americans' attitudes toward the affordable care act

Post on 12-Feb-2022

2 views 0 download

transcript

AMERICANS’ATTITUDESTOWARDTHEAFFORDABLECAREACT:WOULDBETTERPUBLICUNDERSTANDINGINCREASEORDECREASEFAVORABILITY? 

WENDY GROSS1 

TOBIAS H. STARK2 JON KROSNICK3 JOSH PASEK4 GAURAV SOOD5 TREVOR TOMPSON

6 JENNIFER AGIESTA7 DENNIS JUNIUS8  

ThisresearchwasconductedwithsupportfromtheRobertWoodJohnsonFoundation,GfK,StanfordUniversity,andtheAssociatedPress.TheauthorsthankArthurLupiaandMichaelTomzforhelpfulcommentsonthismanuscript.

1GfK.wendy.gross@gfk.com2StanfordUniversityandUtrechtUniversity.t.h.stark@uu.nl3StanfordUniversity.krosnick@stanford.edu4UniversityofMichigan.jpasek@umich.edu5PrincetonUniversity,gsood@princeton.edu6AssociatedPress‐NORCCenterforPublicAffairsResearch.tompson‐trevor@norc.org7AssociatedPress.jagiesta@ap.org8AssociatedPress.djunius@ap.org

AbstractNationalsurveysconductedin2010and2012suggestthefollowingconclusions:

‐ AmericanunderstandingofwhatisandisnotintheACAhasbeenfarfromperfect.

‐ Correctunderstandingoftheelementsofthebillweexaminedvariedwithpartyidentification:Democratsunderstoodthemost,independentsless,andRepublicansstillless.

‐ Olderpeopleandmoreeducatedpeoplehaveunderstoodtheelementsofthebillweexaminedbetterthanhaveyoungerandlesseducatedpeople.

‐ Between2010and2012,publicunderstandingoftheelementsofthebillweexamined

didnotchangenotably.‐ MostpeoplehavefavoredmostoftheelementsoftheACAthatweexamined,butnot

everyonerecognizedthattheseelementswereallintheplan.‐ Mostpeopleopposedpoliciesthatweresometimesfalselythoughttobepartsofthe

ACA..‐ Ifthepublichadperfectunderstandingoftheelementsthatweexamined,the

proportionofAmericanswhofavorthebillmightincreasefromthecurrentlevelof32%to70%.

Takentogether,allthissuggeststhatifeducationeffortsweretocorrectpublicmisunderstandingofthebill,publicfavorabilitymightincreaseconsiderably.

1

IntroductionThePatientProtectionandAffordableCareActof2010(ACA)enactedaseriesofsignificantchangestotheAmericanhealthcaresystem.The900‐page‐longbill,whichelicitedanextremelypartisanreactionandsubstantialnewsmediainterest,amendedtheU.S.codetopreventinsurancecompaniesfromdenyingcoverageforpre‐existingconditions,provideforhealthcareexchangeswhereindividualscouldpurchasecaredirectly,requireallindividualstohavehealthinsuranceorpayafine,andmore.InJune,2012,theU.S.SupremeCourtupheldacentralelementofthislaw.

PublicdebateaboutthebillcalledattentiontomanyaspectsofthelawthatwereincludedintheversionthatCongressapproved.Butduringthecourseofpublicdebate,anumberofinaccurateclaimsweremade,assertingthatthebillincludedprovisionsthatwerenotincludedinthefinalversion.Someofthewidelydiscussedcomponentswerepartofthelegislation,suchastheplantoallowchildrentostayontheirparents’healthplanthroughage26.Butotherwidelydiscussednotionswereneverconsideredforinclusion,suchastheclaimthatapanelofbureaucratscoulddecidewhencoveragewouldbegiven(theso‐called“deathpanels”).Thelegislationincludedavarietyofless‐oftendiscussedprovisions,suchaschargingafeetoinsurancecompaniesthatofferedparticulartypesofinsurance.

Manysurveyswereconductedbothbeforethebill’spassageandafteritsenactmenttogaugetheAmericanpublic’sreactiontoit.Inearly2010,publicopinionwasfairlyevenlysplit.Forexample,accordingtoaKaiserFamilyFoundation(KFF)surveyinAprilofthatyear,46%ofAmericanssaidtheyhadafavorableopinionofthebill,and40%saidtheyhadanunfavorableopinion.Ayearlater,inApril2011,KFFreportedthesetwostatisticstobe41%and41%,respectively.AndinJanuary2012,thefigureswere37%and44%,respectively,perhapssuggestingaslightshiftintheunfavorabledirectionastimehaspassed.InMay2012,thefigureswereidentical:37%and44%,solidifyingevidenceofthatslightshift.AndinAugust2012,thesefigureswere38%and43%,respectively.

AsimilarportraitwaspaintedbyAP‐GfKpolls.InMay,2010,39%ofrespondentssaidtheysupportedtheACA,and46%saidtheyopposedit.InJune,2012,thosenumberswere33%and47%,respectively.Thus,asmalldecreaseintheproportionsupporting,andasmallincreaseintheproportionopposing.

Surveysdonebyotherorganizationsprovidedsimilar,thoughnotidentical,portraitsofthebalanceandtrajectoryofopinions.Forexample,aNBCNews/WallStreetJournalpolldoneinMay,2010,found38%ofrespondentssayingtheythoughttheACAwasagoodidea,and44%saiditwasabadidea.AsofJune,2012,thatorganizationfoundthesefigurestobe35%and41%.Thus,theproportionexpressingapositiveopiniondroppedslightly,liketheKFFpolls.Buttheproportionexpressinganegativeopinionalsodroppedslightly.

Onlyoneprominentnationalsurveyresearchorganization,ThePewResearchCenter,reportedresultssuggestingmovementintheoppositedirection.InJanuary,2011,41%oftheirrespondentsapprovedoftheACA,and48%opposed.Theirmostrecentsurvey,inJune‐July,2012,foundthat47%approved,and43%opposed.

2

Puttogether,mostnationalsurveysduringthelasttwoyearssupporttwoprincipalconclusions:(1)publicopinionhasnotmanifestedasizableandconsistentleaningtowardbeingfavorableorunfavorabletowardtheACA,and(2)aslightshiftinthenegativedirectionmayhaveoccurredsincethelawwaspassed.

ItwouldbeunderstandabletolookatsuchevidenceasanindicationthattheAmericanpublicdoesnotstronglysupportthispieceoflegislation.Afterall,ifaboutasmanypeoplefavoritasopposeit,andifwehaveneverseenamajorityfavoringit,thathardlysendsastrongsignalofsupport.Furthermore,itiseasytoimaginethatsincepassageofthebill,Americanshavehadtimetoconsiderthebillanditsimplicationsinmoreandmoredepth,andifsuchconsiderationleadstoashiftinthenegativedirection,thatcertainlysignalsquitetheoppositeofenthusiasm.Thus,suchdatacouldbetakenasasignalthatRepublicaneffortstorepealthebillwouldbewarmlywelcomedbyagrowinggroupofAmericans.

