The Age of Retirement - Australian Human Rights Commission · AIST-ACFS Research Project 2013 The...

The Age of Retirement

AIST-ACFSResearchProject2013TheAgeofRetirement

2

An independent report prepared for the Australian Institute of Superannuation Trustees(AIST) by the Australian Centre for Financial Studies. The principal author of this report isProfessorKevinDavis,ResearchDirectorattheAustralianCentreforFinancialStudies

TableofContents

ExecutiveSummary................................................................................................................................3

1.Introduction.......................................................................................................................................6

2. TheAustralianSituation................................................................................................................8

2.1TheHistoryofSocialPerspectivesonRetirementIncomesPolicyinAustralia............................8

2.2TheEconomicandSocialSetting................................................................................................10

2.3RetirementRegulations..............................................................................................................17

2.InternationalContext.......................................................................................................................18

3.TheRetirementDecision..................................................................................................................20

4.ThePost-RetirementExperience......................................................................................................25

5.ConclusionsandPolicyIssues...........................................................................................................27

APPENDIX1:CURRENTAGEPENSIONRATESANDCONDITIONS.................................................30

REFERENCELIST............................................................................................................................31


3

ExecutiveSummaryUnlikesomesystemswhichhaveapublicpensionbasedoncontributionsovertheworkinglifeofanindividual, the Australian Age pension is generally viewed as a (means tested) entitlement,regardlessof yearsof labour forceparticipation.Theexistenceof this universal,means-tested, agepension inherently reflects “fairness” considerations, by ensuring that someminimal level of incomesupportisavailabletoallindividualsbeyondsomeage–atwhichtheymaynolongerbeabletogenerateincomethroughemployment.

Thatdoes,however,influenceperceptionsofwhatis“fairness”.Unlikesomeothercountrieswhereearlyretirement ispermissibleafterreachingacertainnumberofyearsofemploymentandcontributionstothenationalscheme(albeitwithalowerannualpension),theAustraliandebateon“fairness”isfocusedonageper se as thecritical issue.Shifting thedebate to incorporate suchconsiderations ismerited,andbecomesparticularlyimportantwhentherelationshipbetweenthePreservationAge(foraccesstosuperannuation savings) and theOfficial RetirementAge (for eligibility for the agepension) is beingconsidered.

Thecurrentgapbetweenthepreservationageof60yearsandtheofficialretirementageof65yearsenables workers to fund early retirement from super balances, shifting their lifetime leisure –consumption profile (with more leisure prior to the official retirement age, and lower lifetimeconsumption).Totheextentthatthisisapurelyvoluntarychoice,withnoconsequencesforgovernmentbudgets, it is difficult to see why that should be a concern for society. Ongoing, trend increases innationalproductivityand real incomes (with realaverageweeklyearnings increasingbyaround30percent since the start of the millennium) should provide the opportunity for older workers, withconsumptionhabitsformedatearlierageswhenreal incomeswerelower,toelecttotakesomeofthegainsasincreasedleisure.Witharoundathirdofretireesprojectedtoreceivethefullagepensionevenwhen the superannuation system reachesmaturity after 2040, there aremany lower incomeworkerswhosevoluntaryearlyretirementwouldhavelittleconsequencesforthegovernmentbudget.Formanyofthese,financialwealthachievediftheyworkedtilltheofficialretirementagewouldstillbebelowtheassetstestrequirementsforreceiptofthefullpension.Theirearlyretirementhaslimitedconsequencesforthegovernmentbudget–otherthanlossoftaxrevenuefromtheirearlyexitfromtheworkforce.

But,therearetwosetsofcircumstanceswherethisbenignscenariodoesnotapply.Oneisinthecaseof“involuntaryretirement”wherehealthorjoblossconsiderationsleadtoearlyretirement.ABSsurveyfigures on reasons for retirement, and data from the HILDA survey suggest that such involuntaryretirementmaycompriseasignificantfractionofretirees.(Onestudyputstheproportionatover25percent).Forsuch involuntaryretirees, thepost-retirementexperience is lesssatisfyingboth in financialtermsandingeneralwellbeing.Thusthereistheneedfortheretirementagedebatetobecognisantofthe fact that a possibly substantial proportion of retirees have not entered that state voluntarily.Measures to facilitatecontinuedworkforceparticipationofolder laid-offworkers,whoseskill setsmaynotmatchavailablejobopportunities,aswellastoensuresupportforthosewithhealthissues(suchasmaybeprovidedforsomebytheNationalDisabilityInsuranceScheme)arethusimportantissues.

The other caveat iswhereearly retirees run down super balances and thus (by virtue of havinglower assets at the official retirement balance) increase their subsequent pension income– at acost to the government budget. Government retirement incomes policy should, arguably, bedesignednottoprovideincentivestotakesuchactions–althoughagainquestionsofwhatismeantby equity become important. Unfortunately, themeans test acts essentially as a disincentive for


4

continuedworkandaccumulationofretirementsavingsbalances,andisfoundbysomeresearchersto have significant disincentive effects on continued workforce participation. And while thetransition to retirement scheme was established with the sensible objective of enabling earlieraccesstopensionsavingstosupportashifttoparttimeworkandultimateretirement,thequestionneeds to be asked whether it is achieving that objective or whether it has been driven by taxarbitragestrategiesofmorewealthyindividuals.(Changestomaximumsuperannuationcontributionlimitsin2012should,atleast,reducetheopportunitiesforsuchtaxarbitrage).

Butthereisscopeforpoliciestodomoretoprovideincentivesfordeferralofretirementforthosewho wish or are able to do so. There are two types of financial disincentive to postponingretirementage.One is theeffectof themeans testwherebyaccumulationof furtherassets fromsavings out of wage income above reduces subsequent pension entitlements. The other is thedeferral of receipt of pension income,which given a constant expected lifetime (and there is noevidence thatageof retirementaffectsmortality), reduces theamountof lifetimepensionwealthwhich the individual can expect to receive. While there is some small financial incentive forcontinuationofworkbeyond theofficial retirementage (theworkbonus test), it falls far shortofbeing “actuarially fair” – such that deferring retirement has no effect on the expected level oflifetimepensionincometobereceived.

Another important consideration is the observation of a significant increase in the proportion ofhouseholds approaching retirement age who still have outstanding mortgage debt. There isanecdotal evidence of this reflecting decisions to fund higher consumption levels pre-retirement(including providing financial support to other family members for house purchase) with theintention to use superannuation balances to pay down the mortgage. Some part of that higherconsumptionmay alsobe “tradingup” to higher valuedwellings to enjoy the amenity benefits inretirement,influencedbytheexclusionofthefamilyhomefromtheassetstestforthepension.Thispotential cause of distortion to asset allocation, and also disincentive to “downsizing” inretirementwarrantsattention.While(significantly)increasingtheallowableamountfortheassetstestandsimultaneouslyincludingthevalueofthefamilyhomewouldbeonewayofrectifyingsuchdistortions,itwouldrequiresignificantpoliticalwill.

Thisreportalsoraisesthequestionofwhatismeantby“retirement”.Aswellasmeaningdifferentthings to different people, it is also important to recognise that a significant number of retireessubsequently become “unretired”, re-entering the workforce for financial or other reasons. InAustralia, that figure is around 10 per cent of retirees (based on ABS survey data) – which isconsiderablylessthanintheUSA,wheretheretirementsafetynetismuchlessrobust.

The report also examines international evidence ondeterminants of actual retirement decisions.The actual retirement age (which on average is generally below the official retirement age) hadbeen declining internationally until prior to theGlobal Financial Crisis, but is now increasing) isinfluencedbychangesintheofficialretirementage.Thatreflectsbothfinancialconsiderationsbutalso“anchoring”ofperceptionsaboutwhatisthe“normal”retirementage.ButinAustralia,thereissome evidence that the preservation age may be more relevant for influencing expectationsregardingageofretirement.Inexaminingthemotivationsforretirement,itwouldappearthatthefinancial incentive is especially strong for males, who indicate that access to pensions andsuperannuationarethestrongestreasontoretire fromage55. It is lesssofor femalesuntilafter


5

age 65, with reasons such as health and family being stronger at earlier ages. The appropriaterelationshipbetween,andlevelsoftheofficialretirementageandthepreservationage,warrantmuchmore study, andmore considerationalsoneeds tobe given toequity and lifestyle choiceconsiderations than is implied by the current focus on primarily national output and budgetaryconsiderations.


6

1.Introduction

One of themajor challenges for governments in developed economies is the need to ensure thesustainability of retirement systems. Demographic forces such as lower birth rates and increasedlongevityhaveresultedinincreasingdemandsongovernmentstoprovidefortheneedsofanageingpopulation. To reduce the rapidly increasing burden on the public purse, and thus improve thesustainabilityofretirementsystems,governmentsaroundtheworldareimplementingoptionssuchas mandating retirement savings contributions, encouraging voluntary contributions and otherforms of private retirement savings, promoting workforce participation for older workers, andincreasingtheretirementageatwhichindividualsbecomeeligibleforapublicpension.

Similar developments have occurred in Australia. Since 1992, the vast majority of the workingpopulation has participated in a mandated retirement savings plan through the superannuationguarantee contribution (SGC) and received tax incentives to encourage further voluntarycontributionstothescheme.

To enhance the sustainability of the Australian system, policy changes have been implementedwhichwilleventuallyleadtoanincreaseinthepensioneligibilityage-hereafterreferredtoastheofficialretirementage(ORA)-to67forbothmenandwomenbornafterJanuary1st1957.Further,mandatorycontributionratesforsuperannuationarescheduledtoincreasefromthecurrent9percentof employee’swages to12per centby July2019. There is alsoongoingdiscussion regardingincreasing the age (and/or rate of withdrawal) at which individuals can access accumulatedsuperannuationsavings.Thatageofaccessisreferredtoasthe“preservationage”(PA).

Thesedevelopmentsraiseanumberofquestionswhichthisreportseekstoaddress:

• Whatisthemeaningandconsequencesof“retirement”forindividuals?(Box1addressestheissueofthemeaningofretirement)

• WhatroledotheORAandPAplayininfluencingindividual“retirement”decisionsrelativetootherconsiderationssuchashealth,employmentoptions,wealthetc.?

