还剩1页未读，继续阅读
文本内容:
第章政策评估模型13
二、单项选择题l.A
2.B
3.D
4.A5,B
6.C
7.A
8.B
9.D
10.Cl.ABCDE
2.BCE
3.ABCDE
4.ABC
5.AD
6.ABCDE
7.ABDE
8.ACDE
9.ABCE
10.ABCDE
三、多项选择题
四、判断题,错误.正确.正确,错误.正确
45123.错误,错.正确.错误.正确910678
五、填空题.因果推断.处理效应.随机实验•选择偏差.反事实
123456.分配机制
7.倾向得分
8.条件独立性假设
9.平行趋势.事前平行趋势检验安慰剂检验10
六、计算题运用基本双重差分模型估计税法改革对企业业绩的处理效应;基本的双重差分模型设定
1.如下+仇q=Groupj+/3Post+OGroup Post+jxit2t i t软件估计结果如下:EviewsdEViewsFileEditObjectViewProcQuickOptionsAdd-insWindowHelpCommandIs qc grouppost group*post□Command]□Capture|国Equation:UNTITLEDWorkfile:HOMEWORK::Homew...ViewProc|ObjectPrintNameFreezeEstimateForecastStats|ResidsDependentVariable:QMethod:PanelLeastSquaresDate:11/12/24Time:14:59Sample20102017Periodsincluded:8Cross-sectionsincluded:4Totalpanelbalancedobservations:32Variable CoefficientStd.Errort-StatisticProb.
5.
7800000.
09561860.
448690.0000GROUP
1.
0000000.
1352257.
3951000.0000POST
0.
1700000.
1352251.
2571670.2191GROUP*POST
0.
4000000.
1912372.
0916500.0457RootMSE
0.252982R-squared
0.863314Meandependentvar
6.465000AdjustedR-squared
0.848669S.D.dependentvar
0.695219S.E.ofregression
0.
2704492.048000Akaikeinfocriterion
0.339005Sumsquaredresid-
1.424078Schwarzcriterion
0.522222Loglikelihood
0.3997365894948Hannan-Quinncriter.F-statistic
0.
0336210.000000Durbin-Watsonstat ProbfF-statisticnrPath=d:\documentsDB=none WF=homework表示在年前后，处理组平均托宾值的增量比控制组的增量高也可以^=
0.40,2014Q
0.40,说，年以后处理组与控制组平均托宾值的差异比年以前提高了这反映了2014Q
20140.40税法改革政策对企业业绩的处理效应，说明税法改革提高了企业业绩运用双向固定效应模型来估计税法改革对企业业绩的处理效应；
2.设定双向固定效应模型如下:q=OGroup Post++匕+Jxit i t软件估计结果如下:Eviews国Equation:UNTITLEDWorkfile:HOMEWORK::Homew...=||EViewProc|ObjectPrint|Name FreezeEstimateForecastStats ResidsDependentVariable:QMethod:PanelLeastSquaresDate:11/12/24Time:15:03Sample:20102017Periodsincluded:8Cross-sectionsinduded:4Totalpanelbalancedobservations:32Variable CoefficientStd.Error t-Statistic Prob.C
6.
3650000.
010458608.
61040.0000GROUP*POST
0.
4000000.
02958013.
522470.0000EffectsSpecificationCross-sectionfixeddummyvariablesPeriodfixeddummyvariablesRootMSE
0.033072R-squared
0.997664Meandependentvar
6.465000AdjustedR-squared
0.996379S.D.dependentvar
0.695219S.E.ofregression
0.041833Akaikeinfocriterion-
3.230266Sumsquaredresid
0.035000Schwarzenterion-
2.680615Loglikelihood
63.68426Hannan-Quinncriter.-
3.048072F-statistic
776.5299Durbin-Watsonstat
1.418719ProbF-statistic
0.000000由双向固定效应模型估计得到处理效应的估计值为与基本双差分模型的估计结果d=
0.400,完全一致.进行事前平行趋势检验；3事前平行趋势检验的模型设定如下:2012q-Groupi Time+/3,ost GroupPost+u+v+与/Z xx5=2010it sitit软件估计结果如下:Eviews回Equation:UNTITLEDWorkfile:HOMEWORK::Homework\|oView!Proc|ObjectI[PrintIName{FreezeIIEstimate!Forecast!Stats]ResidsIDependentVariable:QMethod:PanelLeastSquaresDate:11/12/24Time:15:15Sample:20102017Periodsinduded:8Cross-sectionsincluded:4Totalpanelbalancedobservations:32Variable CoefficientStd.Error t-Statistic Prob.C
6.
3650000.
022687280.
55550.0000GROUP,TIME
101.80E-
140.
0641692.80E-
131.0000GROUP*TIME
111.77E-
140.
0641692.76E-
131.0000GROUP*TIME
121.74E-
140.
0641692.71E-
131.0000GROUP*POST
0.
4000000.
0507307.
8848860.0000EffectsSpecificationCross-sectionfixeddummyvariablesPeriodfixeddummyvariablesRootMSE
0.033072R-squared
0.997664Meandependentvar
6.465000AdjustedR-squared
0.995740S.D.dependentvar
0.695219S.E.ofregression
0.045374Akaikeinfocriterion-
3.042766Sumsquaredresid
0.035000Schwarzcriterion-
2.355702Loglikelihood
63.68426Hannan-Quinncriter.-
2.815024F-statistic
518.6110Durbin-Watsonstat
1.418719ProbF-statistic
0.000000筠、分别对夕、0尸钻进行显著性检验，发现其均不显著，说明事前平行趋势假设成£o2立如果处理效应存在，估计税法改革对企业业绩影响的动态变化
4.动态效应模型设定如下
20122017、+£畔Qu=Time Time+u+v+£Ex Gxs it it5=20105=2014软件估计结果如下:Eviews=Equation:UNTITLED Workfile:HOMEWORK::Homework\|□||B||^3Forecast|Stats ResidsViewProc ObjectPrint NameFreeze EstimateDependentVariable:QMethod:Panel LeastSquaresDate:11/12/24Time:15:26Sample:20102017Periods included:8Cross-sections included:4Total panelbalanced observations:32Variable CoefficientStd.Error t-Statistic Prob.C
6.
3650000.
023146274.
99940.0000GROUP*TIME
101.39E-
140.
0654652.13E-
131.0000GROUP*TIME
111.37E-
140.
0654652.09E-
131.0000GROUP*TIME
121.34E-
140.
0654652.04E-
131.0000GROUP*TIME
140.
4500000.
0654656.
8738640.0000GROUP*TIME
150.
3500000.
0654655.
3463380.0001GROUP*TIME
160.
4000000.
0654656.
1101010.0000GROUP*TIME
170.
4000000.
0654656.
1101010.0000Effects SpecificationCross-section fixeddummy variablesPeriod fixeddummy variablesRootMSE
0.030619R-squared
0.997998Mean dependentvar
6.465000Adjusted R-squared
0.
9955660.695219S.E.of regression
0.046291Akaike infocriterion-
3.009417Sum squaredresid
0.030000Schwarz criterion-
2.184940Log likelihood
66.15067Hannan-Quinn criter.-
2.736126F-statistic
410.4800Durbin-Watson stat
1.103448ProbF-statistic
0.000000由结果可见，分别对原假设万广进行检验，、、、均拒“0”=0s=2014201520162017,绝原假设,说明存在显著的因果效应,且政策实施后各年的处理效应基本稳定在左右
0.40。
个人认证
优秀文档
获得点赞 0
