1

PolicyOptionsforStatePensionsSystemsandTheirImpactonPlanLiabilities*

RobertNovyMarx
UniversityofRochesterandNBER

JoshuaD.Rauh
KelloggSchoolofManagementandNBER

October19,2010

Abstract

Wecalculatethepresentvalueofstatepensionliabilitiesunderexistingpolicies,and
separatelyunderpolicychangesthatwouldaffectpensionpayoutsincludingcostofliving
adjustments(COLAs),retirementages,andbuyoutschedulesforearlyretirement.Liabilitiesif
planswerefrozenasofJune2009wouldbe$3.2trillionifcapitalizedusingtaxablemunicipal
curves,whichcreditstatesforapossibilityofdefaultinthesamestatesoftheworldasgeneral
obligationdebt,and$4.4trillionusingtheTreasurycurve.Underthetypicalactuarialmethod
ofrecognizingfutureserviceandwageincreases,liabilitiesare$3.6trillionand$5.2trillion
usingmunicipalcurvesandTreasurycurvesrespectively.Comparedto$1.8trillioninpension
fundassets,thebaselinelevelofunfundedliabilitiesisthereforearound$3trillionunder
Treasuryrates.AonepercentagepointreductioninCOLAswouldreducetotalliabilitiesby9
11%,implementingactuariallyfairearlyretirementwouldreducethemby25%,andraisingthe
retirementagebyoneyearwouldreducethemby24%.Evenrelativelydramaticpolicy
changes,suchastheeliminationofCOLAsortheimplementationofSocialSecurityretirement
ageparameters,wouldleaveliabilitiesaround$1.5trillionmorethanplanassetsunder
Treasurydiscounting.Thissuggeststhattaxpayerswillbearthelion'sshareofthecosts
associatedwiththelegacyliabilitiesofstateDBpensionplans.

*WethankJeffreyBrown,AndrewBiggs,ChristineJolls,andparticipantsatYaleLawSchoolandatthe
NBERStateandLocalPensionsConferenceinJacksonHole,Wyoming,August2010,forhelpful
commentsandsuggestions.
2

1Introduction
Thefinancialsoundnessofpensionsystemsforstateandlocalgovernmentworkershasbeena
topicofmuchdebate.InrecentworkNovyMarxandRauh(2009,2010)havearguedthatthe
useofGASBrulestodiscountfuturebenefitpaymentsresultsinpresentvaluemeasuresof
liabilitiesthataretoolowtoreflectthetrueeconomicliabilityfacedbystatetaxpayers.Atthe
sametime,anumberofstateshaveenactedchangesdesignedtoreducetheliabilities
associatedwiththeirpensionsystems(Snell(2009,2010)).Mostofthesechangesaffectnew
employeesonly,andhencehavenoimpactonstandardliabilitymeasures,whichdonot
considerfutureemployees.However,somechanges,suchasthereductionsinthecostofliving
adjustments(COLAs)passedbyColoradoandMinnesotathisyear,doaffectexistingplan
membersandhencedoaffecttheeconomicpresentvalueofcurrentstatepensionliabilities.
Inthispaper,weexaminethepresentvalueofstatepensionliabilitiesunderexisting
policiesandseparatelyunderseveralsetsofhypotheticalpolicymeasures.Inparticular,we
considerchangestoCOLAs,fullretirementages,earlyretirementages,andbuyoutratesfor
earlyretirement.
Aswehavepointedoutbefore,thekeyproblemwithGASBaccountingisthediscount
rate.Statesdiscountpensionliabilitiesatanexpectedreturnonpensionplanassets,andthat
rateistypicallyaround8%.Discountingproceduresbasedontheprinciplesoffinancial
economicsrequirediscountingatratesthatreflecttheriskinessoftheliabilities.Thismeans
discountingeitherwithataxablestatespecificmunicipalyieldcurve,whichcreditsstatesfor
thepossibilitytheycoulddefaultonpensionpayments,orwithaTreasuryyieldcurve,which
presentsthebenefitpaymentsasdefaultfree.WereviewtheevidencefromNovyMarxand
Rauh(2010)thatshowsthatevenifbenefitaccrualsowingtofutureserviceandsalarywere
stoppedcompletely,liabilitiesasofJune2009wouldbe$3.2trillionusingtaxablemunicurves
and$4.4trillionusingtheTreasurycurve.Underthetypicalactuarialmethodofrecognizing
futureserviceandwageincreases,liabilitiesare$3.6trillionand$5.2trillionusingmunicipal
curvesandTreasurycurvesrespectively.
Turningtopolicychanges,wefindthatincrementalchangessuchasonepercentage
pointreductionsinCOLAsoroneyearincreasestotheretirementagehaverelativelysmall
3

effectsontotalliabilities.ReducingCOLAsbyafullpercentagepointreducesliabilitiesunder
Treasurydiscountingby11%,whileraisingretirementagesbyoneyearreducesthemby34%.
WealsoconsiderfarreachingmeasuressuchastheeliminationofCOLAsorthe
implementationofretirementageparameterssimilartothoseusedbytheprivatesectoror
SocialSecurity.These"drastic"actionsarelesspoliticallyviablethanmoreincrementalpolicy
measures,andtheirlegalityinthecontextofsomestateconstitutionswouldbeinquestion.
However,eventhesefarreachingmeasuresdonotcomeanywhereclosetoeliminatingstates'
unfundedlegacyliabilities.Thelargestpolicychangesweconsiderstillleavepensionliabilities,
evenundernarrowmeasuresbasedontheageandserviceoftodaysmembers,around$1.5
trillionmorethanplanassetsunderTreasurydiscounting.Unfundedliabilitiesareevenlarger
underbroadermeasuresthatincorporatesomeportionoffutureemployeewagesandservice.
Ouranalysisconsidersonlypartialeffects.Itdoesnotaccountforequilibriumchangesin
retirementbehavior,anddoingsoisbeyondthescopeofthispaper.Totheextentthatbenefit
reductionsimposeawealthshockonemployees,employeeswouldrespondbyworkinglonger,
asleisureisgenerallyviewedasanormalgood.Ontheotherhand,ifthebenefitreductionsare
notfullycompensatedbyincreasesinotherformsofcompensation,reducingbenefitpackages
lowerscompensationandthusreducestherelativepriceofleisure.Thissubstitutioneffect
wouldgiveemployeesincentivestoworkless,butwouldlikelybesmall.Onbalance,
endogenouschangestoretirementbehaviorhaveanambiguoustheoreticalsign.Totheextent
thatworkersmodifytheirbehaviorspecificallytoincreasethevalueoftheirbenefits,this
responsemightwellmutetheeffectsthatwefind.
Someofthepolicyoptionsconsideredinthispaperaremorepoliticallyfeasiblethan
others.AnumberofstateshavealreadyenactedchangestoretirementagesaswellasCOLAs,
butthemajorityofthechangesonlyapplytonewhires.Statesthathavemadechangesthat
affectexistingworkersandretireesareintheminority.MinnesotaandColoradoarefacinglegal
challengesovertheirCOLAreductions,whichaffectallmembers.RhodeIslandhasraised
retirementagesforallemployeeswhoarenotyeteligibleforretirementbenefits,andIowahas
raisedtheagesforallnonvestedemployees.
4

