La lecture en ligne est gratuite
Le téléchargement nécessite un accès à la bibliothèque YouScribe
Tout savoir sur nos offres
Télécharger Lire

Essays in applied microeconometrics [Elektronische Ressource] / Hans-Martin von Gaudecker

De
183 pages
EssaysinAppliedMicroeconometricsHans MartinvonGaudeckerInauguraldissertationzurErlangungdesakademischenGradeseinesDoktorsderWirtschaftswissenschaftenderUniversitätMannheimAbteilungssprecher: Prof. Dr. EnnoMammenReferent: Prof. AxelBörsch Supan,Ph.D.Korreferent: Prof. Dr. ArthurvanSoestTagdermündlichenPrüfung: 15. Juni2007ACKNOWLEDGEMENTSFirst and foremost, I would like to thank my supervisor Axel Börsch Supan forguiding me towards empirical work, creating a truly excellent research environ ment at MEA, and providing me with an exceptionally large number of opportu nitiestomakecontactswithoutstandingresearchersallovertheworld. Themostsalient reflection of my extensive use of these possibilities is the fact that thisthesis is based on joint work with coauthors from London, Mannheim, Rostock,Tilburg, and Lund. I am deeply indebted to each and every one of them: JérômeAdda,JamesBanks,AxelBörsch Supan,RobEuwals,thelateAngelikaEymann,Rembrandt Scholz, Arthur van Soest, and Erik Wengström. Elaborating on eachcoauthor’s contributions would require another thesis chapter. Arthur van Soestalsoagreedtobemysecondadvisor. Iamespeciallygratefulforthis.
Voir plus Voir moins

EssaysinAppliedMicroeconometrics
Hans MartinvonGaudecker
InauguraldissertationzurErlangungdesakademischenGradeseinesDoktorsder
WirtschaftswissenschaftenderUniversitätMannheimAbteilungssprecher: Prof. Dr. EnnoMammen
Referent: Prof. AxelBörsch Supan,Ph.D.
Korreferent: Prof. Dr. ArthurvanSoest
TagdermündlichenPrüfung: 15. Juni2007ACKNOWLEDGEMENTS
First and foremost, I would like to thank my supervisor Axel Börsch Supan for
guiding me towards empirical work, creating a truly excellent research environ
ment at MEA, and providing me with an exceptionally large number of opportu
nitiestomakecontactswithoutstandingresearchersallovertheworld. Themost
salient reflection of my extensive use of these possibilities is the fact that this
thesis is based on joint work with coauthors from London, Mannheim, Rostock,
Tilburg, and Lund. I am deeply indebted to each and every one of them: Jérôme
Adda,JamesBanks,AxelBörsch Supan,RobEuwals,thelateAngelikaEymann,
Rembrandt Scholz, Arthur van Soest, and Erik Wengström. Elaborating on each
coauthor’s contributions would require another thesis chapter. Arthur van Soest
alsoagreedtobemysecondadvisor. Iamespeciallygratefulforthis.
Iappreciatethevaluablecommentsreceivedonversionsofmyworkatwork
shops and seminars at the University of Mannheim, University College London,
theGermanFederalPensionInstitute,TilburgUniversity,theTinbergenInstitute,
aswellasatconferencesinVenice,Paris,Rome,Vienna,Mannheim,andTilburg.
Ishouldliketothankanumberofpeoplewhogavetheirtimetodiscussmywork
withmeandprovideddetailedcommentsthathelpedtoimproveuponit: Syngjoo
Choi, Ralf Himmelreicher, Michael Hurd, Morten Lau, Alexander Ludwig, Jür-
gen Maurer (who deserves a special mention for the countless fruitful discussion
wehadovermanycoffees,dinners,andsharedofficedays),MatthiasParey,Jonas
Radl,DanielSchunk,VladimirShkolnikov,JamesSmith,ErikSørensen,Michael
Stegmann, Andreas Uthemann, and Joachim Winter. My colleagues and fellow
studentsatMEAandCDSEwereacontinuoussourceofadviceandithasbeena
pleasuretoworkthere,notleastbecauseMariaDauer,BrunhildGriesbach,Helga
Gebauer, Isabella Nohe, and Petra Worms Lickteig never interpreted their work
asbeinglimitedtoadministrativematterswhichtheyhandledsuperbly.
