1

Signals,RetirementOptions,PhasedRetirementandRetirementDecisions

AnnaM.Rappaport,FSA,MAAA

AnnaRappaportConsulting

Presentedatthe2008Retirement20/20ConferenceDefiningtheCharacteristicsofthe21stCenturyRetirementSystem

CrystalGatewayMarriot,Washington,DCNovember1718,2008

Copyright2009bytheSocietyofActuaries.AllrightsreservedbytheSocietyofActuaries.Permissionisgrantedtomakebriefexcerptsforapublishedreview.Permissionisalsograntedtomakelimitednumbersofcopiesofitemsinthismonographforpersonal,internal,classroomorotherinstructionaluse,onconditionthattheforegoingcopyrightnoticeisusedsoastogivereasonablenoticeoftheSocietyscopyright.ThisconsentforfreelimitedcopyingwithoutpriorconsentoftheSocietydoesnotextendtomakingcopiesforgeneraldistribution,foradvertisingorpromotionalpurposes,forinclusioninnewcollectiveworksorforresale.

2

AbstractSuccessinretirementdependsonbuildingenoughresources,notretiringuntilwehaveenoughandmanagingassetswellafterretirement.Thispaperfocusesonwhenandhowweretire.Howandwhentoretireisacomplexdecisioninfluencedbydifferentkindsofsignals.Thispaperfocusesonnewknowledgeaboutsignalsandhowtheyinteractwithretirementdecisions.Thepaperdemonstrateshowtheageatretirementinfluencesretirementsecurity.Theprimaryfocusofthepaperishowindividualsdealwiththesystemtoday,andsystemimprovementisasecondaryfocus.SocietyofActuaries(SOA)andotherresearchisusedtosupportthediscussion.IntroductionManyyearsago,whenIwasworkingtoearnanMBAfromtheUniversityofChicago,Itooklaboreconomicsfromaverywellknownlaboreconomist.AtthetimeIhadsubstantialexperienceworkingasanactuaryincludingseveralyearsofpensionconsulting.Theprofessortriedtoteachme(andtherestoftheclass)thatindividualstakejobsearlyinlifebycalculatingthevalueoftheirfuturewagesandbenefitsincludingpensionsandmaximizingthispresentvalue.WehadrepeateddisagreementsonthistopicasIindicatedthatvirtuallynoonecoulddothis,andthatbenefitsmanagersatourclientscouldnotdoiteither.Weleftattheendofthetermwithouteitheroneofuschangingperspectivesabouthoweconomicanalysisaffectshumanbehavior.Duringthatperiod,retirementhadbecomeawellestablishedpartoftheAmericandream,andmiddleclassAmericansgenerallylookedforwardtoaretirementofleisure.RetirementsecurityforAmericanswasgenerallyassumedandnotasubjectinthenewsverymuch.Thatwasmorethan20yearsago,andinthelastdecade,alothaschanged.Retirementisinthenewsalot,andthereisagreatdealofconcernaboutinsecurityinretirement.Whilemanypeoplewillhaveasecureretirement,manyotherAmericanswillnot;andbeingmiddleclassduringworkingyearsdoesnotensureretirementsecurity.Inpartthisisduetothewayindividualsrespondtoprogramsthatrequireindividualresponsibilityandpersonalsavings.ManyAmericansdonotdowellwiththedecisionsthatareneeded.Thereareemerginginsightsaboutthatresponseasadifferentacademicdisciplinehasbeenappliedtounderstandinghumanbehavioranddecisions.Thisdisciplineofbehavioralfinancefocusesonhowpeoplebehave.RichardH.ThalerandCassR.Sunstein,

3

inNudgeImprovingDecisionsaboutHealth,WealthandHappiness,1provideaverydifferentanalysisofhowindividualsmakedecisions.TheydifferentiatebetweenHumansandEcons.Theyintroducedmetotheformalideaofchoicearchitecture.Inmylaboreconomicsclass,welearnedthatallhumanscanbeexpectedtoactlikeEcons.InNudge,welearnedthatitisnowacceptedthatHumansmakedecisionsinverydifferentways.Behavioralfinancetellsusthat:

Inertiaisaveryimportantfactor,anddefaultoptionsaremuchmorelikelytobechosenthanotheroptions.Defaultsaremorelikelytobechosenwhetherthedecisionistrivialorveryimportant.

Nudgescanbeveryimportantindrivinghumanbehavior. Framingmattersthedesignoftherangeofoptionswillpushpeoplein

differentdirections. Wheretherearetwooptionsandcombinationsareallowed,50percent

50percentisalikelychoice. Wheretherearemanychoices,peopledonotchoosewell.Toomany

choicesarebewilderingtothechooser,andoftenitisdifficulttounderstandtheimplicationsofthealternatives.

Decisionframingandchoicearchitectureinfluencemanydifferenttypesofdecisions.

WehavelearnedfromtheSOAresearchonpostretirementneedsandrisksthatmanypeoplehavemisperceptionsandmisunderstandingsaboutretirementrisk,andthattheytendtothinkshortterm.2Publicconfidenceaboutpersonalretirementsecurityexceedstherealityoftheassetsbeingaccumulatedformany.Wehavelearnedfrombehavioralfinancethatweneedtoconsiderchoicearchitectureandhowwedefineourchoices.Ihavetriedtosynthesizetheideasfromlaboreconomics,behavioralfinance,theSOAandotherresearchonhowpeopleperceiveretirementandwhattheydo.Myconclusionsare:

ThewordretirementisaNudgeorsignal.InmyviewaNudgeisasignalwithapushattachedtoit.

