参数估计计,用EEvieews55对上表表数据回回归得：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/229/111 Tiime:: 200:099Sampple:: 19990 20009Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. X10.466468870.0441522711.11899910.00000X21.099940050.522060052.111178810.05546X31.855443330.688374492.711215550.01178X40.099801130.088299981.188091150.25588X50.755908800.0667566111.22353390.00000X6-1.228422790.37786993-3.339133490.00048C-23550.229817211.9227-1.336499230.19954R-sqquarred0.99997006 Meaan ddepeendeent varr1241122..3Adjuusteed RR-sqquarred0.99995771 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion19811.2996 Akaaikee innfo criiterrionn18.2290111Sum squuareed rresiid5103319663 Schhwarrz ccritteriion18.6638661Log likkeliihoood-1755.90011 F-sstattisttic73733.9883Durbbin--Wattsonn sttat1.31138221 Proob(FF-sttatiistiic)0.00000000回归结果果如下：：GDP==-23350..2988+-1..28442799-1.33649926111.11899932..111178552.7712116111.188091111..235540--3.33913354==0.99997706==0.99995571== 73374..0055D.W.=1131338200F=73374..0055 > (6，13)))=2..92((显著性性水平αα=0..05))表明模模型从整整体上看看国内生生产总值值与解释释变量间间线形关关系显著著。

三、检验验及修正正1．经济意义义检验从上述回回归结果果可知：的系系数为负负值，说说明国民民生产总总值随居居民储蓄蓄余额的增增加而减少，这这从理论论上说不不符合我我国的实实际情况况；其他因因素系数数均为正正，均不不与经济济原理相相悖，具具有经济济意义：：各系数数表示国国内生产产总值对对该因素素的弹性性大小2．统计计意义检检验从回归结结果可以以看出，模模型的拟拟和优度度非常好好（=0.99997706）， FF统计量量的值在在给定显显著性水水平α=0..05的的情况下下也较显显著因因为= 73774.000＞ (6，13))，表明明模型的的线性关关系在995%的的置信水水平下显显著成立立.但但是X2、X4的t统统计值均均不显著著3.计量量经济学学检验（1）多多重共线线性检验验①相关系系数检验验：用Eviiewss5求得得解释变变量的相相关系数数矩阵：：GDPX1X2X3X4X5X6GDP 1.00000000 0.99691179 0.99917794 0.99958825 0.77904460 0.99968865 0.99910017X1 0.99691179 1.00000000 0.99489919 0.99510089 0.77525585 0.99500081 0.99608884X2 0.99917794 0.99489919 1.00000000 0.99951181 0.88007763 0.99934493 0.99960078X3 0.99958825 0.99510089 0.99951181 1.00000000 0.88011194 0.99960041 0.99911165X4 0.77904460 0.77525585 0.88007763 0.88011194 1.00000000 0.88021156 0.88300014X5 0.99968865 0.99500081 0.99934493 0.99960041 0.88021156 1.00000000 0.99914482X6 0.99910017 0.99608884 0.99960078 0.99911165 0.88300014 0.99914482 1.00000000 由由此可知知：解释释变量、、、、之间存在在高度正正相关，模模型存在在严重多多重共线线性。

下下面对模模型进行行修正②模型修修正：用用逐步回回归法修修正模型型由相关系系数矩阵阵知解释释变量XX5与GDPP相关性性最强，故故首先选选取X5 做为为基本变变量与GGDP建建立一元元回归模模型： Y==12006.2208++1.11386675(0.44184440)) ((53..451104)) 2=0..99337 F=228577.0114 D.W.=11.11177117依次次引入XX3、、、、X6变量回回归：引入X33 ：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/229/111 Tiime:: 200:19Sampple:: 19990 20009Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C-13993.553730377.4990-0.445877790.65522X33.411066601.822281111.877109990.07786X50.722019990.222454413.200742240.00052R-sqquarred0.99948008 Meaan ddepeendeent varr1241122..3Adjuusteed RR-sqquarred0.99941998 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion72833.9777 Akaaikee innfo criiterrionn20.7762222Sum squuareed rresiid9.022E+008 Schhwarrz ccritteriion20.9911558Log likkeliihoood-2044.62222 F-sstattisttic16288.7442Durbbin--Wattsonn sttat1.26681440 Proob(FF-sttatiistiic)0.00000000引入X33 ，拟合优优度得到到提高，参参数符号号合理且且参数统统计量显显著，故故采纳该该变量。