Atthesametime,theverysamepollingevidencecanbeviewedfromtheoppositeviewpoint.AlthoughamajorityofAmericanshavenotfavoredthebill,itisalsotruethatamajorityhaveneveropposedit,either.Andevenaftertherecentsmallincreaseinopposition,theproportionopposingitdoesnotexceed50%.Sodirectquestionsaskingaboutpositivevs.negativeevaluationsoftheplanhavenotdocumentedamandatefromthepublictorepealthebill.9Takentogether,allthisevidenceportraystheAmericanpublicinwhatmightseematypicalway:splitaboutevenly,andnotprovidingaclearmandatetoelectedrepresentativesonewayoranother.

Itwouldthereforenotbeunreasonableforthoserepresentativestolookatthispollingevidence,reachthatconclusion,andproceedtotakeactionsinkeepingwiththeguardianshipviewofdemocracy:decidingwhattheyfeelisbestforthecountryandtakingaction(ordoingnothing)accordingly,regardlessofpublicopinion.

Suchignoringofpublicopinionmightalsobejustifiedfromanotherperspectiveaswell.Insomanysurvey‐basedinvestigationsoftheAmericanpublicformanydecades,peoplehavebeenfoundtoperformquitepoorlyonquizzesassessingfactualknowledgeaboutdomainsinwhichsignificantlegislationhasbeenconsideredorpassed.Ifmostpeoplelackthefactsneededtotrulyunderstandtheproblemstobesolvedbyapieceoflegislationandthesolutionsofferedbythatlegislation,whyshouldpublicevaluationsofthelegislationbetakenseriously?Thatis,thepublicmightfeelverydifferentlyiftheytrulyunderstoodabill,soopinionsbasedonpartialinformationorsubstantialmisconceptionscancertainlynotbedescribedas“wise”andshouldperhapsthereforebeignoredbylegislators.

Ofcourse,ignoringpublicopinion,evenuninformedpublicopinion,mayplacelegislatorsatriskcomeelection‐time.Evenwhenthepublicdoesnotunderstandapieceoflegislation,membersoftheelectoratemaynonethelessholdstrongopinionsaboutit,eitherfavorableorunfavorable,andthoseopinionsmayshapetheirvotingdowntheroad.Indeed,agreatdealofresearchsuggeststhatpublicopiniononpolicyissuesdoessometimesshapevotechoices(seeAnand&Krosnick,2003;Krosnick,1988).Soalegislatorwhovotesagainstapieceoflegislationthatvotersfavormay9QuestionsaskingwhetherthebillshouldberepealedhavesometimesshownamajorityofAmericansansweringaffirmatively.

3

findhimselforherselflaterpayinganelectoralpriceifthatvotebecomeswellpublicizedbythelegislator’sopponentduringacampaign,evenifpublicunderstandingofthelegislationisseriouslywanting.

Inasituationsuchasthis,legislatorswhowishtoseeapieceoflegislationpassed(oravoiditsrepeal)alwayshavetheoptionofinformingtheAmericanpublicaboutwhatthelegislationwouldtrulydo,inthehopethatbetterunderstandingwouldleadthepublictoofferastrongersignalofsupporttotheirelectedrepresentatives.Butwouldsucheducationindeedleadtomoresupport?Thispresumablydependsonthenatureofthepublic’smisunderstandingsandonthepublic’sevaluationsoftheelementstheybelievecomposethelegislationandoftheelementsthatinfactcomposeit.

Inthispaper,wereportaninvestigationofexactlytheseissueswithregardtotheACA.Usingdatafromtwosurveys(oneconductedin2010andtheotherin2012),weexplored:

1) HowaccuratelyAmericanshaveunderstoodwhatisintheACAandwhatisnot.2) Howtheaccuracyofpeople’sunderstandinghaschangedduringthetwoyearssincethe

billwaspassed.3) HowknowledgeaccuracyisrelatedtofavoringtheACA–thatis,whetherpeoplewho

knowmoreaboutwhat’sintheACAlikeitmoreorlikeitless,4) HowthepublicwouldfeelabouttheACAifeveryoneunderstoodthataseriesofits

centralelementsareindeedincludedintheACAandthataseriesoffrequentlydiscussedbutultimatelyomittedelementsarenotinit.

Alongtheway,weinvestigatedtwootherissues:

1) Thepredictorsofaccurateunderstandingoftheplan‐thatis,whichtypesofpeoplearemoreandlesslikelytoscorewellonaquiz.

2) Thepopularityofvariousspecificelementsthatwereincludedintheplan,andhowpopularareelementsthatwerenotincludedbutweresometimesclaimedtobe.

Incarryingoutthisinvestigation,weimplementedanewapproachtomeasuringpublicunderstandingofapublicpolicyissue.Insurveysdoneduringthelast80years,ithasbeenroutinetotestknowledgebyaskingpeoplefactualquestionsandgradingpeopleaseithercorrectorincorrectbaseduponwhethertheiranswermatchedthefactsornot.Butthisapproachignoresasimpleandunavoidablefact:thatarespondentsayingtoasurveyinterviewerthatmembersoftheRepublicanPartyoutnumbermembersoftheDemocraticPartyintheU.S.HouseofRepresentativesdoesnotnecessarilymeanthattherespondentbelievesthistobetrue.Whenaskedwhichpartyholdsmoreseats,arespondentmightsimplyguessandendupgivingthecorrectanswerbychancealone.Thisresponsewouldnotrevealabeliefthattherespondentgenuinelyholds,norwouldthatpurportedbeliefhaveanyimpactonhisorherthinking,becauseheorshedoesnottrulyholdthatperceptionoftheworld.Guessingseemslikelytoespeciallydistortanswerstoquizquestionsthatofferonlytworesponsechoices,asweusedhere(isthisincludedintheACAornotincludedintheACA?).

4

Onemightimaginethatthisproblemcanbeovercomebyexplicitlyofferingsurveyrespondentstheopportunitytodeclinetoanswerasurveyquestionbysayingthey“don’tknow”theanswerandencouragingthemtodoso.Butagreatdealofresearchsuggeststhatthisstrategyisunwise.Insteadofattractingonlyandallofthepeoplewhotrulydonotholdabeliefonanissue.“don’tknow”responseoptionsattractmanyrespondentswhotrulyholdopinionsandfailtoattractrespondentswhoholdopinionswithlittleornocertainty(forareview,seeKrosnick,2002).

Thesolutiontothisproblemissuggestedinsteadbyaliteratureinpsychologyoncertainty.Theintendedpurposeofofferinga“don’tknow”optionistypicallyexpressedasfilteringoutpeoplewhowouldexpressajudgmentwithnocertaintyatall.Thatis,apersonmightsay“IthinkthattheDemocratsholdmoreseats,butI’mnotatallconfidentaboutthatguess.”Thus,thepreferablesolutionistofirstaskpeopletomaketheirbestguessandthentoaskthemtoratethecertaintywithwhichtheyexpressthatbelief.Thisallowsresearcherstofilteroutpeoplewhoofferopinionswithlittleornocertainty.