Thewordsearlyretirementgivepeoplepermissiontomoveawayfromworktheydontlike,withoutrequiringthemtodoadetailedanalysisoftheirresourcesandneeds.

1 Yale University Press, 2008 2 Society of Actuaries, Public Misperceptions about Retirement Security, 2004 and Public Misperceptions about Retirement Security, Closing the Gaps, 2007 and Spending and Investing in Retirement: Is There a Strategy? 2006.

4

IfwestartanillustrationanddiscussionaboutSocialSecuritybenefitswiththeageofthelargestmonthlybenefitratherthantheearliestbenefit,wewouldprobablygetdifferentdecisionsabouttheageofretirement.

Planningforretirement,particularlyforthelateryearsandlongerterm,isacomplextaskandveryhardformostpeople.

Itisnotuncommonforpeopletomakedecisionsthatwillaffectthemfor25or30yearsormore,eventhoughtheiranalysisdoesnotfocusonmorethanthenextfiveyears.

ThefactthatpensionplanssetretirementagesataparticularageandthatSocialSecuritydoesnotincreaseretirementagesbeyond67,andhasnotchangedthatprovisionforover20years,createsafixedretirementagemindsetratherthananexpectationthatretirementageswillincreasewithlifespans.

Thesignalsaboutretirementfocusmoreonleisureandincomethanontheneedforincomeandlifetimesecurity.

Thesignalsaboutretirementfocusmoreondreamsthanrisksandriskmanagement.

GoalsandContextforthePaperMypurposeistothinkaboutsuccessfulretirementforindividualsandforsocietyasawhole.Successinretirementdependson:

Buildingenoughresources. Notretiringuntilwehaveenoughsinceworkisoftenpartofretirement,

thewhenisreallywhenandhow. Managingassetswellafterretirement.

Thispaperfocusesonwhenandhowweretire,acomplexsetofdecisionsinfluencedbydifferentkindsofsignals.Newknowledgeaboutsignalsandhowtheyinteractwithretirementdecisionsisexamined.Thepaperdemonstrateshowtheageatretirementinfluencesretirementsecurity.Theprimaryfocusofthepaperishowindividualsdealwiththesystemtoday,andsystemimprovementisasecondaryfocus.Thepaperisorganizedinseveralsections:

Thesignalsaboutretirement. Decisionsaboutretirement. Creatingabetterfuture.

5

WhythisPaperIseeaworldinwhichlifespansareincreasinggradually,andtheretirementsystemischangingalot.Wecanthinkabouttheretirementsystemashavingbothspectacularsuccessesanddefiniteweaknesses.Fromanindividualpointofview,itcanbeseenasasuccessorfailuredependingoncircumstances.Thisworkisprovidedwithintheideathatwecanimprovethesystemandtheresultformanypeople.Mypersonalgoalsaresocietalandfocusonseveraldifferentoutcomes:

Moreindividualsthinkinglongtermanddoingbetterpersonalplanningforretirement,withmoreanalysisaspartofthatplanning.

Financialsystems,publicandprivate,tosupportretirementthatworkswithouttoomuchindividualdecisionmaking.Systemsthatsupportbetterworkoptions,laterretirementandretirementsecurityforasmanyAmericansaspossible.Systemsdesignedtorecognizethatmanyindividualswillnotdogoodlongtermplanning.

BetterworkoptionsforthirdageAmericansastheymovetowardfullexitfromthelaborforce.

Laterretirementandfullexitfromthelaborforcethatgraduallyadjuststochanginglifespans,andhelpsmaketheideaofretirementmoreaffordableformorepeople.

Thispaperonretirementagesandoptionsispartofabroaderefforttolookatsignalsandhowtheyaffecttheretirementsystem.Understandingandmodifyingthesignalsaboutretirementageswillnotbringaboutchangedirectly,butitcanhelpustomeetthesegoalsbyhelpingtochangeexpectationsandNudgeustowardsomewhatdifferentsystems.Signalsareimportantinthinkingaboutthegoalsbecause:

Theycreateexpectationsaboutpersonalchoicesanddrivetheoptionsweconsider,andsometimesoverrideanalyticalplanning.

Theymotivateindividualsandemployerstoconsidercertaintypesofworkoptions,provideideasaboutwhatispossible,andtheycreateexpectationsaboutwhatagoodemployershoulddo.

Theyhaveencouragedusnottothinkaboutthegradualincreaseinlifespansaswemoveforward.

Theyinfluencethestructureoffinancialsystemsusedtosupportretirement.

6

TheSignalsaboutRetirementWecanthinkaboutthesignalssurroundingretirementinseveraldifferentways:

Thewordsdescribingretirement. TheprovisionsoftheSocialSecuritysystemandcommonpensionplans. Theadvertisementsandliteraturethatareinthecommonpress. Rolemodels,peersandfamily. Whatisnotsaid.

