Searching for sources of inefficiency in the German health care sector: demand-side, supply-side, and labour-force- status effects on health and health care utilisation [Elektronische Ressource] / vorgelegt von Hendrik Schmitz

Description

Searching for sources of inefficiencyin the German health care sector:demand-side, supply-side, andlabour-force-status effects on healthand health care utilisationD I S S E R T A T I O Nzur Erlangung des akademischen Grades einesDoktors der Wirtschaftswissenschaften(Dr. rer. pol.)durch die Fakultät für Wirtschaftswissenschaften derUniversität Duisburg-EssenCampus Essenvorgelegt vonDipl.-Volksw. Hendrik Schmitzaus RheinbergEssen 2011Tag der mündlichen Prüfung: 12. 1. 2011Erstgutachter: Prof. Dr. Reinhold SchnabelZweitgutachter: Prof. Dr. Stefan Felder2PrefaceI wrote this dissertation while I was a doctoral student at the Ruhr Graduate Schoolin Economics. During these three years I did not only benefit from financial supportby the RGS and the Leibniz Association but also from the infrastructure providedby the RGS. Passing the entire process from course work to finally submitting thedissertation together with my fellow students of the third cohort made this a greattime. I especially thank Daniel Baumgarten, Christoph Braun, Markus Hoermann,and Malte Rieth for being a great team also apart from doing research.A special thanks goes to my supervisor Reinhold Schnabel for his constant support,encouragement, and suggestions. Moreover, I thank Stefan Felder for not hesitatingto be my second supervisor and for his critical remarks. Several people read eithersingle or all chapters of the thesis and improved it with many comments.

Sujets

• Social work

Informations

• Publié par: universitat_duisburg-essen
• Publié le: 01 janvier 2011
• Nombre de lectures: 26
• Langue: English
• Poids de l'ouvrage: 1 Mo

Extrait

Searchingforsourcesofinefficiency
inthedemand-side,Germanhealthsupply-side,caresectoandr:
labandour-force-statushealthcareeffectsutilisationonhealth

DISSERTATION

zurErlangungdesakademischenGradeseines

DoktorsderWirtschaftswissenschaften

ol.)prer.(Dr.

durchdieFakultätfürWirtschaftswissenschaftender

Duisburg-EssenersitätUniv

EssenCampus

onvorgelegtv

hmitzScHendrikolksw.Dipl.-V

Rheinausgerb

2011Essen

agT

der

ündlicmhen

ter:hErstgutac

ter:heitgutacZw

Prüfung:

2

12.

1.

Prof.

Prof.

2011

Dr.

Dr.

Reinhold

Stefan

nabhSc

elderF

el

Preface

IwrotethisdissertationwhileIwasadoctoralstudentattheRuhrGraduateSchool
inEconomics.DuringthesethreeyearsIdidnotonlybenefitfromfinancialsupport
bytheRGSandtheLeibnizAssociationbutalsofromtheinfrastructureprovided
bytheRGS.Passingtheentireprocessfromcourseworktofinallysubmittingthe
dissertationtogetherwithmyfellowstudentsofthethirdcohortmadethisagreat
time.IespeciallythankDanielBaumgarten,ChristophBraun,MarkusHoermann,
andMalteRiethforbeingagreatteamalsoapartfromdoingresearch.

AspecialthanksgoestomysupervisorReinholdSchnabelforhisconstantsupport,
encouragement,andsuggestions.Moreover,IthankStefanFelderfornothesitating
tobemysecondsupervisorandforhiscriticalremarks.Severalpeoplereadeither
singleorallchaptersofthethesisandimproveditwithmanycomments.Theseare
LeilanieBasilio,DanielBaumgarten,AnnikaHerr,MatthiasKeese,AnnikaMeng,
StefanieNeimann,AlfredoPaloyo,andNicolasZiebarth.Allpartsofthethesisalso
stronglybenefitedfromthediscussionsattheRGSWorkshopinEssen,theRWI
therapy-seminar,andseveralGermanandinternationalconferences.

