面板数据分析：英文版

　　《面板数据分析》（第2版）是面板数据分析这一领域的经典之作，《面板数据分析》（第2版）系统地介绍了有关面板数据的基本理论，尤其是对面板数据在控制未观察到的个体或时间偏差，以避免设定误差，改善估计效率方面的应用；并且，《面板数据分析》（第2版）审慎地使用了实证研究的案例，这使得《面板数据分析》（第2版）对经济学、商学、社会学和政治科学的研究生和研究人员非常有用。在1986年第一版的成功基础上，《面板数据分析》（第2版）第二版对第一版进行了丰富的修改，以一种严谨易读的方式将有关面板数据研究的近期进展加入到书中，并且使这部分内容与原有的内容融为一体。第二版特别的修改包括：介绍了贝叶斯方法以及在广义矩方法框架下的估计量的严格外生性的概念，使得各种模型的识别联系起来；对估计离散选择模型的半参数方法提出了直觉解释；以及介绍了面板样本选择模型的估计的成对整理方法（methods of pairwise trimming）等。

　　萧政，是南加州大学经济学教授。本书的第1版已经成为经济学文献中有关面板数据的标准介绍。他还是《经济计量模型、技术与应用》一书的合作者，并与他人共同主编了《面板模型和受限因变量模型分析》及《非线性统计推断》等著作。他是计量经济学会的成员以及《计量经济学杂志》的主编和会员。

