Cognitive determinants of healthcare evaluations : a comparison of Eastern and Western European countries

(1)

ContentslistsavailableatScienceDirect

Health

Policy

jo u rn al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / h e a l t h p o l

Cognitive

determinants

of

healthcare

evaluations

–

A

comparison

of

Eastern

and

Western

European

countries

Simone

M. Schneider

a,∗

_,

_Tamara

_Popic

b

a_Max_Planck_Institute_for_Social_Law_and_Social_Policy,_Germany b_Institute_of_Social_and_Political_Science,_University_of_Lisbon,_Portugal

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received3August2017 Receivedinrevisedform 30November2017 Accepted29December2017 Keywords: Attitudes Publicopinion Healthcare Efﬁciency Performance

a

b

s

t

r

a

c

t

Knowingthepublicopinionofhealthcareisessentialwhenassessinghealthcaresystemperformance; butlittleresearchhasfocussedonthelinksbetweenthepublic’sgeneralattitudetothehealthcare sys-temanditsperceptionsandexpectationsofspecifichealthcare-relatedaspects.Usingdatafromthe fourthroundoftheEuropeanSocialSurvey2008/09,weexplorethecognitivedeterminantsofglobal evaluationsofthehealthcaresystemin12Easternand16WesternEuropeancountries.Wefindthat healthcareevaluationsfollowacoherentcognitivereasoning.Theyareassociatedwith(i)perceptionsof theperformanceofhealthcaresystems(i.e.efficiency,equalityoftreatment,healthoutcomes),(ii) expec-tationsofthegovernment’sroleinprovidinghealthcare,and(iii)reflectionsondemographicpressures (i.e.agingpopulations).Contrarytothegeneralassumptionthatnormativeexpectationsareresponsible fordifferencesinhealthcareevaluationsbetweenEasternandWesternEurope,ourresultssuggestthat regionaldifferencesarelargelyduetoamorenegativeperceptionoftheperformanceofhealthcare sys-temswithinEasternEurope.Toenhancethepublicopinionofhealthcare,policymakersshouldimprove theefficiencyofhealthcaresystemsandtakemeasurestoassureequalityinhealthtreatment.

1. Introduction

Thereis growing scholarly interest in the publicopinion of healthcareand intheassessmentandcomparison ofhealthcare systemperformance[1–11].Globalevaluationmeasuresof health-care,suchassatisfactionratings,areoftenusedtostudythepublic’s generalattitudetothehealthcaresystem[4,5].Theyaddressthe generalpublic–usersandnon-usersalike–andprovidea compos-iteassessmentofthehealthcaresystemmorebroadly[12,13].The morepositivelyindividualsevaluatehealthcareservicesin their country,thehigherthe(subjective)performanceofthesystem.

How the public rates healthcare services provides relevant information on the (mal-)functioning of the healthcare system and informshealth policy makers of areas requiring improve-ment[14–16]. Incontrasttoexpertopinionandotherobjective performance measures,subjectiveevaluations directlyechothe experiences and perceptions of the public[12] and provide an

∗ Correspondingauthorat:MaxPlanckInstituteforSocialLawandSocialPolicy, Amalienstrasse33,80799Munich,Germany.

E-mailaddresses:s.schneider@mpisoc.mpg.de(S.M.Schneider),

tpopic@iscsp.ulisboa.pt(T.Popic).

assessmentofhealthcaresystemsthatisoftendescribedasmore accurate,legitimate,andsensitive[15,17,18].Byholdingthe sys-temaccountablefortheprovisionofhealthcareservicesﬁnancedby thelargerpublic[1,19],suchevaluationscanserveasanimportant indicatoroftheapprovalofhealthcarereforms[12,20].Evaluations ofhealthcareservicesareofwiderpoliticalrelevanceaswell,as theycanaffectthestabilityofpoliticalsystems(e.g.trustin gov-ernment[21,22])andarerelatedtotheutilizationofhealthcare servicesandpopulationhealth[13,15,16].

The background variables that affect healthcare evaluations are important for health policy makers [17,23,24]. Scholars find the institutional design of healthcare systems [20,25–29], an individual’s demographic and socio-economic background [19,26,28,30–33], his/her experiences with healthcare services [29,34], and his/her confidence in findingaffordable and effec-tivecare[35]caninfluencehealthcareevaluations.Thesefindings largelysupporttheoreticalapproacheshighlightingstructural(i.e. interest-based), cultural (i.e. value-based) and contextual (e.g. institutional)factorsasimportantintheformationofwelfare atti-tudes[24,36–38].

Lesswellresearchedishowandtowhatdegreebeliefsabout speciﬁcdimensionsofhealthcareservicesandexpectationsofthe government’s roleinproviding themshape thepublic’sgeneral https://doi.org/10.1016/j.healthpol.2017.12.012

(2)

attitudetowardsthehealthcaresystem[13,23].Althoughhealth outcomes,healthcareefﬁciencyandequityinhealthcarehavebeen addressedasimportantdimensionsintheassessmentofthe‘actual’ or‘objective’ performance of healthcare systems[11], informa-tion onhow individuals perceivethese aspects and how their perceptionsshapethepublic’sgeneralevaluation ofthe health-caresystemismissing.Similarly,anddespiteexistingresearchon theexpectationsofthegovernment’sroleinprovidinghealthcare [25,26,39–43], expectations have often beenstudied separately fromglobalhealthcareevaluations,andlittleisknownabouttheir association.Webelievethatresearchontherelationshipbetween thepublic’sglobalevaluationsofhealthcareand itsperceptions andexpectationsofspeciﬁchealthcare-relatedaspectswillshed lightonthecognitivefactorsdrivinghealthcareevaluations.This, inturn,willhelppolicymakersidentifyprioritiesforactionamidst thedemographicpressuresofagingsocietiesandbudgetary con-straintsfacedbymanyEuropeancountries[35,44].

Cognitiveapproachestoattitudeformationsuggestindividuals formattitudesbasedonconsistentcognitivereasoning[45,46]. Atti-tudesareportrayedasafunctionofdifferentbeliefsaboutanobject [47]:thestrongerthebeliefthattheobjectisconnectedwith cer-tainpositiveattributes,themorepositivetheindividual’sattitude abouttheobject.Ifthis istrue,globalevaluationsof healthcare willbebasedonasetofbeliefsformedaboutspecificattributes ofthehealthcaresystem.Asstatedabove,healthcaresystems pro-motingpopulationhealth,efficiencyindelivery,andequalaccessto thoseinequalneedarepreferredbyhealthpolicyscholarsover sys-temslowonthesedimensions[11].Whethertheseattributesare alsoperceivedrelevantbytheindividualandshapehis/her eval-uationofhealthcaresystemsisanempiricalquestionwhichhas not(yet)beenexamined.Atthispoint,itremainsunclearwhether globalhealthcareevaluationsaremorepositiveifrespondentshave amorepositiveperceptionoftheperformanceofthevarious health-caredimensions,thatis,iftheyperceivefewersickpeopleinsociety, greaterefficiencyinhealthcaredelivery,andmoreequalityin health-caretreatment.

Also critical for attitude formation are expectations [48,49]; thesefunctionasreferentialstandardsagainstwhichindividuals comparetheirperceptionsofthestatusquo[16,50,51].Evaluations areadirectfunctionofthediscrepancybetween(i)expectationsof theobject’sattributesand(ii)theperceptionoftheobject’sactual attributes,the‘statusquo’[52,53].Lackofvalidindicatorsand mea-surementinstrumentsintheareaofhealthcarecanimpedeadirect comparisonofperceptionsandexpectations.Consequently, schol-arsyieldtotheassumptionthatexpectationsaresimplyinversely relatedtoevaluations:thehighertheexpectation ofhealthcare, thelowertheoverallevaluationofthehealthcaresystem[32,54]. Againstthebackdropofrecentreformsmovingtowardsthe privati-sationandmarketizationofhealthcare [55,56],globalhealthcare evaluationsareexpectedtobemorenegativewhenexpectationsof governmentinvolvementarehigher.