IamgratefultotheRWIforallowingmetousetheirofficesandinfrastructure
andtotheBKK-lunchgroupforhavinglaid-backbreakswithdiscussionsaboutall
economics.buttopics

Mostofall,IwanttothankStefanieNeimann,mymotherChristaSchmitz,and
KaiSchmitzforalwayskeepingfaithwithme.Theirencouragingsupportisan
king.bacaluablevin

i

tstenCon

ductiontroIn1

1.1TheProblem...............................

1.2SomeDescriptives.............................

1.3TheData.................................

1

1

7

8

1.4OverviewandSummaryofFindings...................12

2Morehealthcareutilisationwithmoreinsurancecoverage?-Evi-

dencefromalatentclassmodelwithGermandata18

2.1Introduction................................18

2.2InstitutionalBackground.........................24

2.3EmpiricalModel.............................26

2.3.1DataandVariableDescription..................26

2.3.2LatentClassHurdleModel....................30

2.4EstimationResults............................34

2.5Endogeneityconcerns...........................38

ii

3

4

2.6Conclusion.................................41

2.7Appendix.................................45

Practicebudgetsandthepatientmixofphysicians-Evaluating
effectsofremunerationsystemreformsonphysicianbehaviourin
49yGerman

3.1Introduction................................49

3.2PaymentSystemandMajorReforms..................54

3.3Data....................................58

3.4EmpiricalStrategy............................61

3.4.1CountDataHurdleModel....................61

3.4.2EstimationStrategy.......................64

3.4.3ControlforPanelAttrition....................66

3.5EstimationResults............................67

3.6RobustnessChecks............................76

3.6.1SubsamplesandSpecifications..................76

3.6.2TestforSingleSpellAssumption................77

3.7Conclusion.................................81

3.8Appendix.................................83

RiskaversionandadvantageousselectionintheGermansupple-
87insurancehealthtarymen

4.1Introduction................................87

iii

4.2Previoustheoreticalandempiricalliterature..............91
4.3Institutionalbackground.........................92
4.4Data....................................94
4.5Methods..................................97
4.6Results...................................100
4.7Conclusion.................................105
4.8Appendix.................................107

5Whyaretheunemployedsoill?-Thecausaleffectofunemploy-
109healthontmen

5.1Introduction................................109
5.2Data....................................112
5.3EmpiricalStrategy............................116
5.4Results...................................119
5.5Conclusion.................................126
5.6Appendix.................................129

DiscussionsConcluding6

Bibliography

ablesTofList

FiguresofList

iv

136

140

814

151

1Chapter

ductiontroIn

ProblemThe1.1

InGermany,theaveragecontributionratetothesocialhealthinsuranceincreased

from8.2percentin1970to14.9percentin2009(IW,2009).Inashorterperiodof

time,theinsurancecontributionsintheprivatesystemeventripledbetween1985

and2005innominalterms(Grabka,2006).Likewise,Albrechtetal.(2010)report
anaverageincreaseinhealthinsurancecontributionsbetween1997and2008by2.4
percentperyearinthesocialhealthinsuranceandby3.9percentintheprivate

healthinsurance.Whilethelatterisnotyetapublicissue,theformerinducesa
steadypublicconcern.Sincethecontributionstothestatutoryhealthinsuranceare
takenasapayrolltax,theyimposeawedgebetweengrossandnetwageandhave
directimplicationsforthelaboursupplyofindividuals.

Thereasonsforthisevolutionaremanifoldandcanbefoundontherevenueas

wellasontheexpendituresideofthehealthinsurances.Mostimportant,thede-
mographicchange,i.e.,theincreasingaverageageoftheGermanpopulation,has
consequencesforboththerevenuesofthestatutoryhealthinsuranceandthehealth

1

careexpenditures.Adecreasingabsolutenumberofworkingindividualsdecreases
theincomeofthehealthinsurancecompaniesceterisparibussincethesumofcon-
tributionsisdirectlylinkedtothepayrollsum.Atthesametime,anageingsociety
needsmorehealthcareservicesandcausesmoreexpenditures.1Amoreimportant
reasonofhigherexpendituresisthetechnologicalprogress.Thedevelopmentof
newdrugsandmedicalequipmenthelpspatientstosurvivewithchronicconditions
whichwouldhaveledtodeathsomeyearsago.Thishighlywelcomedevelopment,
however,increaseshealthcareexpenditures,andprobablyexplainsthemajorpart
oftheoverallincreaseinthelastdecades(see,e.g.,Newhouse,1992).