• Whataretheequity/distributionalconsequencesofchangestotheORAandPA?

Answerstothesequestionscanhelptoaddressimportantpolicyquestionssuchas:

• Isthecurrentemphasison“retirement”asawell-definedeventinvolvingexitfromtheworkforceappropriate?Iftheagepensionisviewedasanagerelatedentitlementconditionalonwealthorincomecharacteristics,whatistheappropriatesetofconditionswhichmightinduceoptimalcontinuationintheworkforcewhileinreceiptofsomeagepension,orinduceindividualstoretireatan“optimal”age?

• ShouldtheORAdifferinresponsetoindividualcharacteristicssuchas:lengthofworkforceparticipation;gender;oroccupationalcharacteristicswhicharecorrelatedwitheffectiveworkinglifeandexpectedlifespan?

• WhatshouldbetherelationshipbetweentheORAandPAanddrawdownarrangementsorconstraintsonuseofsuperannuationsavings?


7

• WhatarethepotentialconsequencesforothercomponentsoftheGovernmentbudget(suchashealthanddisabilityexpenditures)fromchangingtheORAandPA?

BOX1:WhatisRetirement?

Oneofthecomplexitiesinanalysingretirementissuesliesindeterminingasuitabledefinitionofretirement.Whiletheofficialretirementage(ORA)isrelativelyeasytodetermine–astheageatwhichanagepensionorothergovernmentbenefitsbecomeavailable,identifyingtheactualageofretirementislessstraightforward.Someindividualsbelowtheofficialretirementagewhoareno longer in the workforce may be treated as retired, but may be part of the “hiddenunemployed” (or may be in de facto retirement due to eligibility for other governmentunemployment or health benefits). Others may regard themselves as “retired” by virtue ofhaving left their long term career occupation, and taken up full or part time employment ofsomedifferenttype(andpossiblydrawingdownaccumulatedsuperannuationorothersavings).Conversely, individuals above the ORA may be employed full or part time, possibly drawingdownsuperannuationorothersavings(withsomeparttimeemployeesinreceiptofsomelevelofagepensionbenefits).

An important consideration is whether retirement is voluntary or “involuntary” – effectivelyforcedupon individuals earlier than expecteddue to health problemsor loss of employment.Thepost-retirementexperienceoftheinvoluntaryretiredcanbeexpectedtodiffersignificantlyfromthosehowretirevoluntarily.

Inpractice,formanyindividuals,“retirement”isnotauniqueevent.USdatasuggeststhatonlyaround half of “retirees” cease work permanently – with around a quarter of retireessubsequently becoming “unretired” by returning to the workforce for some further period(Mestas,2010).InAustralia,surveydatacollectedbytheABSin2007indicatedthattherewere310,000 former retirees in the workforce relative to 3.1million retired individuals. Themainreasonsgiven for returningtoworkwere: financialneed (36%formalesand42%for females)andbored(31%formalesand14%forfemales).TheapparentdifferencebetweentheUSAandAustralian “un-retiring” experience is no doubt influence by differences in the nature of theretirementincomesupportsafetynetbetweenthetwocountries.

Such policy related questions are best considered in the context of available evidence aboutrelationships between labour force participation experience, occupational choice, retirement age,income and wealth, gender, post retirement activities, health, and longevity. Consequently theavailable evidence is considered later in this report. Prior to that, however, to provide context,section1providesanbriefoverviewofthehistoricaldevelopmentofAustralianretirementincomespolicyandofthecurrentAustralianeconomicandsocialsettingwithinwhichthesequestionsneedtobeaddressed,while section2provides some international comparisonsofpolicyarrangementsandtrendsregardingretirementage.

Section 3 examines evidence on individual decision-making regarding retirement and the roles oflabour force experience, effective working life, longevity and health. Section 4 considers howretirement age influences post-retirement experience, includingmortality. Section 5 summarizesandconsiderspolicyimplications.


8

2.TheAustralianSituation2.1TheHistoryofSocialPerspectivesonRetirementIncomesPolicyinAustralia1

The Age pension was introduced in Australia around the start of the 20th century reflecting“increased community acceptance of the notion that support for the aged was not solely theresponsibility of the individual, but a collective responsibility for the whole community” (Dixon,1977). A means test (income and/or assets) has been applied since inception reflecting both a“needs”perspectiveandabudgetarycostperspective,andthefamilyhousewasexcludedfromthetest in the early years of Federation.While initially the eligibility agewas common formales andfemales, itwas reduced in1910 to60 forwomen“on thegrounds thatwomengenerallybecame‘incapacitatedforregularworkatanearlieragethanmen’(Kewley,1973,p75)“(Treasury,2001).In1997, the commitment to keep the singlepensionat25per centof averageweeklyearningswasenshrinedinlegislation,andtherateformarriedindividualssetataslightlylowerlevel.

While there were several attempts to introduce a universal national part-contributory pensionscheme (most recently recommended by the Hancock Report in 1976) this approach was neversupported by government. Ultimately the current superannuation scheme, involving (mandatoryandvoluntary)private,tax–favoured,retirementsavingswasintroducedandoperatesalongsideofthegovernment-provided,budgetfunded,agepensionarrangements.

TheAustralianretirementincomearrangementsmeanthatthereisnoactual,orwidelyperceived,link between an individual’s taxation contributions during their working life and pensionentitlements,unlikeinsomeothernations.Pensionpaymentsarefundedoutofcurrentgovernmentbudget revenues (“pay as you go”). 2 The age pension is generally viewed as a (means tested)entitlement,regardlessofyearsoflabourforceparticipation.

Whileeligibilityagehastraditionallybeendefinedby“yearssincebirth”,recentincreasesreflectingincreased longevity could be interpreted as an implicit shift towards definition by “years ofremaining life expectancy” as discussed by Foreman and Chen (2008). While that could beimplementedat anaggregate level, it raises the issueofwhether individualdifferences shouldbetakenintoaccount.ForemanandChenalsonotethattheofficialretirementagecouldalternativelybedefinedusingotherconcepts.Theseinclude:mortalityrisk(ageatwhichriskofdeathwithinoneyear exceeds some critical level such as 2 per cent); or a constant working life to expectedretirement years ratio (e.g. 3 to 1) based on mortality tables (and other information) such thatindividualscommencingworkatanearlieragecould,ceterisparibus,retireearlier.

It is importanttonotethatattitudestowardsretirementpolicymaybeshapedbytheinstitutionalcontextwithinwhich individuals have been brought up. For example, the French pension systeminvolves eligibility for receipt of a full pension after a specified number of years of employmentrelatedcontributionshavebeenmade. InAustralia,where there isno formaldesignationof somepartoflabourincometaxesascontributionstoaretirementprovision,thereislessattentionpaidtolengthofworkinglifeasaneligibilitycriterion,ratherthanageperse.1ThissectiondrawsheavilyonTheTreasury(2001)andNielson(2010)2Infact,Nielson(2010)reportsthattheChifleyGovernmentin1945establishedaNationalWelfareFund,whichreceivedsomepartofbudgettaxreceipts,aspartofplanningforanationalsuperannuationscheme,butthatthisoperatedsolelyasanaccountingdeviceuntilitsabolitionin1985.


9

Particularly relevant to these considerations is the concept of “fairness”. The existence of a universal,means-tested,agepension,inherentlyreflects“fairness”considerations,byensuringthatsomeminimallevelofsupportisavailabletoallindividualsbeyondsomeage–atwhichtheymaynolongerbeabletogenerate incomethroughemploymentor runningdownofaccumulatedsavings.But, it canbeargued,currentarrangementsmeanthatitisnot“actuariallyfair”,potentiallybiasingretirementdecisions.(SeeBox2)

BOX2:ActuariallyFairPensionArrangements

Arelevantconceptforconsiderationsofalternativepoliciesisan“actuariallyfair”pension.Thisisbasedonthinkingofthepensionasa“lifetimeannuity”which,subjecttosomewealthrelatedeligibilityconditions,shouldbeavailableandofequalexpectedvaluetoallrecipients.Theexpectedvalueforanyindividualwilldependonthenumberofyearsfromcommencementofthepensiontoexpecteddeath.Thus,anindividualwhoworkstoanolderageshouldexpecttoreceivealargerannualpensionthanonewhoretiresearlier.(ThistypeofapproachappliesfortheUSSocialSecuritySystemandinreformsinthe1990stoItalianandSwedishschemes,whichadopted“notionaldefinedcontributionschemes”(Hairaultetal,2010)).Compoundingtheinequity,ifthisisnotthecase,individualswhoretirelaterand/orwhohavehadmoreyearsinemploymentcanbeexpectedtohave,onaverage,contributedmoretothetaxationrevenueneededtosustainthepensionsystem.

AnarrangementofthissortwasintroducedbytheHowardGovernmentinthe1997-98Budget,providingthatthosewho“deferredagepensiontake-upwhilecontinuingtoworkaminimumof25hoursperweek,accruedacumulativetaxexemptbonusentitlementof9.4percentofhisorherbasicpensionentitlementforeachfullyearofemploymentpastpensionqualifyingage(maximum5yearsdeferral)”(Treasury2001).Thisschemewasclosedtonewentrantswhohadnotqualifiedfortheagepensionby20September2009.Itwasreplacedbya“WorkBonus”scheme3wherebyanincreasedallowancewasmadeforincomeearnedincalculatingtheincometestforpensioneligibility.(Only50percentofearnedincomeuptoamaximumof$500perfortnightwasincludedinitially.Recentchangestothiswereannouncedin20124).

The Age pension in Australia is now supplemented by superannuation savings. Institutionalisedemployee superannuation began in Australia in September 1985 when the Australian Council ofTradeUnions (ACTU) sought a three per cent employer contribution as part of itsNationalWageCase claim. This principal was approved in 1986 and subsequently new industrial awards wereprogressivelynegotiatedundertheseguidelines.