Further, individuals are unlikely to form global evaluations onhealthcarein a socialvacuumwithout consideringthe pub-licdiscourseonthedemographicpressuresofagingsocietiesand budgetary constraints[57]. Individualswho are awareof these pressuresmaybemorewillingtoexcuse–atleasttosomedegree –shortcomingsofhealthcaresystems.Therefore,globalhealthcare evaluationsareexpectedtobemorepositivewhenawarenessof demo-graphicpressuresonhealthcaresystems,inthiscase,theperceptions ofagingpopulationsasaburdenforhealthcaresystems,isgreater.

In theEuropeancontext, researchonhealthcare evaluations looksateitherWesternEuropean[20,25–27]orEasternEuropean countries[19]. Onlyrecently,andwiththeavailabilityoflarger datasets,havescholarsbeguntoexplorehealthcareevaluations inbothpartsofEurope[28].ThegeneralﬁndingisthatEastern Europeansevaluatehealthcaresystemslesspositivelythan

West-ern Europeans.It isunclear whetherthese regionaldifferences reflectperceptionsoftheactualperformanceofhealthcaresystems orexpectationsthatstemfromsocialisationprocessesin differ-entsocio-politicalcontexts.Despitefundamentalreformprocesses, EasternEuropeanhealthcaresystemsstillscoreloweronobjective performancemeasuresprovidedbyofficialregistries,e.g. financ-ingandprovisionofhealthcareservices[58–62].Recentresearch suggeststheseinstitutionaldifferencesatleastpartlyexplainthe generallylowerevaluationsofhealthcaresystemsinEastern Euro-pean countries[63]. This argument competes with scholarship claimingthattheformationofpreferencestructureswithin differ-entsocio-politicalcontextsexplainsregionaldifferencesinpublic expectationsand evaluations[64–68]. Termed‘legaciesof com-munism’,pastexperiencesandsocialisationprocesseswithinthe former communist regime are declared responsible for higher expectationsofthewelfarestateanditsprovisionofsocialservices inEasternEurope.Oldergenerationshavelivedandexperienced communismforalongerperiodoftime,arguablyexplainingwhy differences in welfare attitudes between Eastern and Western Europeansaremoredistinctamongolderbirthcohorts[66,68].

Theanalysisofexpectationsvis-a-visothercognitivefactorswill yieldnewinsightsontherelevanceofthesocio-politicalcontextof healthcareevaluations.Morespecifically,ifdifferencesin health-careevaluationsstemfrompreferencestructures evolvingfrom socio-politicalcontexts,expectationswillbethedominantcognitive componentexplainingEast-Westdifferencesinhealthcareevaluations. Further,differencesinhealthcareevaluationswill bemore distinct foroldergenerations.Butifdifferencesinhealthcareevaluations betweenEasternandWesternEuropeareduetotheperformance ofhealthcaresystems,perceptionsofhealthcare-relatedaspects(i.e. efficiency,equality intreatment, andhealth outcomes)will bethe dominantcognitivecomponents;wewillnotseesignificantvariation betweenbirthcohorts.

Theaimofthisempiricalstudyistwofold.Firstly,weexamine whetherhealthcareevaluationsfollowaconsistentcognitive rea-soning.Weseektounderstandtheextenttowhichevaluationsof healthcareservicesaredeterminedbythreecognitivefactors:(i) perceptionsoftheperformanceofthehealthcaresysteminterms ofefﬁciency,equality,andpopulationhealth;(ii)expectationsofthe government’sroleinprovidinghealthcare;and(iii)reﬂectionson demographicpressures,i.e.theburdenonhealthcareofanaging society.Secondly,weinvestigatetheextenttowhichthese cogni-tivecomponentsexplaindifferencesinthehealthcareevaluations ofEasternandWesternEuropeans.Theintentionistounderstand whetherEasternEuropeansare,onaverage,morecriticaloftheir healthcareservicesbecauseoftheirhigherexpectationsof govern-ment’srole inprovidinghealthcareorbecauseoftheirnegative perceptionof theactualperformance of thehealthcare system. Fig.1illustratestheresearchmodel.

2. Materialandmethods

2.1. Data

Theempiricalanalysesarebasedonthefourthroundofthe EuropeanSocialSurvey(ESS)from2008/09.TheESSisahigh qual-ity,cross-comparativedatasetthatprovidesbiannualinformation representativeoftheEuropeanpopulationlivinginprivate house-holdsaged15andabove.ThefourthroundoftheESSincludeda specialmoduleonwelfareattitudesandwasthereforeselectedfor theempiricalanalysis.Intotal,thesampleincludes43,460 individ-ualslivinginprivatehouseholdsin28Europeancountries(Eastern Europe:N=12;WesternEurope:N=16)forwhominformationon allvariableswasavailable.

(3)

Fig.1. ResearchModel.

2.2. Measures

2.2.1. Dependentvariable

Theevaluationof healthcareservicesisourmain dependent variable.Respondentswereaskedwhattheythinkoverallabout thestateofhealthcareservicesintheircountryonan11-pointscale rangingfrom0,extremelybad,to10,extremelygood.

2.2.2. Mediatorvariables

We measure perceptions of the performance of healthcare systems [69] with three indicators: the perceived efficiency of healthcareservices,theperceivedequality intreatment, andthe healthstatusofthepopulation.Respondentswereaskedhow effi-cient theythink theprovision of healthcare is intheir country on an 11-point scale, ranging from 0, extremely inefficient, to 10,extremelyefficient.Theywerealsoaskedwhethertheythink doctorsandnursesintheircountrygivespecialadvantagesto cer-tainpeopleordealwitheveryoneequallyusingan11-pointscale, rangingfrom0,givespecialadvantagestocertainpeople,to10, dealwitheveryoneequally.Informationonthestateofpopulation healthwasmeasuredbyaskingrespondentshowmanyofevery 100peopleofworkingageintheircountryarelong-termsickor disabled,usingalistof11categories.Werecodethisiteminto8 categories(groupingthelastfourcategoriesintoone),rangingfrom 0–4,5–9,10–14,15–19,20–24,25–29,30–34,>35personsoutof 100beinglong-termsickordisabled.

Expectations of the healthcare system are measured by the stateresponsibilitytoprovidehealthcareforcitizensandtocure illnesses.Respondentswereaskedhowmuchresponsibility gov-ernments shouldhave in ensuring adequate healthcare for the sick,usingan11-pointscale, rangingfrom0,shouldnotbethe government’s responsibilityat all,to10,shouldbe entirelythe government’sresponsibility.

Thecontextualpressuresonhealthcaresystemsaremeasured bytheperceivedburdenonhealthservicesbecauseofdemographic changes.Respondentswereaskedwhetherornottheythinkpeople over70areaburdenonthecountry’shealthservicesonan11-point scale,rangingfrom0,noburden,to10,agreatburden.

Thesequestionswerenotaskedconsecutively,butwere inte-gratedinvarioussubsectionsthroughouttheinterview.Questions appearedinthefollowingorder:thecoremoduleonpolitics (sec-tionB)includedthequestionontheoverallevaluationofhealthcare services(B29);questionsontheperceivedhealthstatusofthe pop-ulation(D8),expectations ofthegovernment’s responsibilityto providehealthcare (D16),theperceivedefﬁciencyof healthcare

services(D30),theperceivedequalityofhealthtreatment(D32), andtheperceivedburdenofolderpeopleonhealthservices(E12) wereaskedintherotatingmodulesonattitudestowardsthe wel-farestate(sectionD)andageism(sectionE).Fig.2givesanoverview of theaveragescores ofthe dependentand mediator variables acrossEuropeancountries.

2.2.3. Independentvariables–countrylevel

DifferencesbetweenEasternEuropeanandWesternEuropean countriesareofparticularinteresttothisstudy.EasternEuropean countriesarecharacterizedbytheircommunistpastandtheradical transformationsintheirhealthcaresystemssincethefallof com-munism[58,70,71].WiththeregionalexceptionofEastGermany, WesternEuropeancountriesdonotshareacommoninstitutional historyandlacktheexperienceofthepost-communist transfor-mationprocess.Weincludeadummyvariableatthecountrylevel togroupcountriesaccordingtotheirregionalandsocio-political history. Out of the28 countriesin thesample, 12 are consid-ered Eastern European, and 16 are Western European. Eastern EuropeancountriesincludeBulgaria,Croatia,theCzechRepublic, Estonia, Hungary, Latvia, Poland, Romania, the Russian Federa-tion,Slovenia,SlovakiaandUkraine.WesternEuropeancountries includeAustria,Belgium,Cyprus,Switzerland,Germany,Denmark, Spain,Finland,France,theUnitedKingdom,Greece,Ireland, the Netherlands,Norway,Portugal,andSweden.