ConceptstoimproveorstabilisetherevenuesoftheGermanhealthinsuranceare
beyondthescopeofthisthesiswhichfocusesontheexpendituresside.InGermany,
healthcareexpenditures(HCE)asashareoftheGDPhaverisensharplybetween
1970and2006,from6percenttomorethan10percent(seeFigure1.1).How-
ever,whileGermanywasthecountrywiththehighesthealthcareexpendituresin
the1970s,thischangedthereafter.Sincethemiddleofthe1990s(until2006)the
ratiohasbeenstayingfairlystablewhileitincreasedsubstantiallyinmostother
industrialisedcountries.However,Germanystillhasthefourthhighesthealthcare
expendituresofallOECDcountries.ThereasonforthesteadinessinGermany
seemstobetheintroductionofsectoralbudgetsforhospitalexpenditures,ambu-
latoryexpendituresandpharmaceuticalsintheearly1990s.Thesebudgetsarea
meansforthepolicymakertocontaincosts;inmostoftheyearstheyincreased
onlybytherateofthepayrollwageincrease.However,whileatfirstsightitseems
pleasing,thiscomesatacost:ifthedemandforhealthcareconstantlyincreasesdue
todemographicchangeandtechnologicalprogressbuttheexpendituresarecapped,
thiscanonlyberesolvedbyarationingofhealthcareservices.Whileofficiallythere

1Thereis,however,adebateonthispoint.See,e.g.,Zweifeletal.(1999),orWerblowetal.
(2007).

2

isnorationingintheGermanhealthcaresystem,therewasagrowingpublicdebate
about“hiddenrationing”inthelastyears.

Figure1.1:HCEas%ofGDPofselectedcountries

Source:OECDHealthData2008,ownillustration

Sincethedemographicchangeandthetechnologicalprogresswillmostlikelynot
bestopped,theonlypossiblesolutiontothisdilemma-apartfromrationingor
prioritisation-istodetectandreduceinefficienciesinthehealthcaresector.These
inefficienciesarise,forinstance,frominformationasymmetriesbetweendifferent
agentsinthehealthcaremarket.Onesuchinformationasymmetryisbetween
insurancecompanyandinsuree.Sincetheinsurancecompanycannotfullysuper-
wisetheinsured,itisdifficulttosanctionbothanunhealthybehaviourandexcess
demandfortrivialhealthcareservices.Ifindividualsbothexhibitunhealthybe-
haviour(theso-calledex-antemoralhazard)and/orincreasehealthcareutilisation
duetotheirhealthinsurance(ex-postmoralhazard),apossiblesolutioncanbecost-
sharingbetweenhealthinsuranceandinsureeinordertoreduceoverconsumptionof
medicalservices(Breyeretal.,2004).However,demand-sidecost-sharingalsohas
itslimits.Itimposesfinancialriskonthepatients,thus,itinducesawelfarelossin

3

aworldwithrisk-averseindividualsandnomoralhazardbehaviour.Moreover,it
mightbeinconflictwithgoalsofuniversalaccesstohealthcare,especiallyofpoorer
households(seeEllisandMcGuire,1993).

Anotherinformationasymmetryistheonebetweenpatientandphysician.Since,in
general,thephysicianhasmoreinformationaboutthepatient’shealthstatusthan
thepatienthimself,shecan-atleasttoacertaindegree-inducedemandfromthe
patient.This,ofcourse,impliesaninefficientuseofresourcesinthehealthcare
system.Here,theremunerationsystemmightbeagoodinstrumenttogouvernthe
behaviourofthephysicianifsheisnotfullyaltruist