Itwasnotuntil1July1992,thatthesuperannuationguaranteewasintroducedrequiringemployerstomaketax-deductiblesuperannuationcontributionsonbehalfoftheiremployees.Fromabaseofthreeper cent,higher levelsof contributionswerephased-inovera ten-yearperiod, reaching themaximum of nine per cent in 2002/03. Mandatory contribution rates for superannuation arescheduledtoincreasefromthecurrent9percentofemployee’swagesto12percentbyJuly2019.

3Accordingtohttp://www.superguide.com.au/how-super-works/goodbye-pension-bonus-hello-work-bonus4http://jennymacklin.fahcsia.gov.au/node/1974


10

2.2TheEconomicandSocialSetting

SincetheintroductionoftheSGC,economicgrowthandtechnologicalchangehaveledtoincreasedaverage income levels (See Figure 1) andwealth accumulation, suggesting scope for better livingstandards in retirement. As Table I indicates, average household real disposable income hasincreased consistently over the past two decades, before declining slightly with the onset of theGlobalFinancialCrisis.

Figure1HouseholdAverageRealDisposableIncome5:1994-2010

Source:ABS,CAT6523.0HouseholdIncomeandIncomeDistribution,Australia-Detailedtables,2009-10

Theseeconomicgainshavebeenoffsettosomeextentbydemographicchangessuchasincreasedlongevity,alowbirthrate6andconsequentdeclineintheworking-agepopulationrelativetothoseofnon-workingage(youngandelderly).

Figure2:AustralianFertilityRate2001-11

Source:ABS,CAT3301.0BirthsAustralia,2011

Withbetterstandardsofliving,betterhealthcareandlowermortalityrates,averagelongevityoverthepastcenturyhasincreasedsignificantly.Lifeexpectancyhasincreasedfromaround47yearsforamaleborninthe1890sto79yearsforonebornin2007-2009.Thenaturalconsequenceofthishasbeenacommensurateincreaseinyearsoflifepost-retirement,withremaininglifeexpectancyatage

5DisposableincomeismeasuredasGrossincomelessincometax,theMedicarelevyandtheMedicarelevysurcharge6Thefertilityraterepresentsthenumberofchildrenanaveragewomanhasduringherlifetime,andthereplacementrateforadevelopedcountryisestimatedtobearound2.1.ThemostrecentstatisticsfromtheABSshowthatAustralia’sfertilityrateisreasonablyflatataround1.88.

$0

$100

$200

$300

$400

$500

$600

$700

$800

$900

$1,000

1.60

1.65

1.70

1.75

1.80

1.85

1.90

1.95

2.00

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

AustralianFertilityRate:2001-2011


11

65increasingfrom11yearsto18yearsoverthesameperiod.Thechallengeforretirementsystems,therefore,istoprovideforindividualsthroughthisextendednon-workingperiod.

Table2Lifeexpectancy(additionalyearsoflife)forpeopleatselectedyearsofage

Males Females At

birthAtage65

Atbirth Atage65

1881–1890 47.2 11.1 50.8 12.31891–1900 51.1 11.3 54.8 12.81901–1910 55.2 11.3 58.8 12.91920–1922 59.2 12 63.3 13.61932–1934 63.5 12.4 67.1 14.21946–1948 66.1 12.3 70.6 14.41953–1955 67.1 12.3 72.8 151965–1967 67.6 12.2 74.2 15.71975–1977 69.6 13.1 76.6 17.11985–1987 72.7 14.6 79.2 18.61995–1997 75.6 16.1 81.3 19.82004–2006 78.7 18.3 83.5 21.52007–2009 79.3 18.7 83.9 21.8

AsFigure3shows, in theearly1970s therewerearound7workers foreachpersonof retirementage,withthisdecliningto5by2011.Whiletheratioofworkingagepopulationtothoseunder15has increased from 2 to around 3.5 over the same period, this portends a slower growth in theworkingagepopulationoverthelongerterm–whentheagedpopulationisincreasing.

Figure3WorkingAge/Non-WorkingAgePopulation,Australia1961-2011

Source:WorldBank,http://data.worldbank.org/country/australia

Anageingpopulationandhigherlongevityofthoseinretirementmeansthat,withrelativelyfewerof working age, meeting retiree needs involves substantial intergenerational income and wealthconsequences. Family arrangements have also been changing, with later family formation andincreasedlongevityleadingtoincreasedageofinheritancesandbequestsfromparents,andfamily

0.001.002.003.004.005.006.007.008.00

1961

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

2007

2009

2011

WorkingAgetoNon-WorkingAgePopulalon WorkingAgetoOver65's

WorkingAgetoUnder15's


12

restructuringsdue todivorce/separationsand remarriage/re-partnering (Hayesetal,2010provideanoverview).Thesechangesinfamilyarrangements,togetherwiththeageingofthepopulationcanimpacttheextenttowhichfamiliescansupportretireesandsupplementretireeincome.Anotherrelevantfactoristechnologicalchangewhichischangingavailableemploymentoptionsandskill requirements with potentially adverse consequences for employment security, income andalternative employment options for many older workers approaching retirement age. Increasingglobalisationandcompetitionfromlowerwageworkers indevelopingcountriescanaggravatetheadverseemploymentandincomeimpactoftechnologicalchangeonworkers(andbusinessowners)inaffectedoccupationsandindustry,eventhoughsocietyasawholecanbenefitfromthesetrends.Figure4showstrends in labour forceparticipationrateswhile figure5showsrecenttrends in theratioofparttimetofulltimeworkfordifferentagegroups.

Participationratesofmalesintheworkforcefortheagegroups55-59years,60to64yearsandover65years showedsomedeclineover thepast twodecadesbuthave returnedover recentyears tosimilarlevelsasinthelate1970s.Femaleparticipationratesfortheagegroups55-59yearsand60to64years,ontheotherhand,haveshownaconsistentincreasesincethelate1980s,withamarkedaccelerationsincethelate1990s.Participationratesoffemalesover65showsomeincreaseinthepostGFCperiod.

Figure4TrendsinLabourForceParticipationRates1992–2011

Male

Source:ABS,Cat6105.0-AustralianLabourMarketStatistics,July2012

0.010.020.030.040.050.060.070.080.090.0

1978

1979

1981

1983

1984

1986

1988

1989

1991

1993

1994

1996

1998

1999

2001

2003

2004

2006

2008

2009

2011

55-59years 60-64years 65yearsandover


13

Female


Astheratiooffulltimetopart-timemaleworkersshowninFigure5indicates,therehasbeensomedeclineoverthepasttwodecadesinthe55to59and60to64agegroups,possibly indicatingon-goingcasualizationoftheworkforce,Theratiooffulltopart-timeworkershaschangedlittleintheover65yearagegroup.Forfemales,theratiooffulltoparttimeworkersincreasedinthe55to59agegroup,stayedrelativelyconstantinthe60to64agegroup,anddeclinedslightlyintheover65agegroup.Given the increasedparticipation ratesacrossall agegroups itmoredifficult todefineanyimpactofcasualizationamongstthefemaleworkforce.

Figure5Ratiooffull-timetopart-timeworkers:1992-2011

Male


0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

1978

1979

1981

1983

1985

1986

1988

1990

1992

1993

1995

1997

1999

2000

2002

2004

2006

2007

2009

2011

55-59years 60-64years 65yearsandover

0.01.02.03.04.05.06.07.08.09.010.0

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

55-59 60-64 65+


14

Female


Hidden within those aggregate changes are substantial shifts in the composition of employmentshown in Figure 6. Strong growth has occurred in the professional services area, in occupationsrequiringhighereducationtraining,andinminingandconstruction,withlittleornogrowthinareassuch as manufacturing, agriculture, and lower skilled occupations. For older workers in manyindustries these structural shifts can imply redundancies and limited opportunities to obtainemploymentingrowthsectors,therebyhastening“retirement”.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

55-59 60-64 65+


15

Figure6CumulativeChangeinIndustryEmployment(000's)Feb2005-Nov2012

Source:ABS,CAT6291.0LabourForce,Australia,Detailed,Quarterly

Also important, as a driver of policy debate and potential policy changes are the budgetaryconsequencesofthedemographictrends.

Evidence of the fiscal strain of government expenditure on the aged, as a proportion of GDP, isshowninTable7.Overtheperiod2002-3to2011-12,governmentexpenditureontheaged,72%ofwhichisfortheagepension,hasincreasedfrom2.4%ofGDPtoaround3.7%ofGDP.


16

Figure7GrowthinAgedAssistanceExpenseComparedtoGDP:2002-2012

Source:TheTreasuryFinalBudgetOutcomes(2001-2012)andABSCat1350.0

AsshowninTable3,theprojectedincreaseinagepensionexpenditures,healthexpenditures,andtax expenditures (from tax concessions associated with superannuation) create sustainabilityproblems for theFederalGovernmentBudget.Projections to2050 indicate that theproportionofthepopulationagedover65yearwill increasefrom14%to23%,andgovernmentexpenditureonhealth,theagepensionandagedcarewill increasetoalmost13%ofGDPfromitspresentlevelof7.5%.ChangingtheORAand/orthePAareamongtheoptionsavailabletothegovernmenttoreducethesemountingpressuresonthebudget.Howsuchchangeswouldimpactuponothergovernmentexpenditures, suchasdisabilitypensionpayments, isan importantconsideration inassessing theireffectiveness.

Table3DemographicProjections

Year 1970 2010 2050(projection)Dependencyratio(workingage/>65years) 7.5 5 2.7Lifeexpectancyatage60 n.a. Males23.4

Females26.6Males29.2

Females31.4TotalPopulation(mill) 12.5 22.2 35.9Percentage65andover 8.3 13.5 22.7GovernmentSpending%ofGDP Health AgePensions AgedCare

4.02.70.8

7.13.91.8

Source:TheTreasury(2010)

Inthesecircumstances,itisappropriatetoreviewtherationaleforcurrentarrangementsprescribingtheageatwhichindividualscanqualifyforgovernmentprovidedage-relatedincomesupportoraccesstopensionsavingswhichhavebeensubjecttofavourabletaxtreatmentintheaccumulationphase.Currentandmootedpolicychangesregardingtheseissueshavebeenlargelybasedon

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

2002-03 2003-2004 2004-2005 2005-2006 2006-2007 2007-2008 2008-2009 2009-2010 2010-2011 2011-12

CumulativeGrowthinAgedPensionExpenditure CumulativeGDPGrowth


17

macroeconomicanddemographicconsiderations,buttherearepotentiallysignificantdistributionalconsequenceswhichalsowarrantattention.