2.2.4. Independentvariables–individuallevel

We control for demographic and socio-economic character-istics of the individual as these might affect the evaluation of healthservices.Therespondent’ssexandage(mean-centred)are standardcontrolvariables.Wetestfortheu-shapedrelationship betweenageand healthcare attitudesbyincluding thesquared termofage.Tocontrolforhealthneeds,weincludethe respon-dent’sself-reported health status measuredon a 5-pointscale, rangingfromverygoodtogood,fair,bad,andverybad(recoded as −2=very bad, −1=bad, 0=fair, 1=good, 2=very good). We introducesocio-economiccharacteristics,suchasyearsof educa-tion(mean-centred)andthecurrentstatusofemployment(paid work(ref.),unemployed,retired,otherstatus).Theactual house-holdincomeisnotincludedintheanalysisbecausecomparable incomeinformationismissingforthreecountries(Bulgaria, Slo-vakia,Cyprus).Instead,weincludeasubjectiveincomeindicator. Respondents were asked how theyfeel about their household incomeandwhethertheylivecomfortablyontheirpresentincome (ref.),copeontheirpresentincome,ﬁnditdifﬁcultontheirpresent

(4)

Fig.2.Evaluations,PerceptionsandExpectationstowardsHealthcareServicesacrossEurope.

income,orﬁnditverydifﬁcultontheirpresentincome.Tocontrol forotherhouseholdcharacteristics,weincludethehouseholdsize andwhetherchildrenarelivinginthehousehold.

2.3. Statistics

We apply multilevel modelling techniques to estimate the effectsofbothindividualandcountrylevelcharacteristics.Unlike conventional regression analysis,multilevel modelsaccount for thehierarchicalornesteddata structure,whereby observations atthelower (individual) level arenested in higherorder units (countries).Consideringthemultiplelevelsinthecomputation pro-cesstakesintoaccounttheinterdependencyofobservationswithin countries.Withanintraclasscorrelationof0.21forthemain depen-dentvariable(healthcareevaluation),theuseofmultilevelmodels

fortheanalysisishighlyrecommended.Randominterceptmodels allowforthevariationofinterceptsacrosscountries.Variationsin interceptscanbeexplained(i)bycountrylevelpredictorvariables (here:region),explainingthecontextualvariationintheoutcome variable,and(ii)byindividuallevelvariables,explainingthe com-positionalvariation.

Multilevel mediationmodelsoffertheopportunity tomodel complexrelationshipsandtoestimatedirectandindirect relation-shipsbetweenasetofvariableswithina multilevelframework. Weapplya 2-1-1multilevelmediationanalysis(MMA)[72],as ourindependentvariable,region(East/West),islocatedatlevel 2;perceptions,expectations,andevaluationsareindividuallevel characteristicsandlocatedatlevel1.Themediationismeasuredat thebetweenlevel,partitioningthevariancesoftheindividuallevel variablesintoabetweenandwithinlevelcomponent.The

(5)

media-Table1

TheEvaluationofHealthcareServicesinEasternandWesternEurope.

Model1 Model2 Model3

TotalSample EasternEurope WesternEurope

␤ SE ␤ SE ␤ SE Intercept 5.10*** .15 3.83*** .14 5.62*** .15 WithinLevel CognitiveComponents Efﬁciency .55*** .02 .53*** .03 .55*** .02 Equalityoftreatment .10*** .01 .11*** .02 .09*** .01 Prevalenceofsickness −.04*** .01 −.02 .02 −.04*** .01

Expectedgov.involvement −.03*** .01 −.02*** .01 −.04* .02

Burdenofolderpeople .01* .01 .01 .01 .02* .01

Controls

Gender(0=male) −.15*** .04 .04 .06 −.27*** .03

Age −.00* .00 −.01*** .00 −.00 .00

Age-Squared .00*** .00 .00** .00 .00*** .00

Subj.healthstatus .13*** .02 .19*** .02 .10*** .02

Education(inyears) −.02** .01 −.04*** .01 −.01 .01

Subj.income(0=livingcomfortably)

Copingonpresentincome −.11*** .03 −.09 .08 −.11*** .03

Difﬁcultonpresentincome −.22*** .04 −.25** .09 −.21*** .04

Verydifﬁcultonpresentincome −.29*** .06 −.42*** .08 −.14 .08

Labormarketposition(0=paidwork)

Unemployed .04 .04 .09* .04 −.00 .07

Retired .10 .05 .19** .07 .02 .07

Otheremploymentstatus .08* .03 .10* .04 .06 .05

NumberofpersonsinHH .02 .01 .02 .01 .02 .02

KidsinHH(0=nokids) .09* .04 .08 .06 .09 .06

BetweenLevel

EasternEuropean(Ref.WesternE.) −.48** .17 – – – –

Variance(R2₎₋_within _3.20*** _(.42) _3.59*** _(.39) _2.93*** _(.38)

Variance(R2₎₋_between _.22** _(.21) _.07*** _(.00) _.32** _(.00)

Note:Tablepresentsresultsofmultilevelregressionanalysis,unstandardized␤coefﬁcientsandrobuststandarderrors(SE);standardweightsapplied;Source:ESS-4;

N=43460/28;EasternEuropeSample:N=16671/12;WesternEuropeanSample:N=26789/16;*_p_<_0.05,**_p_<_0.01,***_p_<_0.001

tioneffectistheproductof(i)theeffectofregion(East/West)onthe

mediator(perceptions/expectations)and(ii)thesumoftheeffect

ofthemediatoronhealthsystemevaluationatthemacroleveland

themeanoftherandomslopeofthesameeffectatthemicrolevel.

Forouranalyses,weuseMplus,version7[73].Maximum

like-lihoodwithrobuststandarderrorsisusedasanestimator.Atall stages,wecontrolforindividualcharacteristics,i.e.demographic and socio-economic characteristics of the individual, including healthneeds. We meancentreallmetric independentvariables beforeincludingthemintheanalysis.Toassurerepresentative esti-mationsforthecountrypopulations,weapplypost-stratiﬁcation weightsfollowingtherecommendationsoftheESS.

3. Results

3.1. Cognitivedeterminantsofhealthcareevaluationsacross Europe

Multilevelanalysesshowhealthcareevaluationsfollowa coher-entcognitivereasoning(Table1,Model1).Perceptionsofefﬁciency aremoststronglyassociatedwithhealthcareevaluations(␤=0.55, SE=0.02),followedbytheperceivedequalityinhealthtreatment (␤=0.10,SE=0.01).Thisimpliesthatthehighertheperceived efﬁ-ciencyofthehealthcaresystemandequalityofhealthtreatment, themore positive theoverall evaluation of healthcare services. The perceived prevalence of sickness/long term illness among those of working age (␤=−0.04,SE=0.01) and expectations of governmentinvolvementin healthcare(␤=−0.03,SE=0.01)are negativelyrelatedtohealthcareevaluations;inotherwords,the highertheperceivedlevelofsicknessandthehigherthe expecta-tionofgovernmentinvolvement,thelowertheoverallevaluation ofhealthcareservices.Perceptionsofdemographicpressures pos-itivelyaffecthealthcareevaluationssuggesting individualsform

theirevaluationsbyalsoreﬂectingoncontextualfactorsthatput healthcaresystemsunderadditionalpressure,e.g.agingsocieties. However,theeffectsizesuggeststhesereﬂectionsareofonlyminor importance(␤=0.01,SE=0.01).

TheresultsarelargelycomparablebetweenEasternand West-ernEuropeancountries(Table1,Model2and3).However,regional differencesexist;perceivedhealth outcomesand considerations of demographic pressures are not signiﬁcantly associated with healthcareevaluationsinEasternEurope.