Itmustberecognizedthateachstatehasauniqueconstitutionandexistingbodyof
statutorylawthatgrantdifferentprotectionstobenefits(BrownandWilcox(2009)),sothatthe
feasibilityofdifferentmeasuresvariessubstantiallyacrossstates.Changesofsmaller
magnitudesarelikelytofacefewerchallengesthanchangesoflargermagnitudes.
Evensomeoftheincrementaloptionsweexaminehaveonlybeentestedtoalimited
extentbystates.Forexample,wefindthatmakingearlyretirementactuariallyfairreduces
totalliabilitiesby25%dependingonthediscountingprocedure.Thisactuariallyfairearly
retiremententailsreducingbenefitsintheeventofearlyretirementsuchthatthepresentvalue
ofbenefitsequalstothepresentvalueofthebenefitstheemployeewouldhavereceivedif
theyweretodelaycollectingbenefitsuntilthenormalretirementage.Vermonthas
implementedasystemwhereearlyretirementpenaltieswillbeassessedonanactuarialbasis
forworkersbelowacertainageandservicethreshold.IowaPublicEmployeeshavemovedfrom
3%to6%actuarialreductionsperyear,whichgoesinthedirectionofreducingliabilitiesbut
stillallowsearlyretirementunderactuariallygenerousconditions.Wearenotawareofany
stateotherthanVermontthathasimplementedactuariallyfairearlyretirement,despitethe
factthatsuchchangesmightwellprovepoliticallyviable.Generousretirementoptionsmayin
somecasesalsoactuallysavethestatesmoney.Stateworkersgenerallyenjoymuchhigher
levelsofjobsecuritythantheirpublicsectorcounterparts.Earlyretirementincentivescan
thereforereducestatestotalliabilitieswhenusedtoencouragetheretirementofemployees
thatmaynotbefiredbutreceivetotalcompensationinexcessoftheirmarginalrevenue
product.
Thepolicyoptionsconsideredhereallrequirethatexistingplanmembersparticipateto
varyingdegreesinbearingtheburdenoftheunfundedlegacyliabilitiesfacedbythesesystems.
Thesechangesaffectthedistributionoftheburdenamongretirees,publicemployees,and
taxpayers.Totheextentthatthebenefitsbeforethechangesweretrulythepropertyofplan
members,thepolicyoptionscouldbeviewedassimplytransferringwealthfrompublic
employees(andinsomecasesretirees)totaxpayerswhoarecurrentlyliablefortheunfunded
pensiondebt.
5

Thispaperproceedsasfollows.Section2discussesthesampleandexistingmeasuresof
stateliabilitiesundervariousliabilityconcepts.Section3outlinesthepotentialpolicymeasures
weconsider,anddiscussessomeotherpolicyoptionsthatareoutsideofthemodel.Section4
explainshowthecalculationsofpolicychangesareimplementedinthecontextofourmodel.
Section5discussestheresultsforsingleandmultiple(orcomplex)policychanges.Section6
concludes.
2StatePensionLiabilities
Thissectiondescribesthesampleforthispaper,andreviewsexistingworkonliabilityconcepts
andmethodologiesforestimatingtheeconomicmagnitudeoftheseliabilities.
2.1Sample
Thesampleconsistsof116statesponsoredpensionfundsusedinNovyMarxandRauh(2009,
2010).Thisisauniquedatabaseon116stategovernmentpensionplansbuiltfromgovernment
reports.Itincludesallplanswithmorethan$1billionofassetsthataresponsoredbystate
governments,identifiedusingtheU.S.CensusofGovernments.Thedataonplanliabilitieswere
collectedfromtheComprehensiveAnnualFinancialReports(CAFRs)ofeachplan.
Rediscountingofcashflowsunderdifferentactuarialaccrualconceptsanddifferent
yieldcurvesrequiresanestimateofthecashflowsthemselves.Unfortunately,thestate
governmentsdonotprovidethecashflowsthattheyusetoderivetheliabilitiesthatthey
report.Toderiveestimatesofcashflowstreamsbasedontheinformationprovidedinthe
CAFRsthereforerequiresacalibratedmodelandaseriesofassumptions.Weexplainthe
calibrationitselfinsection4.Someoftheserequireadditionaldata,whichweexplainhere.
Onerequiredinputisasetofmatricesthatreflectassumptionsaboutsalaries,yearsof
service,andseparationprobabilitiesbyageforactiveworkers.Inparticular,werequireavector
ofsalarygrowthbyage;avectorofseparationprobabilitiesbyage;thedistributionofplan
participantsbyageandyearsofservice(anageservicematrix),andtheaveragewagesof
employeesineachcell.Toderivetheseinputs,weexaminedtheCAFRsofthe10stateswith
thelargesttotalliabilitiesandtooktheassumptionsfromthereportswheretheywereusable:
NewYork,Illinois,Pennsylvania,Ohio,andTexas.
6

Forretiredworkers,weemployadistributionofretireesbyageandtheaverageannuity
benefitineachagecategory.Thisinformationisonlysporadicallydisclosed,butbyrandomly
samplingtheCAFRsweobtainedanaveragedistributionacross10planscovering
approximately1.2millionretireesoutofthe6.8millionannuitants.Theplansthatwecould
findthatdisclosethisinformationaretheMassachusettsPublicEmployees,GeorgiaEmployees
RetirementSystem,IllinoisTeachersRetirementSystem,FloridaRetirementSystem,Ohio
TeachersRetirementSystem,TennesseeConsolidatedRetirementSystems,Pennsylvania
SchoolEmployeesRetirementSystem,LouisianaStateEmployeeRetirementSystem,NewYork
StateEmployeeRetirementSystem,andtheArizonaRetirementSystem.
2.2LiabilityConcepts
Therearefourdifferentliabilityconceptsthatweconsider:AccumulatedBenefitObligation
(ABO),ProjectedBenefitObligation(PBO),EntryAgeNormal(EAN),andProjectedValueof
Benefits(PVB).Wedescribetheseconceptsinthissection.
ThenarrowestmeasureistheABO.Itreflectsbenefitsalreadypromisedandaccrued.In
otherwords,evenifthepensionplanscouldbecompletelyfrozen,stateswouldstill
contractuallyowethesebenefits.TheABOisnotaffectedbyuncertaintyaboutfuturewages
andservice,asthecashflowsassociatedwiththeABOarebasedoninformationknowntoday:
planbenefitformulas,currentsalariesandcurrentyearsofservice.Onesourceofuncertainty
intheABOisinflation,andinparticularthemagnitudeofCOLAadjustmentsinstateswhere
COLAsarelinkedtoofficialstatisticssuchasCPIinflation.
TheABOisoftenthoughtofasaterminationliability,i.e.theliabilitythatwouldbe
owedtodayevenifplanswerefrozencompletelyorallworkerswerefired.Infact,theABO
actuallycouldbesomewhatlessthanaterminationliability,asitassumesanemployeedoes
notstarttakingbenefitsuntilhisretirementdate,whichmightbelaterthanthefullretirement
age.Aterminationliabilityassumesthatemployeeswilltakebenefitsattheearliest
advantageousdate,whichtypicallywillbeearlierthanthefullretirementagegiventhefact
thatactuarialadjustmentsforearlyretirementaregenerallylessthanactuariallyfair.
Ifworkersreceivetheirmarginalproductintotalcompensation(wagespluspension
benefits),theABOistheonlyconceptthatshouldbeconsideredsinceitmeasuresthebenefits
7

thatemployeeshaveactuallyearned(Bulow(1982),BrownandWilcox(2009)).TheABOisa
narrowmeasureinthatitdoesnotrecognizeanyfuturewageincreasesorfutureservicethat
employeesareexpectedtoprovide,eventhoughsuchwageincreasesandservicearetosome
extentpredictable.Moreover,theABOobligationisindependentofwagerisk,whichsimplifies
thevaluation.
Thethreebroadermeasures(PBO,EAN,andPVB)allaccounttovaryingextentsforthe
factthatbenefitswillcontinuetoaccrueduetothefuturesalaryand/orserviceofexisting
workers.Theyassumethatthepensionsystemwillnotbefrozentodayandallaimtoreflect
someportionofactualexpectedbenefits.
Thebroadestmeasure,thePVB,representsadiscountedpresentvalueofthefull
projectionofthecashflowsactuariesexpectthestatetoowe.ThePVBmethoddoesnotcredit
thestateforthefactthatitmighthavesomeabilitytolimitbenefitaccruals.BoththeEANand
thePBOrecognizeafractionofthePVB.ThePBOandtheEANarethereforeintermediate
measuresbetweentheABOandthePVB.
ThePBOaccountsfullyforexpectedfuturewageincreasesforexistingworkers,butnot
expectedfutureservice.Mathematically,thePBOformularecognizesthePVBinawaythatis
proratedbyservice.NotethatFASBaccountingforpubliclytradedcorporationsrequiresthe
calculationofaPBO.
TheEANisbroaderthanthePBO,butnotasbroadasthePVB.Mathematically,theEAN
methodrecognizesthePVBinproportiontodiscountedwagesearnedtodaterelativeto
discountedexpectedlifetimewages.Inpractice,thisprocedureaccountsforsomeportionof
futurebenefitaccrualsduetobothwagesandfutureservice.
Table2summarizestheliabilityconcepts.Furtherdetails,includingformulas,are
providedinNovyMarxandRauh(2010).Wenotethatnoneofthesemethodsaccountforthe
expectedbenefitsthatwillbeowedtoworkerswhohavenotyetbeenhired.
2.3EstimatesofLiabilitiesinPreviousLiterature
Mostestimatesofliabilitiesthatarenotconductedbyeconomistssimplyadduptheliabilities
thataredisclosedintheCAFRs.Thismethodignorestwoissues.First,itreliesstrictlyonthe
liabilityconceptthatstateactuarieschoose.Inpractice,mostliabilitiesintheCAFRsarestated
8

