Publication#2009-21 4301 Connecticut Avenue,NW,Suite350, Washington,DC20008
Phone202-572-6000 Fax202-362-8420
CHILDREN’SACCESSTOHEALTH INSURANCEAND HEALTH STATUSIN WASHINGTONSTATE:INFLUENTIALFACTORS
ByGregory Matthews,B.A., KristinAnderson Moore, Ph.D., andMary Terzian,Ph.D., M.S.W. May 2009
OVERVIEW
Healthinsurance,andespeciallycoverageforchildren,hasbeenasubjectofrecentpoliticaldebatein WashingtonState,1 aswellasonthenationalstage. Policymakersandhealthcareproviderscanuse high-qualitystate-leveldatatoassesswhichchildrenlackhealthinsuranceanddevisepossiblesolutions toaddressthisneed. Illustratingthevalueofthisapproach, ChildTrendsanalyzed2003datafromthe NationalSurveyofChildren’sHealth(NSCH)2forarepresentativesampleofchildrenfromWashington State. Wefindthatchildren’saccesstohealthinsuranceandtheirhealthstatuswerecorrelatedwithso- cialandeconomicfactors,includingrace/ethnicity, familyincome,andparenteducation.
Ouranalysissoughttoanswerthefollowingresearchquestion:
HowdohealthinsuranceandhealthstatusvaryamongWashingtonchildrenaccordingtotheirsocialand economicbackgrounds?
First,weexaminedwhether parentsreportedthattheirchild:
• Hashealthinsurance,andwhether itwas publicorprivate;
• Is ingood, verygood, orexcellentgeneralhealth;
• Is obese;
• Displays internalizingbehaviors(hasanxiousordepressivesymptoms);and
• Hasaspecialhealthcareneed.
Next,weexaminedwhetherthefollowingsocialandeconomicfactorswereassociatedwiththeproportion ofchildrenwiththesecharacteristics:
• Familyincomeatorbelowtwicetheofficialpovertythreshold;
• Childrace/ethnicity;3
• Childage;
• Childgender;
• Familystructure;and
• Highesteducationallevelofanadultinthehousehold.
ABOUTTHEDATASOURCEAND METHODOLOGYUSEDFORTHISBRIEF
The2003NationalSurveyofChildren’sHealth(NSCH)isatelephonesurveythatwasconducted ineverystate,includingWashington. Thestudysampleof1,932householdswasselectedtobe representativeofthestate’s childpopulationduringthatyear.Forthisanalysis,astatisticalproce- durereferredtoasmultipleimputationwasusedto estimatemissing values for the poverty vari- able. Low-income is defined as income below 200 percent of the Federal poverty line.
We conducted logistic regressions, a multi-variable statistical analysis method, to examine the association of each background characteristic with health insurance and health status, while con- trolling for the other background characteristics. Unadjusted associations that remained significant at the p<.05 level, net of these controls, are reported below.
FINDINGS
GeneralHealthInsurance
•Mostchildren inWashington wereinsured in2003.Alittle morethan 93percentofchildreninthe statehadhealthinsurance in2003,slightlyhigherthanthenationalaverageof91percent.However, whetherchildrencurrentlyhadhealthinsurancevariedby severaleconomicandsocialfactors.
•Childreninlow-incomehouseholds weresignificantly lesslikelytohavehealth insuranceatthe timeofthesurvey.As shown inthetopsegmentof theleft-mostandcentralbars inFigure1, about12 percentofchildreninWashingtonfamilieswithincomesthatwerelessthan200percentofthefederal povertyline4 werecurrentlyuninsured,comparedwithabout4percentofchildreninfamilieswith higherincomes.
•Lowparentaleducationallevelswereassociatedwithalowerlikelihoodofachildbeinginsured.
Amongchildreninhouseholdsinwhichnoparenthadahighschooldiploma, only82percenthadin- suranceatthetimeofthesurvey.Inhouseholdsinwhichthehighestparentaleducationalcredential wasahighschooldiploma,89percentofchildrenwereinsured.Ofchildrenin householdswithparents whohadmorethanahighschooleducation,96percenthadinsurance.Theassociationbetween paren- taleducationallevelsandchildren’saccesstohealthinsuranceremainedsignificantaftercontrolling for familyincome,familystructure,andchildage,gender,andrace/ethnicity.