Across Europe,cognitivedeterminants,together with demo-graphicandsocio-economiccharacteristics,explain42.1%ofthe varianceinhealthcareevaluationsattheindividuallevel;39.2%in EasternEuropeand38.3%inWesternEurope.

Countryspecificregressionanalysesrevealhealthcare evalua-tionsarestronglyandconsistentlyassociatedwiththeperceived efficiencyofthehealthcaresysteminall28Europeancountries (Appendix, Tables A1 and A2). Similarly robust is the associa-tion withthe perceived equalityof health treatment, with the exceptionofthreecountries−Denmark,Hungary,andUkraine– wherethereisnosignificantassociation.Effectsofthethreeother cognitivecomponents–theperceptionofpopulationhealth,the perceivedburdenofolderpeople,andexpectationsofgovernment involvement− varymorestrongly betweencountries. Paradox-ically, expectationsof governmentinvolvement showa reverse effect inAustria and Belgium, suggesting a positive association betweenhealthcareevaluationsandsupportforgovernment inter-vention.

3.2. Regionaldifferencesinhealthcareevaluations:anEast-West comparison

Regional differencesinhealthcare evaluationsbetween East-ernandWesternEuropearesalientandsolid.Aftercontrollingfor

(6)

Fig.3.RegionalDifferencesinHealthcareEvaluations–DirectandIndirectEffects.

Note:Figurepresentsresultsofmultilevelmediationanalyses;unstandardized␤coefﬁcients;robuststandarderrors(SE)inparenthesis;allanalysescontrolfordemographic andsocio-economiccharacteristics;standardweightsapplied;Source:ESS-4;N=43460/28;*_p_<_0.05,**_p_<_0.01,***_p_<_0.001.

demographicandsocio-economiccharacteristics,weﬁndEastern Europeansevaluatehealth services1.38(SE=0.33)pointslower thanWesternEuropeans.Theinclusionofcognitivedeterminants intheanalysisleadstoadecreaseineffectsize,butdifferences remainsigniﬁcant(␤=−0.48,SE=0.17).Intotal, regionexplains 20.6%of thebetween variance inhealthcare evaluations atthe macro/countrylevel(Table1,Model1).

Fig.3presentstheresultsofthemultilevelmediationanalysis which accounts for regional differences in the cognitive com-ponents of healthcare evaluations and tests whether cognitive componentsmediatetherelationshipbetweenregion(East/West) andhealthcareevaluations.Allanalyseswererunseparatelyfor eachmediator,witharandomslopespecificationforeachmediator andcontrolfordemographicandsocio-economiccharacteristics attheindividual-level.Twoofthethreesubjectiveperformance indicators–perceptionsofefficiencyandequalityintreatment– fullyexplainregionaldifferencesinhealthcareevaluationsbetween EasternandWesternEurope.EasternEuropeansperceivehealth systemsasless efficientandmoreunequal thanWestern Euro-peans, and this results in lower overall healthcare evaluations. EasternEuropeansalsoperceiveahigherprevalenceofsickness amongtheworkingagepopulationthanWesternEuropeans,which partlyaccountsforregionaldifferences.Interestingly,wefindno empiricalsupportthatregionaldifferencesinhealthcare evalua-tionsareduetoexpectationsofgovernmentinvolvement,nordo wefindtheperceivedburdenofolderpeopleforthehealthsystem explainsregionaldifferences.

Totestwhetherourresultsonregionaldifferencesarerobust for differentagecohortsand, thus,differentexperiences ofthe socio-politicalcontext(i.e.thecommunistregime),were-runthe multilevel mediationanalysesseparatelyfor thefollowing sub-groups:(i)individualsbornbefore1951whogrewupandlived inthesocialistsystemforthemajorityoftheirlife;(ii)individuals bornbetween1951and1974whowereinfulladulthoodbythe timeofthefallofcommunism;(iii)andindividualsbornafter1974 who,atbest,experiencedcommunismfor14years[66].Results largelysupportourpreviousﬁndingsontheassociationbetween healthcareevaluations,cognitivedeterminants,andEast/West dif-ferencesandindicatethatourresultsarerobustacrossagecohorts (Table2).Thissuggeststhatthelengthoftimeindividualshavebeen socializedindifferentpoliticalsystemsisnotanessential explana-tionofEast-Westdifferencesinhealthcareevaluationsandtheir cognitivedeterminants.

4. Discussion

Thisstudyexploredthecognitivedeterminantsofhealthcare evaluationsin12Easternand16WesternEuropeancountries.We studiedthelinksbetweenglobalevaluationsofhealthcaresystems and(i)perceptionsoftheperformanceofhealthcaresystems(i.e. efﬁciency,inequalityintreatment,health outcomes),(ii) expec-tationsofthegovernment’sroleinprovidinghealthcare,and(iii) reﬂectionsondemographicfactors(i.e.agingsociety)thatput cur-renthealthsystemsunderpressure.Inasecondstep,weexplored whetherthemorenegativeevaluationsofhealthcaresystemsin

(7)

Table2

DirectandIndirectEffectsofRegion(East/West)onHealthcareEvaluation–CohortSpeciﬁcAnalysis.

Efﬁciency EqualityinHealth

Treatment Prevalenceof Sickness ExpectationGov. Involvement BurdenofOlder People ␤ SE ␤ SE ␤ SE ␤ SE ␤ SE Cohort1:>1950 Directeffect .24 .22 −.60 .34 −1.21** .36 −.82* .33 −1.67*** .32

Indirecteffect(viamediator) −1.96*** .34 −.91*** .24 −.41* .20 −.23 .21 .13 .14

Cohort2:1951–1975

Directeffect .05 .19 −.38 .38 −.94* .42 −.94** .28 −1.58*** .36

Indirecteffect(viamediator) −1.54*** .37 −.87** .27 −.37* .16 −.05 .19 .04 .09

Cohort3:<1975

Directeffect .23 .21 .19 .37 −1.05** .34 −.98** .31 −1.45*** .34

Indirecteffect(viamediator) −1.65*** .34 −.98** .32 −.26 .14 −.16 .24 .04 .13

Note:Tablepresentsresultsofmultilevelmediationanalysis,unstandardized␤coefﬁcientsandrobuststandarderrors(SE);allanalysescontrolfordemographicand

socio-economiccharacteristics;standardweightsapplied.Source:ESS-4,Cohort1:N=12984/28;Cohort2:N=19290/28;Cohort3:N=11186/28;*_p_<_0.05,**_p_<_0.01,***

p<0.001.

EasternthanWesternEuropeareduetothenegativeperceptionof

theactualperformanceofhealthcaresystemsinEasternEuropeor

tohighernormativeexpectationsofthegovernment’sinvolvement.

Inlinewithcognitiveapproachestoattitudeformation[45–47],

ourfindingssuggestindividualsholdconsistentviewsonhealthcare systems.Theyalsosuggestthatthepublic’sviewonhealthcare islargelycongruentwithhealth policyresearchthatrecognizes healthcaresystemsas‘highperforming’iftheypromotepopulation health,increaseefficiencyinhealthcaredelivery,andensureequal accesstothoseinequalneed[11].Wefindindividualswhoevaluate healthcareservicespositivelyarealsomorelikelytoperceivethe healthcaresystemasmoreefficientandmoreequalintreatment andtoconsideralowernumberofworkingagepeopletobe long-termsickordisabled.Theperceivedefficiencyofhealthcaresystems isthestrongestcognitivedeterminantofhealthcareevaluations, followedbyequality inhealth treatment;theirassociationwith healthcareevaluationsisrobustandstrongacrossmostEuropean countries.Atthesametime,andinlinewithresearchon expec-tations[16,48–53],we findindividualswho evaluatehealthcare servicespositivelyarealsomorelikelytohavelowerexpectations ofgovernmentresponsibilityinhealthcare.Interestingly,however, expectationsofthegovernment’sroleinprovidinghealthcareare ofonlyminorimportanceinhealthcareevaluations.Thismightbe relatedtothefactthatEuropeansshowa highsupportfor gov-ernment responsibilityin healthcare, withpersonal interestsof remarkablylittleimportance[41]andminorvariationsobserved betweencountries[20,26,65].Further,ourresultsrevealthat indi-vidualswhoevaluatethesystemmorepositivelyarealsomore likelytoseetheagingsocietyasaburdenforcurrenthealthcare systems.Thissuggeststhatevaluationsofhealthcarearenot dis-sociatedfromlargersocietaltrends.Perceptionsofdemographic pressuresonhealthcaresystems causedbyanagingsocietyare acknowledgedandconsideredbyindividualsintheiroverall eval-uationofhealthcaresystems.Ifolderindividualsareperceivedas aburden,individualsadjusttheirevaluationsaccordinglyandare lesscriticalofthecurrentstateofhealthcare.