WordsDescribingDifferentAspectsofRetirementTherearemanydifferentwaystodescriberetirementandretirementages.Howmanypeopleareconsideredtohavethestatusofretirementwillbedifferentifonemeasuresreceiptofbenefits,laborforceactivityorselfdefinition.Wecanthinkofretirementasreflectingoursourcesofincome,ourstateofmind,ouractivitypatternorsomecombinationofthese.Wecanthinkofretirementalsoasbeingthetransitionpointbetweenaccumulationofassetsversususeofthemtosupplementorreplaceincome.Thesearesomeofthedefinitionsfromthereportsonthe2007RisksandProcessofRetirementSurvey.Theyreflectcarefulcraftingandshowsomeoftheissuessurroundingthemeaningofretirementandretirementages:RetirementThisisgenerallydefinedasexitingfromonesjoboroccupation,typicallyatanageatwhichtheindividualhasnoexpectationsofreturningfulltimetothelaborforce.Forthepurposesofthe2007Survey,retirementisbasedonexitfromaprimaryoccupationortheselfdefinitionofrespondents.RetireeTraditionally,apersonwho,havingattainedacertainageoften,butnotnecessarily,normalretirementagehasleftthelaborforce,withnoexpectationofreturning.Today,manyretireesleavefulltimework,butcontinuewithsomework.Asretirementischanging,thereisnocleardefinitionofretirement,andselfdeclarationofstatusproducesvaryingdefinitionswhenbasedonlaborforceparticipationcriteria.Otherstendtoconsiderthemselvesretirediftheyarecollectingretirementbenefits.Forthepurposesofthe2007Survey,respondentswereconsideredretireesiftheyclassifiedthemselvesasretiredinanemploymentstatusquestionorwereemployed,butrespondedthattheyhadretiredfromaprimaryoccupation.

7

FullRetirementAgeDefinedbySocialSecurityastheageatwhichmonthlyretirementbenefitsareavailableinfullwithoutreductionforearlyretirement.Forbirthcohortsthrough1937,thishasbeenestablishedat65.Forthosebornin1938,fullretirementageforSocialSecurityiscurrentlysetat65yearsandtwomonths,increasinganadditionaltwomonthsforeachsubsequentbirthyear,reachingage66forthosebornfrom1943through1954.Itincreasesagainbytwomonthsforeachsubsequentbirthyearafter1954,reachingage67forthosebornin1960andafter.NormalRetirementAgeFormosttraditionaldefinedbenefitpensionplans,thishasbeenestablishedas65,thesameageatwhichfullSocialSecuritybenefitswereavailableforbirthcohortsthrough1937.Seealso,FullRetirementAge.PhasedRetirementThereisnostandardagreedupondefinitionofphasedretirement.Itisusedtodescribesucharrangementsasworkingparttimebeforeretirementorretiringandthentakinganewjob,becomingselfemployedorworkingonalimitedbasisforaformeremployer.Viewedbroadly,itmayincludereducingonesworkschedulebeforeretirement,changingdutiesatnormalretirementage,workingparttimeinretirementorsomecombinationofapproachestograduallyphaseoutofthelaborforce.Thedefinitionofphasedretirementistypicallylimitedtosituationswhereapartialpensionpaymentisavailableortoasituationwheretheindividualisworkingforthesameemployerasbefore.Theauthordoesnottotallyagreewiththisdefinitionofphasedretirementandwouldincludesituationswheretheindividualisworkingforanewemployer,orphasingdownwithoutpartialpensionpayments.Thisdisagreementservestoillustratethattherearemanynuancesinthesedefinitionsandthereisnoconsensusamongexpertsaboutexactlywhattheymean.Notdefinedinthesurveyreportsisearlyretirement,atermusedinmanypensionplanstodefinebenefiteligibilitypriortonormalretirementage.Usuallythereisabenefitreductiontoreflectthelongerperiodofpayment,butthisisnottrueinallsituations.NotethatexceptfortheSocialSecuritydefinition,whichcalledforagradualincreaseinretirementages,agesaregenerallydefinedasafixedageandtheydonotbuildinanyexpectationthattheywillgraduallyincreaseaslifespansincrease.

8

SignalsfromSocialSecurity,EmployersandEmployerPensionPlansSocialSecuritytellsusthatpeopleareexpectedtoretirebetween62and70.Age62istheearliestretirementageforSocialSecurity,andtherearenofurtherbenefitincreasesorreasonstodelayclaimingafterage70.Ifpeopleevaluatedthedifferencesinbenefitscarefully,theymightconcludethatSocialSecuritybenefitssendasignalthatthereisanadvantageinwaiting,butitappearsthatmanydonotdothis.Themonthlyworkerbenefitin2008foraworkerwithmaximumearningsinallyearsbasedonclaimingagein2008isasfollows:

Maximum2008SocialSecurityBenefitsforaWorkerInitiallyClaimingBenefitsin2008

Claimingage Monthlybenefit62 $168265 $203070 $2794

SocialSecurityalsotellsusthatworkisanacceptablepartofretirement.InthecalendaryearspriortofullSocialSecurityretirementage(about66nowandincreasingto67),itispossibletoworkwhilereceivingbenefitswithearningsofuptotheearningstest,$13,560in2008,withoutanyadjustmentinbenefits,andabovethatamountbenefitsarereduced$2forevery$1earned.Afterthefullbenefitage,SocialSecuritybenefitscanbeclaimedwithoutanyrequirementwithregardtofuturework.Thereisnoadjustmentinbenefitsforcontinuedwork.NotethattheleveloftheSocialSecuritymonthlybenefitisveryimportanttotheeconomicsecurityofretiredmiddleandlowerincomeAmericans.Forfouroutof10olderwomenwholivealone,theirSocialSecuritybenefitisvirtuallyallthattheyhave.Manypeoplefeelthatemployersexpectthemtoleaveastheyreachtheirearly60s.Provisionsofretirementplansaswellasbuyoutsreinforcethisexpectation.Arecentsurveytellsusthatiftheywantpeopletostaylonger,Employersmayjustneedtoask:Manyretireesreporttheywouldhavebeenopentoanapproachfromtheiremployeraskingthemtostaylongerwiththecompany.Sixtyonepercentsaytheywouldhaveviewedtheexperiencepositively.Just10percentindicatetheywouldhavereactednegativelytoanapproachaskingthemtodelaytheirretirement.3 3 EBRI 2008 Recent Retirees Survey, Washington, D.C. This study looked at what would have encouraged recent retirees in the aerospace industry to say longer.