undertheEANmethod,anditisinstructivetobeabletocalculateliabilitiesundertheother
methods,particularlytheABOandthePVB.Second,addingliabilitiesdisclosedintheCAFRs
takesasgivenwhateverdiscountratethestateactuarieshavechosen.
AsexplainedinNovyMarxandRauh(2009,2010),thediscountratethestatesuse
underGASBaccountingproceduresdoesnotreflecttheriskoftheliabilities.Statesdiscount
theliabilitiesataflatrate,andusuallythisrateisverycloseto8%.Ofthe116plansinthe
sample,54ofthemuseaflat8%rate,and112usearatebetween7.5%and8.5%.
1
States
justifytheirdiscountrateswiththeargumentthattheyarediscountingliabilitiesatthe
expectedrateofreturnontheassetsintheirpensionfund.Suchaprocedureignorestheriskof
theassetscompletelyandtreatsreturnsabovetheriskfreerateasafreelunch.
Thestateprocedureshavesurvivedcriticisminpartbecauseobservershavenotedthat
manypensionsystemshaveearnedaveragereturnsofaround8%overthepastdecades.But
again,thisassumesthatthe8%wasobtainedwithoutanyrisk.Infact,thesereturnswere
obtainedbytakinginvestmentrisk,andiftheassetshadnotreturned8%,statetaxpayers
wouldhavebeenonthehookforadditionalshortfalls.Ifsystemswanttobeabletotelltheir
employeesthatthebenefitstreamissaferthanaportfolioofstocksandbonds,theyshould
discountthecashflowsinawaythatreflectsthatsafety.
NovyMarxandRauh(2010)employtwoprimarydiscountingprocedures.Thefirstuses
thetaxablemunirate,definedasthestatespecific(orwhereunavailabletheratingcategory
specific)municipalyieldgrossedupforataxpreferenceonmunidebtthatweestimatetobe
25%.ThesecondmethodusestheTreasuryyieldcurve.
Usingthemunirateadmitsandquantifiesaprobabilityofdefault.Theliabilityisa
measurethatcalculatesthepresentvalueofthisdefaultableliabilityfromtheperspectiveof
thetaxpayersundertheassumptionthatthestatewilldefaultonthesepaymentsinthesame
statesoftheworldasontheirgeneralobligationdebtandwiththesamerecoveryrates.
Alternatively,itisthevalueoftheportfolioofstateGObondsthestatewouldneedtodeliver
totheplantodefeasetheobligation.Whenassessingthedifferenceintheliabilityunder

1
The proposals outlined in GASB (2010) would basically maintain this system, with muni rate discounting kicking
in only for liabilities that are not fully funded under the expected return on assets system.
9

differentpolicymeasures,thecomparativestaticsquantifyhowbigtheshiftisinthevalueof
theseuncertainpayments.
UsingtheTreasuryyieldcurvevaluesthepensionbenefitsassecurepromises.The
Treasuryvaluationsbeginfromthepremisethatthebenefitswillbepaid.Totheextentthat
theyarenotpaid,thereisatransferfromparticipantstotaxpayers.Theexpectedvalueof
thesetransferswouldreducethevalueofthepaymentstotheparticipantsbutalsoreducethe
costtothetaxpayer.TheTreasurydiscountingcanthereforebeviewedasvaluingthebenefits
asadefaultfreepromise.Ifstatepensionsystemswanttopresenttotheiremployeestheidea
thatthebenefitsaredefaultfree,theymustdiscountatdefaultfreerates.Ifastatepension
systemwantedtocontractouttheprovisionofthebenefitstoaninsurerwhowouldmakethe
benefitpaymentsevenifthestateinthefuturedefaultedonsomeofitsobligations,the
insurancecompanywouldpresumablyvaluetheliabilityatadefaultfreerate.
OurmeasuresaresummarizedinTable3.Understatechosendiscountrates,theEAN
liabilityis$3.15trillionatameasurementdateofJune2009.Thisisveryclosetoliabilitiesone
wouldobtainbysimplyaddingupthelateststatedisclosures,sincemoststatesystemsusethe
EANmethod.
2
DiscountingonlythecashflowsassociatedwiththeABOlowerstheliabilityto
$2.76trillion,whereasdiscountingthefullexpectedvalueofcashflows(stillunderstate
chosenrates)resultsinaliabilityof$3.75trillion.
ThesecondandthirdcolumnsinTable3showtheresultsundertaxablemuni
discountingandTreasurydiscountingrespectively.UndertheABO,taxablemuniliabilitiesare
$3.20trillionandTreasuryliabilitiesare$4.43trillion.UndertheEAN,themostcommon
accrualmethodusedbystates,thetaxablemuniliabilitiesare$3.62trillionandtheTreasury
liabilitiesare$5.28trillion.UnderthePVB,thetaxablemuniliabilitiesare$4.24trillionandthe
Treasuryliabilitiesare$6.86trillion.
AsofJune2009,weestimatethatstateshad$1.89trillioninassetssetasideinpension
fundsdedicatedtomeettheseobligations.UndertheABO,thismeansthatthereisagap
betweenassetsandliabilitiesof$1.31trillionundertaxablemunidiscountingand$2.54trillion

2
The raw sum of CAFR liabilities regardless of the reporting date was $3.03 trillion. Using the weighted-average
growth rate of the reported plans to extrapolate for missing data and harmonizing all plans to a June reporting date
yielded a total J une 2009 liability of $3.14 trillion on a stated basis. See Internet Appendix Table I of Novy-Marx
and Rauh (2010) for details.
10

underTreasurydiscounting.Thesestatisticssuggestthatexistingassetsfallsubstantiallyshort
ofthepresentvalueofexistingpromisedliabilities.
Furthermore,stateshavebeenfundingundertheEANaccrualmethod,whichsetsaside
moremoneyearlierintheemployeescareerthanfundingundertheABOwould,butless
moneylater.ItthusseemsfairtocompareEANliabilitiestoexistingassets.UndertheEAN
measure,theunfundedliabilityinstatepensionplansis$1.73trillionundertaxablemuni
discountingand$3.39trillionunderTreasurydiscounting.
Thefinaltwocolumnsapplya51bpcorrectionforthepossibilitythatsalariesare
correlatedwithmarketpricingfactorsoverlonghorizons.Benzonietal(2007)arguethatwhile
thecorrelationbetweenearningsgrowthandstockreturnsisnegligibleonashorthorizon,the
correlationishigheronalongerhorizon.LucasandZeldes(2006)discusstheseeffectsinthe
contextofcorporatepensionplanswithamodelinwhichthevalueofhumancapitalandthe
valueofthestockmarkethavepositivecovariance.The51bpcorrectioncomesfromasimple
calibrationdiscussedinNovyMarxandRauh(2010).Thecorrectionisonlyappliedtoactive
workerliabilitiesundertheaccrualconceptsthatrecognizesomefuturesalaryorservice.It
thereforedoesnotaffecttheABOatall,anditdoesnotaffecttheretiredorseparatedworker
liabilities.Thiscorrelationdoesnotsubstantiallyaffectinferenceaboutthemagnitudeof
liabilities.
AnalternativemethodologyisprovidedbyBiggs(2010),inwhichtaxpayersliabilityis
valuedasaputoptionwithastrikepriceequaltotheactuarialliabilitygrossedupatstated
discountratesasinNovyMarxandRauh(2008).Givenhowfarinthemoneythisputoptionis,
itsvalueisalmostequivalent(mechanically)todiscountingtheliabilityattheriskfreerateand
subtractingoffthevalueoftheassets.AsnotedinNovyMarxandRauh(2008),thecostofthe
[taxpayer]insuranceisalmostaslargeasthefundingshortfallbecausethevalueofthe
potentialoverfundingistrivialrelativetothepotentialunderfunding.
3

