eviews面板数据模型详解

《eviews面板数据模型详解》由会员分享，可在线阅读，更多相关《eviews面板数据模型详解（24页珍藏版）》请在装配图网上搜索。

1、1.已知19962002年中国东北、华北、华东15个省级地区的居民家庭人均消费（cp，不变价格）和人均收入（ip，不变价格）居民，利用数据（1）建立面板数据（paneldata）工作文件；（2）定义序列名并输入数据；（3）估计选择面板模型；（4）面板单位根检验。年人均消费（consume）和人均收入（income）数据以及消费者价格指数（p）分别见表9.1，9.2和9.3。表9.119962002年中国东北、华北、华东15个省级地区的居民家庭人均消费（元）数据人均消费1996199719981999200020012002CONSUMEAH3607.433693.553777.413901.8

2、14232.984517.654736.52CONSUMEBJ5729.526531.816970.837498.488493.498922.7210284.6CONSUMEFJ4248.474935.955181.455266.695638.746015.116631.68CONSUMEHB3424.354003.713834.434026.34348.474479.755069.28CONSUMEHLJ3110.923213.423303.153481.743824.444192.364462.08CONSUMEJL3037.323408.033449.743661.684020.8743

3、37.224973.88CONSUMEJS4057.54533.574889.435010.915323.185532.746042.6CONSUMEJX2942.113199.613266.813482.333623.563894.514549.32CONSUMELN3493.023719.913890.743989.934356.064654.425342.64CONSUMENMG2767.843032.33105.743468.993927.754195.624859.88CONSUMESD3770.994040.634143.964515.0550225252.415596.32CON

4、SUMESH6763.126819.946866.418247.698868.199336.110464CONSUMESX3035.593228.713267.73492.983941.874123.014710.96CONSUMETJ4679.615204.155471.015851.536121.046987.227191.96CONSUMEZJ5764.276170.146217.936521.547020.227952.398713.08表9.219962002年中国东北、华北、华东15个省级地区的居民家庭人均收入（元）数据人均收入199619971998199920002001200

5、2INCOMEAH4512.774599.274770.475064.65293.555668.86032.4INCOMEBJ7332.017813.168471.989182.7610349.6911577.7812463.92INCOMEFJ5172.936143.646485.636859.817432.268313.089189.36INCOMEHB4442.814958.675084.645365.035661.165984.826679.68INCOMEHLJ3768.314090.724268.54595.144912.885425.876100.56INCOMEJL3805.5

6、34190.584206.644480.0148105340.466260.16INCOMEJS5185.795765.26017.856538.26800.237375.18177.64INCOMEJX3780.24071.324251.424720.585103.585506.026335.64INCOMELN4207.234518.14617.244898.615357.795797.016524.52INCOMENMG3431.813944.674353.024770.535129.055535.896051INCOMESD4890.285190.795380.085808.96648

7、9.977101.087614.36INCOMESH8178.488438.898773.110931.6411718.0112883.4613249.8INCOMESX3702.693989.924098.734342.614724.115391.056234.36INCOMETJ5967.716608.397110.547649.838140.58958.79337.56INCOMEZJ6955.797358.727836.768427.959279.1610464.6711715.6表9.319962002年中国东北、华北、华东15个省级地区的消费者物价指数物价指数19961997199

8、81999200020012002PAH109.9101.310097.8100.7100.599PBJ111.6105.3102.4100.6103.5103.198.2PFJ105.9101.799.799.1102.198.799.5PHB107.1103.598.498.199.7100.599PHLJ107.1104.4100.496.898.3100.899.3PJL107.2103.799.29898.6101.399.5PJS109.3101.799.498.7100.1100.899.2PJX108.410210198.6100.399.5100.1PLN107.9103.1