2.3RetirementRegulations

TheAustralian retirement incomespolicy frameworkhas four pillars; ameans tested agepensionproviding a safety net; private, mandatory superannuation savings for employees; voluntarysuperannuationsavings;and,voluntarynon-superannuationsavings(includinghomeownership).AsnotedbyTreasury(2012,p66)“Thesepillarswerenotestablishedonasystematicbasisaspartofagrand design. Rather, each pillar emerged and evolved separately”, and a chronology ofdevelopmentscanbefound inNielson(2010).ForolderAustralians,the interrelationshipbetweenthe age pension system, the private superannuation system, and labour force status is relativelycomplex–ifnotaminefield.Thesecomplexitiesinvolve:

• differencesintheageofentitlementtopensionincomeandageofpermittedaccesstouseofprivatesuperannuationsavings;

• theamountofpensionincomereceivablebeingcontingentupon(iemeanstestedagainst)superannuationwealthandotherassetholdingsandincome(fromthoseassetsorlabourincome);

• policiesenablingsomeaccesstosuperannuationbalanceswhilestillinfullorpart-timeemployment(thetransitiontoretirementscheme)andpartialdiscountingoflabourincomeinthemeanstestcalculationforthosereceiptofanagepension(theWorkBonusscheme).

The various recent policy changes and current arrangements for Australia’s retirement incomesystemhavebeendrivenprimarilybymacroeconomicandgovernmentbudgetaryconsiderations–within thecontextofa long standingpolicy commitment toensuring that retireesareguaranteed(viatheagepension)anincomelevelatleastequalto25percentofaverageweeklyearnings.Andpolicyarrangementsarefarfromsettled.

As well as increases in the ORA, there have also been calls7 for an increase in the age at whichindividuals can generally access their superannuation savings (the preservation age) which iscurrently 60 for those born after July 1, 1964.8 Motivating these calls have been concerns toencourage longer participation in the workforce, thus leading to greater accumulation of privatesuperannuation balances (rather than an earlier running down of such balances via “earlyretirement”) and less reliance on, or deferral of receipt of the age pension as a result ofmeanstesting.Others9havesuggestedthatsomepartofsuperannuationsavings,particularlyifearlyaccessispermitted,berequiredtobeconvertedintoalifetimeannuity(eitherimmediateordeferred)tohelpdealwiththeproblemoflongevityrisk.

But current “retirement age” related arrangements and such policy proposals have significantdistributional consequences – in addition to those arising from the generous tax concessions

7See,forexample,theHenryTaxReviewRetirementIncomeReview(2009,Chapter5)andFSCcallsforsuperpreservationagereview:Savingsgapcouldbereducedby$400billion(Kachor,2012)8Itis55forthosebornbeforeJuly1,1960,andincreasesinastepfashionto60forthosebornafterthatdatebutbeforeJuly1,1964.9SeeReallocatesuperandsecurethefuture(Yeow,2012)


18

provided in the accumulation phase of superannuation. Such consequences can be related todifferences in gender, income, occupational choice, and reflect differences in labour forceparticipationexperience,healthconsequences,andlongevityexpectations.Policysettings(regardingpensions, superannuation, disability, health) also have a significant influence upon individualretirementdecisionsandneedtobecarefullysettoencourage“optimal”retirementdecisions.Sincemuchoftheempiricalevidenceoncausesandconsequencesofretirementageisinternational,itisappropriate to consider the international context of retirement incomespolicies to help interpretthatevidence.

2.InternationalContext

Discussionsregardingretirementsavingspolicyneedtorecognizethevarietyofarrangementswhichexistinternationally.TheWorldBankprovidesaframeworkforanalysisbyidentifyingfivepillarsofretirementincomespolicywhichinteractinvariousways.Thecomponentsrangefromthebasicagepension through to voluntary savings outside of any officially supported pension or retirementsavings scheme. Australia has four of the five described pillars, but does not have a mandated,contributory,publiclymanagedpensionplanscheme(seepillar1below).

Table5:TheFivePillarsofRetirementIncomesPolicy10

PillarNo. Form Objective0 Abasicpensionfrompublic

financesthatmaybeuniversalormeans-tested

Fundingtoprovideaminimumleveloffinancialsecurity

1 Amandatedpublicpensionplanthatispubliclymanagedwithcontributionsand,insomecases,financialreserves

Publicprovisionofadditionalfinancialsecuritytoindividualsbasedontheirlengthofemploymentandleveloftaxcontributionsthroughouttheirworkinglives.

2 Mandatedandfullyfundedoccupationalorpersonalpensionplanswithfinancialassets

Aimstotransferatleastsomepartoftheresponsibilityforretirementincomeprovisionfromthestatetotheindividualviacompulsion.

3 Voluntaryandfullyfundedoccupationalorpersonalpensionplanswithfinancialassets

Toencourageadditionalvoluntaryprivateprovision(andprovidesincentivesfortheself-employedandotherstoundertakeretirementsavings.)

4 Avoluntarysystemoutsidethepensionsystemwithaccesstoarangeoffinancialandnon-financialassetsandsupport

Torecognisetheimportanceofnon-pensionassetslikethefamilyhomeandapersonalinvestmentportfolioinretirementandensuringthatqualityadviceisavailableformanagementoftheseassets.

Manydevelopedcountriesare facingsimilardemographicandbudgetarychallengestoAustralia.11Consequentlytherehasbeenwidespreadpolicyemphasison:

• Increasingthestatepensionage/retirementagetoreflectincreasinglifeexpectancyandtoreducethecostofprovidingpubliclyfinancedpensionbenefits,

10(Mercer,2012).11FormoreinformationseetheMelbourneMercerGlobalPensionIndexReport,Mercer2012


19

• Promoting higher labour force participation amongst the elderly to increase retirementsavingsandreducethelengthofretirement,and

• Incentivestoencouragevoluntarysavings.

SimilartoAustralia,thereisevidenceofadeclineinthe“effective”retirementageofbothmenandwomen (the age at which they effectively exit the workforce)12, until the advent of the GlobalFinancialCrisisin2007,withwomenalsotendingtoretireearlierthanmen.(SeeFigures8and9)

Figure8AverageEffectiveAgeofRetirement-Men:1970-2011

Source:OECDStatistics13

Figure9AverageEffectiveAgeofRetirement-Women:1970-2011

Source:OECDStatistics

AsFigure10suggests,thiswasrelatedtoawidespreadtrendforORAstofallthedecadesfollowingtheSecondWorldWar.Thattrendappearstohavebeenreversedinternationallysincearound2000.Projectionsindicatethatthisupwardtrendwillcontinueintotheforeseeablefuture.

12“Theaverageeffectiveageofretirementisdefinedastheaverageageofexitfromthelabourforceduringa5-yearperiod.Labourforce(net)exitsareestimatedbytakingthedifferenceintheparticipationrateforeach5-yearagegroup(40andover)atthebeginningoftheperiodandtherateforthecorrespondingagegroupaged5-yearsolderattheendoftheperiod.”(OECD,http://www.oecd.org/els/employmentpoliciesanddata/Summary_2011+values3dec2012.xls)13www.oecd.org/els/employmentpoliciesanddata/ageingandemploymentpolicies-statisticsonaverageeffectiveageofretirement.htm

56.0

58.0

60.0

62.0

64.0

66.0

68.0

70.0

72.0

74.0

1970 1980 2000 2011

Japan UnitedStates Australia Denmark Greece

54.0

56.0

58.0

60.0

62.0

64.0

66.0

68.0

70.0

1970 1980 2000 2011

Japan UnitedStates Australia Denmark Greece


20

Figure10OECDOfficialRetirementAge-historyandprojections

Source:OECD(2011)14

3.TheRetirementDecision

The retirement decision is inherently bound up with employment, health and lifestyleconsiderations, and influenced by public policy settings which determine the availability ofgovernment financial supportand thusaffect incentives.Figure11shows themain reasonsstatedforretirement.Whiletherelativeimportanceofemployment,health/familyandretirementincomeeligibility is common for both genders, other determinants not shown on the graph are moreimportantforfemales.Theseincludetimingofretirementtocoincidewiththatofpartner,pursuitofleisure/holiday activities, and other (not specified) reasons, while caring for others was relativelymoreimportantwithinthehealth/familyreasonscategory.

Figure11RetirementReasons:Australia2010-11

Source:AustralianBureauofStatistics:62380DO005_2010201106RetirementandRetirementIntentions,Australia,July2010toJune2011

14OECD(2012),PensionableAgeandLifeExpectancy,1950-2050,PensionsataGlance2011:Retirement-incomeSystemsinOECDandG20Countries.

60

61

62

63

64

65

1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050

Pensionableage(years)

Men Women

0%10%20%30%40%50%60%70%

Lessthan55 55–64 65andover

RefrementReasons,Males:2010-2011

EmploymentRelated Pension/SuperEligibility

HealthFamily

0%

10%

20%

30%

40%

50%

60%

Lessthan55 55–64 65andover

RefrementReasons,Females:2010-11

EmploymentRelated Pension/SuperEligibility

HealthFamily


21

Manyeconomists,suchasGruberandWise (1997),arguethatakeydeterminant forremaining inthelabourforceiswhetheranadditionalyearofworkwillprovidetheindividualwithanincreasedpresent value of retirement wealth. In this regard, the structure of tax and social securityarrangements can be expected to play amajor role in the retirement decision. For example, taxincentives, or increases in the ORA, which encourage longer workforce participation (and saving)increaseretirementwealth,whereasmeanstestingofpensionsisessentiallyataxonpre-retirementearnings and savings and provides a disincentive for continued workforce participation. Thisdisincentiveeffect is compoundedby thedeferralof receiptofavailablebenefits (suchas theagepension–unlessthebenefitisadjustedforthisdeferral).