During my graduate studies, I had the chance to spend prolonged periods of
timeatotherinstitutionsandIthankthemfortheirhospitality: UniversityCollege
London, the German Federal Pension Institute in Berlin (FDZ RV), and Tilburg
University. Financial support from the German Research Foundation (DFG), the
European Commission Marie Curie program, the Land Baden Württemberg, and
theGermanInsuranceAssociation(GDV)isgratefullyacknowledged.
Last but certainly not least, I would like to thank my family and friends for
theirsupportduringtheyearsofmydissertation.TABLEOFCONTENTS
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. LifetimeEarningsandLifeExpectancy . . . . . . . . . . . . . . . . . 7
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1 TheGermanPublicPensionSystem . . . . . . . . . . . . 10
2.2.2 DescriptionoftheDataset . . . . . . . . . . . . . . . . . 12
2.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
pers2.3.1 RemainingLifeExpectancyatAge65byEP . . . . . 15
pers CP2.3.2 MortalitybyEP andEP . . . . . . . . . . . . . . . 18
pers2.3.3 LifeExpectancyatAge65byEP andPlaceofResidence 20
2.3.4 InternationalComparisons . . . . . . . . . . . . . . . . . 23
2.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.5 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3. TheImpactofIncomeShocksonHealth . . . . . . . . . . . . . . . . . 33
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.2 EmpiricalStrategy . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.1 StochasticProcessforIndividualIncome . . . . . . . . . 37
3.2.2forIndividualHealth . . . . . . . . . . 38
3.2.3 Aggregation . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.2.4 Identification . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2.5 Estimation . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44iv TableofContents
3.3.1 TheFamilyExpenditureSurvey(FES) . . . . . . . . . . . 45
3.3.2 TheGeneralHouseholdSurvey(GHS) . . . . . . . . . . 46
3.3.3 TheHealthSurveyforEngland(HSE) . . . . . . . . . . . 47
3.4 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4.1 TheVarianceofIncomeShocks . . . . . . . . . . . . . . 48
3.4.2 PermanentIncomeShocksandMortality . . . . . . . . . 49
3.4.3ShocksandHealth . . . . . . . . . . . 50
3.4.4 PermanentIncomeShocksandBehaviour . . . . . . . . . 51
3.4.5 RobustnessofResults . . . . . . . . . . . . . . . . . . . 53
3.4.6 RelationtotheLiterature . . . . . . . . . . . . . . . . . . 53
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.6 Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.7 Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4. ExperimentalElicitationofPreferences . . . . . . . . . . . . . . . . . 73
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.2 DataandExperimentalSetup . . . . . . . . . . . . . . . . . . . . 76
4.2.1 TheMultiplePriceListFormat . . . . . . . . . . . . . . . 76
4.2.2 TheCentERpanelExperiment . . . . . . . . . . . . . . . 78
4.2.3 TheLaboratory . . . . . . . . . . . . . . . . 80
4.3 SelectionEffectsintheCentERpanelExperiment . . . . . . . . . 81
4.3.1 DescriptiveStatisticsandParticipation . . . . . . . . . . . 83
4.3.2 Perseverance . . . . . . . . . . . . . . . . . . . . . . . . 87
4.3.3 OverallSelectionandConstructionofSamplingWeights . 90
4.4 ErrorsandPreferences . . . . . . . . . . . . . . . . . . . . . . . 92
4.4.1 ErrorsandInconsistenciesintheLabvs. theCentERpanel 92
4.4.2 PreferencesintheLabvs. theCentERpanel . . . . . . . . 95
4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5. RiskPreferencesintheSmallforaLargePopulation . . . . . . . . . . 103
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104TableofContents v
5.2 TheoreticalFramework . . . . . . . . . . . . . . . . . . . . . . . 107
5.2.1 ASimpleModelofChoiceUnderRisk . . . . . . . . . . 107
5.2.2 PreferencestowardstheResolutionofUncertainty . . . . 108
5.3 DataandExperimentalSetup . . . . . . . . . . . . . . . . . . . . 110
5.3.1Design . . . . . . . . . . . . . . . . . . . . 111
5.3.2 DescriptiveEvidence . . . . . . . . . . . . . . . . . . . . 113
5.4 EconometricSpecification . . . . . . . . . . . . . . . . . . . . . 114
5.5 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
5.5.1 Evidence from a Model not Allowing for Individual Het
erogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . 118
5.5.2 Evidence from a Model of Risk Aversion with Individual
Heterogeneity . . . . . . . . . . . . . . . . . . . . . . . . 120
5.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.7 CertaintyEquivalents . . . . . . . . . . . . . . . . . . . . . . . . 124