9、99.398.699.910098.9PNMG107.6104.599.399.8101.3100.6100.2PSD109.6102.899.499.3100.2101.899.3PSH109.2102.8100101.5102.5100100.5PSX107.9103.198.699.6103.999.898.4PTJ109103.199.598.999.6101.299.6PZJ107.9102.899.798.810199.899.1（1）建立面板数据工作文件回凶首先建立工作文件。打开工作文件后，过程如下:Torkfile:UHTITLED代tuh司D白亡t亡&nr5日币灰|他円眉匚1

10、|：裁蚯词print5已卡亡D亡tdik十卜showRange:Sample:Ec立residHewObject.-jen已r迅t已吕已riez.EredeLinks.DisplayFilter:*FetchfromIlE.UEnlateEelecteiifromDB.Stcifh!thidt-：iDB.EerL：anieselected.Bfejlti+fc!EtjltjiZ+tjilFi_iritSelected建立面板数据库。KctOtjeclTjieofobjectFcq1|EquitionJraphGtoupLcigLMatr1st-Vector-Coe:ModelPoolSJiff

11、ipleSeriezStEiteLinkSsriesjVLphaEEpiCt；Syst已ITlTableTextvaMaj.畑ff：ameforobjectUK在窗口中输入15个不同省级地区的标识。(2) 定义序列名并输入数据?Vev2一Pnnl=COMSTITETorkTile;IIWTITLED:UntitledIZI回13贸旦|乔百Iph旦匚t|PrintJMmm旦|巨庁tlmmtE口mfInm|PcicilGEnF亏hEEtlCrossSectionIdentiTiers:(Enteridentifiersbelowthisline)AHBJFJHE：HLJJLJSJXLNNMGSDS

12、HSXTJEJ产生3*15个尚未输入数据的变量名。这样可以通过键盘输入或黏贴的方法数据数据。(3) 估计、选择面板模型打开一个pool窗口，先输入变量后缀(所要使用的变量)。点击Estimate，打开估计窗口。A混合模型的估计方法左边的Common表示相同系数，即表示不同个体有相同的斜率。得到如下输出结果:DependentVariable:CP?Method:PooledLeastSquaresDate:0702/00Time:13:13Sample:19962002Includ已dobservations:7Cross-sectionsincluded:15Totalpool(balari

13、cud)observations:105VariableCoeffi匚Std.Errort-StatisticProb.C129.631363.692592.0352650.0444IP?07587260.00952279.681890.0000R-squared0.984036Meandependentvar4917.608AdjustedR-squared0.983881S.D.dependentvar1704.704S.E.ofregression216.4270Akaikeinfocriterion13.61125Sumsquaredresid4824588.Schwarzcriter

14、ion13.66180Loglikelihood-712.5905F-statistic6349.204Durbin-Watsonstat0.7B4107Prob(F-statistic).000000相应的表达式是:CP129.63+0.76IPitit(2.0)(79.7)R20.9&SSE=4824588r上式表示15个省级地区的城镇人均指出平均占收入的76%。B个体固定效应回归模型的估计方法将截距项选择区选Fixedeffects(固定效应)得到如下输出结果:DependentVariable:CP?Method:PaoledLeastSquaresDate:07702/08Time:

15、13:36Sample:19962002Includedobservations:7Cross-sectionsincluded:15Totalpool(balancrcI)observations:105VariableCoefficientStd.Errort-StatisticProb.C515.E14281.596656.3190610.0000IP?0.6975610.01269254.960290.0000FixedEffects(Cross)AH-C-36.305B3BJ-C537.5660FJ-C-47.E4515HB-C-154.2367HM-C-169.7015JL-C24

16、.60427JS-C-35.195B7JX-C-319.6960LN-C106.4273忖MG-C-209.5484SD-C-134.1145SH-C266.9859SX-C-74.88901TJ-C47.22940ZJ-C198.6202EffectsSpecificationCrass-s已匚tionfi)(已d(dumrri/variables)R-squar已M0.992488Meandep已ncl已ntvar4917.608AdjustedR-squar已d0.991222S.D.dep已ndntvar1704704S.E.ofregression159.7184Akaikeinfo