引入：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/229/111 Tiime:: 200:255Sampple:: 19990 20009Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C27511.955813222.72232.088052240.05539X32.200678810.755797712.911143320.01102X50.633445520.099244466.866291150.00000X10.355470030.033840029.233647730.00000R-sqquarred0.99991880 Meaan ddepeendeent varr1241122..3Adjuusteed RR-sqquarred0.99990226 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion29833.7445 Akaaikee innfo criiterrionn19.0016660Sum squuareed rresiid1.422E+008 Schhwarrz ccritteriion19.2215775Log likkeliihoood-1866.16660 F-sstattisttic64999.4888Durbbin--Wattsonn sttat1.35548997 Proob(FF-sttatiistiic)0.00000000引入，拟拟合优度度再次提提高，参参数符号号合理且且参数统统计量显显著，故故采纳该该变量。

引入：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/229/111 Tiime:: 200:366Sampple:: 19990 20009Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C-2077.3883223000.6551-0.009011410.92294X32.933479960.866705593.388477710.00041X50.677013370.099167737.311010070.00000X10.355788850.033690069.699729920.00000X2-0.551111510.33314441-1.554222080.14439R-sqquarred0.99992992 Meaan ddepeendeent varr1241122..3Adjuusteed RR-sqquarred0.99991004 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion28622.9669 Akaaikee innfo criiterrionn18.9969442Sum squuareed rresiid1.233E+008 Schhwarrz ccritteriion19.2218336Log likkeliihoood-1844.69942 F-sstattisttic52955.1661Durbbin--Wattsonn sttat1.52237332 Proob(FF-sttatiistiic)0.00000000引入，拟拟合优度度虽然得得到了提提高，但但是参数数符号为为负值，表表示GDDP随财财政支出出增加而而减少，与与实际情情况相悖悖，故将将该变量量剔除。

剔除，引引入Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/229/111 Tiime:: 211:05Sampple:: 19990 20009Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C26788.000912211.17702.199298850.04445X32.277678800.700036623.255086610.00054X50.644834440.088560057.577362290.00000X10.355060080.033549999.877649930.00000X4-0.113300840.06683661-1.994677830.07705R-sqquarred0.99993445 Meaan ddepeendeent varr1241122..3Adjuusteed RR-sqquarred0.99991771 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion27533.3333 Akaaikee innfo criiterrionn18.8891333Sum squuareed rresiid1.144E+008 Schhwarrz ccritteriion19.1140226Log likkeliihoood-1833.91133 F-sstattisttic57255.5663Durbbin--Wattsonn sttat1.60069445 Proob(FF-sttatiistiic)0.00000000引入，拟拟合优度度再次提提高，但但是参数数符号为为负值表表明我国国GDPP随税收收收入增增加而减减少，与与实际情情况相悖悖，所以以将之剔除除。

引入：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/229/111 Tiime:: 211:22Sampple:: 19990 20009Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd..Errrort-SttatiistiicProbb. X10.400854420.0330811113.22596650.00000X32.800521140.566660054.955091120.00002X50.733455550.0771222210.33135560.00000X6-0.660755290.15528007-3.997577870.00012C-26998.443616699.7443-1.661600790.12269R-sqquarred0.99996001 Meean deppenddentt vaar1241122..3Adjuusteed RR-sqquarred0.99994994 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion21500.2995 Akaaikee innfo criiterrionn18.3396992Sum squuareed rresiid6935565225 Schhwarrz ccritteriion18.6645885Log likkeliihoood-1788.96692 F-sstattisttic93899.6772Durbbin--Wattsonn sttat1.50000771 Proob(FF-sttatiistiic)0.00000000引入，拟拟合优度度虽再次得到到提高，但但是参数数符号为为负值表表明我国国GDPP随上期期GDPP增加而而减少，与与实际情情况相悖悖，所以以将之剔除除。