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6. RiskAttitude,Impatience,andAssetChoice . . . . . . . . . . . . . . . 129
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
6.2 DataandDescriptiveEvidence . . . . . . . . . . . . . . . . . . . 131
6.2.1 DefinitionofAssets . . . . . . . . . . . . . . . . . . . . 132
6.2.2ofWealth . . . . . . . . . . . . . . . . . . . . 133
6.2.3 MeasuringRiskandTimePreferences . . . . . . . . . . . 135
6.2.4 DeterminantsofAssetChoice,WealthandPreferences . . 140
6.3 ModelandEmpiricalStrategy . . . . . . . . . . . . . . . . . . . 141
6.3.1 Wealth, Preferences and Asset Choice: A Structural Ap
proach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
6.3.2 IdentificationintheModelwithContinuousOutcomes . . 145
6.3.3oftheCovarianceParameters . . . . . . . . 148
6.3.4 IdentificationinthePresenceofDiscreteMeasurements . 149
6.3.5 EstimationoftheStructuralModel . . . . . . . . . . . . . 150
6.4 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
6.4.1 MeasuringWealthandAttitudes . . . . . . . . . . . . . . 151
6.4.2 TheDeterminantsofWealthandAttitudes . . . . . . . . . 155vi TableofContents
6.4.3 TheDeterminantsofAssetChoice . . . . . . . . . . . . . 157
6.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
6.6 AdditionalTables . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1661. INTRODUCTION2 Chapter1.Introduction
Thefieldofeconometricsseekstogiveempiricalcontenttomodelsfromeco
nomictheory. Arguablythemostfrequentlyencounteredstumblingblockonthis
questistheavailabilityofsuitabledata. Variables thatare centralfor thetheories
under scrutiny are often unobserved, measured with error, or simultaneously de
termined with other relevant characteristics. A large part of econometric theory
can be characterised by searching for ways to identify and estimate the parame
tersofinterestfortypicaldataconfigurations. Startingfromthisbasis,theapplied
researcher needs to match these methods with a dataset in which the correspond
ingassumptions–someofwhichwillbefundamentallyuntestable–canbemade
with reasonable confidence. This data driven view of empirical work forms the
threadthatholdsthefiveself containedchaptersofthisdissertationtogether.
In Chapter 2, Rembrandt Scholz and I exploit a novel dataset in order to
study life expectancy differentials among elderly German men. Unlike in many
other countries, publicly available German census records do not include socio
economicvariables. ItwasonlywiththeinceptionoftheResearchDataCentreof
the German Federal Pension Institute in 2004 that it became possible to analyse
data that satisfy the heavy requirements of nonparametric life table estimators.
Utilisingameasureoflifetimelabourincomeforasampleof3.5millionpension
ersin2002,wecanputalowerboundofsixyearsonthelifeexpectancydifference
at age 65 between the groups at the extremes of the lifetime earnings distribu
tion. Furthermore, our analysis shows that the lower overall life expectancy in
theeasternpartofGermanyappearstobedrivenbyacompositioneffect: Condi
tional upon socio economic status, we cannot detect any mortality differences. It
is rather the distributions of earnings that are quite distinct from each other with
relativelymorepersonspopulatingthehighcategoriesintheWestthanintheEast.
Moving beyond a mere data description, the analysis performed in the third
chapter aims at estimating a causal effect of income on health based on a syn
thetic cohort approach. In joint work with Jérôme Adda and James Banks, we
model both variables as stochastic processes for the population between 30 and
60 years of age. The aggregation to the cohort level is motivated by two obser-
vations. First, there are no panel data currently available that contain both health
andsocio economicvariablesinlongenoughtimeseriesastofacilitateatrulydy
namicanalysisattheindividuallevel. Second,wearguethatpermanentshocksto
cohort income are driven by other factors than health innovations, allowing us to
disregard reverse causality issues. We also present some robustness checks with
respecttodeviationsfromthisassumption. Otherfactorsthathavebeenshownto
influence both income and health trajectories – such as education and childhood
circumstance–willbecapturedbytheinitialconditions. Henceweareconfident
thatwecanidentifythecausaleffectthatweaimfor.
Interestingly,wefindapositiveandstatisticallysignificanteffectonmortality,
i.e. positive income innovations lead to higher subsequent mortality. There are

Un pour Un
Permettre à tous d'accéder à la lecture
Pour chaque accès à la bibliothèque, YouScribe donne un accès à une personne dans le besoin