17、criterion13.12414Sumsquaredresid2270386.Schwarzcriterion13.52855Loglikelihood-673.0173F-statistic783.8902Durbin-Watsonstat1.609517Prab(F-statistic)0.000000相应的表达式为:CP515.6+0.70IP36.3D+537.6D+.+198.6Ditit1215(6.3)(55)R20.99,SSE=2270386r其中虚拟变量D,D，,D的定义是：1215f1,如果属于第个个体,1,2,.,15i0,其他15个省级地区的城镇人均指出平均占收入7

18、0%。从上面的结果可以看出北京市居民的自发性消费明显高于其他地区。接下来用F统计量检验是应该建立混合回归模型，还是个体固定效应回归模型。H:=。模型中不同个体的截距相同（真实模型为混合回归模型）。0iH:模型中不同个体的截距项不同（真实模型为个体固定效应回归模型）。1iF统计量定义为:(SSESSE)/(NTk1)，(NTNk)_(SSESSE)/(N1)SSE/(NTNk)SSE/(NTNk)uu其中SSE表示约束模型，即混合估计模型的残差平方和，SSE表示非约束模型，即ru个体固定效应回归模型的残差平方和。非约束模型比约束模型多了N-1个被估参数。所以本例中：(4824588-227386

19、)/(15-1)2270386/(105151)8.1F(14,89)1.80.05所以推翻原假设，建立个体固定效应回归模型c时点固定效应回归模型的估计方法更合理。将时间选择为固定效应。PoolEsti-MationSp&cificatiotlOt-11otleFiseii:miFuELTLiion-!EsiimailctlrnetholDeperLilentv：=lti=bleFeriollorie3djCi-j5s_se：ti口weii7WeighFlnB：-il：=LrLC*!S：dTTiplWEs+irTi：i+i-:-nset+irLgsMethod：L5一LeastSqu：irez

20、C：ndATLJS.ple2002取消得到如下输出结果：DeperidentVariable:CP?Method:PooledLeastSquaresDate:0712Time:U:44Sample:19962002Includedobservations:7Cross-sectionsincluded:15Totalpool(balanced)abservaiions:105YaribleCoefficientStd.Errort-StatisticProb.C2.57723468.805480.0374570.9702冋0.77BB600.01043874.619110.0000Fiice

21、dEffects(Period)1996-CW5.92B51997-C134.05581990-C54.040281999-C-37.074042000-C-7.1103592001-C-156.74742002-C-93.09200EffectsSpecificationPeriodfixed(dummyvariables)R-squared0.986669Meandependentvar4917.608AdjustedR-squared0.986707S.D.dependentvar1704.704S.E.ofregressior203.8001Akaikeinfocriterion13.

22、64629Sumsquaredresid402S843.Schwarzcriterion13.74749Loglikelihood-703.1276F-statistic1025.643Durbiri-Watsonstat0.786376ProbfF-statistic)0.000000相应的表达式为:CP2.6+0.78IP+105.9D+134.1D+.93.9Ditit127(76.6)R20.987,SSE4028843其中虚拟变量D,D，,D的定义是：127八1,如果属于第个截面,t0,其他D个体随机效应回归模型估计截距项选择Randomeffects（个体随机效应）得到如下部分输出

23、结果:相应的表达式是：CP345.2+0.72IP-2.6D+367.0D+.+106.1Ditit1215(68.5)R20.9&SSE=2979246其中虚拟变量D,DD的定义是：1215八1,如果属于第个个体D=i|0,其他接下来利用Hausman统计量检验应该建立个体随机效应回归模型还是个体固定效应回归模型。H:个体效应与回归变量（IP）无关（个体随机效应回归模型）0itH:个体效应与回归变量（IP）相关（个体固定效应回归模型）1it分析过程如下：75.472170.0105724.57359968.53814DescriptibE.UnitrijijtLabelBJ-C367.043

24、9FJ-C-54.24006HB-C-104.8367HLJ-C-101.7680JL-C54.90671JS-C-32.27868JX-C-223.9519LN-C112.1152MhfIP-r：一1阳n77vF：Qth二c:KdocuiTi&rLtE：=LTLdsettingszhuyisimydocuniHntEDB=nijneHF=casel3得到如下检验结果:CorrelatedRandomEffects-HausmanTestPool:POOU02Testcross-secticmrandomeffectsTestSummaryChi-Sq.StatisticChi-Sq.d.f.