•Wefoundnosignificantdifferencesininsurance statusbyraceorethnicbackgroundaftertherelevant controlswereincludedinthemodel.
Figure1:Childrenin Low-IncomeHouseholds
AreMoreLikelytoBeUninsuredandMoreLikely toRelyonPublic HealthInsurance
100%
80%
60%
40%
20%
0%
47
NoInsurance PrivateInsurance PublicInsurance
LowIncome HigherIncomeTotal
PublicInsurance
•Around 28percent ofchildren inWashington reliedonpublichealthinsurancein2003, as shown inthebottomsegmentoftheright-mostbarinFigure1.Asshowninthebottomsegmentoftheleft andcentralbars,childreninlow-income householdsweremuchmorelikelytohavepublicinsurance (61 percent)thanwerethoseinhouseholds withhigherincomes(9 percent).
• Also morelikelytorelyon publichealthinsurancefor theirchildrenwere:
•Singlemothers(56percent,comparedwith21percentofparentsinhouseholdsheadedbytwo biologicalandadoptiveparents);and
•Parentswithlesseducation(87percentofthosewithlessthanahighschooleducation,com- paredwith54 percentof thosewithhighschooldiplomasand21 percentof thosewithmorethan a highschooleducation).
•Comparedwithwhitechildren,black,Latino,andAsianchildrenwereslightlymorelikelytohave publichealthinsurance,butthedifferencewasnotstatistically significantwhencontrolling forthese otherfactors.
HealthStatus
•Mostchildren werehealthy atthetimeofthesurvey.ThehealthofchildreninWashingtonState wasratedjustslightlyhigherthanthehealthofchildrennationwide. Mostparentsratedtheirchild’s overallhealth asexcellentorverygood(88percentinWashington;84percentnationally;significantly higherinWashington)orgood(10percentinWashington;13percentnationally;significantlylowerin Washington).Only3percentofparentsinWashingtonandtheUnitedStatesasawholesaidthattheir child’shealthwas justfairor poor.5
•However,health statusdiffered forchildren withdifferent ethnicbackgroundsandfamilystruc- tures.Comparedwith parentsofwhitechildren,parentsofLatinochildrenwerelesslikelytorate their children’s healthasexcellent,verygood,orgood.Andchildreninfamiliesheadedbysinglemothers wereratedlowerinoverallhealth,comparedwithchildreninfamiliesheadedbytwobiologicalor adoptive parents.Blackchildren inthesurveyhadlowerhealthratingsthandidtheirwhitecounter- parts,butwedonotknowifthisdifferencewassignificantforallofWashington Statebecausethe numberof blackchildrenwas toosmallfor statisticaltests.
Weight Status
•Washington children betweentheages10and17werelesslikelytobeobesethan werechildren inthenation asawhole.Atthetimeofthesurvey,around11percentofchildreninthestatehada bodymassindexthatwouldindicateobesity,comparedwithanationalaverageof15percentinthis agegroup.6
Internalizing Behavior
Childrenwereratedasdisplayinginternalizingbehaviorproblemsiftheirparentdescribedthemasfeeling worthless,inferior,unhappy,sad,depressed,withdrawn, oruninvolvedwithothers,atleastsomeofthe time.
• Around 13 percent of Washington children displayedinternalizingbehavior.
•Internalizingbehavior problems were related to family structure.Childrenfromtwo-parent, blendedfamilies(thatis,familieswithonestep-parent andonebiologicalparent)weresignificantly morelikelytodisplayinternalizingbehaviorthanwerechildrenwholivedwiththeirbiologicalor adoptiveparents.
•Internalizingbehavior problems didnotvarysignificantlybylow-incomestatus,parentaleduca- tion,race/ethnicity,gender,orage.