Acollateralbenefitofthisapproachisthatcertaintystronglycorrelateswithuseofbeliefsduringdecision‐making.Peoplewhoholdabeliefwithconfidenceareinclinedtouseitwhenmakinghighlyrelevantdecisions.Incontrast,peoplewhoholdabeliefwithminimalconfidenceareunlikelytouseit(forareview,seePetty&Krosnick,1995).Thus,givingsurveyrespondents“credit”foraccuratelypossessingabeliefonlywhentheyexpresshighcertaintyallowsustoidentifythosebeliefsthatarealsolikelytohaveshapedpeople’soverallevaluationsoftheACA.

Therefore,inkeepingwiththisperspective,whenweadministeredquizquestionsassessingpublicunderstandingoftheACA,eachquestionwasfollowedbyaquestionaskingrespondentshowsuretheywereabouttheiranswertothepriorquestion.Peoplewhoexpressedhighdegreesofconfidencewhengivingacorrectanswerweretreatedasholdinganaccuratebelief,andpeoplewhogaveacorrectanswerwhileexpressedlowdegreesofconfidencewerenotcreditedashavinganaccuratebelief,norwerepeoplewhoansweredthequizquestionsincorrectly.

TheelementsoftheACAthatwereaddressedbythequizquestionswereselectedcarefullytocovermostofthecentralelementsoftheplan.Intheirdocumententitled“FocusonHealthReform:SummaryofNewHealthReformLaw”(Publication#8061;www.kff.org),theKaiserFamilyFoundationprovidedwhattheycalleda“summaryofthelawandchangesmadetothelawbysubsequentlegislation.”WereliedonthissummarytoselecttheelementsoftheACAtoaskaboutinoursurvey.WealsoaskedaboutanadditionalsetofpoliciesthatwerenotultimatelyincludedintheACAbutwerediscussedduringthepublicdebateofit.

DataandMethodsThedataforthisstudycomefromtwocross‐sectionalsurveysofnationallyrepresentativesamplesofAmericanadultsconductedviatheInternetbyGfK(formerlyKnowledgeNetworks).RespondentsweredrawnfromtheKnowledgePanel®‐anationallyrepresentativepanelrecruitedviarandomdigitdialingandbyaddress‐basedsampling.Thesamplingdesigncovers97%oftheAmericanpopulation,includinghouseholdsthatdonothaveInternetaccessoralandlinetelephone.Allpanelistswereremuneratedfortheirparticipation;peoplewhodidnotalreadyhave

5

eitheracomputerorInternetaccesswereprovidedthem.Uponjoiningthepanel,panelistsfirstcompletedacoreprofilequestionnairethatcapturedinformationabouttheirrace,gender,age,income,education,andmore.Foreachsubsequentsurvey,panelmemberswereselectedusingaprobabilityproportionaltosize(PPS)weightedsamplingdesign,producingasamplethatisrepresentativeoftheAmericanpopulation.

ThefirstsurveyforthisprojectwasconductedbetweenAugust31andSeptember7,2010.Arandomsampleof1,815adultswasinvitedtoparticipate,and1,271completedthesurvey(completionrate=70%).Themediantimespentcompletingthequestionnairewas26minutes.

ThesecondsurveywasconductedbetweenAugust3and13,2012.GfKinvited2,344Americanadultstoparticipate,and1,334completedthequestionnaire,acompletionrateof57%.Themediantimespentcompletingthequestionnaire(whichwasmuchshorterthanthequestionnaireusedin2010)was17minutes.

Allanalysesreportedbelowwereconductedusingweightstoadjustforunequalprobabilityofselectionandpost‐stratifyingbasedondemographics.

Table1displaysvariouspercentagesneededforthecalculationofresponseratesforthesurveyanddisplaysvariousthebreakdownfortheresponseratesforbothsurveys.

Table1:ResponseRatesRate 2010 2012

A. PanelRecruitmentResponseRate(AAPORResponseRate3)

17.2% 17.2%

B. HouseholdProfileRate

61.5% 61.5%

C. HouseholdRetentionRate

35.0% 35.0%

D. SurveyCompletionRate

70.0% 57.3%

E. ActiveRate 99.2% 99.2%ORR1(A*B*C*D*E) 2.6% 2.1%ORR2(A*B*D) 7.4% 6.1%ORR3(A*D) 12.0% 9.9%

AppendixBdisplaysthedemographicprofilesofbothsamples.

6

Measures

KnowledgeandCertaintyRespondentswereasked18knowledgequizquestions,probingwhetherspecificprovisionswereinthehealthcarebill.Respondentsreadonedescriptionatatimeandindicatedwhethertheythoughttheprovisionwas“inthebill”or“notinthebill”thatCongresspassedin2010.

TwelveoftheelementswereprincipalprovisionsoftheACA.Theremainingsixelementswerenotinthebillbuthadbeenfrequentlydiscussedinpublicdebate;theseelementswereidentifiedbyexpertsattheAssociatedPressandresearchersatStanfordUniversity.AppendixAshowstheinstructionsforrespondents,thefulllistofquestions,andthecorrectanswerforeachitem.

Followingeachquizquestion,respondentswereasked,“Howsureareyouaboutthis?”Theanswers“extremelysure”and“verysure”werecoded1(certain),andtheanswers“moderatelysure”,“slightlysure”,and“notsureatall”werecoded0(uncertain).

Tomeasureeachrespondent’slevelofknowledge,wefirstcomputedthepercentofthe12provisionsoftheACAthattherespondentscorrectlyidentifiedassuchwithhighconfidence.Next,wecomputedthepercentofthe6provisionsnotintheACAthattherespondentcorrectlyidentifiedassuchwithhighconfidence.Then,weaveragesthesetwopercentagestoyieldafinalknowledgescoreforeachrespondent.Consequently,thefactthattwiceasmanyquestionstappedknowledgeaboutprovisionsinthebillastappedprovisionsnotinthebilldidnotcausethefinalindexscoretobebasedmoreonunderstandingoftheelementsinthebillthanonunderstandingoftheelementsnotinthebill.Thefinalknowledgescorerangedfrom0%forpeoplewhodidnotidentifyanyprovisioncorrectlywithhighconfidence,to100%forpeoplewhoproperlyidentifiedallprovisionswithhighconfidence.

EvaluationoftheACAInthe2012survey,respondentswereasked,“Ingeneraldoyoufavor,oppose,orneitherfavornoropposethelawchangingthehealthcaresystemthattheU.S.CongresspassedinMarch2010?”Theresponses“favorstrongly”and“favorsomewhat”werecoded1(indicatingfavoring),andtheresponses“neitherfavornoroppose”,“opposesomewhat”,and“opposestrongly”werecodedas0(notfavoring).

SupportforACAPlanElementsRespondentswerealsoaskedtoindicatewhethertheyfavoredoropposedeachofthe18provisionsaddressedbythequizquestions.Oneachscreen,respondentswereasked“Doyoufavoropposethischange?”alongwithastatementdescribingtheprovision.Responseswerecodedinthesamefashionasforthegeneralfavorabilityquestion,“favor”versus“notfavor”.

PartisanshipTwodummyvariableswerecreatedtodistinguishRepublicansandDemocratsfrompeoplewithoutapartyaffiliation.RespondentswerecodedtobeaDemocratoraRepublicaniftheyanswered“Democrat”or“Republican”tothequestion“DoyouconsideryourselfaDemocrat,aRepublican,an

7

independent,ornoneofthese?”Allotherrespondentsweretreatedasindependents.Twenty‐sevenrespondentsrefusedtoanswerthisquestionandweretreatedasmissinginanalysesusingthisvariable.