9

Employerdefinedbenefit(DB)planstellusthatweareexpectedtoretirebynormalretirementage.UnlikeSocialSecurity,employerDBplansaredesignedtosignalallornothingretirement,andtheygenerallydonotformallysupportphasedretirement.Mostphasedretirementcomeseitherbyworkingparttimebeforeretirementorreturningtoworkafterretirement,eitheratthesameemployeroranotheremployer.Earlyretirementagealsosignalsthatitistimetothinkaboutretirement.Thenormalformofpayment,lifeincomewithsurvivorbenefitsformarriedparticipants,signalsusaboutthinkingaboutthepensionplanasincomereplacement.Evenplanswithlumpsumsprovidethesesignalssincetheycommunicatebenefitsasincome.Employerdefinedcontribution(DC)planstellusmuchlessaboutwhenorwhetherweareexpectedtoretire.Manyofthempaybenefitsprimarily(andinsomecasesonly)asalumpsum.Theyprovideasignalaboutretirementsavingsandresources,butleavemuchmoreopenexpectationsaboutwhentoretireandabouthowtousefunds.Theymakeplanningfarmoreimportant.AdvertisementsandLiterature:TheNewRetirementCommonadvertisementsforfinancialservices,retirementhomes,vacationsandotherproductsarefullofsignalsandshowretirementasacontinuousvacation,fishingtriporgolfgame.Theyfocusonleisureandshowhappy,healthypeople.Theytendtocreatesignalsandimpressionsaboutleisure.Theyforgetthatvacationisabreakfromwhatwenormallydo,andthatwecannotbeonvacationallofthetime.Peoplewhoretiredreamingofapermanentvacationoftenfindtheyareunpreparedtodealwithanewsetofdailyrealities.Whatwasexcitingasabreakfromworkcaneasilybecomeboringifitisadailyactivity.Recently,expertsfocusedonsuccessfulplanningareincreasingthinkingaboutthreedimensionsofplanningfinancial,healthandactivities.Thefocusonactivitiesconsidersnotjustleisure,butwork,volunteerism,hobbies,familyandbuildingasatisfyinglifeinthethirdageandbeyond.Recently,therehasbeenalotoffocusontheliteratureonhowretirementischanging,andonworkasapartofretirement.Severalbooksprovideterminologyandfocusonthenewretirementdesignedtocreatesignalsaboutretirement.Theauthornotesthatmuchofthisworkfocusesontheearlyyearsofretirement.Someofthisliteratureignorestheneedforriskmanagementandplanningforlimitations

10

Herearesomeexamples:DontRetire,Rewire!byJeriSedlarandRickMiners.Thecoverpromisesusfivestepstofulfillingworkthatfuelsyourpassion,suitsyourpersonalityorfillsyourpockets.PortfolioLife,byDavidCorbett.ThecoverpromisesusThenewpathtowork,purposeandpassionafter50.Encore,byMarcFreedman.Freedmandiscussesworkthatmattersinthesecondhalfoflife,andtellsstoriesofpeopleinencorecareers,whoworkbothformoneyandthepromiseofworkthatmatters.Otherbooksprovidealongertermandmorebalancedapproachlookingatchallengesandrisksaswellasnextsteps.ProjectRenewment,byBerniceBratterandHelenDennis.Thisworkoffersatrailblazingretirementmodelforthegenerationofwomenjustnowleavingtheworkforce.Thesewomenwerethepreboomersandbabyboomerswhopioneeredthenewterritoryofwomenintheworkplaceandworklifebalance.Theauthorsemphasizethepossibilityofpositivechange,enlightenmentandadventure.Theyalsodiscusschallengesandmoredifficulttodiscusstopicssuchascaregiving,lossofaspouseandwhattodowhenyounolongerhaveyouridentity.SmartWomenDontRetire,TheyBreakFree,byGailRentsch.Thisbookdescribeshowwomenarereinventingtheirlivesandcareersinthesecondhalfoflife.TheQuest,fromSteveVernon.ThismaterialisaretirementplanningprogramintheformofaDVD.ThereisadifferenttwistonthistopicinapaperfromtheAustralianInstituteofActuariestitledItstimetoabolishretirement(andhereshowyoudoit)byDarrenWickman.4ThispaperfocusesonthereplacementofretirementbyaLifetimeIncomesystemthatfocusesondisability,notretirement,andalsosupportscyclinginandoutofwork,trainingandleisure.Governmentfunds