3PotentialPolicyMeasures

3
To be precise, Novy-Marx and Rauh (2008) found that the present value of the possibility that pensions could end
up overfunded, if there assets' performance greatly exceeds their expected returns, is only worth $16 billion today.
The surplus would only come in low marginal utility states of the world. This means that even if the government is
unlikely to squander any overfunding (or promise it to beneficiaries as in Bodie (1990)), the value of that possible
overfunding , taking the state pricing into account, is almost trivial to taxpayers.
11

Inthissectionwediscussthepotentialpolicymeasuresthatweconsidertotrytoaddress
shortfallsinstatepensionfunding.
3.1FullRetirementAge
Thefirstpolicychangeweanalyzeisanincreasetheageof"fullretirement,"i.e.,the
ageatwhicharetireecanbegintakingpaymentswithoutincurringabenefitreduction.This
change,whichdoesnotaffectcurrentannuitants,mayprovepoliticallyfeasiblealthoughfew
stateshaveactuallyraisedretirementagesforexistingworkers.Changestotheretirementage
arecurrentlybeingphasedintoSocialSecurity.Theyhavealsobeenimplementedbysome
stateplans,thoughatthestatelevelthesechangeshavebeenwrittentoaffectnewworkers
only,andthushavenoimpactonthepensionliabilitiesofthecurrentworkforce.The
implementationweconsiderappliestocurrentaswellasfutureworkers,andalsoappliesto
thevestedseparatedworkersnotyetreceivingbenefits.Wemodelaoneyearincreaseinthe
fullretirementagebyassumingthatparticipantsareeligibletoreceivefullbenefitsattheage
of61,not60,andthattheyarefirsteligibletoreceiveearlyretirementwithreducedbenefitsat
56,not55.
Wedonotattempttoquantifytheimpactofthispolicychange(oranyother)on
workersretirementbehavior.Changesinretirementbehaviormayhaveadditionaleffectson
theliabilityasmeasuredbythebroaderliabilityconcepts(PBO,EAN,PVB),butwouldhaveonly
athirdordereffectontheABOliability.
3.2ActuariallyFairEarlyRetirement,orNoBenefitsUntilFullRetirementAge
Anotherpolicychangeaffectingtheretirementage,whichhasbeenlessdiscussedbut
maybepoliticallyfeasible,isanadjustmenttotherulesgoverningearlyretirement.Currently
theplansalsoallowforearlyretirementontermsthatareextremelyfavorabletoplan
participants.Themostcommonrulesallowaplanparticipanttobegintakingbenefitsuptofive
yearsearly,byincurringthe6percentperyearlinearreductioninthebenefittheyreceivefor
eachyeartheyretirebeforethefullretirementage,andsomeplansofferearlyretirementon
termsthatareevenmoregeneroustoplanparticipants.BecauseCOLAs,whichaverage
betweentwoandahalfandthreepercentperyear,donotapplytoretirees'benefitsuntilthey
aretakingpayments,theeffectivecostofearlyretirementisonlyabenefitreductionofthree
12

tothreeandahalfpercentperyear.Thatis,aparticipantcangetanextrayearofbenefitsin
exchangeforareductioninfuturebenefitsofthreetothreeandahalfpercent.Thisisroughly
halfthemarketvalueofanimmediatelifeannuityforasixtyyearoldman,ortwothirdsthe
costofanimmediatejointlifeannuityforasixtyyearoldmarriedcouple,suggestingthatplans
rewardeachyearofearlyretirementwitha"gift"equaltoroughlyfourtosixmonthsworthof
benefitpayments.
Actuariallyfairearlyretirementinvolvesabenefitreductionformulaforearly
retirementthatequatestheexpectedpresentvalueofbenefitsunderearlyandregular
retirement.Becauseactuariallyfairearlyretirementmeans,bydefinition,thattheNPVof
expectedbenefitpaymentsisunaffectedbyearlyretirement,thesimplestwaytoconsiderthe
impactofAFERonpensionliabilitiesistoassumeparticipantsdonotretireearly,i.e.,bysetting
theearlyretirementageequaltothefullretirementage.Weassumeagainthatthispolicy
changeappliestocurrentaswellasfutureworkers,andalsothevestedseparatedworkersnot
yetreceivingbenefits.Itofcoursehasnoimpactontheannuitants.
Thecostoftheearlyretirementoptiontotheemployer,whichisalsothevalueofthe
earlyretirementoptiontotheemployees,doesdependonthediscountrate.Theearly
retirementoptionisparticularlyvaluableunderhighdiscountrates,becauseearlyretirement
givestheemployeeearly,largelyundiscountedcashflowsinexchangeforgivingupsomelater,
heavilydiscountedcashflows.Higherdiscountrateshardlyreducethevalueofthebenefitof
theearlycashflows,butgreatlyreducethevalueoftheforegonelatercashflows,increasing
thenetbenefitofearlyretirement.Wethereforewouldexpecttofindanimplementationof
AFERaffectingliabilitiesmoreunderdiscountratesthatreflectgreaterdegreesofrisk.
4

3.3COLAReductions
COLAreductionsareoneoftheonlypolicymeasuresavailablethathavethepotentialto
forcecurrentannuitants,aswellascurrentandfutureworkers,tobearsomeoftheburden
associatedwiththecurrentfundingsituation.Thismakesthemattractiveinsomeways,
becauseretireesarethebiggestbeneficiariesofthelegacyliabilities,andsomefeelthatthey

4
For example, a 6% buyout rate is very actuarially generous under stated or muni rates, but it is closer to actuarially
fair under risk-free rates. In a rough estimate, a 6.5% buyout rate would appear actuarially fair under Treasury
discounting.
13

shouldsharesomeofthecosts.Theymaybedifficulttoimplementpolitically,however,aslegal
protectionsinmanystatespreventmeasuresthatimpairthevalueofpensionpromises,and
publicemployeeunionsareespeciallyresistanttomeasuresthatdamagetheclaimsoftheir
seniormembership.Inatleasttwostates(ColoradoandMinnesota)lawshavebeenpassed
thatentailmodestCOLAreductionsforevencurrentannuitants.Thesecurrentlyfacelegal
challenges,butseemlikelytoholdupincourt,andarepopularwithtaxpayersinthestatesin
whichtheyhavebeenpassed.
5

3.4DramaticMeasures:MovingtoSocialSecurityRetirementAges
Wealsoconsiderseveralmoredramaticchanges,whichshiftagreaterportionofthe
legacyliabilityfromtaxpayerstoplanparticipants.Thefirstoftheseinvolvesbringingtherules
governingretirementintolinewiththeprivatesector,byraisingthefullretirementageto65
andmakingearlyretirementactuariallyfair.ThesecondadditionallyreducesCOLAsbyone
percentperyear.Thelastconsiderspolicychangesthatbringtheretirementrulescloserto
thosecurrentlygoverningSocialSecurity,afullretirementageof67andearlyretirementthatis
closetoactuariallyfair.
6

3.5ExtremeMeasures:CuttingCOLAsCompletely
Toourknowledge,nostatehaseliminatedCOLAscompletelyforanyplans.Atleastone
state,Georgia,haseliminatedCOLAsfornewworkers,withtheexplicitintentofprotecting
existingandlegacybenefits.
7
IllinoishascappedCOLAsathalfofCPIfornewworkers.Sucha
measureclearlytaxesnewemployeesrelativetothepreviousbaselineinordertosubsidize