MediaUseRespondentswereasked,“Howoftendoyougetnewsfromeachofthefollowing?”andpresentedwithalistof“LocalTVNews,”“FoxNewscablechannel,”“MSNBCcablenews,”“CNNcablenews,”“NationaleveningnetworktelevisionnewsonCBS,ABC,orNBC”,“Radionews”,and“newsfromtheinternet.”Responsestothesequestionswerecoded1iftherespondentsanswered“extremelyoften”or“veryoften”.Theresponses“moderatelyoften”,“rarely”,and“never”werecoded0.

DemographicInformationAgewascodedtorangefrom0to1.DummyvariablesdistinguishedbetweenWhite,Black,Hispanicrespondentsandthosewhoindicatedtheybelongedtoanotherethnicgroup.Variablesindicatingeducationseparatedpeoplewithahigh‐schooldegreeorlessfrompeoplewhoindicatedhavingattendedsomecollegeeducationbutnodegreeandpeoplewhograduatedfromcollege.Finally,twodummyvariablesdistinguishedthreeequallylargegroupsofpeoplewhoindicatedhavinglowincome(lessthan$39,999),moderateincome(between$40,000and$84,999),orhighincome(morethan$85,000).

Results

Favoringvs.OpposingtheACAin2012In2012,32%ofrespondentssaidtheyfavoredtheACA,36%saidtheyopposedit,and32%saidtheyneitherfavorednoropposedit.Thefactthatnegativeresponsesslightlyoutnumberedpositiveresponsesresemblesresultsproducedbyvariousotherpollingorganizationsinrecentmonths.

Notsurprisingly,evaluationsofthebillvariedaccordingtopoliticalpartyidentification.AmongDemocrats,51%favorthebill;29%ofindependentsexpressedthesameopinion,and8%ofRepublicansdidso.Thepartisangapof43percentagepointsisnotatypicalinAmericanevaluationsofpiecesoflegislationthesedays.

KnowledgeabouttheACAin2012In2012,whenimplementingthemethodusedinmostpaststudiesofgivingcredittorespondentswhogavecorrectanswersregardlessofcertainty,frequencyofcorrectanswerswasstrikinglyhighforsomeelementsoftheplan(seecolumn2ofTable2).Forexample,80%saidthatchildrencouldbecoveredbytheirparents’policy.And80%saidthatcompanieswithmorethan50employeeswererequiredtoprovidehealthinsurancetotheiremployees.Majorities,andoftenlargemajorities,gavecorrectanswersregardingmostoftheprovisionsinthebill.Foronlytwoofthetwelveprovisionsweaskedabout(newfeestobechargedtohealthinsurancecompaniesandpharmaceuticalcompanies)didmajoritiesgivetheincorrectanswer(only44%and37%gavecorrectanswers,respectively).

8

However,thesenumbersshouldbetakenwithagrainofsalt,foratleasttworeasons.First,becausethesequestionsofferedjusttwoanswerchoices(inthebillvs.notinthebill),randomguessingwouldbeexpectedtoyield50%ofanswerscorrectbychancealone.Andsecond,manyofthecorrectanswerspeoplegavemayhavebeengivenwithlittleornocertainty,soitmightbeinappropriatetosaythatthesepeoplepossessedthesebeliefs.

Whentakingintoaccountpeople’scertaintyratings,weobservedmuchlowerlevelsofaccurateknowledge(seecolumn1ofTable2).Indeed,onlyoneprovisionwascorrectlyidentifiedwithhighcertaintyasbeingpartoftheACAbyamajorityofrespondents.52%ofrespondentscorrectlysaidwithhighcertaintythatchildrenundertheageof26couldgethealthinsurancebybeingincludedontheirparents’healthinsurancepolicies.Allotherprovisionsofthelawwerecorrectlyidentifiedwithhighcertaintybylessthan40%ofAmericans.Only11%correctlybelievedwithhighcertaintythatdrugcompanieswouldpaynewfeesunderthelaw,andonly10%correctlybelievedwithhighcertaintythatinsurancecompanieswouldpaysuchnewfees.

Table 2: Accuracy of Knowledge About Provisions That Were in the ACA– 2012 Data 

Policy 

% Thinking policy is in law with high 

certainty 

% Thinking policy is in law regardless of 

certainty Children under 26 can be included in parents’ insurance  52.2%  80.1% Large companies have to provide health insurance to employees   38.7%  80.1% U.S.  citizens  without  health  insurance  have  to  pay fine if they don’t have specific reasons  36.3%  69.4% Insurance companies have to sell health insurance to people with preexisting conditions   32.6%  72.2% Insurance companies have to continue insurance as long as no rules are broken   28.8%  77.1% 

Make insurance for sale for any American  28.7%  69.2% Prevent limiting amount paid for person’s health care costs  23.5%  60.5% Discounts on prescriptions to seniors with high drug costs  20.0%  68.4% Federal tax credits for small companies that buy insurance for their employees  17.5%  66.8% Subsidize health insurance for U.S. citizens with low income  16.8%  51.4% 

New fees for health insurance companies   11.1%  44.3% 

New fees for companies that make drugs  10.2%  37.0% 

Asimilarportraitemergedwithregardtothe6provisionsweaskedaboutthatwerenotactuallyintheACA.Majoritiesgavecorrectanswerstothequizquestionsforallbutoneoftheseprovisions

9

(seecolumn2ofTable3).Butsomeofthesewereonlyslightmajorities.Forexample,54.3%saidthatso‐called“deathpanels”arenotpartofthelaw.Thisisonlyveryslightlymorethanwouldbeexpectedbychancealone.Andwhencertaintywastakenintoaccount,minorities(andsometimesverysmallminorities)believedwithhighconfidencethattheywerenotincludedinthelaw(seecolumn1ofTable3).Forexample,only25.6%wascertainthatjobapplicantswouldnothavetodisclosepreviousillnessestonewemployers.And“deathpanels”weresaidwithcertaintytonotbepartofthelawbyonly17%oftherespondents.

Table 3: Accuracy of Knowledge About Provisions That Were Not in the ACA– 2012 Data 

Policy 

% Thinking policy is NOT in law with high 

certainty 

% Thinking policy is NOT in law regardless 

of certainty Job applicants have to disclose previous illnesses to employer  25.6%  75.3% Restaurants with unhealthy food must pay fee to government  23.2%  73.1% Committees decide whether people get medical care (“death panels”)  16.8%  54.3% 

Smokers have to pay additional $1,000 a year  14.1%  64.0% Health care ID card needed to get treatment at hospital  13.3%  56.5% 

Require treatment of illegal immigrants for free  10.5%  41.7% 

Lookedatfromanothervantagepoint:78%ofrespondentsansweredmorethanhalfofthe18quizquestionscorrectly(seecolumn4ofTable4),butonly44%answered13ormoreofthe18questionscorrectly.Andthesefigureswerestrikinglylowerwhentreatingascorrectonlyanswersprovidedwithhighcertainty(seecolumn2ofTable4):14%and3%,respectively.Notasinglerespondentansweredeveryquizquestioncorrectlywithhighcertainty.