4 Paper presented to the Institute of Actuaries of Australia Biennial Convention, September 2007.

11

wouldnotbeusedforincomereplacementforpeoplewhoareabletoworkinthisapproach.Thetypicallifecyclepatternisevolving,butretirementcertainlyremainspartofthelifecycleintheUnitedStates,Canada,Europeandmanyothercountries.Anumberofpeoplehavesuggestedlookingforanewtermfortheevolvingpattern,andsomehavesuggestedreplacingthesystemwithaverydifferentone.Itistheauthorsopinionthatwewillneverreachagreementaboutterminologytoreplacethewordretirementandthatfurther,mostpeoplewontreallybeabletoworkafterage75,sothatwhilewemightretirelater,retirementisunlikelytogoaway.RoleModels,PeersandFamilyAsetofsignalsandexpectationsiscreatedbywhatothersdo.Weareparticularlyinfluencedbyfamily,peersandpubliclydefinedrolemodels.Anextremeexampleofthiswillservetoillustrate.DuringarecenttriptotheDominicanRepublic,atourguideexplainedhowthesuccessofSammySosa,amajorleagueU.S.baseballplayeroriginallyfromthatcountry,hadchangedtheaspirationsofmanyyoungpeople.Formanyofthem,Sosassuccesshadledthemtothinkbeyondtheworldaroundthemandtodreamaboutandreachforfargreatergoalsinlife.Theimpactofpeersondecisionswasafocusatthe2008RetirementResearchCentersConference.Oneofthepresenterssaidthatthisinfluencedhispersonalresearchagendawhenheaskedhismotherhowshedecidedaboutwhentoretire,andsherepliedthatshehadstartedoutbytalkingwithherfriends.Someoftheresearchpresentedwasstructuredtoexplorehowpeersinfluencedindividualsastheymakeretirementdecisions.ThepaperWhoDeterminesWhenYouRetire:PeerEffectsandRetirementfindsIndeed,ourmainfindingisthatindividualdecisionsaboutwhentoretirestronglycorrelatedwiththeretirementtimingdecisionsoftheirpeers,whowedefinedtoberetirementeligiblecoworkerswithinthesameemployer.Theauthorspointoutthatthisiseconomicallylogicalinsomesituationswherethecircumstancesarethesame,butillogicalinotherswherethecircumstancesarenotthesame.5Theysuggestfurtherresearchonpeereffectsandtheirupsidesanddownsides.

5 Chalmers, John M.R., University of Oregon, Johnson, Woodrow T. U.S. Securities and Exchange Commission, and Reuter, Jonathon, Boston College, Who Determines When You Retire? Peer Effects and Retirement, presented at the 2008 Center for Retirement Research Conference.

12

Picturesofwhatwecandoandbeareimportantaswethinkaboutemergingpatternsofthirdageactivity.Thisisthefirsttimeinhistorythatsomanypeoplearelivingtohigherages,andthatpeoplearehealthyandactiveattheseages.Formanypeople,familypatternsbeforethemprovidenohelpinshapingtheirownthirdage.Weneedtofindmorerolemodelstohelpusthinkabouthowtobuildournewlives.Severalofthebooksreferencedaboveareaboutlookingatpersonalrolemodelsandindividualstories.Rolemodelsandexamplesareimportantforindividualsbutalsoforemployersastheythinkaboutphasedretirementandnewjoboptions.Employersseekoutcasestudies:AcasestudythatbridgestheorganizationalandindividualperspectivesisBonSecoursHealthSystem,whichusesthreedifferentapproachestophasedretirement.6

Employeesmayretireatage65butcontinueworkingparttime(notmorethan24hoursperweek)whilecollectingtheirfullpension.

Employeeswhoworkpastage70.5beginreceivingapensioncheckinAprilofthefollowingyear,regardlessofwhethertheycontinuetoworkforBonSecours.

Employeeswhoprefertoretirealtogethermaycomebacktoworkafteranabsenceofatleastthreemonths.Theycontinuetoreceiveretirementbenefitsaftertheyarerehired,regardlessofhowmuchtheywork.

Employeeswhochooseanyoftheoptionscontinuetoearnpensioncreditiftheyworkover1,000hoursinayear.Theirbenefitisthenrecalculatedwhentheystopworking.DawnMalonefromBonSecoursHealthSystemdescribedsomecasestudiesofemployeesandhowtheyhadmovedfromjobtojob.Hereisoneexample:

Inthisnextscenario,welllookatNettiescareer.Shebegannursingin1957.Sheworkedonthreeunits.Shewasoneofthefirstworkingmotherstorequestflexschedulingtoaccommodatechildcareissues.Herhusbandwasinthemilitaryandwasgoneformonthsatatime.Shewasoriginallyhiredtowork3p.m.to11p.m.However,withsmallchildrenathomeandchildcareissues

6 Rappaport, Anna and Young, Mary, Phased Retirement After the Pension Protection Act, The Conference Board, New York, 2007.

13

interferingwithherworkschedule,NettielobbiedtheNursingDirectortoallowhertoflexherschedule.Sheworked7a.m.to7p.m.formanyyears.In1975,shetransferredtoEmployeeHealth.Duringthistime,shealsoworkedPRNevenings,andweekendsontheunits.Thismadeherthefirstemployeeallowedtoworkinmorethanonecostcenteranotherflexschedulingmilestone.In1999,sheretired.TheninJanuary2000,shereturnedtoworkforEmployeeWellness.Amongotherduties,sheperformsTBskintestsonemployees.Shehasgraduallyreducedherhourssinceretirement.Shecurrentlyworkstwodaysperweek.7

Thisandseveralotherexamplesmadephasedretirementmuchmoreunderstandableanddemonstratedhowindividualsusearangeofoptionstoapplyskillsindifferentroles.HiddenSignalsThingsThatAreNotSaidSincethereisalotofinformationprovidedaboutretirementinmanydifferentways,therearealsohiddensignalsthatwecanfindinwhatisnotsaid.ThebiggapsinthemessagesthatIhearmostoftenarethefailuretostresstheimportanceofmakingmoneylastalifetime,theuncertaintyoflifespan,thefactthatsomepeoplewilllivetoage100andtheneedtothinkaboutriskprotection.Thereisquitealotoffocusonhowmuchonecansafelywithdrawfrominvestments,butmuchlessonthechancethatinvestmentswilldeclineinvalueandthattheywillrunout,orbeusedupforlongtermcareexpenses,leavingthecoupleor,moreoften,thewidow,withnothingbutSocialSecurityorSocialSecurityandahouse.Thegapsareaggravatedbecausethereisquitealotofnegativepublicityaboutannuities,muchofitreferringtodeferredvariableannuitiesasaninvestment.However,thedistinctionbetweenimmediateanddeferredannuitiesisoftenmissing,asisthedistinctionbetweenbuyingindividualannuitiesandselectingincomeoptionsinemployersponsoredplans.