Withregardtoregionaldifferencesin healthcareevaluations betweenEasternand WesternEurope,ourresultsindicate that EasternEuropeansperceivetheirhealthcaresystemsasless efﬁ-cientandmoreunequal.Atthesametime,theyperceiveahigher prevalenceofsickpeopleintheirsociety.Whileallthree dimen-sionsofperformanceperceptionmediatetherelationshipbetween regionandhealthcareevaluations,perceptions ofefﬁciencyand inequalityoftreatmentfullyexplainregionaldifferences.Put oth-erwise,EasternEuropeanshavenegativeperceptionsoftheactual performanceofhealthcaresystemsintheircountries;these percep-tionsexplainwhyEasternEuropeansevaluatehealthcaresystems

less positively than Western Europeans. Thus, public’s percep-tionsareconsistentwithresearchthatshowsEasternEuropean healthcaresystemsstillscoreloweronobjectiveperformance mea-sures,e.g.financingandprovisionofhealthcareservices[58–62]. Theseinstitutionaldifferencesmatterforthepublic’sevaluationof healthcaresystems[63].Despitetheradical,oftenmarket-oriented reformsimplementedduringtransition,policyoutcomeshavenot always metthe concerns of thebroader public;nor have they helpedtoimprovetheefficiencyandequalityofpost-communist healthcaresystems[74].Corruption,forexample,commonunder communism,remainsamajorconcerninmanyEasternEuropean countries[75–77]andmayexplainwhyEasternEuropeansoften perceiveinefficienciesandinequalitiesinhealthtreatment. Erad-icatinginformalpaymentsandintroducingmeasurestoimprove efficiency andequalityin accessing and obtainingmedicalcare mayresultinmorepositivehealthcareevaluations,particularlyin post-communistcountries.Interestingly,andincontrasttoprior researchonEast-Westdifferencesinwelfareattitudes[64,66–68], wefindnoempiricalsupportfordifferencesinexpectationsof gov-ernmentinvolvementasanexplanationofEast-Westdifferences in healthcare evaluations.Cohortspecific analysesyield similar results. Asstatedabove,governmentinvolvement inhealthcare enjoysstrongandfairlysimilarsupportacrossEasternand West-ernEuropeancountries[65],remindinghealthpolicymakersofthe needtotakeactionagainstmarketforcesandpressurestoprivatize. Thesefindingsraisealargerquestion.Arethepublic’sgeneral attitudes to the healthcare system related to objective charac-teristicsofEuropeanhealthcaresystems[78]?Inlinewithprior research[20,26],ourfindingssuggestEuropeansfavourastrong stateinvolvementinhealthcare;whilethegeneralevaluationand perceptions of specifichealthcare-related aspects vary strongly betweencountriesandhealthcaresystems.Forexample,in con-tinentalEurope,SocialHealthInsurance(SHI)systemsfavouringa contribution-basedmodelofhealthcarefinancingindependentof thegeneralgovernmentbudget[79]receivethehighest health-careratings.Belgium,Switzerland,Austria,theNetherlands,and FrancerankamongthesixtopratedhealthcaresystemsinEurope. SHI systems are also characterizedby comparatively high effi-ciencyratings,lowerexpectationsofgovernment’sinvolvementin healthcare,andacomparativelyhighawarenessofdemographic pressures.AnexceptiontothisgeneralpictureisGermany.Withits distincthealthcaresystem,increasinghealthinsurance contribu-tions,andgrowingout-of-pocketpayments[80],Germanyisonly ratedmoderatelywellbythepublic.Unlikeothercountriesof conti-nentalEuropeanddespiteitsawarenessofdemographicpressures, theGermanpublichashighexpectationsofstateinvolvementin

(8)

healthcareandperceivestheefﬁciencyofitshealthcaresystemand theequalityinhealthtreatmentratherpoorly.

Incontrast, NationalHealthcare Service(NHS) systems ofthe NordicandAnglo-Saxoncountries[81,82]arerecognizedfortheir egalitariancharacter.Ourfindingssupportthisnotion;Denmark, GreatBritain,Finland,Norway,Sweden,andIrelandrankamongthe tenEuropeancountrieswiththehighestperceptionsofequalityin healthtreatment.However,NHScountriesoftenscorelowerthan SHIcountriesinperceivedefficiencyandperformancein promot-ingpopulationhealth.Twoexceptionsexistandarenoteworthy. TheFinnishhealthcaresystem,withitscomparativelyhighquality ofhealthcare servicesat reasonablecosts[83],is rated particu-larlyhigh, anditsefficiencyand equalityin healthtreatmentis acknowledgedbytheFinnishpublic.Irelandisdifferent,ranking belowotherNordicandAnglo-SaxonNHScountriesinterms of efficiencyandequality,aswellasgeneralhealthcareratings;the distinctfeaturesoftheIrishhealthcaresystem,amixtureofa uni-versaltax-financedpublichealthserviceandafeebasedprivate system[84–86],mayexplainthisdifference.Further,high fertil-ityratesinIrelandmayexplainwhytheagingofsocietyis not perceivedasapressingtopicfortheIrishhealthcaresystem.

Interestingly,theratingsofSouthern EuropeanNHSsystems arelowerthantheratingsofNordic andAnglo-SaxonNHS sys-tems.Thismaybeexplainedbythelowpublicinvestments,weak administrativecapacitiesandlessdevelopedinfrastructures[87]. ThehealthcaresystemsofPortugalandGreece,countrieshithard bytheeconomiccrisis,areperceivedasparticularlyinefﬁcientand unequalinhealthtreatmentandareratedlowerthansomeofthe EasternEuropeancountries.

Overall,ourresearchfindingscontributetothegeneraldebate onhealthcareattitudes.Scholarsoftenusestructural,cultural,and institutionalfactorstoexplainhealthcareattitudes[19,20,24–33]. Theyexplore differentcomponentsof healthcare attitudes sep-arately [6,8,20,26] or use composite measures, combining, for example,perceptionsofefficiencywiththeoverallevaluationof healthcare services [28]. In contrast, we focused onthe cogni-tivedeterminantsofhealthcareevaluations.Ourfindingssuggest healthcareevaluationsaredistinctandcontextspecific.Individuals areabletoprovidemeaningfulanswerstosurveyquestionsrelated tovariousaspectsofhealthcare.Theirviewsseemtobecongruent andcognitivelyconsistent.

4.1. Studyrecommendations

Our findings suggest that in order to improve the public’s generalattitudetothehealthcaresystem,policymakersshould payparticularattention totheimprovement of healthcare sys-temefficiencywithoutsacrificingequality[88].Costcontainment strategies that reduce the costsfor unnecessary administrative tasksandinternaloperations[27]andcontrolthepayment mech-anisms of healthcare providers are therefore preferable over strategiesthat shifthealthcare financing ontousers and poten-tiallyincreaseinequityinaccessingmedicalcare[3,30,89–92].At thesametime,strategiesthatdetect,eradicate,andprevent infor-malpaymentswithinthehealthsector,whichputafinancialstrain ontheusers,mustberadicaltoreducetheunequaltreatmentof patients[75].Equitycanbeattainedonlyifindividualsaretreated accordingtotheirhealthneedsirrespectiveoftheirdemographicor socio-economiccharacteristics[90,93].Atthesametime,thehigh supportforthegovernment’sroleinhealthcareacrossEuropean countriessuggeststhatpolicymakersshouldrethinkstrategiesof privatisationandmarketizationwhichdiminishthestate’srolein thefinancinganddeliveryofmedicalcare.