Inthisregard,WarrenandOguzoglu(2010)findthat“theAustralianretirementsystemdoesprovidean incentivetoretireearly.However,menaremuchmore likely thanwomentorespondtothesefinancial incentives.” This is reflected in the gender difference in relative importance ofsuper/pensioneligibilityasadeterminantofretirementdecisionsshowninFigure11.

ThereisconsiderableinternationalevidencethatretirementdecisionsareinfluencedbychangesintheORA–eventhougheffectiveretirementage(permanentdeparturefromtheworkforce)typicallydiffers from theORA.Wise andGruber (1997) find, in amulti-country study, a strong correlation(dependent on the nature of pension arrangements) between the age at which social securitybenefitsareavailableandageofretirementfromthe labourforce. InGermany,areduction intheofficialagetoreceiveearlyretirementbenefitsin1972from65to60correspondedwithareductionintheaverageageofretirementofwhitecollarworkersby5.5years.(AsimilarexampleisgivenforFrance).In1961,asimilarrevisiontotheallowableageforearlyretirement(atwhichSocialSecuritybenefitscouldbeaccessed)wasimplementedintheUS.Whilethisledtoareductionintheaverageageofearlyretirement,significantincentivesforcontinuedwork(suchasanupwardadjustmentforfuture pension benefits)meant that the reductionwas not as pronounced as that ofGermanyorFrance.

AmorerecentstudybyGruberandWise(2005)involvingsimulationsacross12differentcountriessuggests that increasing the eligibility age for pension payments by three years reduced theproportionofworkers that retire fromtheworkforcewithin fouryearsof theeffectiveretirementagebyanaverageof47%.Theirsimulationsalsoshowthatonaveragethechangeinpolicyresultedin a 27per cent reduction in governmentbenefit payments to the agednet of tax revenues. Thereduction comes from two elements they label as the mechanical effect (the reduction causedpurelybybenefitsnotbeingpaiduntil threeyears later)andthebehaviouraleffect (thereductioncaused primarily through increased tax receipts and contributions to the scheme as a result ofcontinuedworkforceparticipation).

BernalandVermuelen(2012)findthatanincreaseintheofficialageinretirementof65to67intheNetherlandsmay increase theeffective retirementageby (only)around4months (sixmonths formenandtwomonthsforwomen).Thisstudy(likemanyonretirement)onlydiscriminatesbetweenworkingandnot-workinganddoesnot take intoaccount theamountofworkundertaken. In thatregard,HurdandMcGarry (1993) findthat individuals thathavetheability toreducehoursor thelevel of responsibility at a given job as one approaches retirement have a higher effective age ofretirement. This suggests potential benefits from policies that provide incentives for people ofretirementagetotransitiontoretirement.


22

IntheAustraliancontext,thepreservationage(PA)canalsobeexpectedtoplayaroleindecisionsregardingearly retirement. Indeed, Felminghametal (2006) state thatbasedon responses to the2003waveof theHILDAsurvey,over50per centof theagegroupbetween50and60 state thatreachingthePAisanimportantfactorindeterminingretirementintentions.Thisishigherthanthenumber that state reaching theORA is important. Andwithover1/3of superannuationbalancesdrawndownbeforeage65(TaxReview–RetirementIncomeStrategicIssuesPaper)reachingthePAclearlyhasaninfluenceonretirementbehaviour.15

Hardinget al (2009) estimated that aoneoff shift in thepreservationage from55 to6016wouldhave significant effect on the retirement incomes of “generation X” (those who would be agedbetween 54 and 68 in 2029). An induced increase in retirement age and higher superannuationbalancesatage65 (due togreatercontributionsand lowerwithdrawals)wouldmean thatprivateincome in retirement would increase by around 30 per cent and age pension receipts woulddecreasebyaround25percent.

But thereareclearlypotentialdistributionaldifferences in theconsequencesof increasing thePA.For many individuals working till the ORA will still mean that superannuation (and otheraccumulatedwealth)willbebelowtheamountspecifiedfortheassettestforreceiptofthefullagepension. Early retirement, after reaching the PA, means that such individuals (households) enjoyincreasedleisure(inyearspriortotheORA)andavoluntaryreductionintheirlifetimeconsumptionlevel (lesswage income over remaining years to theORA and use of superannuation balances tofundconsumptionpriortotheORAratherthanafter),butwithnoconsequencesfortheamountoflifetimeagepensiontheyreceive.

Eventhoughfinancialincentivesclearlyplayarole,whethertheretirementdecisionisbasedontherationalcalculationsfavouredbyeconomistsisopentodispute.Apartfromtheroleofotherfactorssuchashealthandemploymentopportunities,thereislittleevidencethatmostindividualsactivelyplanahead for retirement.This is illustratedby thesurvey findingsofAgnew,BatemanandThorp(2012) who find “that more than half of Australians in their 50s and 60s have not planned keyaspectsofretirement.”

Furthermore, Australians appear to have optimistic expectations about when they may havesufficientfinancialresourcestobeabletoretire.Figure12,derivedfromHILDAsurveydatashowsthat both males and females generally anticipate retiring well before the ORA, although as theyapproachthatage,realityappearstosetin!

15SomepartofthisdrawdownmayreflectusageofTransitiontoRetirementpensions,introducedonJuly12005whichenableindividualsaged55andovertowithdrawpartofaccumulatedsuperannuationbalanceswhileremainingintheworkforce.Theconversionofearningsonthosesuperbalancesnowindecumulationmodetotax-freestatusforthoseaged60andover,andtheabilitytocontinuecontributingintoanaccumulationaccountprovideataxarbitrage,limitedonlybythesizeofmaximumallowablecontributionrates.

16ThePAhasbeen55,butischangingto60forthosebornafterJuly1,1964


23

Figure12RetirementIntentions2010

Source:ACFS-AIST:Superannuationoverthepastdecade:Individualexperiences,2012

In practice, expected and actual retirement ages differ significantly, due to financial, health, andemployment availability factors. Dorn and Sousa-Poza (2010), for example, examine the extent towhich retirement is “voluntary” or “involuntary” in a study of 19 industrialized countries (notincluding Australia). They find that while involuntary retirement was low (15 per cent or less ofretirees) in a number of countries (USA, Canada, Denmark, Japan, Norway) “involuntary earlyretirementseemstobetheruleratherthantheexceptioninsomecontinentalEuropeancountries”.Highunemploymentratesappeartoincreaseinvoluntaryretirements,asdostrongerjobprotectionpolicies-byreducingwillingnesstohire.

How much retirement in Australia is involuntary. Figure 11, showing reported reasons forretirement, suggests it might be substantial. Barrett and Brzozowski (2010) study “involuntary”retirement inAustralia, drawingon theHILDA surveydata. If involuntary retirement is definedbyretirees stating that they retiredbecause theywere laid-off, left their job formedical reasons, orbecausetheywereself-employedandthebusinessclosed,around27percentofretireeswouldbeclassified as involuntary. This is relatively high by international standards, raising questions aboutwhatcharacteristicsoftheAustralianlabourmarketorhealthcarearrangementsmightaccountforthis finding. Also important for policy consideration iswhether there should be differential policyapproaches to involuntaryandvoluntary retirees–sincethe former face lesssatisfactory financialandwellbeingconditionsinretirement(asdiscussedlater).

Withregardtohealth,Pitetal(2010),confirmtheimportanceofhealthfactorsshowninFigure11.Theyfindfromasurveyofretireesagedover45thatwomenwhohadexperiencedseriousproblems“weretwiceaslikelytohaveretiredearlyduetoill-healthasthosewithoutthesehealthproblems.Thenumberofhealthproblemsassociatedwithearly retirementdue to ill-healthappeared tobeslightly greater for men than for women. .In men and women, the strongest association withretirementduetoill-healthwasinself-reportedhealthstatus.”

Retirement Intentions 2010

60.062.0

64.066.068.0

70.072.0

43 45 47 49 51 53 55 57 59 61 63 65

Age

Inte

nded

Ret

irem

ent A

geMalesFemales


24

Buthealthconsiderationsalsoinfluenceretirementexpectations.Sargent-Coxetal(2012)findusingdata fromtheHILDA (Australia)andHRS (USA)surveys that“Australianworkers reportedyoungerexpectedageofretirementcomparedtotheU.S.sample.ReportingpoorhealthwasmorestronglyassociatedwithyoungerexpectedretirementageintheUnitedStatesthaninAustralia”.“Australianwomeninparticularhavemuchyoungerexpectedagesatretirement…mostlikelytheresultofthecurrentyoungerageofeligibilityforsuperannuationbenefits”,aresultwhichtheauthorsspeculatemaydisappearasORAsaremovedtoequality.

ForemanandChen(2008)argue,fortheUS,that“Allinall,poorhealthisoneofthemostimportantfactors influencing individual decisions to retire, perhaps even more important than financialconsiderations”.

Ontheotherhand, financial incentives throughtheavailabilityofdisabilityschemesmay facilitateearlyretirementduetohealthreasonsformaleswhoretirebeforeage55.WhateffecttheplannedintroductionoftheNationalDisabilityInsuranceScheme,willhaveinthisregardremainstobeseen.Itmayhelp some individuals remain in employment but also give incentives to others to exit theworkforce.

One commonly observed characteristic of labour force experience internationally is thatemployment rates of older, pre-retirement, workers are lower than for other age groups in thework-force (See Figure 13). This has significant implications for the ability of such workers toaccumulatefurtherretirementsavings,butalsoraisesthequestionofwhysuchloweremploymentratesareobserved.