25、Prob.Cross-sectionrandom1478756310.0001Cross-sectionrandomRffEctstest匚ompariscins:VariableFixedRandomVar(Diff.)Prob.IP?0.69766107246690.0000490.0001由检验输出结果的上半部分可以看出，Hausman统计量的值是14.79,相对应的概率是00001,即拒接原假设，应该建立个体固定效应模型。检验结果的下半部分是Hausman检验中间结果比较。个体固定效应模型对参数的估计值为0697561,随机效应模型对参数的估计值为0724569。两个参数的估计量的分布

26、方差的差为0000049。综上分析，19962002年中国东北、华北、华东15个省级地区的居民家庭人均消费和人金收入问题应该建立个体固定效应回归模型。人均消费平均占人均收入的70%。随地区不同，自发消费(截距项)存在显著性差异。(4) 面板单位根检验以cp序列为例。首先在工作文件窗口中打开cp变量的15个数据组。回国单位根检验过程如下:Eie4853.4415197.0415314.5215522.7626094.3366665.0053868.3193896.7784104.2814361.5554457.4635120.4853077.9893289.9903596.8393890.580

27、4159.0874493.5353286.4323477.5603736.4084077.9614281.5604998.874445；491E507E531；548E609N-ii1ayTabulation.CerrelatiotleCov：di-i:iTLCeeFrincip:ilCamponente.Correlijgr：iiTi(.1.).CrR呂eCorrelation(2.).CuirL宜titjiLTezt.HuitRootTest.G；r:iiLgerC：ue：ility.LabelVLFath=c:docijjrierLtsandse11ingszhuyiximydociji

28、tienDB=neneWF=casel3得到如下检验结果：GrciiwunitmcrttESt:Summ日Date:07?02/DBTime:15:40Sample:19962002Series:CPAH.CPBJ,CPFJ,CPHB,CPHLJ,CPJL,CPUS,CPJX,CPLN,CPNMG,CPSD,CPSH,CPSX,CPTJ,CPZJExogenousvariables:IndividualeffectsAutomaticselectionofmaximumlagsAutomaticselectionoflagsbasedonSIC:0to1Newey-Westbandwidths

29、electionusingBartlettkernelMethodStatisticProb.*CrosssectionsObsNull:Unitroot(assumescommonunitroolprocess)Levin,LinChu1*9.697781.00001684Null:Unitroot(assumesindividualunitrootprocess)Im,PesaranandShinW-stat6.463801.00001584ADF-FisherChi-square3.049971.00001584PP-FisherChi-square5.243351.00001590Nu

30、ll:Nouritroot(assumescorrimoriunitrootprocess)HadriZ-stat7.583830.000015106fProbabilitiesforFishertestsarecompuiedusinganasympoticChi-squaredistribution.Allothertestsassumeasymptoticnormality.从上面的检验结果可以看出来，6种检验方法的结论都认为15个cp序列存在单位根。选择IPS检验方法进行单位根检验。检验结果如下：NullHypothesis:Unitroot(individualunitrootpro

31、cess)Date:07/02/08Time:15:44Sample:19962002Series:CPAH,CPBJ,CPFJ,CPHB,CPHU.CPJL,CPUS.CPJX,CPLN,CPNMG,CPSD,CPSH,CPSX,CPTJ,CPZJExogenousvariables:IndividualeffectsAutomaticselectionofnriaxiiriJiriImgsAutomaticselectionoflagsbasedonEIC:0to1Totalnumberofobservations:84Cross-sectionsincluded:15MethodStat