7 Several examples are published in Pension Section News, the Society of Actuaries, January 2008, Some Interesting Information About Phased Retirement, by Anna M. Rappaport. Stories based on comments made during Webcast sponsored by The Conference Board in June 2007.

14

DecisionsaboutRetirementSomeoftheMostImportantDecisionsWeWillEverMakeThedecisionsaboutwhentoretireandhowtoorganizethedistributionofourretirementassetsareamongthemostimportantlifedecisionswewillevermake.Arelatedsetofdecisionsincludeswhetherwewillcontinuetoworkonsomebasisaspartoftheretirementperiod.Newinsightsreinforcehowimportantsignalsaretovitaldecisions.Wearestilllearningaboutthebalancebetweeninertia,intuition,signals,economicanalysisandriskpreferencesindecisionmaking.SOAResearch,SpendingandInvestinginRetirementconcludedItcanbearguedthatthedecisionofwhetherandwhentoretireisoneofthemostimportantfinancialdecisionsapersonmakes.Itcaneasilyimpactapersonsfinancialsecurityforaperiodof25yearsormore.Butrelativelyfewfocusgroupparticipantsretiredwiththebenefitofhavingundertakenacarefulanalysistodeterminewhethertheirfundswouldsustainthemthroughouttheretirementperiod.Andnonereportedthattheyretiredafterachievingacalculated(dollar)targetlevelindicatingthattheycouldaffordtoretire.8Thesefocusgroupswithretireeshelpusunderstandmoreaboutthedecisionspeoplemakeattimeofretirement.9

Ithoughtyouweresupposedtoretirewhenyouwere65andIthoughtIwouldtryit.Ididntconsiderhowlongretirementwouldbe.

TheseriesofSOAsurveysonPostRetirementRiskregularlyhasshown,ashasotherwork,thatpeopleexpecttoworktoolderagesthantheretirementagesofcurrentretirees.Twentyeightpercentofthepreretireesinthecurrentsurveyindicatethatretirementdoesnotapplytothem,andtheydonotseemtoexpecttoretire.10Manypeopletellmethattheywillnotretire,andthenwhenIaskthemiftheyplantoworkbeyondage75,theyusuallysayno.Myimpressionis

8 LIMRA and the Society of Actuaries, Spending and Investing in Retirement: Is There a Strategy? 2006, page 6. 9 Quotations from Spending and Investing in Retirement, LIMRA and Society of Actuaries, 2006. Focus groups of individuals from 2 to 10 years after retirement designed to understand decisions made about when to retire, its affordability and how money was invested in retirement. Participants had significant assets in 401(k) or similar plans. 10 2007 Survey of Risks and the Process of the Retirement, Society of Actuaries

15

thatwhenpeoplesaytheydonotplantoretire,thatmeanstheyplantoretirelater.Someofthethingsthatweknowaboutthisdecisionare:

PeoplewithonlyDCplansretirelaterthanthosewhohaveDBplans. ThemajorityofpeopleclaimSocialSecuritybenefitsattheearliest

possibleage.Thosewholivetohigheragescertainlyoverage90wouldhavebeenbetterofftowaitaslongaspossibletoclaimbenefits.

Inrecentyears,commonadviceprovidedbymanyfinancialplannersandemployeesofSocialSecuritywastoclaimbenefitsasearlyaspossible.

Themajorityofrespondents(54%)inarecentsurveyintheaerospaceindustryconsideredtheirretirementtimingfor18monthsorlessbeforetheyretired,and72%consideredthetimingfor2yearsorless.11

Mostoftheretireesinthestudyoftheaerospaceindustrydidnothaveagooddiscussionaboutplanswiththeiremployersbeforetheyretired,andtheywouldhavebeenwillingtostaylongerifaskedandifmadetofeelthattheywerevalued.

Fouroutof10olderwomenwholivealone,themajorityofwhomarewidows,havevirtuallynoincomeotherthanSocialSecurity.

Itisnotuncommonforpeopletoretirewithoutdoingalongertermfinancialanalysis.

However,notallindividualschoosewhentheywillretire.Historicallyaboutfouroutof10peopleretiredearlierthantheychoseto,and,in2008,thishadincreasedtofiveoutof10.12Manyoftheparticipantsinthefocusgroupsdiscussedabovealsodidnotchoosewhentheywouldretirefromtheirlongtermjobs.TheTimingofRetirementandWhyitisImportantWhenonedecidestoretireisextremelyimportantinretirementsecurity.Thereareseveralreasonsforthis:

Medicarestartsatage65,andemployercoverageoftenisnotcontinuedtoretireesbeforeage65.Forpeopleinpoorhealth,itisextremelydifficulttogetindividualhealthcoverageandveryexpensiveifitisavailableatall.

SocialSecuritybenefitsincreasewithclaimingage.Themonthlybenefitatage70willbeabout75%largerthanthebenefitatage62.Whilethe

11 EBRI Issue Brief No. 319, EBRI Recent Retirees Survey: Report of Findings, July 2008, page 4. 12 EBRI Retirement Confidence Study series.

16

shorterperiodofpayoutoffsetsthehigherbenefit,thepresentvaluemaystillincrease,particularlywhenwidowsbenefitsareconsidered.

Forfouroutof10widowsoverage65,SocialSecurityisessentiallytheonlyincomethattheyhave.