PrefacetotheSecondEdition
PrefacetotheFirstEdition
Chapter1.Introduction
1.1AdvantagesofPanelData
1.2IssuesInvolvedinUtilizingPanelData
1.2.1HeterogeneityBias
1.2.2SelectivityBias
1.3OutlineoftheMonograph
Chapter2.AnalysisofCovariance
2.1Introduction
2.2AnalysisofCovariance
2.3AnExample
Chapter3.SimpleRegressionwithVariableIntercepts
3.1Introduction
3.2Fixed-EffectsModels:Least-SquaresDummy-VariableApproach
3.3Random-EffectsModels:EstimationofVariance-ComponentsModels
3.3.1CovarianceEstimation
3.3.2Generalized-Least-SquaresEstimation
3.3.3MaximumLikelihoodEstimation
3.4FixedEffectsorRandomEffects
3.4.1AnExample
3.4.2ConditionalInferenceorUnconditional(Marginal)Inference
3.4.2.aMundlak'sFormulation
3.4.2.bConditionalandUnconditionalInferencesinthePresenceorAbsenceofCorrelationbetweenIndividualEffectsandAttributes
3.5TestsforMisspecification
3.6ModelswithSpecificVariablesandBothIndividual-andTime-SpecificEffects
3.6.1EstimationofModelswithIndividual-SpecificVariables
3.6.2EstimationofModelswithBothIndividualandTimeEffects
3.7Heteroscedasticity
3.8ModelswithSeriallyCorrelatedErrors
3.9ModelswithArbitraryErrorStructure-ChamberlainπApproach
Appendix3A:ConsistencyandAsymptoticNormalityofthe
Minimum-DistanceEstimator
Appendix3B:CharacteristicVectorsandtheInverseofthe
Variance-CovarianceMatrixofa
Three-ComponentModel
Chapter4.DynamicModelswithVariableIntercepts
4.1Introduction
4.2TheCovarianceEstimator
4.3Random-EffectsModels
4.3.1BiasintheOLSEstimator
4.3.2ModelFormulation
4.3.3EstimationofRandom-EffectsModels
4.3.3.aMaximumLikelihoodEstimator
4.3.3.bGeneralized-Least-SquaresEstimator
4.3.3.cInstrumental-VariableEstimator
4.3.3.dGEneralizedMethodofMomentsEstimator
4.3.4TestingSomeMaintainedHypothesesonInitialConditions
4.3.5SimulationEvidence
4.4AnExample
4.5Fixed-EffectsModels
4.5.1TransformedLikelihoodApproach
4.5.2Minimum-DistanceEstimator
4.5.3RelationsbetweentheLikelihood-BasedEstimatorandtheGeneralizedMethodofMomentsEstimator(GMM)
4.5.4Random-versusFixed-EffectsSpecification
4.6EstimationofDynamicModelswithArbitaryCorrelationsintheResiduals
4.7Fixed-EffectsVectorAutoregressiveModels
4.7.1ModelFormulation
4.7.2GeneralizedMethodofMoments(GMM)Estimation
4.7.3(Transformed)MaximumLikelihoodEstimator
4.7.4Minimum-DistanceEstimator
Appendix4A:DerivationoftheAsymptoticCovarianceMatrixoftheFeasibleIVIDE
Chapter5.Simultaneous-EquationsModels
5.1Introduction
5.2JointGeneralized-Least-SquaresEstimationTechnique
5.3EstimationofStructuralEquations
5.3.1EstimationofaSingleEquationintheStructuralModel
5.3.2EstimationoftheCompleteStructuralSystem
5.4TriangularSystem
5.4.1Identification
5.4.2Estimation
5.4.2.aInstrumental-VariableMethod
5.4.2.bMaximum-LikelihoodMethod
5.4.3AnExample
Appendix5A
Chapter6.Variable-CoefficientModels
6.1Introduction
6.2CoefficientsThatVaryoverCross-SectionalUnits
6.2.1Fixed-CoefficientModel
6.2.2Random-CoefficientModel
6.2.2.aTheModel
6.2.2.bEstimation
6.2.2.cPredictingIndividualCoefficients
6.2.2.dTestingforCoefficientVariation
6.2.2.eFixedorRandomCoefficients
6.2.2.fAnExample
6.3CoefficientsThatVaryoverTimeandCross-SectionalUnits
6.3.1TheModel
6.3.2Fixed-CoefficientModel
6.3.3Random-CoefficientModel
6.4CoefficientsThatEvolveoverTime
6.4.1TheModel
6.4.2PredictingβtbytheKalmanFilter
6.4.3MaximumLikelihoodEstimation
6.4.4TestsforParameterConstancy
6.5CoefficientsThatAreFunctionsofOtherExogenousVariables
6.6AMixedFixed-andRandom-CoefficientsModel
6.6.1ModelFormulation
6.6.2ABayesSolution
6.6.3AnExample
6.6.4RandomorFixedParameters
6.6.4.aAnExample
6.6.4.bModelSelection
6.7Dy.namicRandom-CoefficientModels
6.8AnExample-LiquidityConstraintsandFirmInvestmentExpenditure
Appendix6A:CombinationofTwoNormalDistributions
Chapter7.DiscreteData
7.1Introduction
7.2SomeDiscrete-ResponseModels
7.3ParametricApproachtoStaticModelswithHeterogeneity
7.3.1Fixed-EffectsModels
7.3.1.aMaximumLikelihoodEstimator
7.3.1.bConditionsfortheExistenceofaConsistentEstimator
7.3.1.cSomeMonteCarloEvidence
7.3.2Random-EffectsModels
7.4SemiparametricApproachtoStaticModels
7.4.1MaximumScoreEstimator
7.4.2ARoot-NConsistentSemiparametricEstimator
7.5DynamicModels
7.5.1TheGeneralModel
7.5.2InitialConditions
7.5.3AConditionalApproach
7.5.4StateDependenceversusHeterogeneity
7.5.5TwoExamples
7.5.5.aFemaleEmployment
7.5.5.bHouSeholdBrandChoices
Chapter8.TruncatedandCensoredData
8.1Introduction
8.2AnExample-NonrandomlyMissingData
8.2.1Introduction
8.2.2AProbabilityModelofAttritionandSelectionBias
8.2.3AttritionintheGaryIncome-MaintenanceExperiment
8.3TobitModelswithRandomIndividualEffects
8.4Fixed-EffectsEstimator
8.4.1PairwiseTrimmedLeast-Squaresand
Least-Absolute-DeviationEstimatorsfor
TruncatedandCensoredRegressions
8.4.1.aTruncatedRegression
8.4.1.bCensoredRegressions
8.4.2ASemiparametricTwo-StepEstimatorfortheEndogenouslyDeterminedSampleSelectionModel
8.5AnExample:HousingExpenditure
8.6DynamicTobitModels
8.6.1DynamicCensoredModels
8.6.2DynamicSampleSelectionModels
Chapter9.IncompletePanelData
9.1EstimatingDistributedLagsinShortPanels
9.1.1Introduction
9.1.2CommonAssumptions
9.1.3IdentificationUsingPriorStructureoftheProcessoftheExogenousVariable
9.1.4IdentificationUsingPriorStructureoftheLagCoefficients
9.1.5EstimationandTesting
9.2RotatingorRandomlyMissingData
9.3Pseudopanels(orRepeatedCross-SectionalData)
9.4PoolingofaSingleCross-SectionalandaSingleTime-SeriesDataSet
9.4.1Introduction
9.4.2TheLikelihoodApproachtoPoolingCross-SectionalandTime-SeriesData
9.4.3AnExample
Chapter10.MiscellaneousTopics
10.1SimulationMethods
10.2PanelswithLargeNandT
10.3Unit-RootTests
10.4DatawithMultilevelStructures
10.5ErrorsofMeasurement
10.6ModelingCross-SectionalDependence
Chapter11.ASummaryView
11.1Introduction
11.2BenefitsandLimitationsofPanelData
11.2.1IncreasingDegreesofFreedomandLesseningtheProblemofMulticollinearity
11.2.2IdentificationandDiscriminationbetweenCompetingHypotheses
11.2.3ReducingEstimationBias
11.2.3.aOmitted-VariableBias
11.2.3.bBiasInducedbytheDynamicStructureofaModel
11.2.3.cSimultaneityBias
11.2.3.dBiasInducedbyMeasurementErrors
11.2.4ProvidingMicroFoundationsforAggregateDataAnalysis
11.3EfficiencyoftheEstimates
Notes
References
AuthorIndex
SubjectIndex