5
Anecdotally, at least some unions and public officials seem to understand that taxpayers are bankrolling generous
retirement plans. For example, a Minnesota executive director of a public employee union has been quoted in the
Wall Street Journal as saying that his board was accepting COLA reductions for current retirees so that we don't
kill the goose that lays that golden egg (Neumann (2010)).
6
Specifically, we allow early retirement as early as age 62, but use a linear buyout schedule with a buyout rate six
percent higher than the COLA rate employed by each plan. Social Security uses a non-linear buyout schedule that is
roughly six percent per year, but unlike the state plans has COLAs, tied to wage growth, that apply prior to the date
one begins receiving benefits. On this dimension Social Security is more generous than the state plans, and the fact
that the unions do not seem particularly interested in fighting for COLAs that apply prior to the date of the first
benefit payment suggests that they do not well represent plan participants, either current or future, that are separated
and vested but not receiving benefits.
7
The General Assembly is desirous of providing an established annual cost-of-living adjustment to all current
active and retired members of the Employees' Retirement System of Georgia, the Georgia Legislative Retirement
System, and the Georgia J udicial Retirement System. In order to do so, limiting future liability of the systems by
adjusting the retirement expectations of persons who are newly employed is a regrettable but necessary step toward
fiscal soundness. (Georgia House Bill 452/ Act 82). The bill passed the House March 9, 2009 (157-0), passed the
Senate March 26, 2009 (35-10) was signed by the governor, and went into law on J uly 1.
14

existingemployeesandexistingretirees.DramaticCOLAreductionsareextremelyeffectivein
reducingthestates'obligations,becauseCOLAsaffectannuitantsaswellasworkers,and
increasethestates'liabilitybyroughlyathirdrelativetowhattheywouldbeabsentCOLAs.For
thissamereasondrasticCOLAreductionsmaybeinfeasiblepolitically.
3.6OtherMeasuresOutsidetheModel
Severalstateshavetakenstepsoverthelasttwoyearsthatattempttoaddresstheirpension
fundingshortfalls,usingpolicychangesthatareoutsidethescopeofourmodel.Thesesteps
includerulestoprevent"spiking"(largeendofcareersalaryincreasesthatincreaselifetime
pensionbenefits),employeecontributionincreases,increasesinvestingperiods,andlimitson
benefitpayments.
VirginiaandRhodeIslandhaveincreasedfromthreetofivethenumberofyearsusedto
calculatethefinalaveragesalary(FAS),whileGeorgia,ColoradoandVermonthavecappedthe
increasesinsalariesthatcounttowardbenefitcalculations.Colorado,Iowa,Minnesota,
Mississippi,Nebraska,NewMexico,VermontandWyominghaveincreasedemployeepension
contributionsforbothcurrentandfutureemployees,whileNewHampshireandTexashave
increasedthesefornewemployeesonly.Whiletheseincreasesdonotreducethestates'future
pensionliabilities,theydodivertaportionofworkers'salaryintothepensionfund,thereby
increasingthefundassetsandthusimprovethefundingsituation.
8
Iowa,Minnesota,and
Missourihaveincreasedthelengthoftimeanemployeemustworktobeeligibleforthestates'
DBplans.Thesechangeshavealmostnoimpact,qualitatively,asworkersthatseparatewithout
vestingtenttobeyoung,andthevalueoftheirbenefits,hadtheyvested,wouldhavebeen
extremelysmallanywaybecausetheplansCOLAsdonotapplyuntilretirement.Michiganand
IllinoishaveintroduceddollarcapsontheFASthatmaybeusedinsomeplansbenefit
calculations.

4ImplementationofPolicyChangesintheModel

8
These contribution increases forceyoungerworkers,andespeciallynewemployees,tosubsidizeolderworkers
andtheretired.Thisfactpresentsaconflictofinterestsforstateworkers'unions,whichnegotiatethe
compensationandbenefitsforfutureworkerswhilerepresentingtheinterestsoftheircurrentmembership.
15

Inordertodeterminetheimpactofthesepolicychangesonstates'pensionliabilities,
weforecasteachplan'scashflowstoplanparticipantsusinginformationfromtheplans'CAFRs.
Wecalibratethesesothateachplan'sliabilityunderitschosenaccountingmethodology,
discountedusingitsstateddiscountrate,matchesthestatedliabilityinitsCAFR.Policychanges
arethenimplementedinthismodel,andtheimpactofthesechangesonthestates'pension
liabilitiescanbequantifiedunderanyaccountingmethodologyusinganydiscountrates.
TheinitialcalibrationproceedsasinNovyMarxandRauh(2010),whichprovidesamore
detailedaccountofthemethodology.Thefirststepofthiscalibrationemploysourmodelto
forecasteachplan'sliabilitieseachyearinthefuture.Themodelusesplanlevelinformation
regardingthenumberofactive,retiredandseparatedworkers,aswellasthebenefitfactor
(i.e.,thefractionofsalarywhich,whenmultipliedbyyearsofservice,determinesaparticipant's
initialbenefit),costoflivingadjustment(COLA)andinflationassumptionemployedbytheplan.
Italsoemploysassumptionsregardingtherelativenumberofemployeesandaveragewagesby
ageandyearsofservice(anageservicematrix),aswellassalarygrowthandseparation
probabilitiesbyage,andtherelativenumberandaveragelevelofbenefitsforannuitantsof
eachage.Thebenefitcalculationsassumeafullretirementageof60,andthatyoungerretirees
canstarttakingbenefitsuptofiveyearsearlybyincurringalinear6%benefitreductionfor
eachyearaparticipantretiresbefore60.
9

BenefitsareprojectedassumingmortalityratesfromtheRP2000tables,whichare
employedbymanystates.Weusethetables'combined(employee/retired)healthyrates,and
assumethatparticipantsareevenlydividedbygender,that60percentaremarriedatthetime
theyretiretoaspouseofthesameage,andthatplansallowfor50percentsurvivorbenefits.
Wethencalibrateeachplan'scashflowsbyadjustingtheaveragesalarylevelofthe
employedandtheaveragebenefitsofthenonactivemembers.Theyarecalibratedto
simultaneouslymatchboth1.)theplansstatedaccountingliabilitywhencapitalizedatthe
statechosendiscountrateusingtheactuarialmethod(EANorPUC)employedbythestate;and

9
This benefit reduction schedule is common practice in state public pension systems, and if anything conservative;
the Florida Retirement System uses a 5% reduction for each year below the full retirement age, while the New
J ersey Public Employees and Teachers systems use 1-3% reduction for each year.
16

2.)theplansexpectedfirstyearcashflow,whichweestimateat107%ofthecashflowforthe
yearendingJune2009,basedonrecenthistoricalcashflowgrowth.
Themodel,usingthecalibratedaveragesalaryleveloftheemployedworkersandthe
calibratedaveragebenefitsofthenonactivemembers,togetherwiththeparametersusedin
thecalibration,servesasour"baseline."Weinvestigatetheimpactofeachofthedifferent
policychangesdiscussedearlierbyalteringtheparametersemployedbythemodelinamanner
consistentwiththepolicychanges,andusingittogeneratetheplans'newforecastliabilities.
Thesecanthenbecapitalized,underwhicheveraccountingconceptwewishtoemploy,using
eitherthetaxablemuniyieldcurveortheTreasuryyieldcurve.