SpecialHealthCondition
Childrenwereconsidered ashavingaspecialhealthconditioniftheirparentsreportedthattheirchildhad experiencedanyofthefollowingcircumstances:(a)usedmedicineprescribedbyadoctorforacondition; (b)usedspecialmedicalcare,mentalhealth,oreducational services;(c)hadlimitedabilities;(d)hadun- dergone specialtherapy, suchasphysical,occupational,orspeechtherapy; or(e)hadundergonetreatment orcounseling foranemotional, developmental,orbehavioral problem.Tocountasaspecialhealthcondi- tion,atleastoneof theaboveconditionsmusthavelastedor hadbeenexpectedtolastatleast12 months.
•Around 18percent ofchildren hadaspecialhealth conditionthat hadlastedorwasexpectedto lastatleast12 months.
Bothofthedifferencesreportedbelowarestatisticallysignificant(p<.05),after controllingforbackground characteristics.
•Olderchildrenweresignificantlymorelikelythanwereyoungerchildrentobereportedtohaveaspe- cialhealthcondition.
•Comparedwithchildrenwithbiologicaloradoptiveparents,childrenwithonebiologicalparentand onestep-parentweremorelikelytohaveaspecialhealthcondition.
DISCUSSION
Governmentagencies,suchastheNationalInstitutesofHealthandtheCentersforDiseaseControland
Prevention,havemadetheeliminationofhealthdisparitiesapriorityforbuildingahealthierpopulation.7
However,socialandeconomicdifferencesinaccesstohealthinsuranceandinhealthoutcomespersistfor largenumbersofAmericans.8Resultsofthe analysespresentedinthis ResearchBriefindicatethatdispari- tiesinhealthandaccesstoinsurancestillexistinWashington State. Findingsindicatethatpovertylevel andparentaleducationarestronglyrelatedtowhetherchildrenhavehealthinsurance.Further,our findings indicatethat racial/ethnicbackgroundisassociatedwithhealthoutcomesandinsuranceaccess,butmostof theseassociationsarenotstatisticallysignificantwhendifferencesinsocioeconomic statusareaccounted for.9 Thus,inWashington State,racialdisparitiesinhealthinsuranceandhealthstatusmayreflectdiffer- encesinearnings,familystructure,andparentaleducation.
Thedatausedforthisanalysiswerecollectedbeforethecurrenteconomicdownturn.Censusdatashow thattherateofuninsuredAmericansroseslightlyduringtherecessionof2001.10 Arecentreportfrom WashingtonKIDSCOUNTestimatesthatnearly40,000morechildrenwillliveinpovertyin2009ifun-
employmentinthestatereaches9percent.11Thepresenteconomicclimatewilllikelyreducethenumber offamiliesabletoprovidetheirchildrenwithhealthinsurance andmayhaveanegativeeffectonchild healthoutcomesoverall.
Ontheotherhand,stateandfederalpoliciesalsoaffecthealthlevelsanddisparities.Congressrecently liftedrestrictions ontheStateChildren’sHealthInsuranceProgram(SCHIP)thathadlimitedthestate’s abilitytospenditsallotment. Thispolicychangeisoneofseveralthatcouldreducethenumber ofunin- suredchildreninWashingtonState. Moreresearchwillbeneededtomonitortrendsandascribecause- and-effectlinksbetweenspecificstateorfederalpoliciesandchildhealthdisparities, andtoexaminethe relativeeffectsof theeconomiccrisisandpolicychangeson healthdisparitiesamongchildren.
SUMMARY
Mostchildren inWashingtonStatearehealthyandarecovered byhealthinsurance,butchildren fromcer- tainsocialandeconomicbackgroundsaremorelikelytobeuninsuredortohavepoorhealthoutcomes thanareothers.Childrenfromlower-income andlower-educationhouseholdsweredisproportionatelyun- insuredanddisproportionately relyonpublicinsurance.Latinoparentswerelesslikelytoratetheirchil- dren’shealthasexcellent,verygood,orgood.However,aftercontrolling forotherfactors,therewereno otherstatisticallysignificantdifferencesbyrace/ethnicityineitherhealthoutcomesorhealthinsurance coverage.Childreninmother-onlyfamiliesweremorelikelytouse publicinsuranceandless likelytobein goodhealth. Andchildreninstep-familiesappear tobeatahigher riskthanareotherchildrenforinternal- izingbehaviorsandspecialhealthconditions.