Table 4: Number of Correctly Answered Quiz Questions With and Without Taking Certainty into Account (N = 1344) – 2012 Data 

  Requiring High Certainty 

  Regardless of  Certainty 

Number of correctly answered questions  Percent 

Cumulative percent    Percent 

Cumulative percent 

All 18  0.0%  0.0%  0.3%  0.3% 17 out of 18  0.0%  0.0%  1.5%  1.8% 16 out of 18  0.1%  0.1%  6.0%  7.8% 15 out of 18  0.8%  1.0%  10.9%  18.8% 14 out of 18  1.3%  2.3%  12.5%  31.3% 13 out of 18  1.1%  3.4%  12.9%  44.1% 

10

12 out of 18  3.2%  6.6%  14.4%  58.5% 11 out of 18  2.7%  9.3%  11.9%  70.4% 10 out of 18  4.2%  13.5%  7.7%  78.1% 9 out of 18  5.4%  18.9%  6.1%  84.2% 8 out of 18  5.6%  24.5%  5.6%  89.7% 7 out of 18  5.9%  30.4%  3.0%  92.7% 6 out of 18  6.7%  37.1%  2.8%  95.5% 5 out of 18  5.9%  42.9%  0.9%  96.4% 4 out of 18  7.2%  50.1%  0.6%  97.0% 3 out of 18  8.8%  58.9%  0.3%  97.3% 2 out of 18  8.6%  67.5%  0.1%  97.5% 1 out of 18  10.4%  77.9%  0.3%  97.8% 0 out of 18  22.1%  100.0%  2.2%  100.0% Total  100.0%  100.0%  0.3%  0.3% 

PredictingLevelsofKnowledgeabouttheACAAsshownbytheestimatesofparametersofordinaryleastsquares(OLS)regressionspredictingtheknowledgescore(computedasdescribedinthemethodsectionabove),levelsofknowledgevariedsystematicallywithsomecharacteristicsofrespondents.Whenusingdemographicsandpartyidentificationaspredictors(seecolumn1ofTable5),wefound:

‐ Democratsweresignificantlymoreaccuratethanwereindependents.

‐ IndependentweresignificantlymoreaccuratethanRepublicans.‐ Accuracyincreasedasrespondentageincreased.‐ Accuracyincreasedasrespondenteducationincreased,perhapsatleastpartlythe

resultofthefactthatmoreeducatedpeoplegenerallypaymoreattentiontoinformationaboutpoliticsinthenews.

‐ Sex,race,andincomewerenotsignificantlyrelatedtoknowledgeaccuracy.10

Whenweaddednewssourcesaspredictorsintheregressionequation,weobservedaseriesofstrikingfindings:

‐ Frequentexposuretowhatmightbecalled“mainstream”newssources(CNN,CBS,ABC,NBC,andlocaltelevisionnewsprograms)appearstohavehadnosignificanteffectontheextentofaccurateknowledgepossessedbyrespondents.Thatis,thesenewsorganizationsmaynothaveconferredenhancedunderstandingontheirviewers.

10Representingincomeinfivegroupsinsteadofthreedidnotchangetheseresults.

11

‐ FrequentexposuretoMSNBCmayhaveenhancedaccurateknowledgemorethandidfrequentexposuretoanyothernewssourceweexamined.

FrequentexposuretoFoxNewsmayalsohavehadapositiveeffectonunderstandingthebill.

‐ Frequentexposuretoradionewsandinternetnewsalsoappeartohavesignificantlyenhancedunderstandingofthebill.

Table 5: OLS Regression Predicting Percent Correct Answers to Knowledge Quiz Questions with High Certainty – 2012 Data 

Predictor  Analysis 1  Analysis 2 

Democrat  3.90*  3.00 Republican  ‐4.69**  ‐5.77*** Age  9.18**  5.70 Female  ‐1.86  ‐1.85 Black  ‐.93  ‐1.11 Hispanic  ‐.37  ‐0.41 Other Race  3.22  3.75 Some college education  4.10*  3.59* College graduate  8.59***  7.04*** High income  2.11  1.95 Middle income  ‐.47  ‐1.05 Fox News exposure  ‐  3.81* MSNBC exposure  ‐  8.05** CNN exposure  ‐  ‐1.30 Exposure to news on CBS, ABC, or NBC 

‐ 1.07 

Exposure to local TV news 

‐ 1.01 

Radio news exposure  ‐  4.40* Internet news exposure  ‐  3.77* Intercept  14.94***  13.42*** 

N  1316  1222 

*** p < 0.001, ** p< 0.01, * p< 0.05 

TheRelationofKnowledgetoFavoringtheACAAsrevealedbyestimatesoftheparametersofalogisticregressionequationpredictingopinionsabouttheACAusingknowledgescores,moreknowledgewasassociatedwithahigherprobabilityoffavoringthebill(seeTable6).Putdifferently,themoreaccurateaperson’sbeliefswereaboutthe18elementsthatweaskedabout,themoreheorshelikedtheACA.

12

AccordingtothecoefficientsinTable6,anindependentwithperfectknowledgeoftheplanelementsweexaminedhada15.7timeshigherchanceoffavoringthebillthandidsomeonewhodidnotanswerasinglequizquestioncorrectly(log‐odds=.03,odds‐ratiofor100%knowledgeversusnoknowledge=15.73).Whenusingtheparameterestimatesfromanequationexcludingpartyaffiliationasapredictor,ahypotheticalpersonwithperfectknowledgeoftheplanelementsweexaminedhadaalmost21timeshigherchanceoffavoringthebillthandidsomeonewhodidnotanswerasinglequizquestioncorrectly.

Whencontrollingforknowledgelevelandallotherbackgroundvariables,partyidentificationwasstillsignificantlyrelatedtofavoringtheACA.Democratslikeditmorethanindependents,andindependentslikeditsignificantlymorethanRepublicans.ThechancesthataDemocratwouldfavorthebillwasabout2.5timeshigherthanthechanceforanindependent(log‐odds=.94,odds‐ratio=2.56).AndaRepublican’schancesoffavoringthebillwasonly0.26timesthechancethatanindependentwoulddoso(log‐odds=‐1.36,odds‐ratio=0.26).Thus,althoughthesethreegroupsdifferedintermsoftheirknowledgelevels,thegapbetweentheminapprovalremainedevenwhenstatisticallyequatingthesegroupsintermsoftheirknowledgelevels.

CollegegraduatesandpeoplewhowerefrequentlyexposedtonewsonCNNhadahigherlikelihoodoffavoringtheACAthanothers.Incontrast,peoplewhofrequentlywatchedFoxNewsweresignificantlylesslikelytofavorthebillthanothers.AllotherdemographicvariableswerenotrelatedtotheprobabilityoffavoringtheACA.