Figure13WorkforceParticipationRatebyAgeBracketComparison

Source:ABS,Cat4125.0-GenderIndicators,Australia,Jan2012,Table1

Thereareanumberofpotentialexplanations.One isthat it isaquasi-voluntarydecisionbyolderworkers who because of age related health reasonsmay be unable to continue working in theirpreviousoccupation,andareabletoaccessgovernmentbenefitsorpersonalsavingstosubsist(SeeFigure 11). A second possible reason is that technological changemay impactmost adversely onemployment opportunities for older age groups whose education experience makes them lessadaptable to new technology. A third possible reason is that proximity to retirementmakes suchworkerslessattractivetoemployersbecauseofperceivedhighercostsoflabourforceturnover.

Inresponsetothislastpoint,onthe23rdofJune2004theAustralianFederalGovernmentbroughtinto forceTheAgeDiscriminationAct; the first federal level legislationagainstagediscrimination.TheActseeksequalityof rights inareas includingemploymentregardlessofage.Oneobjectiveof

76.4 73.4 75 78.4

54.6

12.6

8391.8 91.6 88.8

71.9

24.5

20–24 25–34 35–44 45–54 55–64 65–74

2010-2011Females2010-11 Males2010-11

77.370.8 71.5 71.2

38.3

5.9

85.991.4 91.4 87.8

61.4

15.2

20–24 25–34 35–44 45–54 55–64 65–74

2001-2002Females2001-02 Males2001-02


25

the Act is to respond to Australia’s ageing population by increasing the ease at which olderAustralian’s can participate in the workforce. Between 2004 and 2010, 69.2 per cent of allcomplaintsmade under theAct have been in relationship to employment (TheAustralianHumanRightsCommission,2011).However,despitethislegislationa2012reportpreparedfortheFinancialServicesCouncilsuggeststhatolderworkerscontinuetobetreateddifferentlyandthatemployersneed to better explore the ways in which the skills of older workers can be utilised. (WestfieldWright,2012)

Hairault et al (2010) analyse the possibility that proximity to retirement age is a determinant oflabourforceexperienceofolderworkers.Ifworkforceturnovercostsaresignificant,employerswillprefer to hire workers with a longer expected duration in the job, giving younger workers anadvantageincompetingforjobs.Intheirview,itisnotbiologicalagepersewhichisthekeyvariable,butagerelativetoofficialretirementdate.Thiscouldexplainthepositivecrosscountryrelationshipbetweenworkforce participation rates of olderworkers and official retirement age.Whereas thenormalexplanationforsucharelationshipmightemphasizesupplyof labourfeatures(theneedtokeep working until the official retirement age is in sight) they note a demand side aspect whichsuggeststhatproximitytoretirementisperceivedasanegativecharacteristicbyemployers.

Ultimatelybothdemandandsupplyfactorsarerelevant.UsingpaneldataforFrance,Hairaultetalinvestigate determinants of labour force experience and find that adding distance to expectedretirementasanadditionalexplanatoryvariabletousualexplanatorssuchasage,education,skills,occupationalsector,location,familystatus,improvessignificantlythepredictabilityofparticipationstatusformalesover55.Theyarguethatreducingthetaxoncontinuedlabourforceactivity(suchas by actuarially fair pensions or less generous older age unemployment benefits) has a doubledividend – encouraging postponement of retirement and increasing pre-retirement workforceparticipation.

Therearearangeofotherfactorsaffectingtheretirementdecisionsuchassocio-economicfactors(Skugoretal2012),physical jobdemand(ModrekandCullen2012),psychologicalandbehaviouralbiases (Knoll 2011), and hyperbolic discounting (Knoll 2011). In general these studies find moregenerousretirementbenefitsarelikelytohavealargerimpactontheearlyretirementoflow-skilledworkers compared to high-skilled workers, workers in more demanding jobs retire earlier,individualsdisplay lossaversionratherthanriskaversionsuggestthatindividualsaremorelikelytodiscount potential future retirement benefits gained from delayed retirementmore heavily giventhe potential “lost” years of retirement, and that individuals tend to place an excessive (non-constant) discountondistant events, thereby leading to short-sightednesswhenonemustdecidebetweentheimmediatebenefitsofretirementandlongertermfinancialconsequences.

4.ThePost-RetirementExperience

How does the age and causes of retirement affect post-retirement experience? Does involuntaryretirement have particularly adverse consequences? How does the financial situation of retireesaffectsubsequentwell-being?Istherearelationshipbetweenageofretirementandremaininglifeexpectancy?


26

Involuntary retirement can be expected to lead to a reduction inwellbeing in retirement. This isborneoutbyresearchbyBarrettandBrzozowski(2010)whoanalyseHILDAdatatodeterminethereasons for an observed drop in household consumption after retirement –which is inconsistentwith optimal life-cycle planning. In aggregate, they find “an economically significant decline inexpenditures on groceries and food with retirement… comparable to that found for othercountries.”However, by classifying retirees into voluntary and involuntary retirees, they find that“the observed retirement effect is in fact due to retiring unexpectedly earlier than planned -households that retire as planned report no significant changes in these basic expenditurecategories.”

Itiswellknownthatrelativelyfewretireeswillhaveaccumulatedenoughsuperannuationorothersavings to achieve “comfortable” living standards in retirement – where that is measured byreference to a replacement ratio (post/pre-retirement income) of (say) 60 per cent or an incomelevel needed to achieve some calculated level of consumption. This is particularly true forinvoluntary retirees, but shown also to be relevant for voluntary retirees by various simulationmodellingexerciseswhichcalculateexpectedfinancialwealthattheORA.

Onerelevanttrendinthisregardistheincreasingnumberofretireeswhohaveasignificantlevelof(generallymortgage)debtattheageofretirement.TheKellyReportforCPAAustralia(2012)notes,usingHILDAdata,that30.7%ofthe60-64agegroupin2010hadamortgagecomparedto16.5%in2002.Arangeoffactorsmayberelevanttothatsituation(suchaslowerinterestratesorincreasedborrowings to finance other investments such as investment properties). But there is anecdotalevidence (or widely held suspicions) that such mortgage debt is financing higher pre-retirementspendingthanultimatelysustainable,withtheintentiontousesuperbalancestopay-offthedebtatthe timeof retirement. Suchpre-retirement spending canbe either for own consumption, or cantake the formof transfers to familymembers (suchas childrenunable to independently generatesufficient savings to financehousepurchase).A further factorof relevance is theexclusionof thefamilyhomefromtheassettestfortheagepension,givingincentivesforhouseholdsto“tradeup”to higher value housing to enjoy the associated lifestyle benefits (albeit at the expense of lessdisposable income in retirement). Unfortunately, there is little hard evidence on thesematters,makingitanimportantissueforfurtherstudy.

Turning to mortality issues, the question of how the age of retirement affects subsequent lifeexpectancy is confoundedby the roleofother factors suchashealth,wealth,andsocio-economicstatuswhichaffectmortality.“Avastliteratureshowsthatindividualswithlowincome,lowwealth,low education, or low social status often die younger than those who are better off or bettereducated;andthisistrueformanycountriesandformany(ifnotall)periods.”(Cutleretal,2006).While some part of that relationshipmay involve poor health leading to both lower income andearlier death, there does not appear to be any effect of different behaviours associated withsocioeconomic status. “Healthgradientsby socioeconomic statuspersistevenwhendifferences insmoking,drinking,andotherfactorsaretakenintoaccount.”(Cutleretal,2006).

Kalwij et al (2009) examine the relationship between income and remaining life expectancy atretirementfortheNetherlands.Theynotethatnumerousstudieshavefoundalinkbetweensocio-economic factors andmortality, such that lower income groups have lower life expectancy. Oneconsequenceisthatlowerincomegroupshaveaworsedealintermsofpensionentitlements,due


27

to the lower number of years they can expect to receive such payments. They find that “[t]hedifferenceinremaininglifeexpectancyatage65betweenlowerincomeindividualswhoonlyhaveapublicoldagepensionandhigherincomeindividuals,definedastwotimesmedianincome,isaboutthreeyearsforbothmenandwomen.”

SimilarfindingsarereportedbySullivanandvonWachter(2009)basedonasampleofUSworkers.Theyfindthataverageincomeintheyearsjustpriortoretirement(assumedtobe60)issignificantlynegativelycorrelatedwithremaininglifeexpectancy,andthathighervariabilityofincomealsohasanegativeeffect.

How does age of retirement influence remaining life expectancy? Kuhn et al (2010), review theliteratureandstatethat“Insum,theoverallconsequencesofearlyretirementarenotatallclear”,evenwhenearly retirementduetopoorhealth iscontrolled for.That is reflected in theresultsofvarious studies. Coe and Lindebloom (2008) concur on the lack of clarity from earlier studies.However, for their sample data, they find no adverse effect of early retirement on subsequenthealth, but Kuhn et al find an increase in early post-retirement mortality. As well as mortality,retirementmaybeassociatedwithchanges inhealthorcognitive facilities,althoughthe literaturehereisalsofarfromreachingdefinitiveconclusions.Amongawiderangeofstudies,Ekerdtetalfindthatthereisnosignificantdifferenceinthehealthofretireesandnon-retirees.DeGripetal(2012)find that retirement can have a negative impact on the information processing speed of retireeshoweverotherfacetsofmentalcapacitysuchascognitiveflexibilitymayactuallyimprove.Jokelaetal(2010)findaslightincreaseinbothphysicalandmentalhealthafterbothstatutoryandvoluntaryretirement.

Perhaps the lastwordonthis issueofhowretirementageaffectssubsequentmortalityshouldbeleft toHernaesetal (2012)whostudyhowdifferentialchanges in theORAfordifferentgroupsofindividuals in Norway (thereby enabling the researchers to control for the influence health andincome)affectedmortality.Noapparentrelationshipcouldbefoundleadingthemtoconcludethat“mortality considerations should not have a prominent place in policy considerations of theretirementage”.