Whenonestartstousesavingsatalaterage,thereisashorterperiodduringwhichtheyneedtobeused.Itbecomesmuchmorecosteffectivetobuyanannuityatahigherageaswell.

Whenonestartstousesavingsatalaterage,youhavemoreyearstosaveandmoreyearstoearninvestmentincome.

Addedyearsofworkusuallymeanmoreyearsofemployerprovidedhealthinsurancecoverage,whichisparticularlyimportantforpeoplewhoarenotyetMedicareeligible.

Addedyearsofworkforindividualswhoarecoveredbypensionplansusuallymeanmoreyearsofpensioncredit.Iftraditionalpensionbenefitsarepaidbeforenormalretirementage,themonthlybenefitsareoftenadjustedtoreflectthefactthatbenefitswillbepaidformoreyears.

ResearchonFinancialImpactofDelayingRetirementThesefactorsallinteract.Hereisanexampleconsideringseveralofthefactorstogether.ACongressionalBudgetOffice(CBO)Study,RetirementAgeandtheNeedforSaving,May2004,illustratestheimpactofchoosingretirementatages62,66and70forsinglepersonsandmarriedcouples.Theage62coupleneeds8to11timesannualincomeinassetsinadditiontoSocialSecurityiftheyaretoretireandreplace80percentofincome.Atage70,theamountneededisto2timesincome.MuchofthedifferencestemsfromhigherSocialSecuritybenefits,immediateaccesstoMedicare,feweryearsofretirement,andmoretimetosave.Thisreinforcesthebenefitofconsideringbothtimingandneedsforplanningpurposes.TheCBOstudyemphasizesthefactthatthegapbetweenwhatisprovidedbySocialSecurityandwhatisneededforareasonablelevelofretirementincomeisverylargeatage62,anditgraduallydecreasesovertimeasSocialSecuritybenefitsprovidemoreofwhatisneeded.Thisissueisbeingrecognizedinvariouscircles.AliciaMunnellandStevenA.Sasstellus:Americanswillneedtoworktwotofouryearslongerthantheydoatpresenttoensureareasonablycomfortableoldage.Astheaverageretirement

17

ageformenhasbeen63forthepastquartercentury,theymustnowremainemployeduntiltheirmidtolate60s.13ResearchonPublicKnowledgeaboutDelayingRetirementTheSocietyofActuaries2007RisksandProcessofRetirementSurveyexploredwhatthepublicknowsabouttheimplicationsofdelayingretirement.InresponsetothequestionSupposeyouhadretired/youweretoretirethreeyearslaterthanyouactuallydid/arecurrentlyplanning.Doyouthinkthiswouldmakeyourretirementfinancially.?

Retirees PreretireesAlotmoresecure 15% 14%Alittlemoresecure 30% 54%Nomoresecure 51% 30%Dontknow 4% 3%

Thesurveywentontoaskrespondentshowvariouselementsofchangewouldmakethemmoresecureiftheyretiredthreeyearslater.Resultsareasshowninthefollowingexhibit.

32%59%

15%28%

15%17%

12%17%

11%16%

58%

48%61%

48%54%

40%52%

41%54%

28%14%

47%

16%17%

A lot A little

How much, if at all, would each of the following have increased your financial security in retirement [if you retired three years later]? (Among those providing retirement age from primary occupation; Retirees, n=385; Pre-retirees, n=266)

Source: Society of Actuaries, 2007 Risks and Process of Retirement Survey

(48%)(77%)

(63%)(78%)

(70%)(51%)

(63%)(83%)

Having 3 more years to make contributions to investments

Continuing to receive health insurance from your employer

Relying on your savings for a shorter period of time

Increasing the amount you receive from a traditional pension plan (retirees n=242; pre-retirees n=207)

Increasing the amount you receive from Social Security

Increasing the amount your [surviving] spouse will receive from Social Security (if married)

RetireesPre-retirees

RetireesPre-retirees

RetireesPre-retirees

RetireesPre-retirees

RetireesPre-retirees

RetireesPre-retirees

(63%)(86%)

(52%)(70%)

13Munnell, Alicia H. and Sass, Steven A., Working Longer: The Solution to the Retirement Income Challenge, , Brookings, 2008, page 117.

18

Theresponsesindicatedthatthebiggestexpectedimpactamongpreretireesiscontinuingtoreceivehealthinsurancefromyouremployer,with59percentsayingitwouldincreasesecurityalotand17percentalittle.Theotherresponsesthatgotthemostattentionwerehavingthreemoreyearstomakecontributionstoinvestments,with28percentofpreretireessayingitwouldincreasetheirsecurityalotand58percentalittle;andincreasingtheamountyouwillreceivefromatraditionalpensionplan,with28percentsayingitwillincreaseitalot,and54percentsayingitwillincreaseitalittle.Preretireesfocusedontheimportanceofmaintainingemployerhealthinsurance,buttheydidnotplacealotofimportanceordifferentiateverymuchbetweentheimpactofneedingretirementassetsforashortertimeandgettingmoreSocialSecurityorpensionsatalaterage.Thefocusonhealthcarecoverageisveryrealisticsinceitisverydifficult,expensiveandsometimesimpossibletobuycoverageasanindividualontheopenmarket.Otherresearchhasrepeatedlyshownthataccesstohealthinsuranceisaveryimportantfactorinmanyindividualsdecisionsaboutwhentheywillretireandwhethertoseeknewwork.Theresponseswithregardtowaitingandotherfactorsarenotsurprisingwhenconsideredwithotherthingsthatweknowaboutpublicknowledgeincludingthelackoflongertermplanning,andthefailureofmanypeopletodoacarefulanalysisofthesituation.However,formostpeople(allbutthosewithmajorhealthexpenses),theeconomicimpactofnotneedingtouseretirementresourcesforthreemoreyearswillbemuchgreaterthanthevalueofthreemoreyearsofemployersponsoredhealthinsurance.Iftherespondentshaddoneacarefulanalysisoftheimpactofthealternativefactors,wewouldexpectdifferentresponsestothesurvey.PreretireestendtounderestimatetheimportanceofSocialSecurity,andthesamethingshowsuphere.TheauthornotesthatthereisnoparticularfocushereonthesurvivingspousebenefitfromSocialSecurityeventhoughfouroutof10widowshavevirtuallynothingotherthanSocialSecurity.Anotherquestioninthesurveyfocusedontheawarenessoffinancialchangeonlossofaspouse.Theresultsindicatedrelativelylowawarenessofthelikelydeclineineconomicstatusonwidowhood.

