Angrist_-_Mostly_Harmless_Econometrics.pdf

《Angrist_-_Mostly_Harmless_Econometrics.pdf》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、MostlyHarmlessEconometrics:AnEmpiricist’sCompanionJoshuaD.AngristJörn-Ste¤enPischkeMassachusettsInstituteofTechnologyTheLondonSchoolofEconomicsMarch2008iiContentsPrefacexiAcknowledgmentsxiiiOrganizationofthisBookxvIIntroduction11QuestionsaboutQuestions32TheExperimentalIdeal92.1TheSelec

2、tionProblem........................................102.2RandomAssignmentSolvestheSelectionProblem........................122.3RegressionAnalysisofExperiments.................................16IITheCore193MakingRegressionMakeSense213.1RegressionFundamentals..............................

3、........223.1.1EconomicRelationshipsandtheConditionalExpectationFunction...........233.1.2LinearRegressionandtheCEF...............................263.1.3AsymptoticOLSInference..................................303.1.4SaturatedModels,MainE¤ects,andOtherRegressionTalk...............363.2R

4、egressionandCausality......................................383.2.1TheConditionalIndependenceAssumption........................383.2.2TheOmittedVariablesBiasFormula............................443.2.3BadControl..........................................473.3HeterogeneityandNonlinearity...

5、................................513.3.1RegressionMeetsMatching.................................51iiiivCONTENTS3.3.2ControlforCovariatesUsingthePropensityScore....................593.3.3Propensity-ScoreMethodsvs.Regression.........................633.4RegressionDetails..................

6、........................663.4.1WeightingRegression....................................663.4.2LimitedDependentVariablesandMarginalE¤ects....................693.4.3WhyisRegressionCalledRegressionandWhatDoesRegression-to-the-meanMean?.803.5Appendix:Derivationoftheaveragederivativeformula.

7、....................814InstrumentalVariablesinAction:SometimesYouGetWhatYouNeed834.1IVandcausality...........................................844.1.1Two-StageLeastSquares..................................894.1.2TheWaldEstimator.....................................944.1.3GroupedDataand