5.ConclusionsandPolicyIssues

AmongtheimportantissuesincurrentretirementpolicydebateareappropriatesettingsoftheORAandthePAtodealwithincreasinglongevity,motivatedlargelybytheimplicationsofalargerratioofaged(non-working)toworkingpopulationforthegovernmentbudget.InternationalstudiessuggestthatincreasingtheORAleadstoincreasesinactualretirementage(whicharegenerallylessthantheORA),butnotnecessarilybythesameextent.Thesearisefromboththefinancialeffectsaswellasthebehaviouraleffects–suchasanchoringofperceptionsontheORAasthe“normal”retirementage. The distributional consequences of such a change warrant consideration, and take severalforms.First,thosefromlowerincomeand/orsocio-economicgroupswithlowerexpectedlifespanssuffera larger relative reduction in theirexpected life-timepensionwealth (given the laterageofaccess).Second,thereisnodifferentiationintherelativetreatmentofthosewhohaveenteredtheworkforce at earlier or later ages – unlike in many other countries where universal contributorypension/socialsecurityschemesrelateallowableretirementage(andamountofbenefits) toyearsofemploymentandcontributions.Third, therearesignificant,adverse, implications for thosewho


28

are“involuntary”earlyretireesbecauseofhealthorworkforceconditions–andthusneedtorelyonaccumulatedsavingsorothergovernmentbenefitsforlongerbeforeaccessingtheagepension.

However, the setting of the PA may be a more significant issue than the setting of the ORA –particularlygiventhatthere issomeevidencethat it isamorerelevant influenceuponretirementageexpectations.

The Henry Tax Review, Retirement Income Review, recommended that “The superannuationpreservationageshouldalsobe increased, toalignwith theAgePensionage.” However, itnotedtheconsequencesforsomeinvoluntaryretireesbyarguingthat“Anincreaseinthesuperannuationpreservationage shouldbe coupledwithmechanisms toallowearly access to superannuation forthose above age 60 years who are unable to continue to participate in the workforce due todisability.”(Henry,2009).Whetherthatwouldbeasufficientmechanismtodealwiththosewhoareunabletofindalternativeworkfollowinglay-offandthusconsiderearlyretirementisdebateable.

Whilethereareclearlybudgetaryconsequencesfromearlydrawdownofsuperbalanceswhichleadto increased levels of age pension receipt (through nowmeeting the income and assets test), itshould be asked what proportion of early retirees fall into this category. Projections of futurepensiondependencyimplythataroundathirdofindividualswillcontinuetoreceivethefullpension,whenthesuperannuationsystemreaches“maturity”aroundthemiddleofthiscentury.Forthose,generally lower life-time income groups, a voluntary early retirement decision involving access tosuperannuation savings is a life-style choice involving increased leisure and lower consumption intheremainingyearspriortotheORAaswellaslowerconsumptionposttheORAduetouseofsupersavings in the pre-ORA years. Where accumulated super is sufficiently low that the full pensionwouldhavebeenreceivedanyway,this involvesnoconsequencesforthegovernmentbudget,andthe rationale for changing PA arrangements to prevent this life-style choice is open to question.With the current assets test settings (seeAppendix 1) a single homeownerwould receive the fullpensionwithassets(otherthanthefamilyhome)of$192,500.RunningdownsuchbalancesbyearlyretirementcouldfinancethreetofouryearsofconsumptionpriortotheORAatlevelsequivalenttothat from average weekly earnings, before accessing an unchanged age pension entitlement onreachingtheORA.

An important issue in this regard is equity. Lower incomeworkers are likely tohave spent longertimeintheworkforce(duetolesseryearsineducationandearlierentrytotheworkforce).Shouldequity be interpreted as access to the age pension at the same age for all (and regardless ofexpectedremaininglifetime)oraccesstosomepartofretirementsavingsaftersomegivennumberof years of workforce participation and of contributions? These are issues whichwarrant furtherdebate, recognising that views are often coloured by the particular system with which we arefamiliar. Thenotion that retirementage could varydependingonyearsworkedand contributionsmadeismorereadilyacceptabletothoseincountrieswhereacontributorynationalpensionschemeoperates.

But in that regard, an important feature is onewhich Australia has largely eschewed. That is thelinkingofannualentitlementstotheageofaccess.IncentivestoworkbeyondtheORAarelimited–withagepensionentitlementsunaffectedbyageofaccess. Indeed, there isan implicit taxvia theoperationofmeanstestingarrangements.Theythusdonotmeetthecriteriaof“actuarialfairness”,whichwouldinvolvehigherratesofpensionforlaterretirees.Whilethereissomeincentive(viathe


29

workbonustestwhichreducestheimpactofwageincomeuptosomelevelonmeetingtheincometest for the pension) it is, arguably, ofminor impact and relatively cumbersome administratively.LinkingthesizeoftheagepensionpositivelytoageofaccessinanactuariallyfairmannercouldbeexpectedtoprovidegreaterincentivesforworkforceparticipationbeyondtheORA(eitherfulltimeorparttimesupplementedbydrawingdownofsuperannuationsavings).

The one area in which some such arrangement has been introduced is via the transition toretirement incomeschemewhichallowspeopleabovethepreservationagetoaccessaportionoftheir superannuation balances prior to retirement. These were introduced by the FederalGovernment inJuly2005.TheobjectiveoftheschemewastoprovideolderAustralianswithmoreflexibility in planning for retirement and to encourage participation in part-time work prior toretirement. (Commonwealth of Australia, 2004). However, the transition to retirement incomeschemeappears tohavebeenwidelyusedasa tax-effectivemethod forboosting superannuationsavings prior to retirement. The extent to which this strategy can be utilised was significantlyreducedonthe1stof July2012whenthemaximumconcessionalcontributioncapwasreducedto$25,000forallindividuals.Butformanylowerincomeworkers,aslongasthePAremainsbelowtheORA,theschemeprovidesamechanismforfacilitatingearlypartialretirement,whichispotentiallyquitevaluableforthose“involuntary”retireesunabletomaintainfulltimeemploymentforhealthorjobavailabilityreasons.

Finally, the age of retirement debate needs to take into account not just the budgetaryconsequences of the ageing population, but also the consequences of significant increases inproductivityandrealincomesovertime,andtheopportunitythisprovidesforanenhancedvarietyofwork-leisurechoicesoverthelifetime.Sincethestartofthemillennium,averagerealhouseholddisposable income has increased by over 30 per cent from around $600 to $850 (see Figure 1).Similarlyaverageweeklyearnings,towhichthepensionislinked(at25percentofthatfigureforthesingle pension), have increased in real terms by around 30 per cent over that period. For olderindividualswhoseconsumptionhabitshavebeendevelopedoverseveraldecades,earlyretirementprovidesanopportunitytotakesomeoftheproductivitygainsintheformofincreasedleisure.Thatisnotnecessarilyabadsocialoutcome.


30

APPENDIX1:CURRENTAGEPENSIONRATESANDCONDITIONS

AsofMarch2013,themaximumfortnightlyagepensionpaymentis$712forasingleindividualand$536.70 for a couple (See table 4). However, according to projections made by the ProductivityCommissionin200817,only55percentofallretireesreceiveafullpension.Anumberoftestsareperformedtofirstdetermineeligibilityfortheagedpensionandthentodeterminewhatpercentageofthemaximumrateanindividualisentitledto.Thefirstcriteriatodetermineeligibilityiswhethertheindividualhasreachedtheofficialageofretirement.Theofficialageofretirementforbothmenand women is currently 65. However, in accordance with the increasing life expectancy ofAustralians,thiswillbegraduallyincreasedto67bythe1stofJuly2023.Thisincreasewilltakeplaceover6yearsbeginninginJuly2017.

Oncean individualhasreachedtheofficial rangeof retirement, twoadditional testsknownas theincometestandtheassettestareundertaken.Thefortnightlyincomeandtotalassetsofapotentialagedpensionrecipientarecomparedto thereduction free thresholds. Ifanapplicant’s fortnightlyincomeorassessableassetsexceedthesethresholds,therateofpaymentisgraduallyreducedby40centsforeachdollaroverthefortnightlyincomethresholdandby$1.50forevery$1000abovetheasset threshold. It is important tonote thatwhilemostassetsare consideredassessableassets, amajor exception is the family home. There has been some suggestion that the exception of thefamilyhomefromtheassettestmaybeoneimportantreasonforthefamilyhomemakingupsuchalargeproportionoftheaverageAustralian’stotalwealth.18

Table4:AgePensionMeansTestCriteria

Incometest Assettest

FreethresholdPartpayment

cut-offFree

thresholdPartpayment

cut-off

$perfortnight $perfortnight $ $

HomeownerSingle 152 1,697.20 192,500 707,750

HomeownerCouple 268 2,597.60 273,000 1,050,000

Non-homeownerSingle 152 1,697.20 332,000 847,250

Non-homeownerCouple 268 2,597.60 412,500 1,189,500

MaximumRatesofPayment Single Couple

Fortnightly $712.00 $536.70(each)

PerAnnum $18,512 $27,908.4(combined)

Source:Centrelink(2012)

17ProductivityCommission(2008).TrendsinAgedCareServices:someimplications.18ABS,CAT6554HouseholdWealthandWealthDistribution,Australia,2009-10


31

REFERENCELIST

AgnewJ,BatemanHandThorpS(2012).Work,money,lifestyle:PlansofAustralianretirees’,CentrefortheStudyofChoice,WorkingPaperSeriesNo.12-006.AustralianHumanRightsCommission(2011).TheRoadSoFar:theAgeDiscriminationAct2004(Cth).Retrievedfromhttp://humanrights.gov.au/age/publications/ADA_road_2011.pdf

BarrettGandBrzozowskiM(2010).InvoluntaryRetirementandtheResolutionoftheRetirement-ConsumptionPuzzle:EvidencefromAustralia,UniversityofSydney,EconomicsWorkingPaperSeries2010-10

BernalNandVermuelenF(2012).Betweendreamandreality:Theimpactofanincreaseinthelegalretirementageontheeffectiveretirementage.PaperpresentedattheNetsparPensionDayEvent,9November2012.

