《健康經(jīng)濟(jì)學(xué)》課件Chapter15

MarketsforhealthcareandinsuranceWehavelearnedabouttheproblemsthatariseinboththesemarketsMarketsforhealthcareservicesOligopolypricingMonopolyrentsfordoctorsandspecialistsMedicalarmsracesHealthinsurancemarketsAdverseselectionandunderinsuranceMoralhazardandtechnologyoveruseHealthpolicytriestodealwiththeseproblems(buteverynewpolicycreatesnewproblemsorexacerbatesoldones).Ch15|ThehealthpolicyconundrumARROW’SIMPOSSIBILITYTHEOREMArrow’simpossibilitytheoremThetaskofdesigninganationalhealthsystemisatitsheartanoptimizationproblem,notdissimilartothetaskofanindividualintheGrossmanmodel.Societiesmustdecidehowmuchtimeandmoneytheywanttospendonimprovingtheirhealth,andhowmuchtimeandmoneytheywanttospendonothernationalpriorities–likeeducationandthemilitary.Thentheymustchartastrategyforachievingthelevelofhealththeywant,bothcheaplyandefficiently.Arrow’simpossibilitytheoremButthisanalogybetweenanindividualandasocietyisnotquiteright.Asociety–composedofmanypeople–isfundamentallydifferentfromasingleperson.Forexample,eachindividualinasocietyispresumedtohaveconsistentandtransitivepreferences.Withouttransitivepreferences,welfareeconomicsfallsapart.However,economistKenArrowhasprovedthatsocietiesdonotnecessarilyhavetransitivepreferences,evenwheneveryoneinthemdoes.HisfindingisknownasArrow’simpossibilitytheorem.ImplicationsofArrow’stheoremArrow’stheoremsuggestsitdoesnotmakesensetospeakofan“optimal”healthpolicyforacountrybecausesocietiesmaynothavepreferencesthatcanbeoptimizedinthetraditionalsense.Nevertheless,politicaldecisionsdogetmadeandvariousnationalhealthpolicieshaveemerged.Wewillassessthesepoliciesbyanalyzinghowwelltheymeetthreebroadgoals:health,wealth,andequity.Theseassessmentscannotrevealwhichpoliciesare“optimal,”buttheyallowustostudythetradeoffsinherentinhealthpolicy.Ch15|ThehealthpolicyconundrumTHEHEALTHPOLICYTRILEMMAThehealthpolicytrilemmaNationshavethreebroadgoalsinmindwhendesigninghealthpolicyAnyattemptbyanationtomoveclosertooneofthesethreegoalsnecessarilyinvolvesatradeoffthatmovesthatnationfurtherawayfromanothergoal.Forexample,anyhypotheticalpolicythatcombatsadverseselectionandincreasesequitywouldeitherincreasecostsorlowerhealthforsome.ThehealthpolicytrilemmaTherewillalwaysbetradeoffs,sotherewillneverbeaperfecthealthcaresystemwhereallthreegoalsaremaximized.ThehealthpolicytrilemmaFurthermore,peopledisagreeabouthowimportanteachofthesethreeis.Somecountriesvaluesocialequityveryhighly,andarewillingtopaymoreintaxestoachieveit.Othersplaceahighervalueonhealth,andarewillingtotoleratemoremoralhazardormonopolypricingtosecureit.KeyhealthcarepolicychoicesEverypolicychoiceinvolvesatradeoffbetweenhealth,wealthandequity(otherwiseitwouldbeobviousandprobablyimplementedalready).Thepolicyoptionsthatfollowarethuspresentedasanswerstothethreebroadquestionsthatanynationalhealthcaresystemmustanswer:Howshouldinsurancemarketswork?Howshouldmoralhazardbecontrolledinpublicinsurance?Howshouldhealthcareprovidermarketsberegulated?Ch15|ThehealthpolicyconundrumHOWSHOULDHEALTHINSURANCEMARKETSWORK?Howshouldhealthinsurancemarketswork?Severaloptions:CompletelyprivateinsurancemarketsUniversalpublicinsuranceCompulsoryinsuranceEmployer-sponsoredinsuranceMeans-testedhealthinsuranceThesearenotmutuallyexclusive,andmanynationsemployseveralatonceOption1:PrivateinsurancemarketsTheRothschild-Stiglitzmodelmodelpredictsthatinprivatemarkets,onlythefrailcustomersareinsuredfullyandmuchofthepopulationisunderinsured.Undercertainconditions,acompletelyprivatemarketcanunravelcompletely,leadingtouninsuranceforeveryone.Thisoptionminimizesgovernmentinvolvement,butitresultsinmaximaladverseselection.Taxpayersarehappywithlowtaxbills,butmanycitizenscannotbuyfullinsurance.Insteadtheyfretaboutthemedicalbillstheymightrackupiftheybecomeillorinjured.Option2:UniversalpublicinsuranceThegovernmentprovidesinsurancetoallcitizens,andfinancesitwithtaxes.Thispolicyoptionisappealingbecauseitside-stepsadverseselectionandendsuninsurance.Italsofurthersthegoalofequitybecausethepoorpaylittleornothingforcoverage.However,withuniversalpublicinsurance,stepsmustbetakentocontrolmoralhazard,whichcanexplodethegovernmentbudgetifleftunchecked.Option2:UniversalpublicinsuranceHighertaxesarethemaincostofpublicinsurance.Further,mosttaxesdistortbehaviorbydiscouraginglaborandcommerce,sotheentireeconomymaybecomelessefficientasaresult.Butsomearguethatuniversalpublicinsuranceismoreefficientthanprivateinsurancemarketsbecauseoflowoverheadcosts.Highertaxesareonecostofpublicinsurance.Note:thisis“single-payer”healthcarebecauseoneentity(thegovernment)paysforallcare.Option3:CompulsoryinsuranceAmandate(alegalrequirementthateveryoneinapopulationpurchaseprivateinsurance)confrontsadverseselectionbyeffectivelybanningit.Forexample,evenhealthycustomerswhowouldprefertooptoutarelegallyrequiredtobuyintothesystem.Butamandateisnotfreeforgovernmentsanddoesnotabsolvethemofregulatingthemarket.Amandatecanbeexpensive,andmanycitizenscannotaffordit.Thus,mandatesareusuallyeithercoupledwithsubsidiestothepoororpaidforwithpayrolltaxes.Amandatemustalsobecarefullydefinedoritmaybecompletelyineffective.Option4:Employer-sponsoredinsuranceUndersuchasystem,employersarerequiredorencouragedtoofferaprivateinsurancecontracttoalloftheiremployees.Job-specifichumancapitalprovidesastrongincentiveforhealthyemployeeswithalowriskofillnesstopoolwithhighrisk,unhealthyemployees.Thismitigatedadverseselection.Drawbacks:cancreatelabormarketinefficiencies,andnotappropriateforunemployedpopulations(children,retirees,disabled).Option5:Means-testedinsuranceSubsidizedhealthcareforthepoor.Example:MedicaidintheU.S.Itattemptstoimproveequitybyprovidinghealthcaretothosewhootherwisecouldnotaffordit.Thecostsofexpandingsubsidizedinsuranceforthepoorarebasicallyidenticaltothecostsofexpandingpublichealthinsuranceinotherways:highertaxburdensandgreatermoralhazard.Ch15|ThehealthpolicyconundrumHOWSHOULDMORALHAZARDBECONTROLLED?Howshouldmoralhazardbecontrolled?Inaprivatemarket,privateinsurerscompetetooffercustomerstheoptimalmixofinsurancecoverageandmoralhazardcontrol.Butwhengovernmentsentertheinsurancemarket,lawmakersandpolicymakersassumeresponsibilityforthesetoughdecisions.Theexperienceofmanycountrieshasshownthatmoralhazardcontroliscontroversialandpoliticallytreacherous.Howshouldmoralhazardbecontrolled?Policymakersdonothaveaneatlylabeledfiguretohelpthemmakethesedecisions,sotheymustmakesomeeducatedguessesabouthowmuchmoralhazardtoeliminate.Howshouldmoralhazardbecontrolled?Again,severaltoolsavailable:Healthtechnologyassessment(HTA)CostsharingGatekeepingandqueuingProspectivepaymentsOption1:Cost-effectivenessanalysisCEAentailsgatheringinformationabouttreatmentoptionsanddeterminingwhichoptionsproducethemostadditionalhealthfortheleastcost.CEAlimitsmoralhazardbyreducingspendingoninefficient,costlytreatments.ButCEAalsomakesinsurancecontractsless“full”forpatients,becausesomeservicesarenolongercovered.Thistradeoffcanbeworthwhilebecauseitmakestheentiresystemcheaper.Option1:Cost-effectivenessanalysisButdenyingcoverageforsometreatmentsmaybeunappealingforpoliticalreasons.WhilesomegovernmentshaveembracedCEA,othershavereactedbyshunningitaltogether.Example:U.S.MedicareisforbiddenbylawfromusingCEAinitscoveragedecisions.Medicarecoversanymedicallyeffectivetreatment,nomatterhowexpensive.Thisstrategyobviatesgut-wrenchingdecisionsabouttreatingsickpatients,butitalsoallowsmoralhazardtoflourish.Option2:CostsharingCostsharingmaybeaccomplishedthroughtheuseofdeductibles,coinsurance,andcopayments.Theseareout-of-pocketcoststhatinsuredpatientspaywhentheyreceivehealthcare.CostsharingcontrolsmoralhazardinawaythatissometimesmorepoliticallypalatablethanCEA,butitalsomakeshealthcarelessaffordableforpatients.Thiscanundermineequity.Option2:CostsharingExample:theUSMedicaresystemdoesnotfullycoverpatientcosts.Asof2012,Medicareenrolleesmustpaythefirst$1,156dollarsofexpensesforeachhospitalvisitandthefirst$140ofoutpatientclinicexpenseseachyear.Theymustalsostartpaying$289perdayonceahospitalstaylastslongerthan60days.Thisforcesenrolleestoeithereconomizeorpurchasesupplementalprivateinsurance.Option3:GatekeepingandqueuingGate-keepingentailsatieredsystemofdoctorsthatpatientsmustvisitinaspecifiedorder.Thiskeepscostsdownbyeliminatingfrivolousappointmentsandfocusinglimitedresourcesonpatientswhotrulyneedcare.Publicinsurancesystemsalsocontrolcostsbylimitingthetotalnumberofspecialistsavailable.However,whendemandforspecialists’servicesoutstripssupply,queuesresult.Option3:GatekeepingandqueuingUsuallyqueuesareasignofamarketinefficiency,butinthepresenceofmoralhazard,queuesmightbeanindicationofinflateddemand.Ifso,limitingthenumberofspecialistmaysavemoneywithoutsacrificinghealth.Thehassleofwaitinginlineconstitutesanon-financialcostthatpatientsmust“pay”forcare.Queue-basedsystemsmaybemoreequitablethanacost-sharingsystemifitmeansthatrichandpooralikemustwaitforcare.Butqueuingsystemsriskprovokingpoliticalbacklash.Option4:ProspectivepaymentsThetraditionalmethodforpayingforhealthcareisretrospectivepayments.Suchpaymentsaremadeafteraserviceisrendered,andtheamountpaiddependsonhowmuchhealthcareisreceived.Inafee-for-servicesystem,doctorshavenoreasontodenypatientsaservicebecausethecostsaretoohigh.Thissystemfosterstrustbetweenpatientsanddoctors,butcreatesincentivesforphysician-induceddemand.Option4:ProspectivepaymentsAnalternativesystemdesignedtoreducemoralhazardisprospectivepayments.Withprospectivepayments,paymentsaremadetodoctorsorhospitalsbeforehealthcareisdelivered.Chargesarenotbasedonproceduresperformed,butontheconditionofthepatientwhoisadmitted.Example:Aprospective-paymentssystemwillpayhospitalsafixedamountfortreatinganyheartattackpatient.Thisgiveshospitalsincentivestoeconomizeintheirtreatmentofheartattackpatients,becausetheynolongerreceiveextrapaymentsfordoingextrawork.Option4:ProspectivepaymentsSincetheearly1980s,governmentsaroundtheworldhaveembracedprospectivepaymentschemesasaneffectivewaytoreducemoralhazardandphysician-induceddemand.Butprospectivepaymentsystemscomeataprice.Doctor-patientrelationshipsturningadversarialInonestudyconductedaftertheU.S.Medicareprogramimplementedprospectivepaymentsin1984,patientmortalityincreasedsignificantlyatasubsetofhospitalsinthemonthsafterthetransition.Ch15|ThehealthpolicyconundrumHOWSHOULDHEALTHCAREPROVISIONBEREGULATED?Howshouldhealthcareprovisionberegulated?Recallthemaladiesofprivatehealthcaremarkets:OligopolypricingMonopolyrentsfordoctorsandspecialistsMedicalarmsracesPhysician-induceddemandWhatsetsofpoliciescancombatthesemaladieswithoutcreatingnewinefficienciesthatareevenworse?Howshouldhealthcareprovisionberegulated?Approach1:Nationalizehealthcareprovision.Underthisapproach,hospitalsaregovernment-runandfinancedbytaxes,andphysiciansareemployedbythegovernment.Thisapproachcouldreducecostsofmedicalcareandimprovequalityofcareifthemaladiesofhospitaloligopolyorinefficientqualitycompetitionaresufficientlysevere.Itmayalsolimitinefficientqualitycompetition.Howshouldhealthcareprovisionberegulated?AvailableoptionsPublicprovisionPrivatehospitalmarketsGovernment-setpricesOption1:PublicprovisionUnderthisapproach,hospitalsaregovernment-runandfinancedbytaxes,andphysiciansareemployedbythegovernment.Thisapproachcouldreducecostsofmedicalcareandimprovequalityofcarebybanishingoligopolypowerandmedicalarmsraces.Somealsosuggestthatnationalizedsystemsarelessefficientthanprivatemarkets.Governmentsarevulnerabletoagencyproblems,becausegovernmentworkersmayhavelessincentivethanprivateworkerstoensurethesuccessoftheirhospital.Governmentsystemsalsolackclearfeedbackmechanismstocorrectthemiftheyarenotsucceeding.Option1:PublicprovisionEmpirically,countrieswithnationalizedsystemsseembetteratcontrollinghealthcarecosts.However,thecommonchargeagainstgovernment-runhospitalsisthattheyofferlowerqualityhealthcare.Example:Countrieswithpublichospitalssufferlongqueues.Option2:PrivateprovisionThisapproachallowsforcompetitionamonghospitalsandpreservestheincentivesforhospitalstooperateefficiently.However,inprivatemarkets,toolittlecompetitionleadstomarketpowerandtheaccompanyingsociallossduetohighpricesandunderprovision.Conversely,toomuchcompetitioncanexacerbateinefficientqualitycompetition,leadtoamedicalarmsrace,andincreasehealthcarecosts.Option2:PrivateprovisionAnotherconcernisthatsomepopulations–likethepooranduninsured–willlackaccesstocare.Onesolutionistogivetaxbreakstonon-profithospitals,whichhistoricallyhaveattendedtothepoorandthevulnerable.Mostdevelopedcountriesalsorequirehospitalstoprovideemergencycaretoincomingpatientsregardlessoftheircitizenshipstatusorabilitytopay.Such“l(fā)astresort”lawspromoteequitybyensuringaminimumlevelofcareforeveryone,buttheyalsoimposecostsanddeterhospitalsfrombuildingemergencyroomsandtraumacenters.Option3:Government-setpricesBysettingprices,governmentsaimtopreventprivateprovidersfromexercisingmarketpowerandkeephealthcareaffordable.Intheory,suchpricecontrolscouldcontainhospitalcosts,butgovernmentsetpricescouldalsoinducesomeperverseincentives.Unlesspricesaresetproperly,treatmentspricedbelowmarginalcostsmaynotbeoffered,whilethemostprofitableservicesmaybeover-prescribed.Ch15|ThehealthpolicyconundrumCOMPARINGNATIONALHEALTHPOLICIESThreehealthpolicymodelsEverycountryhasitsownsetofpolicies,buteachcountry’sapproachfitsinoneofthreebroadgroups:BeveridgemodelBismarckmodelAmericanmodelBeveridgemodelBeveridgemodelSingle-payerinsurancePublicprovisionofhealthcare(physiciansaregovernmentemployees)VerylittlecostsharingatpointofserviceEmphasisonequityExamples:UK,Scandinavia,Canada,Australia,NZBismarckmodelAmericanmodelBismarckmodelBeveridgemodelBismarckmodelCompulsoryprivateinsurancePrivatehospitalsanddoctorsStrictpricecontrolssetbygovernment(sometimesinnegotiationwithdoctorsandhospitals)BalancesequityandlibertyExamples:Germany,Japan,Switzerland,NetherlandsAmericanmodelAmericanmodelBeveridgemodelBismarckmodelAmericanmodelPrivatemarketsinacentralroleNomandateforuniversalinsuranceNopricecontrolsPublicinsuranceforselectgroups:elderlyandpoorEmphasisonlibertyExamples:uniquetotheUS(othernationshadsimilarsystemsinpreviousdecades)ComparingnationalhealthpoliciesComparingahealthoutcome(likelifeexpectancy)andexpenditurescanserveasakindofreportcardforanation’shealthcaresystem.Howmuchhealthisachievedathowmuchcosttowealth?Doesnotprovideameaningfulassessmentofequity.ComparingnationalhealthpoliciesButwhatif:differentcountrieshavedifferentinherentlevelsofhealth?peopleindifferentcountriesvaluehealthdifferently?countriesdifferinhowmuchhealthinequalityexists?Ifcountriesdifferinanyoftheseways,thentheprecedinggraphisinsufficienttorenderjudgmentontheeffectiv