19

WorkPlansandtheDecisionMakingProcessResearchhasrepeatedlyshownthatmanymorepeople(oftensevenoutof10)saytheyplantoworkinretirementthanactuallyworkinretirement.Therefore,workexpectationsarelikelytobeovervaluedinthedecisionmakingprocess.RetirementTimingandExpectationsResearchalsoshowsthatpeopleareretiringearlierthantheyexpectto.The2007RisksandProcessofRetirementSurveyshowedthatonceagainpreretireesexpecttoworklongerthanretireesactuallydid,withtheresultsillustratingdramaticdifferencesbetweenthetwogroupsforcertainresponses.Halfofretireesretiredbeforeage60(52percent),whileonlyonein10preretireescurrentlyintheworkforceexpectstoretirepriortoage60(10percent).Thelargestshareofpreretireesintheworkforcereporttheyplantoretireatorafterage65(36percent),comparedto17percentoftheretireeswhoretiredat65orolder.Amongthoseintheworkforce,nearlythreein10preretireessayretirementdoesnotapplytotheminotherwords,theyexpecttoneverretire(28percent).Theseresultsshowthatsomepreretireeshaveexpectationsaboutworkthatwillnotberealized.Somewillbeforcedtoretireearlierthantheyexpectduetopoorhealth,layoffsorcaregivingresponsibilities.HistoricalfindingsfromtheRetirementConfidenceSurveyillustratethatfourtofivein10workersretirebeforetheyhadexpected,whileresultsfroma2006LIMRAInternationalsurveyindicatethatsixin10retireeshadretiredearlierthanplanned.14

ChangingtheSignals:DirectionsforaBetterFutureSowhatmightwedotochangesignalsinawaythatwillsupportourgoals?

IndexSocialSecurityandprivateplanretirementagesoratleastincreasethemretirementagesthatchangegraduallywithchangesinlifespanswouldcreateverydifferentexpectations.NotethattheAmericanAcademyofActuariesissuedastatementonAug.4,2008callingforanincreaseinSocialSecurityretirementages.

14 Society of Actuaries, Key Findings and Issues: The Phases of Retirement and Planning for the Unexpected, 2007 Risks and Process of Retirement Survey.

20

ShowinformationaboutSocialSecuritybenefitsbystartingwiththeagewherethemonthlybenefitisthelargestratherthantheearliestageatretirement.

DiscussretirementasatimewhenwagesandearningsarereplacedbySocialSecurity,pensionsifanyandtheuseofretirementassetsfocusthoughtaboutretirementintermsofsourcesoffinancialsupport.

Trytobuildacultureofanalysisandimprovefinancialliteracyencourageindividualstodomoreanalyticalworkinretirementplanning.Createsituationswherepeerstalkaboutthisandwherepeergroupsencourageit.Manytoolsarealreadyavailable,andmorearecomingonthemarketregularly.

Balancemessagesaboutleisure,workinginretirement,newretirementwithmessagesaboutrisk,longlifeandtheneedforretirementincome.

Buildandformalizenewandcreativeworkoptionsforthethirdagewhilemanypeopleworkinthethirdageandmanyfindnewwaystowork,therearerelativelyfewformalemployerprogramsforphasedretirementorformalprogramsforcreativeworkoptions.

RemoveandreducebarrierstophasedretirementintheUnitedStates,therearearangeofbarrierstophasedretirement.Theyincludecomplexitiesinpensionlawsanduncertaintiesaboutandbarrierstotherehireofretirees.Thesebarriersshouldbeaddressedsothatphasedretirementcanbecomeanacceptedoption.

Changetheterminologyaboutretirementageswhileitdoesnotseempracticaltogetanentirelynewterm,itissuggestedthatthetermsnormalandearlyretirementarenothelpfulinworkingtowardadifferentworld.

Buildexpectationsaboutemergingpatternsofretirementandworkinthethirdagebuildinformationandcasestudiesaboutwhatbothemployersandindividualsaredoing.

Improvecommunicationsbetweenemployersandpotentialretireessothatbothareawareofsituationswhereitwouldbehelpfulforpeopletoretirelater.

Recognizethelimitationssurroundingfinancialliteracyandincludeappropriatedefaultsinprograms.

PostscriptAlthoughthispaperdealsprimarilywiththeperiodaroundretirementage,thechoicesandresourcesavailableatthattimeareveryinfluencedbywhatwas

21

donetosaveandbuildresourcesfromentryintothelaborforcetotimeofretirement.Theimportanceofsavingandplanningearlyareveryimportantandshouldnotbeforgotten.Themorethatisdoneearly,thebettertheoptionswillbethatareavailableaswereachretirementage.