Table 6: Logistic Regression Predicting Favoring the ACA – 2012 Data 

Predictor  Analysis 1 

Percent correct answers  .03*** Democrat  .94*** Republican  ‐1.36*** Age  .08 Female  ‐.23 Black  .21 Hispanic  .01 Other Race  .36 Some college education  ‐.08 College graduate  .55* High income  .27 Middle income  .01 Fox News exposure  ‐1.31*** MSNBC exposure  .40 CNN exposure  .72** Exposure to news on CBS, ABC, or NBC  .15 Exposure to local TV news  .11 

13

Radio news exposure  ‐.10 Internet news exposure  .28 Intercept  ‐1.95*** 

N  1316 

*** p < 0.001, ** p< 0.01, * p< 0.05 

ProjectingFavorabilitywithPerfectKnowledgeHowwouldthenationhavefeltabouttheACAifeveryonehadunderstoodalltheprovisionsweaskedaboutcorrectlyandwithconfidence?

Weansweredthisquestionintwoways.First,wecalculatedthedegreeoffavorabilityamonghypotheticalpeopletoillustratehowknowledgeaffectedsupportersofthetwopartiesandindependentsdifferently.Figure1showshowtheprobabilityoffavoringthebillchangedbasedontheresultsinTable6forahypotheticalwhitefemale,45yearsold,withhighschooleducationorless,ahighincome,andwhodidnotreceiveinformationoftenfromanyofthenewssources.

ThisfigureillustratesthatifthishypotheticalpersonwereaRepublicanwhoanswerednoneofourknowledgequizquestionscorrectlywithconfidence,shewouldhaveonlya4%probabilityoffavoringthebill.Andisshehadansweredallofourquestionswithhighconfidence,shewouldhavehada37%chanceoffavoringtheACA.

IfthepersonwereaDemocratorconsideredherselftobeanindependent,boththeincreaseintheprobabilityoffavoringthebillwithincreasingknowledgewouldbegreater.Ahypotheticalindependentwhogavenocorrectanswerswithhighcertaintywouldhavehada13%chanceoffavoringthebill,andthisnumberwouldincreaseto70%ifsheansweredallquestionsaccuratelywithhighconfidence.ForahypotheticalDemocratwhoansweredallquestionsaccuratelywithhighconfidence,theprobabilityoffavoringthebillwas86%.

Toapproachthisissueinasecondway,wecalculatedforeveryrespondenttheprobabilitythatheorshewouldfavortheplanifheorshehadansweredallquestionsaccuratelywithhighconfidence.

Thismethodyieldedthepredictionthat70%ofpeoplewouldfavorthebilliftheyallansweredthequestionsaccuratelywithhighknowledge.Thiscontrastsdramaticallywiththefactthat32%ofrespondentsinthesurveysaidthattheyfavoredthebill,basedontheircurrent,farfromperfectlevelsofcurrentactualunderstandingoftheprovisionsweaskedabout.Thus,thisanalysissuggeststhatincreasedunderstandingmighthavetransformedthe32%favoringinto70%favoring.

Thismethodsuggestedthatabout88%ofDemocratswouldfavorthebill,74%ofindependentswoulddoso,and40%ofRepublicanswoulddoso.

14

Figure1:TheRelationofKnowledgetoFavoringtheACA:Dems,Inds,andRepsSeparately

FavoringtheElementsoftheACATheresultsofthissimulationsuggestthateliminatingmisunderstandingsabouttheACAmightleadtomorefavorableevaluationsofit.Inorderforthistobetrue,itwouldbenecessarythatpeoplegenerallylikedtheprovisionsthatwereactuallyincludedintheACAbutthattheyoftendidnotbelievewithconfidencewere.Then,learningoftheirinclusionwouldincreaseoverallfavorability.Andlikewise,itmayalsobethecasethatpeoplegenerallydislikedtheprovisionsthattheysometimesthoughtincorrectlywereinthebill,solearningthattheywerenotinthebillwouldalsoincreaseoverallfavorability.

Infact,directquestionsassessingevaluationsofthespecificplanelementsconformedtothoseguesses.AsshowninTable7,majoritiesofrespondents,andsometimeshugemajorities,favoredtheelementsthatwereactuallyincludedintheplan.OnlythreeplanelementswerenotfavoredbyamajorityofAmericans:chargingafeetocitizenswhodonothavehealthinsurance,andchargingnewfeestohealthinsurancecompaniesandpharmaceuticalmanufacturers–werenotfavoredbyamajorityofAmericans

0 20 40 60 80 100

020

4060

8010

0

Predicted percentage favoring the bill by knowledge

Percentage correct answers with high certainty

Pre

dict

ed p

erce

ntag

e in

favo

r

DemocratIndependentRepublican

15

Table 7: Percent Favoring the Elements of the ACA ‐ 2012 Data 

Element Percent Favoring the 

Element  Insurance companies have to continue insurance as long as no rules are broken  81.8% 

Discounts on prescriptions to seniors with high drug costs  79.2% 

Make insurance for sale for any American  77.9% Federal tax credits for small companies that buy insurance for their employees  71.2% 

Children under 26 can be included in parents’ insurance  70.4% 

Prevent limiting amount paid for person’s health care costs  63.9% Insurance companies have to sell health insurance to people with preexisting conditions  62.0% Large companies have to provide health insurance to employees  55.3% 

Subsidize health insurance for U.S. citizens with low income  47.3% 

New fees for companies that make drugs  31.1% U.S. citizens without health insurance have to pay fine if they don’t have specific reasons  27.5% 

New fees for health insurance companies  24.3% 

Evenmorestrikingly,theprovisionsthatwerenotintheACAwerefavoredbyminorities,andoftensmallminorities,ofAmericans(seeTable8).Aminorityof35%favoredtheideathatsmokersshouldhavetopayanadditional$1,000peryearfortheirinsurance.Notsurprisingly,“deathpanels”werefavoredbyfewerthan20%.Thus,thesefiguressuggestthatcorrectingmisunderstandingsabouttheseelementsmightleadtoincreasedpublicfavorabilityabouttheACA.

Table 8: Percent Favoring of Elements That Were NOT in the ACA – 2012 Data 

Element Percent Favoring the 

Element  

Smokers have to pay additional $1,000 a year  35.1% 

Health care ID card needed to get treatment at hospital  27.1% Restaurants with unhealthy food must pay fee to government  26.1% 

Require treatment of illegal immigrants for free  20.8% Committees decide whether people get medical care (“death panels”)  19.1% Job applicants have to disclose previous illnesses to employer  15.5% 

16

ChangeinKnowledgeAccuracybetween2010and2012DuringthetwoyearssincetheACAwaspassedbyCongress,publicunderstandingofthelawhasincreasedslightlyforsomeplanelementsandhasnotchangedformost.Wewereabletoassessthisbycomparingtheresultsofour2012knowledgequizwiththeresultsobtainedbyadministeringanidenticalquizinour2010survey.

Withregardtothe12elementsthatareincludedintheACA,thepercentofpeoplewhocorrectlyrecognizedthatfactwithhighconfidenceincreasedsignificantlyfor5ofthe12elements(seethelastcolumnofTable9).Thisincreasewasmostsizablefortheplanelementsrequiringthatcitizenswithoutinsurancepayafine(13.9%increase),allowingchildrenunder26tobeincludedonparentsinsuranceplan(9.5%increase),andrequiringlargecompaniestoprovidehealthinsurancetotheiremployees(8.4%increase).

Correctunderstandingdecreasedsignificantlyforoneplanelement,thoughslightly.Specifically,fewerpeopleareawareofthefactthatsmallcompaniesthatbuyhealthinsurancefortheiremployeescangetfederaltaxcredits(decreaseof‐3.4%).Correctunderstandingheldsteadyfortheremaining6elements.