通过逐逐步回归归，剔除除了变量量、、，得到新新模型：： YY=27751..9588+0..35447033+2..20667811+0..634445222.08805224 9..23664733 2.99114432 6..862291552=0.99990026 F==64999.4488 D.W.=11.35548997(2) 异方差检检验：①a. 画出残残差e^^2与GDPP的散点点图：观观察散点点图，可可知模型型存在异异方差，下下面从理理论上加加以说明明b.利用用Whiite--检验模模型是否否存在异异方差性性：Whitte HHeteerosskeddastticiity Tesst:F-sttatiistiic7.89989228 Proobabbiliity0.00016770Obs**R-ssquaaredd17.5533662 Proobabbiliity0.04409887Testt Eqquattionn:Depeendeent Varriabble:: REESIDD^2Methhod:: Leeastt SqquarresDatee: 006/330/111 Tiime:: 122:555Sampple:: 1 20Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C157110522811988779901.311054440.21193X126.6637114230..753370.111543350.91104X1^220.011848810.000618842.988856600.01136X1*XX3-0.337011190.26605885-1.442033390.18859X1*XX50.022675570.033549990.755373370.46684X3-88330.5545145112.551-0.660844780.55564X3^222.866580003.700943300.777257710.45576X3*XX5-0.551633790.72228009-0.771444060.49913X5817..9677016399.00070.499906630.62285X5^220.022572220.033354460.766676600.46609R-sqquarred0.87766881 Meaan ddepeendeent varr712221944.Adjuusteed RR-sqquarred0.76656993 S.DD. ddepeendeent varr824559177.S.E.. off reegreessiion399114544. Akaaikee innfo criiterrionn33.5544006Sum squuareed rresiid1.599E+114 Schhwarrz ccritteriion34.0041993Log likkeliihoood-3255.44406 F-sstattisttic7.89989228Durbbin--Wattsonn sttat1.74477669 Proob(FF-sttatiistiic)0.00016770 结果：：因为nnR2 =17..533362>>=166.9119，故故拒绝原原假设，即即模型存存在异方方差性。

②异方差差的修正正： a.采采用对数数变换法法对模型型进行修修正：用用GENNR产生生对数序序列：genrr lnngdpp=loog(ggdp))genrr lnnx1==logg(x11)genrr lnnx3==logg(x33)genrr lnnx5==logg(x55) 然后用用OLSS方法对对新序列列回归，结结果如下下：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/330/111 Tiime:: 133:333Sampple:: 1 20Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C-107784885.1355563..8-7.995555560.00000LNX11-394403..27369332.006-1.006699120.30018LNX333187707..3780552.2214.088325580.00009LNX55-11993633.5541669.668-2.220355110.04426R-sqquarred0.92244333 Meaan ddepeendeent varr1241122..3Adjuusteed RR-sqquarred0.91102664 S.DD. ddepeendeent varr956223.117S.E.. off reegreessiion286444.880 Akaaikee innfo criiterrionn23.5540119Sum squuareed rresiid1.311E+110 Schhwarrz ccritteriion23.7739333Log likkeliihoood-2311.40019 F-sstattisttic65.2244332Durbbin--Wattsonn sttat0.31168669 Proob(FF-sttatiistiic)0.00000000GDP = --107784885.4432 - 3394003.2267664*LLNX11 + 31887077.29961**LNXX3 -- 11193663.550188*LNNX5-7.99555556--1.0066991244.08832558-22.20035111=0.99244433==0.99102264== 655.2444322D.W.=0..31668699F=655.2444322 > ((6，13)))=2..92((显著性性水平αα=0..05))表明模模型从整整体上看看国内生生产总值值与解释释变量间间线形关关系显著著。