32、isticProb.*Im,PesaranandShinW-stat6.463001.0000*Probabilitiesarecomputedassumingasym卩oticnormalityIntermediateADFtestresultsMaxSeriest-StatProb.Eft)EYar)LaqLaqObsCPAH5.55420.9999-1.5562.648115CPBJ0.56130.9695-1.5472.332016CPFJ0.6969077111.5472.332016CPHB0.53040.8U91.5472.332016CPHU2.65880.99911.5472

33、.332016CPJL0.63240.9732-1.5472.332016CPJS-1.36320.6305-1.5472.332016CPJX0.53560.9681-1.5472.332016CPLN1.12760.9893-1.6472.332016CPNMG3.07660.9992-1.6682.648116CPSD-0.33890.8450-1.6682.648116CPSH0.78710.9754-1.5582.648115CPSX2.11120.9980-1.5472.332016CPTJ-0.10.BB27-1.5582.648115CPZJ2.00920.9966-1.558

34、2.648115Average1.0E53-1.5522.459从上面的结果可以看出，cp面板存在单位根，同时每个个体都存在单位根。2收集中国20002005年各地区城镇居民人均可支配收入X和消费指出Y统计数据如表9.4。数据是6年的，每一年都有32组数据，共192组观测值。人均可支配收入和消费支出数据（单位：元）地2000可支2001可支配收消费2002200320042005可支配收消费可支配收消费可支配收消费可支配收消费配收消费区入支出入支出入支出入支出入支出入支出XYXYXYXYXYXY全6279.949986859.553097702.860298472.265109421.6718

35、210493.7942国8.008.010.880.941.1003.88北10349.849311577.892212463.102813882.111215637.122017652.1324京69.4978.72924.60623.84840.40954.20天8140.561218958.769879337.5719110312.786711467.880212638.9653津0.040.226.9691.5316.4455.26河5661.143485984.844796679.650697239.054397951.358199107.06699北6.472.758.286.77

36、1.189.67山4724.139415391.041236234.347107005.051057902.856548913.96342西内1.875.016.963.386.151.63内蒙5129.039275535.841956051.048597012.954198122.962199136.76928古5.759.620.880.149.269.60辽5357.743565797.046546524.553427240.560778007.565439107.57369宁9.061.422.648.926.285.27吉4810.040205340.443376260.149737

37、005.154927840.660688690.66794林0.876.226.887.101.992.71黑八、龙4912.838245425.841926100.544626678.950157470.755678272.56178江8.447.366.080.191.531.01上11718.886812883.933613249.104614867.110416682.126318645.1377海01.1946.10804.00490.34821.03033.41江6800.253237375.155328177.660429262.4670810481.733212318.8621

38、苏3.180.744.606.5893.2657.82浙9279.1702010464.795211715.871313179.971214546.106316293.1225江6.2267.3960.0853.89386.14773.74安5293.542325668.845176032.447366778.050647511.457118470.66367徽5.980.650.523.343.338.67福7432.256388313.060159189.366319999.5735611175.816112321.8794建6.748.116.684.2637.1531.41江5103.

39、536235506.038946335.645496901.449147559.653378619.66109西8.562.514.322.554.846.39山6489.950227101.052527614.355968399.960699437.8667310744.7457东7.008.416.321.350.7579.31河4766.238305267.441106245.445046926.149417704.952948667.96038南6.712.170.682.600.197.02湖5524.546445855.948046788.556087321.959638022.7

40、63988785.96736北4.508.792.928.255.524.56湖6218.752186780.555466958.555747674.260828617.468849523.97504南3.796.226.720.628.617.99广9761.5801610415.809911137.898812380.963613627.106914769.1180东7.9119.6320.4843.27654.79949.87广5834.448526665.752247315.354137785.057638689.964459286.77032西3.313.732.444.509.73

41、0.80海5358.340825838.843676822.754597259.255027735.758028123.95928南2.564.852.645.438.404.79重6275.955696721.058737238.063608093.671189220.9797310243.8623庆8.849.694.247.066.0546.29四5894.248556360.451766610.854137041.857597709.863718385.96891川7.787.170.087.217.146.27贵5122.242785451.942735944.045986569.2