5Results
Thefirstpartofthissectionconsidersthepolicychangesthatonlyaffectonepensionplan
parameter.Thesecondpartconsidersmorecomplexchangesthataffectmultipledimensions
ofthesystemparameters.
5.1SinglePolicyChanges
Table5showstheeffectofsinglepolicychangesundertaxablemunidiscounting.Table
6showstheeffectunderTreasurydiscounting.Thefirstcolumnofeachtableshowsthe
baselineresultswithoutanypolicychanges,asapointofcomparison.Figure1showsa
graphicalrepresentation.
ThesecondcolumnsofTables5and6showtheeffectsofaonepercentagepointCOLA
reductionforallsystemmembers(includingcurrentretirees)onliabilitiesundertaxablemuni
andTreasurydiscountingrespectively.Undertaxablemunidiscounting,theeffectonthetotal
ABOisadecreaseof$280billion,ofwhichtheannuitantportionis$160billionoraround60%.
UnderTreasurydiscounting,theeffectonthetotalABOisadecreaseof$470billion,ofwhich
theannuitantportionis$220billionoraround50%.
Overall,aonepercentagepointCOLAreductionlowerstheliabilitymeasuresbyalmost
9%underthetaxablemunidiscountingandbyalmost11%underTreasurydiscounting.The
percentageeffectislargerforactiveparticipantsthanforannuitantsinbothcases.COLA
reductionshaveamoreimmediateeffectonannuitants,butthisisoffsetbythefactthatCOLA
17

reductionswillaffectcurrentworkersonaverageforalongertime.Currentworkershavetheir
fullretirementaheadofthemwithdecreasedCOLAs,whilecurrentannuitantsfaceonlythe
futurepartofitwithdecreasedCOLAs.NotethatreducingCOLAsreducestheeffective
durationoftheliabilities.
ThethirdcolumnsofTables5and6showtheliabilitiesifAFERisimplemented.Thefirst
thingtonoteisthatthispolicychangedoesnotchangetheannuitantliabilityatall,sincethese
participantsarealreadyretired.ImplementingAFERreducesactiveparticipantliabilityby89%
undertaxablemunidiscounting,butbyonly23%underTreasurydiscounting.Theeffecton
totalliabilitiesisevenmoremutedbecauseofthezeroimpactontheannuitantliability.
Overall,implementingAFERreducestotalliabilitiesby35%undertaxablemunidiscountingbut
only12%underTreasurydiscounting.
ImplementingAFERwouldthereforegeneratelowerliabilitiesunderstateddiscount
ratesandalsoundertaxablemunirates.Butifthepensionpromisesaredefaultfree,
implementingAFERdoesnotaffectthevalueoftheliabilitynearlyasmuch.Itisinstructiveto
comparetheseeffectstopotentialchangesintheFRAundertheABOaccrualmethod.The
generosityofthebuyoutscheduleforearlyretirementisworthmorethana1/2yearreduction
intheFRAunderriskfreediscounting,butalmost11/2yearsundertaxablemunidiscounting.
ThefourthcolumnsofTables5and6showtheimpactofincreasingtheretirementages
(boththeFRAandtheERA)byoneyear.Relativetothebaseline,totalliabilitiesarereducedby
around24%.Thebroadermeasures(e.g.PVB)areaffectedmorethanthenarrowermeasures
(e.g.ABO).Thispolicymeasuredoesnotaffectannuitants,sothedecreasesinliabilitiesare
achievedlargelyattheexpenseofthecurrentworkers.Inthenextsectionweconsiderthe
effectsofmoredramaticretirementageincrease.
ThefinalcolumnshowswhatwouldhappentoliabilitiesifCOLAswereeliminated
entirely.Thereductionisaround22%undertaxablemunidiscountingand26%underTreasury
discounting.TheeliminationofCOLAswouldstillleaveABOliabilitiesat$3.27trillionunder
Treasurydiscounting.ThegapbetweenassetsandABOliabilitieswouldstillbealmost$1.5
trillion,andthegapbetweenassetsandEANliabilitieswouldbe$2trillion.
5.2MultiplePolicyChanges
18

Tables7and8showtheeffectsofthemultiplepolicychangepackagesonliabilities,and
Figure2showsagraphicalrepresentation.Theeffectsofbothincreasingtheretirementagesby
oneyearandimplementingactuariallyfairearlyretirementistoreduceliabilitiesby6%9%
undertaxablemunidiscountingand46%underTreasurydiscounting.Aswasthecaseabove,if
thepensionpromisesaredefaultfree,implementingAFERdoesnotaffectthevalueofthe
liabilityasmuchasunderthetaxablemunidiscounting.Theeffectsofcombiningthesetwo
policychangesareslightlylargerthanthesumofthepercentageeffectsofthetwoeffects
separately.Thisisbecauseoftheinteractionbetweenthetwopolicymeasures.
Settingthefullretirementageandearlyretirementagebothto65reducesliabilitiesby
1524%.Theeffectisgreaterunderthetaxablemunidiscounting,forthesamereasonsthat
introducingactuariallyfairearlyretirementisgreaterundertaxablemunidiscounting.Setting
bothretirementagesto65isequivalenttoanincreaseintheretirementagecombinedwith
theintroductionofearlyretirement.ReducingtheCOLAby1percentagepointandsettingboth
retirementagesto65reducesliabilitiesby2330%.MovingtotheSocialSecurityparameters
reducesthemby2231%.Inallofthesecases,thereductionisgreaterforthebroader
measuresthanforthenarrowermeasures.
Allofthesepolicymeasuresstillleavesubstantialgapsbetweenassetsandliabilities.
EvenunderamovetoSocialSecurityparameters,whichwouldbethemostdrasticofthese
measures,theABOatTreasuryratesisstill$3.38trillion,leavingagapbetweenassetsand
liabilitiesof$1.5trillion.
6.Conclusions
Thenotionthatthehistoricalequitypremiumprovidesacostlesssolutionisincorrect.Even
underthemostconservativemeasures,publicpensionliabilitiesarecurrentlyover$1trillion
largerthanplanassets.Usingdiscountratesthatactuallyreflectthepromiserevealsshortfalls
of$2.5billionforaccumulatedbenefitsonlyandover$3trillionforbroadermeasures.This
shortfallhastobebornebysomeparty:taxpayersorpublicemployees,betheypast,current,
orfuture.
19

Inessence,then,thedebateoverthesolutionisovertransfers.Thecurrentsituationis
oneinwhichbeneficiariesviewtheirbenefitsassecurepromisesandtaxpayersdonotperceive
thattheywillbeheldaccountableforguaranteeingthosepromises.
Eachofthepolicychangesweconsiderreflectsareductioninperceivedwealthfeltbya
groupofparticipants.Almostallofthemwouldfallunderthecategoryof"goodstarts"for
addressingstates'underfundedpensionliabilities.However,weshowthatthereisno"magic
bullet"asfaraspolicychangesareconcerned.EvenimplementingSocialSecurityparameters,
andundertherelativelynarrowaccountingoftheABO,thedifferencebetweenassetsand
liabilitiesisstill$1.5trillion.Wespeculatethattaxpayerswilllikelynotsucceedinforcingpublic
employeestobeartherestofthisgap,andthatultimatelytaxpayerswillhavetocomeupwith
sumsofthismagnitudetopayforlegacyliabilities.Ifunfundedliabilitiescontinuetogrow,the
bailoutscouldbeevenlarger.
Theexaminationoftheimpactofpolicychangesonliabilitiesinthispaperisdonein
partialequilibrium.Wehavenotmodeledanyendogenousresponsesbyworkersintheir
retirementbehavior.Wehaveassumedthatthesechangescanbeimposedonworkerswithout
beingoffsetbyanyotherchangesthatmightincreasetheliability.Wehavenotanalyzedthe
plancharacteristicsthatmightmaketheliabilitiesofsomeplansmoresensitivetothesepolicy
changesthanothers.Finally,wehavenotmodeledanycostsavingsthatearlyretirementmight
achieveforstatesifworkersclosetoretirementarecompensatedmorethantheirmarginal
product.Theseareimportantavenuesforfutureresearch.