4.2. Studylimitations

Thisstudyisnotwithoutlimitations.Wewerenotableto empir-icallytestintra-individualchangesintheevaluationofhealthcare servicesanditscognitivecomponentsacrosstime,asourdataare cross-sectional.Anyassumptionsoncausalityremainspeculative and require theinclusion of relevant indicators in longitudinal surveystudies.Nordo thedataallowus todeterminethetype of servicesindividuals considerin theirevaluations and to dif-ferentiate between the availability and quality of primary and secondaryhealthcareservicesthatmayhavedifferentimplications forhealthcareevaluations[94,95].Atthesametime,ourresultsdo notaccountforindividualexperienceswithhealthcareservicesor respondents’insurancestatus(publicorprivate),somethinglikely toinﬂuencetheirjudgmentofhealthcareservices[34].

Becauseofthelimitationsofavailablecognitiveindicators,we werealsonotabletostudytheinteractionbetweenexpectations andperceptionsortoquantifythediscrepancybetweenthetwo. Perceptiveandnormativeindicatorsdifferinourstudy,depending onthehealthcaredimensionstheyrelateto.Indicatorsof percep-tionstargetthe efﬁciencyof healthcare systems,theinequality of health treatment, and theprevalenceof sickness amongthe working age population, while the indicatoron the normative component is directed at expectations of government involve-mentinhealthcaremoregenerally.Theimplementationofsuitable indicators in large scale populationsurveys is therefore highly recommendable.Itwillallowscholarstoempiricallyexplore differ-encesinperceptionsandexpectationofhealthcarespeciﬁcaspects andtocalculate itsdiscrepancywithpossibleconsequences for globalevaluationsofhealthcare.

4.3. Implicationsforfutureresearch

Moreresearchiswarranted,particularlyoncross-country vari-ationsinthedegreetowhichcognitivecomponentsarerelated toglobalhealthcareevaluations.Our ﬁndingsshowthat notall cognitivefactorsareequallyimportantforhealthcareevaluations acrossEuropeancountries.Previousresearchsuggeststhis varia-tionisrelatedtosystematicdifferencesintheinstitutionaldesign ofhealthcaresystems[35].Researchinthisareawillhelphealth policymakersdevelopcountryspeciﬁcstrategiestoaddresspublic concernsabouthealthcaresystems.

Countryspecificanalysesshouldbecomplementedby cross-comparativeresearchontheinfluenceoftheinstitutionalcontext on people’s perceptions and expectations of healthcare. Even thoughexisting researchhasshown institutionalcharacteristics ofthehealthcaresystemmatterforglobalhealthcareevaluations [20,25–29],littleisknownaboutthecontextsinwhichperceptions andexpectationsofcertaindimensionsofthehealthcaresystem areformed.Moreresearchisneededonhowspecificinstitutional arrangements(e.g.accessregulations, availability ofhealth ser-vices)shapeperceptionsandexperiencesofhealthcare services, asthismayexplain differencesinhealthcare evaluationsacross Europeancountries.Thiswillprovidemoredetailedpolicyadvice onhowtheactualinstitutionaldesignofhealthcaresystemsand system-specifictreatmentproceduresinfluenceperceptionsof effi-ciencyandequalityamongthebroaderpublic,aspectswhichwill improvethepublic’sglobalevaluationofhealthcareservices.

Equallyimportantissubgroup-specificanalysisandthe identi-ficationofvulnerablegroups.Previousresearchshowsthatpeople supportwelfarearrangementsfromwhich theyexpectto bene-fit[96]orarrangementscomplyingwiththeirideologicalbeliefs aboutfairnessordeservingness[38].Giventheparticularitiesof thevarioushealthcareinstitutionsandaccessregulations,people ofdifferentsocio-economicandideologicalbackgroundsmaynot onlyperceivetheperformanceofhealthcaresystemsdifferently

(9)

butalsoattributedifferentimportancetothevarious subcompo-nents.Thisrequiresfurtherresearch.

5. Conclusion

We conclude that healthcare evaluations follow a coherent cognitivereasoning.Individualsbasetheirevaluationsontheir per-ceptionsoftheperformanceofthehealthcaresystem(i.e.efficiency, equality,healthoutcomes),theirexpectationsofthegovernment’s role inproviding healthcare, and theirunderstandingof demo-graphicpressures.Whileallfactorsseemrelevant,theperceptionof theefficiencyofthesystemandequalityofhealthtreatmentarethe mostimportant components influencing healthcare evaluations acrossEuropeancountries.Analysesofthecognitivedeterminants ofhealthcareevaluationshelpscholarsandhealthpolicymakers understand why Eastern and Western Europeans come to dif-ferent conclusions. EasternEuropeans aremore critical as they perceivetheirsystemsaslessefficientandmoreunequalinterms oftreatment.Toimprovethepublic’sgeneralattitudetothe health-caresystem,policymakersshouldpayparticularattentiontothe improvementofhealthcaresystemefficiencywithoutsacrificing equality.

Conﬂictofinterest

The authors certify that they have NO affiliations with or involvementinanyorganizationorentitywithanyfinancialor non-financialinterestinthesubjectmatterormaterialsdiscussedinthis manuscript.

Funding

Thisresearch is partof the NORFACEWelfare StateFutures funded research project ‘The Paradox of Health State Futures’ (HEALTHDOX)(ECERA-NetPlusfunding,grantagreementnumber 618106,FileNumber462-14-070).

Acknowledgements

Wesincerelythanktheanonymousreviewersfortheirvaluable commentsonearlierdraftsofthisarticle.Forallstatementsoffact, dataanalysesandinterpretationofresultstheauthorsalonebear responsibility.

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttps://doi.org/10.1016/j.healthpol.2017.12. 012.

References

[1]SmithPC,MossialosE,PapanicolasI,LeathermanS.Performance Measure-mentforHealthSystemImprovement:Experiences,ChallengesandProspects. Cambridge:CambridgeUniversityPress;2009.

[2]RobertsMJ,HsiaoW,BermanP,ReichMR.GettingHealthReformRight.AGuide toImprovingPerformanceandEquity.NewYork:OxfordUniversityPress; 2004.

[3]WorldHealthOrganization.TheWorldHealthReport.Geneva:WHO;2000.

[4]BusseR.Understandingsatisfaction,responsivenessandexperiencewiththe healthsystem.In:PapanicolasI,SmithPC,editors.Healthsystemperformance comparison.Anagendaforpolicy,informationandresearch.NewYork:Open UniversityPress;2013.p.255–79.

[5]BusseR,ValentineN,LessofS,PrasadA,vanGinnekenE.Beingresponsiveto citizens’expectations:theroleofhealthservicesinresponsivenessand satis-faction.In:FiguerasJ,McKeeM,editors.HealthSystems:Health,wealthand societalwell-being.Assessingthecaseforinvestinginhealthsystems.New York:OpenUniversityPress;2012.p.175–208.

[6]BlendonRJ,BensonJ,DonelanK,LeitmanR,TaylorH,KoeckC,etal.Who hasthebesthealthcaresystem?Asecondlook.HealthAffairs(Millwood) 1995;14:220–30.

[7]BlendonRJ,LeitmanR,MorrisonI,DonelanK.Satisfactionwithhealthsystems intennations.HealthAffairs(Millwood)1990;9:185–92.

[8]BlendonRJ,SchoenC,DesRochesC,OsbornR,ZapertK.Commonconcerns amiddiversesystems:healthcareexperiencesinﬁvecountries.HealthAffairs (Millwood)2003;22:106–21.

[9]BlendonRJ,KimM,BensonJM.ThepublicversustheWorldHealthOrganization onhealthsystemperformance.HealthAffairs(Millwood)2001;20:10–20.

[10]DonelanK,BlendonRJ,SchoenC,DavisK,BinnsK.Thecostofhealth sys-temchange: public discontent inﬁve nations. HealthAffairs(Millwood) 1999;18:206–16.

[11]PapanicolasI,CylusJ. Comparisonofhealthcare system performance.In: KuhlmannE,BlankRH,BourgeaultL,WendtC,editors.ThePalgrave Inter-nationalHandbookofHealthcarePolicyandGovernance.NewYork:Palgrave Macmillan;2015.p.116–32.