42、49487322.054948151.16159州1.281.908.283.985.453.29云6324.651856797.752527240.558277643.560238870.868379265.96996南4.311.606.927.568.010.90西7426.355547869.159948079.169528765.480459106.083389431.18617藏2.426.392.445.347.218.11陕5124.242765483.746376330.853786806.356667492.462338272.06656西4.673.744.045.547

43、.072.46甘4916.241265382.944206151.450646657.252987376.759378086.86529肃5.471.314.244.914.302.20青5169.941855853.746986170.550426745.354007319.657588057.86245海6.732.592.522.247.955.26宁4912.442005544.145956067.451046530.453307217.858218093.66404夏0.507.404.928.347.384.31新5644.844226395.049316899.656367173

44、.555407503.457737990.16207疆6.934.404.404.612.625.52首先建立工作文件，打开工作文件后，过程如下:dorhfile:UKTITLED回函灯陰网|p忙迫Ei?底j|IPfiit|De=tai卜伍卜1门刈|Fistth忖口比|血皑氓归的”|5drriple|Range:令mpl巳Ec之residDisplayFilter:*Upiila-leeelecte1romDB.StoreselectedtoDB.Copy.Belete-electeiiPrintSelectedk风UivtitledXNewPage/建立面板数据库，并命名为XY。输入不同省市

45、（包括全国）的标识，如下:EVieTs-Pool:XTTorkfile:UNTITLED:Untitled口回111FileEditObjectViewFrocQuick0+ionsWindowHelp-D1XViewProcobjectPrintNameFreezeEstimateDefinePoolGenrSheetCrossSectionIdentifiers:(Enteridentifiersbelowthisline)00QGJ1BJJ2TJJ3HBJ4SX35NMGJ6LNJ7JLJ8HUJ9SH10JS11ZJ12AH13FJ14JX15SD16HN17HUB18HUN19GD2

46、DGXAVF：th=c:VdocuirierLt吕andEttdccuiTiHrLDB=rLuneHF=utlIitied点击sheet键，定义变量X和Y。点击Edit+/-后，在数据窗口键入数据即可。对模型进行估计，建立个体固定效应回归模型，过程如下:3EViesPool:XTorkfile:UNTITLED:UntitledFileEditObjectViewFrocQuickOptionsWindowHelpie训PrmObject,Print,Name,FreezeEdi田一Order+/-|Smpl+卜,FormatJTitleEstimateDefinePcicilGenr5amp

47、le4998obs2GXZ-2DCI2GXZ-2DCI2EXZ-2002GXZ-2DCI2GXZ-20CI27SHX-20(2阳HX-252阳HX-2527SHX-2012殆HX-25刀SHX-2口肝GS-2圧28G5-2H28GS-20C2EGB2圧2EGB2圧29QH-2圧-iccu-inrititled得到如下输出结果:DependentVariable:Y?Method:PooledLeastSquaresDate:07/U8J08Time:19:12Sample:20002005Includedobservations;6Cross-sectionsincluded;32Totalp

48、ool(balanced)obsemtians;192VariableCoefficientStd.Errort-StatisticProb.c634.3534X?0.699629FistedEffects(Cross)00QG-C-45.9141501BJ-C566.338802TJ-C40.1904603HB-C-295.061704SX-C-338.85665NMG-C-155.449406LN-C188.093107JL-C-11.0738208HU-C-292.515409SH-C117.831810JS-C-385.983611ZJ-C-54.1720612AH-C-164.883913FJ-C-348.122514JX-C-563.378616SD-C-428.007416HN-C-462785817HUB-C126.052818HUN-C163.660619GD-C500.365320GX-C-160.124521HN-C-240730322CQ-C712.423723SC-C212.307786.217677.3575000.00000.01053366.421030.0000从估计结果可以看出，对于32个省市来说，虽然它们的城镇居民消费倾向相同，但是其城镇居民的自发消费存在显著的差异，其中重庆的城镇居