b. 利利用Whhitee-检验模模型是否否存在异异方差性性：Whitte HHeteerosskeddastticiity Tesst:F-sttatiistiic3.18819225 Proobabbiliity0.04428223Obs**R-ssquaaredd14.8823666 Proobabbiliity0.09958995Testt Eqquattionn:Depeendeent Varriabble:: REESIDD^2Methhod:: Leeastt SqquarresDatee: 006/330/111 Tiime:: 133:355Sampple:: 1 20Inclludeed oobseervaatioons:: 200VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C-1.998E++119.588E+110-2.006488300.06659LNX11-3.007E++103.766E+110-0.881577480.43336LNX11^2-6.006E++095.000E+009-1.221155520.25535LNX11*LNNX32.333E+0091.822E+1100.122844430.90003LNX11*LNNX51.233E+1101.744E+1100.700858890.49948LNX33-2.117E++119.899E+110-2.119422930.05529LNX33^22.099E+1104.644E+1100.455008840.66622LNX33*LNNX5-1.771E++106.055E+110-0.228266560.78832LNX552.466E+1119.633E+1102.544962230.02289LNX55^2-9.994E++091.733E+110-0.557333820.57791R-sqquarred0.74411883 Meaan ddepeendeent varr6.566E+008Adjuusteed RR-sqquarred0.50082448 S.DD. ddepeendeent varr8.622E+008S.E.. off reegreessiion6.055E+008 Akaaikee innfo criiterrionn43.5585114Sum squuareed rresiid3.666E+118 Schhwarrz ccritteriion44.0083001Log likkeliihoood-4255.85514 F-sstattisttic3.18819225Durbbin--Wattsonn sttat2.14478004 Proob(FF-sttatiistiic)0.04428223结果：因因为nRR2 =14..823366<<=166.9119，故故没有理理由拒绝绝原假设设，即修修正后的的模型不不存在异异方差性性。

3）序列性相相关性检检验：①用拉格格朗日乘乘数方法法检验（GGB检验验法）：：取检验验水平为为0.005 假设模模型随机机干扰项项存在pp阶序列列相关：：，从pp=1开开始，经经过逐次次高阶检检验，并并利用各各残差项项前参数数的显著著性判断断序列相相关性，得得到模型型在p==2时结结果：Breuuschh-Goodfrrey Serriall Coorreelattionn LMM Teest::F-sttatiistiic2.77796117 Proobabbiliity0.10096663Obs**R-ssquaaredd6.43313006 Proobabbiliity0.04401229Testt Eqquattionn:Depeendeent Varriabble:: REESIDDMethhod:: Leeastt SqquarresDatee: 006/330/111 Tiime:: 166:088Pressampple misssinng vvaluue llaggged ressiduualss seet tto zzeroo.VariiablleCoeffficcienntStd.. Errrorrt-SttatiistiicProbb. C1039927..25172280..50.200091110.84448LNX11-93777.4423131447.112-0.771322680.49920LNX33254881.990600008.2240.422464400.68801LNX55-208887..31477446.777-0.443744600.67711AR(11)0.088629900.300339900.288441190.78819AR(22)-0.007399800.28807009-0.226355460.79975RESIID(--1)0.300454420.366455590.833537720.42230RESIID(--2)-0.771377980.34471220-2.005633440.06668R-sqquarred0.35572995 Meaan ddepeendeent varr-2.449E--07Adjuusteed RR-sqquarred-0.00925599 S.DD. ddepeendeent varr55066.9992S.E.. off reegreessiion57566.3118 Akaaikee innfo criiterrionn20.4455009Sum squuareed rresiid3.311E+008 Schhwarrz ccritteriion20.8850881Log likkeliihoood-1766.09958 F-sstattisttic0.79941776Durbbin--Wattsonn sttat1.75506997 Proob(FF-sttatiistiic)0.60090776 由于nnR2 =6.44313306>>=5..9911，则拒绝绝原假设设，即认认为存在在自相关关性。

n=3时时：Breuuschh-Goodfrrey Serriall Coorreelattionn LMM Teest::F-sttatiistiic2.19947226 Proobabbiliity0.15583110Obs**R-ssquaaredd7.60048441 Proobabbiliity0.05549225 由于nnR2 =7.66048841<<=7..8155，则没没有理由由拒绝原原假设，即即认为不不存在三三阶自相相关性②自相关关的修正正：采用广义义差分变变换法修修正模型型：在LLS命令令中加上上AR((1)、AAR(22),使使用迭代代估计法法估计模模型，回回归结果果如下：：Depeendeent Varriabble:: GDDPMethhod:: Leeastt SqquarresDatee: 006/330/111 Tiime:: 155:199Sampple (addjusstedd): 3 220Inclludeed oobseervaatioons:: 188 affterr addjusstmeentssConvverggencce aachiieveed aafteer 11。

点击阅读更多内容