[12]BlendonRJ,BensonJM.Americans’viewsonhealthpolicy:aﬁfty-yearhistorical perspective.HealthAffairs(Millwood)2001;20:33–46.

[13]HudakPL,WrightJG.Thecharacteristicsofpatientsatisfactionmeasures.Spine 2000;25:3167–77.

[14]PapanicolasI,SmithPC.HealthSystemPerformanceComparison.AnAgenda forPolicy,InformationandResearch.NewYork:OpenUniversityPress;2013.

[15]FitzpatrickR.Surveysofpatientsatisfaction:importantgeneralconsiderations. BritishMedicalJournal1991;302:887–9.

[16]CrowR,GageH,HampsonS,HartJ,KimberA,StoreyL,etal.Themeasurementof satisfactionwithhealthcare:implicationsforpracticefromasystematicreview oftheliterature.HealthTechnologyAssessment(Rockv)2002:6.

[17]ParkK,ParkJ,KwonYD,KangY,NohJW.Publicsatisfactionwiththehealthcare systemperformanceinSouthKorea:universalhealthcaresystem.HealthPolicy (NewYork)2016;120:621–9.

[18]AharonyL,StrasserS.Patientsatisfaction:Whatweknowandwhatwestill needtoexplore.MedicalCareReview1993:49–79.

[19]FootmanK,RobertsB,MillsA,RichardsonE,McKeeM.Publicsatisfactionasa measureofhealthsystemperformance:astudyofninecountriesintheformer SovietUnion.HealthPolicy(NewYork)2013;112:62–9.

[20]MossialosE.Citizens’viewsonhealthcaresystemsinthe15memberstatesof theEuropeanUnion.HealthEconomics1997;1166:109–16.

[21]KumlinS.ThePersonalandthePolitical.HowPersonalWelfareState Expe-riencesAffectPoliticalTrustandIdeology.NewYork:PalgraveMacmillan; 2004.

[22]KumlinS.Overloadedorundermined:europeanwelfarestatesintheface ofperformancedissatisfaction.In:SvallforsS,editor.Polit.Sociol.Welf.Stat. Stanford,CA:StanfordUniversityPress;2007.

[23]JudgeK,SolomonM.PublicopinionandtheNationalHealthService: pat-terns and perspectives in consumer satisfaction.Journal ofSocial Policy 1993;22:299–327.

[24]WendtC,MischkeM,PfeiferM.WelfareStatesandPublicOpinion.Perceptions ofHealthareSystems,FamilyPolicyandBeneﬁtsfortheUnemployedandPoor inEurope.Cheltenham:EdwardElgar;2011.

[25]KohlJ,WendtC.Satisfactionwithhealthcaresystems.acomparisonofEU countries.In:GlatzerW,vonBelowS,StoffregenM,editors.Challengesfor QualityofLifeintheContemporaryWorld.Dordrecht:KluwerAcademic;2004. p.311–31.

[26]WendtC,KohlJ,MischkeM,PfeiferM.HowdoEuropeansperceivetheir health-caresystempatternsofsatisfactionandpreferenceforstateinvolvementinthe ﬁeldofhealthcare.EuropeanSociologicalReview2010;26:177–92.

[27]KotzianP.Determinantsofsatisfactionwithhealthcaresystem.OpenJournal ofPoliticalScience2009;2:47–58.

[28]MissinneS,MeulemanB,BrackeP.ThepopularlegitimacyofEuropean health-caresystems:amultilevelanalysisof24countries.JournalofEuropeanSocial Policy2013;23:231–47.

[29]BorisovaLV,MartinussenPE,RydlandHT,StornesP,EikemoTA.Public eval-uationofhealthservicesacross21Europeancountries:theroleofculture. ScandinavianJournalofPublicHealth2017;45:123–39.

[30]NikoloskiZ,MossialosE.Corruption,inequalityandpopulationperceptionof healthcarequalityinEurope.BMCHealthServicesResearch2013;13:472.

[31]HavilandMG,MoralesLS,DialTH,PincusHA.Race/ethnicity,socioeconomic status,andsatisfactionwithhealthcare.AmericanJournalofMedicalQuality 2005;20:195–203.

[32]AliFMH,NikoloskiZ,RekaH.Satisfactionandresponsivenesswithhealth-care servicesinQatar-evidencefromasurvey.HealthPolicy2015;119:1499–505.

[33]KimMK,BlendonRJ,BensonJM.Whatisdrivingpeople’sdissatisfactionwith theirownhealthcarein17LatinAmericancountries?HealthExpectations 2013;16:155–63.

[34]BleichSN,OzaltinE,MurrayCJL.Howdoessatisfactionwiththehealth-care systemrelatetopatientexperience?BulletinoftheWorldHealthOrganization 2009;87:271–8.

[35]PapanicolasI,CylusJ,SmithPC.Ananalysisofsurveydatafromelevencountries ﬁndsthatsatisfactionwithhealthsystemperformancemeansmanythings. HealthAffairs(Millwood)2013;32:734–42.

[36]vanOorschotW,OpieklaM,Pfau-EfﬁngerB.CultureandWelfareState:Values andSocialPolicy.In:ComparativePerspective.Cheltenham,UK:EdwardElgar; 2008.

[37]SvallforsS.ThePoliticalSociologyoftheWelfareState:Institutions,Social Cleavages,andOrientations.Stanford,CA:StanfordUniversityPress;2007.

(10)

[38]VanOorschotW.Whoshouldgetwhat,andwhy?Ondeservingness crite-riaandtheconditionalityofsolidarityamongthepublic.Politics&Policy 2000;28:33–48.

[39]KikuzawaS,OlafsdottirS,PescosolidoBA.Similarpressures,differentcontexts: publicattitudestowardgovernmentinterventionforhealthcarein21nations. JournalofHealthandSocialBehavior2008;49:385–99.

[40]GeversJ,GelissenJ,ArtsW,PublicMR.Publichealthcareinthebalance: exploringpopularsupportforhealthcaresystemsintheEuropeanUnion. InternationalJournalofSocialWelfare2000;9:301–21.

[41]JensenC,PetersenMB.Thedeservingnessheuristicandthepoliticsofhealth care.AmericanJournalofPoliticalScience2017;61:68–83.

[42]JensenC,NaumannE.Increasingpressuresandsupportforpublichealthcare inEurope.HealthPolicy(NewYork)2016;120:698–705.

[43]WendtC,MischkeM,PfeiferM,ReiblingN.Conﬁdenceinreceivingmedicalcare whenseriouslyill:aseven-countrycomparisonoftheimpactofcostbarriers. HealExpect2011;15:212–24.

[44]BlendonRJ,HuntK,BensonJM,FleischfresserC,TrendsBuhrT. Understand-ingtheAmericanpublic’shealthpriorities:a2006perspective.HealthAffairs (Millwood)2006;25:w508–15.

[45]FishbeinM,AjzenI.PredictingandChangingBehavior.TheReasonedAction Approach.NewYork:PsychologyPress;2011.

[46]AjzenI.Operationofattitudes.AnnualReviewofSociology2001;52:27–58.

[47]AjzenI,FishbeinM.Attitudesandtheattitude-Behaviorrelation:reasonedand automaticprocesses.EuropeanReviewofSocialPsychology2000;11:1–33.

[48]ThompsonAGH,SunolR.Expectationsasdeterminantsofpatientsatisfaction: concepts,theoryandevidence.InternationalJournalforQualityinHealthCare 1995;7:127–41.

[49]KravitzRL.Patients’expectationsformedicalcare:anexpanded formula-tionbasedonreviewoftheliterature.MedicalCareResearchandReview 1996;53:3–27.

[50]Conner-SpadyBL,SanmartinC,JohnstonGH,McGurranJJ,KehlerM, Nosewor-thyTW.Theimportanceofpatientexpectationsasadeterminantofsatisfaction withwaitingtimesforhipandkneereplacementsurgery.HealthPolicy(New York)2011;101:245–52.

[51]RossCK,FrommeltG, HazelwoodL, ChangRW.Theroleof expectations inpatientsatisfactionwithmedicalcare.JournalofHealthCareMarketing 1987;7:16–26.