Theratesofaccuracywithregardtoelementsnotintheplanshowedevenlessimprovement(seethelastcolumnofTable10).Onlyonesuchelementmanifestedasignificantincreaseinaccuracyfrom2010to2012.Therestshowednochange.Thus,withthepassageoftime,mythsabouttheseelementshavenotbeennotablydiscreditedinthepublic’smind.

Table 9: Change in Accuracy for Policies that are in the ACA Bill between 2010 and 2012 

Element 

% Thinking the Policy Was in the ACA with High 

Certainty Change from 2010 to 2012 

  2010  2012   

Children under 26 can be included in parents’ insurance  42.6%  52.2%  +9.5%*** Large companies have to provide health insurance to employees   30.3%  38.7%  +8.4%*** U.S.  citizens without  health  insurance  have  to  pay fine if they don’t have specific reasons  22.4%  36.3%  +13.9%*** Insurance companies have to sell health insurance to people with preexisting conditions   24.8%  32.6%  +7.8%*** Insurance companies have to continue insurance as long as no rules are broken   27.0%  28.8%  +1.8% 

Make insurance for sale for any American  26.7%  28.7%  +2.0% Prevent limiting amount paid for person’s health care costs  19.5%  23.5%  +3.9%* Discounts on prescriptions to seniors with high drug costs  21.7%  20.0%  ‐1.7% 

17

Federal tax credits for small companies that buy insurance for their employees  20.8%  17.5%  ‐3.4%* Subsidize health insurance for U.S. citizens with low income  19.7%  16.8%  ‐2.9% 

New fees for health insurance companies   11.1%  11.1%  +0.0% 

New fees for companies that make drugs  10.0%  10.2%  +0.2% 

*** p < 0.001, ** p< 0.01, * p< 0.05 

Table 10: Change in Accuracy for Policies that are NOT in the ACA Bill between 2010 and 2012 

Element 

% Thinking the Policy Was NOT in the ACA with High 

Certainty 

Change from 2010 to 2012 

  2010  2012   

Job applicants have to disclose previous illnesses to employer  26.9%  25.6%  ‐1.2% Restaurants with unhealthy food must pay fee to government  19.6%  23.2%  +3.6%* Committees decide whether people get medical care (“death panels”)  17.2%  16.8%  ‐0.4% 

Smokers have to pay additional $1,000 a year  13.9%  14.1%  +0.2% Health care ID card needed to get treatment at hospital  14.5%  13.3%  ‐1.2% 

Require treatment of illegal immigrants for free  12.3%  10.5%  ‐1.8% 

*** p < 0.001, ** p< 0.01, * p< 0.05 

Inlinewiththepreviouslydescribedfindings,peoplegavemorecorrectanswerswithhighcertaintyin2012thanin2010.Table11showsthatonly14.9%ofAmericansanswered9quizquestionscorrectlywithhighconfidencein2010,whereas18.9%didsoin2012.Themediannumberofcorrectanswersofferedwithhighconfidenceincreasedfrom3to4between2010and2012,astatisticallysignificantincreasethatwasaccompaniedbyanon‐significantincreaseinthepercentofpeoplefavoringthehealthcarebill,from29.9%in2010to33.6%in2012

18

Table 11: Number of Correctly Answered Quiz Questions with High Certainty in 2010 and 2012 

  2010    2012 

Number of correctly answered questions  Percent 

Cumulative percent    Percent 

Cumulative percent 

All 18  0.0%  0.0%  0.0%  0.0% 17 out of 18  0.2%  0.2%  0.0%  0.0% 16 out of 18  0.1%  0.2%  0.1%  0.1% 15 out of 18  0.6%  0.9%  0.8%  1.0% 14 out of 18  0.6%  1.5%  1.3%  2.3% 13 out of 18  1.8%  3.4%  1.1%  3.4% 12 out of 18  2.5%  5.8%  3.2%  6.6% 11 out of 18  2.6%  8.5%  2.7%  9.3% 10 out of 18  2.9%  11.4%  4.2%  13.5% 9 out of 18  3.5%  14.9%  5.4%  18.9% 8 out of 18  5.2%  20.1%  5.6%  24.5% 7 out of 18  5.4%  25.4%  5.9%  30.4% 6 out of 18  5.8%  31.3%  6.7%  37.1% 5 out of 18  7.2%  38.4%  5.9%  42.9% 4 out of 18  8.5%  46.9%  7.2%  50.1% 3 out of 18  8.6%  55.5%  8.8%  58.9% 2 out of 18  8.3%  63.8%  8.6%  67.5% 1 out of 18  12.2%  75.9%  10.4%  77.9% 0 out of 18  24.1%  100.0%  22.1%  100.0% Total  100.0%  100.0%  100.0%  100.0% N  1251  1344 

ConclusionsTakentogether,thesefindingsreinforcetwomajorconclusions:

‐ AmericanunderstandingofwhatisandisnotintheACAhasbeenfarfromperfect.

‐ Correctunderstandingoftheelementsofthebillthatweaskedaboutvariedwithpartyidentification:Democratsunderstoodthemost,independentsless,andRepublicansstillless.

‐ Olderpeopleandmoreeducatedpeoplehaveunderstoodtheelementsweaskedaboutbetterthanhaveyoungerandlesseducatedpeople.

‐ Between2010and2012,publicunderstandingofthebilldidnotchangenotably.

19

‐ MostpeoplehavefavoredmostoftheelementsoftheACAthatweexamined,butnoteveryonerecognizedthattheseelementswereallintheplan.

‐ MostpeopleopposedtheelementsweaskedaboutthatwerenotintheACA,butsome

peoplethoughttheseelementswereintheplan.‐ Ifthepublichadperfectunderstandingoftheelementsthatweexamined,the

proportionofAmericanswhofavorthebillmightincreasefromthecurrentlevelof32%to70%.

Takentogether,allthissuggeststhatifeducationeffortsweretocorrectpublicmisunderstandingofthebill,publicevaluationsmightincreaseconsiderablyinfavorability.

20

ReferencesAnand,S.,&Krosnick,J.A.(2003).Theimpactofattitudestowardforeignpolicygoalsonpublic

preferencesamongpresidentialcandidates:Astudyofissuepublicsandtheattentivepublicinthe2000U.S.Presidentialelection.PresidentialStudiesQuarterly,33,31‐71.

Krosnick,J.A.(1988).Theroleofattitudeimportanceinsocialevaluation:Astudyofpolicy

preferences,presidentialcandidateevaluations,andvotingbehavior.JournalofPersonalityandSocialPsychology,55,196‐210.

Krosnick,J.A.(2002).Thecausesofno‐opinionresponsestoattitudemeasuresinsurveys:Theyarerarelywhattheyappeartobe.InR.M.Groves,D.A.Dillman,J.L.Eltinge,&R.J.A.Little(Eds.),Surveynonresponse.NewYork:Wiley.

Petty,R.E.,&Krosnick,J.A.(Eds.).(1995).Attitudestrength:Antecedentsandconsequences.

Hillsdale,NJ:Erlbaum.