Especiallyinatimewhenthestatebudget isstrained,thesegroupdifferencesmaybeusefulinsettingpri- oritiesfor usingscarcegovernmentresources.
ACKNOWLEDGEMENTS
TheauthorswishtothankRichardWertheimer, Ph.D.,forhishelpfulsuggestions onthisResearchBrief. ThisresearchwasfundedbytheAnnieE.CaseyFoundation. Wethankthemfortheirsupport,butac- knowledgethatthefindings andconclusionspresented inthisreportarethoseoftheauthorsalone,anddo notnecessarilyreflecttheopinionsof theFoundation.
Editor:HarrietJ. Scarupa
REFERENCES
1 SomeWashingtonstatelegislatorsandadvocacyorganizationshaveagoalof completechildhealthinsuranceby 2010. This year’sstatebudgetshortfallhasreducedthestate’scontributiontochildhealthinsuranceprogram,butfederalfundswereused tomakeupthedifference.(
childrens_advocacy_group_march.html)
2 ChildandAdolescentHealthMeasurementInitiative(CAHMI),2003NationalSurveyofChildren'sHealthIndicatorDataSet, DataResourceCenterfor ChildandAdolescentHealth,
3We use theterm“white”torefertonon-Hispanicwhitechildren,“black”torefertonon-Hispanicblackchildren,and Latinoto refertochildrenof Hispanicor Latinoethnicity.
4 Thefederalpovertythresholdiscalculatedbasedonfamilysizeandnumberofchildren.In2003,thefederalpovertythreshold for afamilyof threewithtwo childrenwas $14,824. For afamilyof four withtwo children,itwas $18,660 (
5 Thisstudydidnotanalyzenationaldatawithcontrols,sothesedifferencesarewithoutcontrols.Itispossiblethatsignificant differencesbetweenstateandnationalestimatesof childhealthstatusmaybeduetosocialandeconomicfactors,such as higher incomesor educationlevels.
6 Groupdifferencesinobesitywerenotestimatedatthestatelevelbecausethenumbersinthesecategoriesweretoosmall.
7 U.S. Departmentof HealthandHumanServices.Healthypeople2010:Whatareitsgoals?RetrievedMarch26, 2009, from
8 CentersforDiseaseControlandPrevention.Addressinghealthdisparities.RetrievedFebruary25,2009,from
9 AnexceptionwastheparentratingofgeneralhealthforLatinochildren,whichwassignificantlylesslikelytobegoodorbetter thangood, evenaftercontrollingfor earnings.Also, modelestimatesof generalhealthfor African-Americanchildrencouldnot be estimatedbecausethereweretoofewAfrican-Americanchildreninthesample.
10 SeeFigure6:DeNavas-Walt,C.,Proctor,B.D.,Smith,J.C.U.S.CensusBureau,CurrentPopulationReports,P60-235, Income,Poverty,andHealthInsuranceCoverageintheUnitedStates:2007, U.S. GovernmentPrintingOffice,Washington,DC,
2008.RetrievedMarch26,2009,from
11 Pfingst,L.,Stutman,T.J.,Magarati,M.,Loeb,H.(2009).“ThestateofWashington’schildren:Povertyandthefutureof childrenandfamiliesinWashingtonState.”HumanServicesPolicyCenter,Universityof Washington.Seattle,WA:Washing- tonKIDSCOUNT.RetrievedMarch31,2009,from
ChildTrends isanonprofit,nonpartisanresearchcenterthatstudies childrenatallstagesofdevelopment.Ourmission istoim- prove outcomesforchildrenbyprovidingresearch,data,andanalysistothepeople andinstitutionswhose decisionsandactions affectchildren.Foradditionalinformation onChildTrends,includingpublications availabletodownload,visitourWebsiteat Forthelatest informationonmorethan 100key indicatorsofchild and youthwell-being,visittheChild TrendsDataBankat tionsfor children,visit
©2009 ChildTrends.Maybereprintedwithcitation.