[52]MichalosAC.Multiplediscrepanciestheory(MDT).SocialIndicatorsResearch 1985;16:347–413.

[53]JassoG.Onthejusticeofearnings:anewspeciﬁcationofthejusticeevaluation function.AmericanJournalofSociology1978;83:1398–419.

[54]MurrayCJL,KawabataK,ValentineN.People’sexperienceversuspeople’s expectations.HealthAffairs(Millwood)2001;20:21–4.

[55]GingrichJR.Makingmarketsinthewelfarestate:thepoliticsofvaryingmarket reforms.Cambridge:CambridgeUniversityPress;2011.

[56]JensenC.Marketizationviacompensation:healthcareandthepoliticsofthe rightinadvancedindustrializednations.BritishJournalofPoliticalScience 2011;41:907–26.

[57]Taylor-GoobyP.WelfareStatesunderPressure.London:SagePublications; 2001.

[58]RechelB,McKeeM.HealthreformincentralandeasternEuropeandtheformer SovietUnion.Lancet2009;374:1186–95.

[59]RechelB,RichardsonE,McKeeM.TrendsinhealthsystemsintheformerSoviet countries.EuropeanObservatoryonHealthSystemsandPolicies2014.

[60]StucklerD,BasuS,SuhrckeM,CouttsA,McKeeM.Thepublichealtheffect ofeconomiccrisesandalternativepolicyresponsesinEurope:anempirical analysis.Lancet2009;374:315–23.

[61]EuropeWHO.CoreHealthIndicatorsintheWHOEuropeanRegion.HEalth2020 targetindicators.Copenhagen:WHORegionalOfﬁceforEurope;2014.

[62]EuropeWHO.TheEuropeanHealthReport.Healthandhealthsystems. Copen-hagen:WHORegionalOfﬁceforEurope;2009.

[63]PopicT,SchneiderSM.AnEast-Westcomparisonofhealthcareevaluations inEurope:Doinstitutionsmatter?JournalofEuropeanSocialPolicy2018,

http://dx.doi.org/10.1177/0958928717754294.

[64]AndreßHJ,HeienT.Fourworldsofwelfarestateattitudes?Acomparison ofGermany,Norway,andtheUnitedStates.EuropeanSociologicalReview 2001;17:337–56.

[65]LipsmeyerC,NordstromT.EastversusWest:comparingpoliticalattitudes andwelfarepreferencesacrossEuropeansocieties.JournalofEuropeanPublic Policy2003;10:339–64.

[66]SvallforsS.Policyfeedback,generationalreplacement,andattitudestostate intervention:EasternandWesternGermany,1990–2006.EuropeanPolitical ScienceReview2010;2:119.

[67]KulinJ,MeulemanB.HumanvaluesandwelfarestatesupportinEurope:an East-Westdivide?EuropeanSociologicalReview2015:1–15.

[68]AlesinaA,Fuchs-SchündelnN.Good-ByeLenin(orNot?):TheEffectof Commu-nismonPeople’sPreferences.AmericanEconomicReview2007;97:1507–28.

[69]SmithPC,MossialosE,PapanicolasI.Performancemeasurementforhealth sys-temimprovement:experiences,challengesandprospects.Copenhagen:World HealthOrganisation;2008.

[70]KutzinJ,CashinC,JakabM.Implementinghealthﬁnancingreform:lessonsfrom countriesintransition.EuropeanObservatoryonHealthSystemsandPolicies 2010.

[71]MarreeJJ,GroenewegenPP.BacktoBismarck:EasternEuropeanHealthCare SystemsinTransition.Avebury:Aldershot;1997.

[72]PreacherKJ,ZyphurMJ,ZhangZ.AgeneralmultilevelSEMframeworkfor assessingmultilevelmediation.PsychologicalMethods2010;15:209–33.

[73]MuthénLK,MuthénBO.MplusUser’sGuide.StatisticalAnalysisWithLatent Variables.LosAngeles,CA:Muthén&Muthén;2015.

[74]NemecJ,KolisnichenkoN.Market-basedhealthcarereformsinCentraland EasternEurope:lessonsaftertenyearsofchange.InternationalReviewof AdministrativeSciences2006;72:11–26.

[75]LewisM.InformalhealthpaymentsinCentralandEasternEuropeandthe formerSovietUnion:issues,trendsandpolicyimplications.In:MossialosE, DixonA,FiguerasJ,KutzinJ,editors.Fundinghealthcare:optionsforEurope. Buckingham:OpenUniversityPress;2002.p.184–205.

[76]EnsorT.Informalpaymentsforhealthcareintransitioneconomies.Social Science&Medicine2004;58:237–46.

[77]Stepurko T,PavlovaM,GrygaI, GrootW.Makingpatients pay:informal patientpaymentsinCentralandEasternEuropeancountries.FrontPublicHeal 2015;3:1–4.

[78]MarmorT,WendtC.Conceptualframeworksforcomparinghealthcarepolitics andpolicy.HealthPolicy(NewYork)2012;107:11–20.

[79]SaltmanRB,BusseR,FiguerasJ.Socialhealthinsurancesystemsinwestern Europe.NewYork:OpenUniversityPress;2004.

[80]BusseR,RiesbergA.HealthCareSystemsinTransition.Germany,Copenhagen: WorldHealthOrganisation;2004.

[81]WendtC.MappingEuropeanhealthcaresystems:acomparativeanalysisof ﬁnancing,serviceprovisionandaccesstohealthcare.JournalofEuropeanSocial Policy2009;19:432–45.

[82]WendtC,FrisinaL,RothgangH.Healthcaresystemtypes:aconceptual frame-workforcomparison.SocialPolicy&Administration2009;43:70–90.

[83]VuorenkoskiL.Finland:Healthsystemreview.HealthSystemsinTransition 2008;10:1–168.

[84]McDaidD,WileyMM,MaressoA,MossialosE.Ireland:Healthsystemreview. HealthSystemsinTransition2009;11:1–298.

[85]SmithS.TheIrishhealthbasket:abasketcase?TheEuropeanJournalofHealth Economics2010;11:343–50.

[86]ThomsonS,JowettM,MladovskyP.Healthsystemresponsestoﬁnancial pres-suresinIreland.Policyoptionsinaninternationalcontext.Copenhagen2014.

[87]WendtC.Changinghealthcaresystemtypes.SocialPolicy&Administration 2014;48:864–82.

[88]OECD.HealthCareSystems:GettingMoreValueforMoneyOECDEconDep PolicyNotes,No2;2010.

[89]WagstaffA,vanDoorslaerE,vanderBurgH,CalongeS,ChristiansenT,Citoni G,etal.Equityintheﬁnanceofhealthcare:somefurtherinternational com-parisons.JournalofHealthEconomics1999;18:263–90.

[90]OliverA,MossialosE.Equityofaccesstohealthcare:outliningthefoundations foraction.JEpidemiolCommunityHealth2004;58:655–8.

[91]RobinsonR.Userchargesforhealthcare.In:MossialosE,DixonA,FiguerasJ, KutzinJ,editors.FundingHealCareOptionsEur.Buckingham:Buckingham: OpenUniversityPress;2002.p.161–83.

[92]GwatkinDR,BhuiyaA,VictoraCG.Makinghealthsystemsmoreequitable. Lancet2004;364:1273–80.

[93]vanDoorslaerE,WagstaffA,vanderBurgH,ChristiansenT,DeGraeveD, Duch-esneI,etal.EquityinthedeliveryofhealthcareinEuropeandtheUS.Journal ofHealthEconomics2000;19:553–83.

[94]ClearyPD,Edgman-LevitanS,RobertsM,MoloneyTW,McMullenW,Walker JD,etal.Patientsevaluatetheirhospitalcare:anationalsurvey.HealthAffairs (Millwood)1991;10:254–67.

[95]ClearyPD,Edgman-LevitanS,McMullenW,DelbancoTL.Therelationship betweenreportedproblemsandpatientsummaryevaluationsofhospitalcare. QualityReviewBulletin-Journals1992;18:53–9.

[96]NaumannE,BussC,BährJ.Howunemploymentexperienceaffectssupport forthewelfarestate:arealpanelapproach.EuropeanSociologicalReview 2016;32:81–92.