21

AppendixA:QuestionWordingsofKnowledgeQuestions

Respondentswereshownthefollowingtextononescreen:

We’dliketofindoutpeople’simpressionsaboutthelawthattheU.S.CongresspassedbackinMarch2010tochangetheU.S.healthcaresysteminmanyways.

We’dliketofindoutpeople’simpressionsaboutwhatthatlawsayswillhappen.

BeforethelawwaspassedbytheCongress,therewasalotoftalkinthenewsaboutthingsthattheplanmightormightnotdo.

Next,youwillreadalistofthesethings,oneatatime.

Mostthethingsyou’llreadweretalkedaboutaspossiblybeinginthelaw.

Butonlysomeofthethingsyou’llreadareactuallyinthelawthattheCongresspassedinMarch2010.

We’dliketolearnyourbestguessesaboutwhichofthesethingsareinthelawandwhicharenot.

Wewillalsoaskyouhowsureyouarethateachansweryougiveiscorrect.

It’sfineifyouaresureofananswerorifyouarenotsureofananswer.Wejustwanttofindoutyourbestguesses.

WewanttoknowwhatpeoplethinkwithoutaskingsomeoneelsefortheanswersandwithoutlookinguptheanswersontheInternetorinanyotherway.Sopleasedonotdoanyofthesethings.Pleasejustgiveusyourbestguesses.

Onaseriesofscreens,respondentswereasked:

Doyouthinkthatthenewlawwillorwillnotdothefollowingafterthelawisfullyineffect?

[STATEMENTAPPEAREDHERE]

Howsureareyouaboutthat?

Theorderoftheitemswasrotatedacrossrespondents.Theitemsare:

RequirethatifaU.S.citizendoesNOThavehealthinsurance,thatpersonwillhavetopayafineonhisorherfederalincometaxesunlessheorsheisallowednottohavetheinsuranceforaseriesofspecificreasons,suchashavingaverylowincome.(INTHEPLAN)

Requirecompanieswith50ormoreemployeestoprovidehealthinsurancetotheiremployeesorpayafinetothefederalgovernmentiftheydonot.(INTHEPLAN)

GivemoneytopayforhealthinsurancetopeoplewhoareU.S.citizensandhaveverylowincomes.(INTHEPLAN)

22

Givefederaltaxcreditstosomeverysmallcompaniesiftheybuyhealthinsurancefortheiremployees.(INTHEPLAN)

Requirecompaniesthatmakedrugstopaynewfeestothefederalgovernmenteachyear.(INTHEPLAN)

Requirecompaniesthatsellhealthinsurancetopaynewfeestothefederalgovernmenteachyear.(INTHEPLAN)

Preventahealthinsurancecompanyfromlimitingtheamountofmoneythatitwillpayforaperson’shealthcarecostsduringhisorherlife.(INTHEPLAN)

RequirehealthinsurancecompaniestosellhealthinsurancetoU.S.citizensandlegalimmigrantswhodon’thavehealthinsuranceandhaveaseriousmedicalproblem.(INTHEPLAN)

Allowyoungadultstogethealthinsurancebybeingincludedintheirparents’healthinsurancepoliciesuntiltheyturn26.(INTHEPLAN)

Requireahealthinsurancecompanytocontinueaperson’shealthinsuranceaslongasheorshepaysforitandhasnotbrokenanyrulesofthehealthinsuranceplan.(INTHEPLAN)

MakehealthinsuranceavailableforsalesothatanyAmericancanbuyifheorshewantsto.(INTHEPLAN)

Providediscountsonprescriptionstoseniorswithhighdrugcosts.(INTHEPLAN)

Requirethatanyoneapplyingforajobmusttelltheemployerifheorshehaseverhadanyseriousdiseases.(NOTINTHEPLAN)

Requirethatfastfoodrestaurantsthatsellunhealthyfoodordrinkstopayafeetothefederalgovernment.(NOTINTHEPLAN)

Requireinsurancecompaniestochargeanadditionalfeeof$1,000yeartoanyonewhobuysinsurancefromthemandsmokescigarettes.(NOTINTHEPLAN)

Createcommitteesofpeoplewhowillreviewthemedicalhistoriesofsomepeopleanddecidewhethertheycangetmedicalcarepaidforbythefederalgovernment.(NOTINTHEPLAN)

RequireeveryAmericantoshowagovernmenthealthcareidentificationcardinordertogetmedicalcareatahospital.(NOTINTHEPLAN)

Requiresomedoctorsandhospitalstotreatillegalimmigrantsfreeofchargeiftheycannotaffordtopay.(NOTINTHEPLAN)

23

AppendixB:SampleDemographics

2010Sample

Table B.1: 2010 Sample Demographic Characteristics 

Unweighted(%)

Target(%)

Weighted(%)

Target–Weighted

Age18‐2930‐4445‐5960+

13.228.331.227.3

22.026.127.824.1

21.226.528.226.1

0.8‐0.4‐0.4‐2.0

GenderMaleFemale

50.749.3

48.351.7

48.351.7

0.00.0

EducationLessthanhighschoolHighschoolSomecollegeBachelorsdegreeorhigher

7.224.430.138.3

13.031.328.027.7

12.031.728.328.0

1.0‐0.4‐0.3‐0.3

Race/EthnicityWhite,Non‐HispanicBlack,Non‐HispanicOther,Non‐HispanicHispanic2+Races,Non‐Hispanic

78.17.83.38.42.5

67.911.55.514.01.1

68.511.65.513.21.2

‐0.6‐0.10.00.8‐0.1

RegionNortheastMidwestSouthWest

18.022.813.228.3

18.421.822.026.1

18.122.021.226.5

0.3‐0.20.8‐0.4

Note:TargetdataaretakenfromtheJuly, 2010, CurrentPopulationSurvey.

24

2012SampleTableB.2:2012 Sample Demographic Characteristics

Unweighted(%)

Target(%)

Weighted(%)

Target–Weighted

Age18‐2930‐4445‐5960+

15.921.628.933.6

21.725.627.425.4

21.325.727.625.4

0.4‐0.2‐0.20.0

GenderMaleFemale

48.951.1

48.251.8

48.251.8

0.00.0

EducationLessthanhighschoolHighschoolSomecollegeBachelorsdegreeorhigher

7.729.529.233.6

12.330.928.728.2

12.230.828.828.3

0.10.1‐0.1‐0.1

Race/EthnicityWhite,Non‐HispanicBlack,Non‐HispanicOther,Non‐HispanicHispanic2+Races,Non‐Hispanic

73.28.85.09.43.6

66.211.56.114.91.3

66.711.56.114.41.3

‐0.50.00.00.50.0

RegionNortheastMidwestSouthWest

19.022.735.023.3

18.321.537.023.2

18.321.537.023.1

‐0.10.00.00.1

IncomeUnder$25,000$25,000‐$49,999$50,000‐$74,999$75,000andabove

15.824.119.141.0

19.323.218.938.6

19.323.219.038.5

0.00.0‐0.10.1

Note:TargetdataaretakenfromtheJuly, 2012,CurrentPopulationSurvey (age,gender,education,race/ethnicity,andregion)andtheMarch,2011,CurrentPopulationSurveySupplement(income).