20

References
Benzoni,L.,PierreCollinDufresne,andR.Goldstein(2007)PortfolioChoiceovertheLifecycle
whentheStockandLaborMarketsareCointegrated.JournalofFinance62,21232167.
Biggs,Andrew(2010)TheMarketValueofPublicSectorPensionDeficits.AmericanEnterprise
InstituteOutlookSeries.http://www.aei.org/outlook/100948
Bodie,Zvi(1990).TheABO,thePBO,andPensionInvestmentPolicy.FinancialAnalystsJournal
46,2734.
Brown,JeffreyandDavidWilcox(2009)DiscountingStateandLocalPensionLiabilities.
AmericanEconomicReview99(2):538542.
GovernmentAccountingStandardsBoard(2010)PreliminaryViewsoftheGovernment
AccountingStandardsBoardonMajorIssuesRelatedtoPensionAccountingandFinancial
ReportingbyEmployers.
http://www.gasb.org/cs/ContentServer?c=Document_C&pagename=GASB/Document_C/GASB
DocumentPage&cid=1176156938122
Lucas,DeborahandStephenP.Zeldes(2009)HowShouldPublicPlansInvest?American
EconomicReview99(2).
Neumann,Jeannette(2010)StateWorkers,LongResistant,AcceptCutsinPensionBenefits.
WallStreetJournal,29June2010.
NovyMarx,Robert,andJoshuaD.Rauh(2008)TheIntergenerationalTransferofPublicPension
Promises.NBERWorkingPaper#14343.
NovyMarx,Robert,andJoshuaD.Rauh(2009)TheRisksandLiabilitiesofStateSponsored
PensionPlans.JournalofEconomicPerspectives,23(4):191210.
NovyMarx,Robert,andJoshuaD.Rauh(2010)PublicPensionPromises:HowBigAreTheyand
WhatAreTheyWorth?http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1352608
Snell,Ronald(2009)PensionsandRetirementPlanEnactmentsin2009StateLegislatures.
http://www.ncsl.org/?tabid=17594
Snell,Ronald(2010)PensionsandRetirementPlanEnactmentsin2010StateLegislatures.
http://www.ncsl.org/?TabId=20255

21

Table1:SummaryofStateSponsoredPensionPlans
Thetoppanelsummarizesthenumbersofindividualmembersineachofthreemaincategories:activeworkers,
annuitants,andthosewhoarevestedbutnolongerinpublicemployment.Thesampleisthe116majorpension
planssponsoredbythe50U.S.stategovernments.Thebottompanelliststhesedataforthe10statesponsored
pensionplansthatarethelargestbytotalmembers.

MemberCounts(NumberofPlans=116)
Separated& Active
Active Annuitants Vested Total Share
SummaryStatistics
Total 12,920,361 6,813,294 2,592,462 22,326,118 58%
Mean 111,382 58,735 22,349 192,466 55%
Median 61,240 35,678 6,144 103,483 58%
StdDev 153,683 77,012 44,838 259,096 16%
Largest10Plans
CalPERS 821,113 492,513 313,284 1,626,910 50%
TexasTeachers 946,405 283,018 59,622 1,289,045 73%
FloridaRetirementSystem 668,416 321,137 87,284 1,076,837 62%
NewYorkEmployees 530,081 339,122 54,426 923,629 57%
CalSTRS 461,579 228,969 31,179 721,727 64%
OhioPublicEmployees 360,186 169,333 123,673 653,191 55%
NorthCarolina(Teachers&State) 325,689 153,929 96,435 576,052 57%
WisconsinRetirementSystem 265,779 146,484 149,258 561,521 47%
PennsylvaniaPublicSchool 273,119 178,939 103,691 555,749 49%

22

Table2:DescriptionofMethodsforRecognizingAccruedLiabilities
Thistablesummarizesthefourmainmethodsforrecognizingpensionliabilities.Themethodsdifferintheir
treatmentofexpectedfuturesalaryincreasesandservicethatisyettobeperformed.Themethodsarelistedin
increasingorderofbroadness,startingwiththemethodthatonlyreflectscurrentserviceandsalaryandending
withthemethodthatreflectsafullprojectionofbenefitsthatareexpectedtobepaid.

AccumulatedBenefitObligation(ABO) Representspromisedbenefitsundercurrentsalary
andyearsofservice.Oftenusedinterchangeably
withtheconceptsofterminationliability,or
liabilityiftheplanwerefrozen,althoughthereare
somedifferences(seetext).
ProjectedBenefitObligation(PBO) Takesprojectedfuturesalaryincreasesinto
accountincalculatingtodaysliability,butnot
futureyearsofservice.UsedinFASBaccounting
forcorporations.
EntryAgeNormal(EAN) Reflectsaportionoffuturesalaryandserviceby
allowingnewliabilitiestoaccrueasafixed
percentageofaworkerssalarythroughouthis
career.
PresentValueofBenefits(PVB) Fullprojectionofwhatcurrentemployeesare
expectedtobeowediftheirsalarygrowsandthey
work/retireaccordingtoactuarialassumptions.

23

Table3:SummaryofBaselineResultsfromNovyMarxandRauh(2010)
ThistablereviewsandsummarizesthebaselineresultsfromNovyMarxandRauh(2010).Liabilitiesaremeasured
asofJune2009.Thefirstcolumnshowsthepresentvalueofliabilitiesunderstatechosenrates.Thesecond
columnshowsliabilitydiscountedattaxablemunirates,basedontaxablemunicipalzerocouponyieldcurvesasof
June30,2009.Themunicipalyieldcurvesarestatespecificwhereavailable,otherwisetheyarematchedtothe
ratingcategoriesofthestates,andthetaxadjustmentisdonebygrossingupyieldsby25%.Thethirdcolumn
discountsthecashflowsusingtheTreasuryzerocouponyieldcurveasofJune30,2009.Thefourthcolumn
discountswiththetaxablemunirateplusa51bpsalaryriskpremium,andthelastcolumndiscountswiththe
Treasuryrateplusa51bpsalaryriskpremium.Thesalaryriskpremiumisonlyaddedtotheactiveworker
component,anditisalsonotaddedtotheABOsincetheABOisinvarianttofuturesalarygrowth.SeeNovyMarx
andRauh(2010)forfurtherdetails.
State
Chosen
(Usually
8%)
Taxable
Muni Treasury
Taxable
MuniPlus
WageRisk
Where
Applicable
Treasury
PlusWage
Risk
Where
Applicable
Total(Active+Annuitants+Separated)
ABO $2.76 $3.20 $4.43 $3.20 $4.43
PBO $3.07 $3.53 $5.11 $3.41 $4.86
EAN $3.15 $3.62 $5.28 $3.49 $5.01
PVB $3.75 $4.24 $6.86 $4.04 $6.36
ActiveParticipants
ABO $1.09 $1.23 $1.94 $1.23 $1.94
PBO $1.40 $1.56 $2.62 $1.45 $2.37
EAN $1.48 $1.65 $2.80 $1.53 $2.52
PVB $2.08 $2.27 $4.37 $2.07 $3.87
Annuitants $1.58 $1.87 $2.28 $1.87 $2.28
SeparatedNotYet $0.09 $0.10 $0.20 $0.10 $0.20
ReceivingBenefits

24

Table4:DescriptionofBaselinePoliciesandPolicyChanges

Baseline
FullRetirementAge(FRA)=60
EarlyRetirementAge(ERA)=55
BuyoutRateforEarlyRetirement=6%perannum

SinglePolicyChanges
(1) COLA1% CostofLivingAdjustment(COLA)reducedbyonepercentagepoint
relativetowhateachstatesystemcurrentlyuses.
(2) AFER ImplementsActuariallyFairEarlyRetirement(AFER).Ifanemployee
retiresearly,benefitsarereducedsothatthepresentvalueofbenefits
equalsthepresentvalueiftheemployeeweretodelaycollecting
benefitsuntilthenormalretirementage.
(3) FRA/ERA+1year Raisesthenormalretirementageandtheearlyretirementagebyone
yeareach.
(4) EliminateCOLA CostofLivingAdjustment(COLA)reducedtozero.