Borowski,A.(2008).Backatthecrossroads:TheslipperyfishofAustralianretirementincomepolicy.AustralianJournalofSocialIssues,43,311–334.CoeNandLindebloomM(2008).DoesRetirementKillYou?EvidencefromEarlyRetirementWindows,CentERDiscussionPaperSeriesNo.2008-93CommonwealthofAustralia(2004).Amoreflexibleandadaptableretirementincomesystem.Retrievedfromhttp://demographics.treasury.gov.au/content/_download/flexible_retirement_income_system/flexible_retirement_income_system.pdf

CommonwealthofAustralia.(2010).Australiato2050:Futurechallenges:the2010intergenerationalreport,CommonwealthofAustralia.Retrievedfromhttp://archive.treasury.gov.au/igr/igr2010/report/pdf/IGR_2010.pdfCutlerD,DeatonAandLleras-MuneyA(2006).TheDeterminantsofMortality,JournalofEconomicPerspectives20:3,97–120

DavidsonP(2009).WhenRaisingRetirementAges,Don'tStartwiththePoorest[online].Impact,Winter2009:14.

DeGripA,DupuyA,JollesJandvanBoxtelM(2012).RetirementandCognitiveDevelopment:AretheRetiredReallyInactive?,ResearchCentreforEducationandtheLabourMarket,ROAResearchMemorandum

Dixon,John(1977).Australia’sPolicyTowardstheAged:1890-1972,CanberraCollegeofAdvancedEducation,Canberra.

DornDandSousa-PozaA(2010).Voluntary’and‘involuntary’earlyretirement:aninternationalanalysis,AppliedEconomics,42:427–438

DubnerS(2012).RetirementKills:aNewMarketplacePodcast,Freakonomics.Retrievedfromhttp://www.freakonomics.com/2012/05/17/retirement-kills-a-new-marketplace-podcast/


32

EkerdtD,BadenL,BosseRandDibbsE(1983).Theeffectofretirementonphysicalhealth,JournalofPublicHealth73(7)779-783FormanJandChenY(2008).OptimalRetirementAge,Presentedatthe2008Retirement20/20ConferenceDefiningtheCharacteristicsofthe21stCenturyRetirementSystem,WashingtonDC,November17-18,2008

GruberJandWiseD(1997).SocialSecurityProgramsandRetirementaroundtheWorld,NationalBureauofEconomicResearch,WorkingPaper6134

GruberJandWiseD(1998).Socialsecurityandretirement:Aninternationalcomparison.TheAmericanEconomicReview,88,158–163.Gruber J and Wise D (2005). Social Security Programs and Retirement around the World: FiscalImplications,NationalBureauofEconomicResearch,WorkingPaper11290.Hairault J, Langot, F and Sopraseuth T (2010). Distance to Retirement and Older Workers’Employment: The Case For Delaying the Retirement Age, Journal of the European EconomicAssociation8,5(2010)1034-1076HardingA,KellyS,PercivalRandKeeganM(2009).PopulationAgeingandGovernmentAgePensionOutlays:UsingMicrosimulationModelstoInformPolicyMaking,ESRIInternationalCollaborationProject.Retrievedfromhttp://www.esri.go.jp/jp/workshop/0907/a_02_01.pdfHayesA,WestonR,QuLandGrayM(2010).Familiesthenandnow:1980-2010,AustralianInstituteofFamilyStudies,FactsSheet,2010.Retrievedfromhttp://www.aifs.gov.au/institute/pubs/factssheets/fs2010conf/fs2010conf.html

HenryK(2009).RetirementIncomeStrategicIssuesPaper,CommonwealthofAustralia.Retrievedfromhttp://www.taxreview.treasury.gov.au/content/StrategicPaper.aspx?doc=html/publications/Papers/Retirement_Income_Strategic_Issues_Paper/chapter_5.htm

HernaesE,MarkussenS,PiggottJandVestadO(2012).DoesRetirementAgeImpactMortality?,DepartmentofEconomics,UniversityofOslo,MemorandumNo19/2012

HoldenK(1988).PhysicallyDemandingOccupations,HealthandWorkAfterRetirement:FindingsfromtheNewBeneficiarySurvey,SocialSecurityBulletin,Vol.51(11)

HolzmannRandHinzR(2005).OldAgeIncomeSupportinthe21stCentury,TheWorldBank,WashingtonDC.

HumpelN,O’LoughlinK,WellsYandKendigH.(2009).AgeingbabyboomersinAustralia:Evidenceinformingactionsforbetterretirement.AustralianJournalofSocialIssues,44,399–415.

HurdMandMcGarryK(1993).TheRelationshipBetweenJobCharacteristicsandRetirement,NBERWorkingPaperSeries,WorkingPaperNo.4558


33

JacksonN,WalterM,FelminghamBandSpinazeA(2006).Willolderworkerschangetheirretirementplansinlinewithgovernmentthinking?Areviewofrecentliteratureonretirementintentions.AustralianBulletinofLabour,32(4),315.JacksonR,HoweNandNakashimaK(2010).TheGlobalAgingPreparednessIndex,TheCentreforStrategicandInternationalStudies.

JokelaM,FerrieJ,GimenoD,ChandolaT,ShipleyM,HeadJ,VahteraJ,WesterlundH,MarmotMandKivimakiM(2010).FromMidlifetoEarlyOldAge:HealthTrajectoriesAssociatedwithRetirement,Epidemiology,Volume21,Number23

KanchorK(2012).FSCcallsforsuperpreservationagereview:Savingsgapcouldbereducedby$400billion,InvestorDaily,August2012.

KalwijA,AlessieRandKnoefM(2009).IndividualIncomeandRemainingLifeExpectancyattheStatutoryRetirementAgeof65intheNetherlands,NetsparDiscussionPaper09/2009–033.Retrievedfromhttp://media.leidenuniv.nl/legacy/voordr-2010-02-individual-income-and-remaining-life-expectancy-dp092009-033.pdf

KaramP,MuirD,PereiraJandTuladharA(2011).BeyondRetirees,InternationalMonetaryFund,FinanceandDevelopment,Vol48-2.

KnollM(2011).BehaviouralandPsychologicalAspectsoftheRetirementDecision,SocialSecurityBulletin,Vol71,No.4,15-32.

KuhnA,WuellrichJandZweimüller(2012).FatalAttraction?Accesstoearlyretirementandmortality,Vox.Retreivedfromhttp://www.voxeu.org/article/fatal-attraction-access-early-retirement-and-mortality

LeslieNielson(2010).ChronologyofsuperannuationandretirementincomeinAustralia,EconomicsSection,ParliamentaryLibrary,AustralianParliamentHouse,1June2010,2009–10.Retrievedfromhttp://aphnew.aph.gov.au/binaries/library/pubs/bn/eco/chron_superannuation.pdf

MaestasN(2010).BacktoWorkExpectationsandRealizationsofWorkafterRetirement,J.HumanResources,Summer2010vol.45no.3718-748

McGearyK(2009).HowdoHealthShocksInfluenceRetirementDecisions?,ReviewofEconomicsoftheHousehold,Vol7–Issue3.

Mercer(2012).MelbourneMercerGlobalPensionIndex,AustralianCentreforFinancialStudies,Melbourne.

ModrekSandCullenM(2012).PhysicalJobDemandandEarlyRetirement,Presentedatthe14thAnnualJointConferenceoftheRetirementResearchConsortium,August2-3,2012,Washington,D.C.Retrievedfromhttp://www.nber.org/programs/ag/rrc/rrc2012/summaries/3.1%20Modrek,%20Cullen.pdf

OECD(2009).Ageoflabourforceexit,SocietyataGlance2009:OECDSocialIndicators,Pages:82–83,OECDPublishing.Retrievedfromhttp://www.keepeek.com/Digital-Asset-


34

Management/oecd/social-issues-migration-health/society-at-a-glance-2009/age-of-labour-force-exit_soc_glance-2008-14-en

OECD(2011).Pensionableyears,SocietyataGlance2011:OECDSocialIndicators,OECDPublishing.Retrievedfromhttp://dx.doi.org/10.1787/soc_glance-2011-14-enO’ReillyH(2008).Optimisingatransitiontoretirement incomestream,JASSATheFinsiaJournalofAppliedFinance,Issue1,2009.

PitS,ShresthaR,SchofieldDandPasseyM(2010).Healthproblemsandretirementduetoill-healthamongAustralianretireesaged45–64years,HealthPolicy,94:175–181

Sargent-CoxK,AnsteyJ,KendigHandSkladzienE(2012).DeterminantsofRetirementTimingExpectationsintheUnitedStatesandAustralia:ACross-NationalComparisonoftheEffectsofHealthandRetirementBenefitPoliciesonRetirementTimingDecisions,JournalofAging&SocialPolicy,24:3,291-308SicklesRandTaubmanP(1984).AnAnalysisoftheHealthandRetirementStatusoftheElderly,NationalBureauofEconomicResearch,Massachusetts

SkugorD,MuffelsRandWilthagenT(2012).LabourLaw,SocialNormsandtheEarlyRetirementDecision,NetsparDiscussionPaperSeries,DP11/2012-046

Staubli,SandZweimüller,J(2011).DoesRaisingtheRetirementAgeIncreaseEmploymentofOlderWorkers?,CEPRDiscussionPaperno.8510,London,CentreforEconomicPolicyResearch.Retrievedfrom:http://www.cepr.org/pubs/dps/DP8510.asp.

SullivanDandvonWachterT(2009).AverageEarningsandLong-TermMortality:EvidencefromAdministrativeData,TheAmericanEconomicReview,Vol.99,No.2,2009,133-138.

TheTreasury(2001).TowardshigherretirementincomesforAustralians:ahistoryoftheAustralianretirementincomesystemsinceFederation.Retrievedfromhttp://archive.treasury.gov.au/documents/110/PDF/round4.pdf

Warren,D.andOguzoglu,U.(2010).RetirementinAustralia:ACloserLookattheFinancialIncentives,AustralianEconomicReview,43:357–375.

WestfieldWright(2012).AttitudestoOlderWorkers,PreparedfortheFinancialServicesCouncilJanuary2012.

YeowL(2012).Reallocatesuperandsecurethefuture,AustralianFinancialReview,December2012.

***

Date post:	06-Jun-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

The Age of Retirement - Australian Human Rights Commission · AIST-ACFS Research Project 2013 The...

Documents