MultiplePolicyChanges
(5) AFER,FRA/ERA+1year (2)and(3)combined
(6) FRA=ERA=65 RaisesbothFullRetirementAge(FRA)andEarly
RetirementAge(ERA)to65years
(7) COLA1%,FRA=ERA=65 (1)and(6)combined
(8) SocialSecurityRetirementAges FRA=67,ERA=62,andimplementsearlyretirement
agebuyoutsthataresimilartothoseusedbySocial
Security(whichareclosertoactuariallyfair)

25

Table5:SinglePolicyChangesunderTaxableMuniDiscounting

Baseline COLA1% AFER

FRA/ERA+
1Year
Eliminate
COLA
PanelA:TrillionsofDollars
Total(Active+Annuitants+Separated)
ABO $3.20 $2.92 $3.08 $3.12 $2.49
PBO $3.53 $3.22 $3.38 $3.43 $2.75
EAN $3.62 $3.30 $3.46 $3.51 $2.82
PVB $4.24 $3.86 $4.02 $4.09 $3.29
ActiveParticipants
ABO $1.23 $1.12 $1.13 $1.16 $0.96
PBO $1.56 $1.42 $1.43 $1.47 $1.21
EAN $1.65 $1.50 $1.52 $1.56 $1.28
PVB $2.27 $2.06 $2.08 $2.13 $1.76
Annuitants $1.87 $1.71 $1.87 $1.87 $1.47
SeparatedNotYet $0.10 $0.09 $0.08 $0.09 $0.07
ReceivingBenefits
PanelB:PercentRelativetoBaseline
Total(Active+Annuitants+Separated)
ABO 8.7% 3.7% 2.6% 22.1%
PBO 8.7% 4.1% 2.9% 22.1%
EAN 8.7% 4.2% 3.0% 22.1%
PVB 8.9% 5.0% 3.5% 22.3%
ActiveParticipants
ABO 9.2% 8.1% 5.9% 22.5%
PBO 9.2% 8.2% 5.9% 22.5%
EAN 9.2% 8.2% 5.9% 22.4%
PVB 9.3% 8.6% 6.1% 22.6%
Annuitants 8.2% 0.0% 0.0% 21.6%
SeparatedNotYet 11.1% 17.9% 11.0% 28.4%
ReceivingBenefits

26

Table6:SinglePolicyChangesunderTreasuryDiscounting

Baseline COLA1% AFER

FRA/ERA+1
Year
Eliminate
COLA
PanelA:TrillionsofDollars
Total(Active+Annuitants+Separated)
ABO $4.43 $3.96 $4.37 $4.31 $3.27
PBO $5.11 $4.56 $5.03 $4.95 $3.76
EAN $5.28 $4.71 $5.20 $5.12 $3.89
PVB $6.86 $6.10 $6.74 $6.61 $5.03
ActiveParticipants
ABO $1.94 $1.72 $1.89 $1.84 $1.41
PBO $2.62 $2.32 $2.56 $2.48 $1.91
EAN $2.80 $2.47 $2.73 $2.65 $2.04
PVB $4.37 $3.86 $4.27 $4.14 $3.17
Annuitants $2.28 $2.06 $2.28 $2.28 $1.72
SeparatedNotYet $0.20 $0.18 $0.20 $0.19 $0.14
ReceivingBenefits
PanelB:PercentRelativetoBaseline
Total(Active+Annuitants+Separated)
ABO 10.6% 1.4% 2.7% 26.2%
PBO 10.8% 1.5% 3.0% 26.4%
EAN 10.8% 1.5% 3.1% 26.4%
PVB 11.0% 1.7% 3.7% 26.7%
ActiveParticipants
ABO 11.5% 2.6% 5.3% 27.1%
PBO 11.5% 2.5% 5.3% 27.2%
EAN 11.5% 2.5% 5.2% 27.2%
PVB 11.6% 2.4% 5.4% 27.4%
Annuitants 9.6% 0.0% 0.0% 24.7%
SeparatedNotYet 13.7% 5.0% 8.3% 34.4%
ReceivingBenefits

27

Table7:MultiplePolicyChangesunderTaxableMuniDiscounting

Baseline
AFER,and
FRA/ERA
+1Year
FRA=
ERA=65
COLA1%
andFRA=
ERA=65
Social
Security
Parameters
PanelA:TrillionsofDollars
Total(Active+Annuitants+Separated)
ABO $3.20 $2.99 $2.64 $2.42 $2.50
PBO $3.53 $3.27 $2.83 $2.60 $2.66
EAN $3.62 $3.35 $2.89 $2.65 $2.71
PVB $4.24 $3.86 $3.24 $2.98 $2.99
ActiveParticipants
ABO $1.23 $1.05 $0.72 $0.67 $0.60
PBO $1.56 $1.33 $0.92 $0.85 $0.76
EAN $1.65 $1.41 $0.97 $0.90 $0.80
PVB $2.27 $1.92 $1.33 $1.22 $1.09
Annuitants $1.87 $1.87 $1.87 $1.71 $1.87
SeparatedNotYet $0.10 $0.07 $0.04 $0.04 $0.03
ReceivingBenefits
PanelB:PercentRelativetoBaseline
Total(Active+Annuitants+Separated)
ABO 6.5% 17.5% 24.3% 21.9%
PBO 7.4% 19.8% 26.3% 24.6%
EAN 7.5% 20.2% 26.7% 25.2%
PVB 8.9% 23.5% 29.8% 29.3%
ActiveParticipants
ABO 14.9% 41.3% 45.9% 51.6%
PBO 15.0% 41.2% 45.9% 51.6%
EAN 14.9% 41.0% 45.7% 51.3%
PVB 15.4% 41.6% 46.3% 52.0%
Annuitants 0.0% 0.0% 8.2% 0.0%
SeparatedNotYet 26.6% 54.0% 57.9% 64.7%
ReceivingBenefits

28

Table8:MultiplePolicyChangesunderTreasuryDiscounting

Baseline
AFER,and
FRA/ERA
+1Year
FRA=
ERA=65
COLA1%
andFRA=
ERA=65
Social
Security
Parameters
PanelA:TrillionsofDollars

Total(Active+Annuitants+Separated)
ABO $4.43 $4.24 $3.73 $3.37 $3.38
PBO $5.11 $4.86 $4.20 $3.79 $3.74
EAN $5.28 $5.03 $4.32 $3.90 $3.84
PVB $6.86 $6.47 $5.39 $4.87 $4.65
ActiveParticipants
ABO $1.94 $1.78 $1.32 $1.19 $1.01
PBO $2.62 $2.40 $1.79 $1.62 $1.37
EAN $2.80 $2.57 $1.91 $1.73 $1.47
PVB $4.37 $4.01 $2.99 $2.70 $2.28
Annuitants $2.28 $2.28 $2.28 $2.06 $2.28
SeparatedNotYet $0.20 $0.18 $0.13 $0.11 $0.09
ReceivingBenefits
PanelB:PercentRelativetoBaseline
Total(Active+Annuitants+Separated)
ABO 4.3% 15.8% 23.9% 23.6%
PBO 4.8% 17.8% 25.8% 26.7%
EAN 4.8% 18.2% 26.1% 27.3%
PVB 5.7% 21.4% 29.0% 32.1%
ActiveParticipants
ABO 8.4% 31.9% 38.5% 47.7%
PBO 8.3% 31.7% 38.3% 47.6%
EAN 8.2% 31.5% 38.1% 47.3%
PVB 8.3% 31.7% 38.3% 47.7%
Annuitants 0.0% 0.0% 9.6% 0.0%
SeparatedNotYet 12.6% 38.6% 44.7% 56.6%
ReceivingBenefits

29

Figure1:EffectsofPolicyChangesonLiabilities,EvaluatedattheTaxableMuniRate

Figure2:EffectsofPolicyChangesonLiabilities,EvaluatedattheTreasuryRate

$0.00
$1.00
$2.00
$3.00
$4.00
$5.00
$6.00
$7.00
ABO PBO EAN PVB
baseline
COLA 1%
AFER
RetirementAge+1year
EliminateCOLA
$ trillions
$0.00
$1.00
$2.00
$3.00
$4.00
$5.00
$6.00
$7.00
ABO PBO EAN PVB
baseline
COLA 1%
AFER
RetirementAge+1year
EliminateCOLA
$ trillions