PUMA/SBO(2001)4

TABLEOFCONTENTS

1. Macrobudgeting:FromStimulustoConstraint 4

2. MacrobudgetaryFutures:NeutralizingthePoliticalPressuretoSpend 8

3. DistributiveBudgeting:FromAllocationtoReallocation 13

4. DistributiveBudgetFutures:WhoWillGetWhat? 17

5. OperationalBudgeting:FromControltoPerformance 22

6. OperationalFutures:GovernmentasaProducer 27

7. DoesBudgetingHaveaFuture? 30

2
 PUMA/SBO(2001)4

DOESBUDGETINGHAVEAFUTURE?

by Allen Schick

Budgetingisaworkinprogress.Theprocessisneverquitesettledbecausethosewhomanageitarenever
fully satisfied. To budget is to decide on the basis of inadequate information, often without secure
knowledgeofhowpastappropriationswereusedorofwhatwasaccomplished,oroftheresultsthatnew
allocationsmayproduce.Mostpeopleinvolvedinbudgetinghaveexperiencedthefrustrationofhaving
theirpreferencescrowdedoutbythebuiltincostofpastactions.Budgetingisadeadlinedrivenprocess,
inwhichsuboptimaldecisionsoftenarethenormbecausegovernmentdoesnothavetheoptionofmaking
nodecisions.Whenonecycleends,thenextbegins,usuallywithlittlerespiteandalongthesamepaththat
wastrodtheyearbefore.Theroutinesofbudgetingdullconflict,buttheyalsoareabreedinggroundfor
frustration.

Withinthispressuredworld,thosewhomakebudgetsorareaffectedbythemyearnforamorerationaland
orderlyprocess.ReformistheHolyGrailofbudgetpeople,theirunendingquestforabetterwaytoparcel
outmoneyandplantheworkofgovernment.Sometimestheyembracebigbangreforms,suchasplanning
programming budgeting systems (PPBS) and zerobase budgeting; usually, however, they strive for
incremental adjustments in one or another element of the process. Tinkering is ongoing because the
adoptedchangesrarelyproducethepromisedimprovements.Inbudgeting,thefailureofonereformbegets
anotherreform.

Ifbudgetingisaworkinprogress,doesithavemuchprogresstoshow?Haveproceduralreformschanged
thewaybudgetsarecompiledorimplemented?Theclearevidenceisthattherehasbeensignificantchange
bothinthebudgetpracticesofmostOECDcountriesandintheirbudgetpolicies.Inmostcountries,items
ofexpenditurehaverecededinprominenceandnowareconsolidatedintobroadcategories.Nowadays,the
budgethasmoreinformationonprogramsandperformanceanditstimehorizonhasbeenextendedbeyond
asinglefiscalyeartothemediumterm.Overthepasttwodecades,budgetinghasbeenmoreclosely
integrated with other financial management processes including accounting systems and financial
statements.Onthepolicyfront,spendinggrowthisslowerthanitwasduringthepostwarperiod,but
entitlementsclaimalargershareofcentralgovernmentfinancialresources.Nevertheless,revenuesand
expendituresareclosertobalanceinmostcountries,thoughfewhavemanagedtosustainabalanced
budgetthroughanentireeconomiccycle.Reformhasmadeadifference,butnotaverybigone.

Mytaskinthispaperistoimaginehowthepracticeofbudgetingmightevolveintheyearsahead.As
someonetrainedindataanddetail,Ifinditdifficulttocontemplatethefuturewithoutconnectingittothe
presentandthepast.Budgetingisincurablyincremental,notonlyintheamountsallocatedforparticular
purposes,butalsointheadjustmentsmadefromtimetotimeinitsoperatingrulesandprocedures.To
comprehendwherebudgetingisheading,onemustknowwhereitisnowandhowitgothere.Tofacilitate
discussion,thelaunchofSBOin1980isusedasabreakpointintheevolutionofbudgetsystems.The
establishmentofSBOnotonlyprovidedaforumfortheexchangeofideas,butwasitselfpartofthe

3
PUMA/SBO(2001)4

movementtoreconsiderbudgetpractices.Ifbudgetinghaschanged,itisnotbecauseSBOexists;rather,
SBOexistsbecauseoftheimpulsetochangebudgeting.

Inconsideringhowbudgetinghaschanged,itisusefultodistinguishbetweenthreetypesofinnovation:
macrobudgetary, distributive, and microbudgetary reform. Macrobudgeting deals with the budget
aggregatesandwiththemaintenanceoffiscaldiscipline;distributionalissuespertaintotheallocationof
costs and benefits through budget decisions; microbudgeting is concerned with the operation of
governmentprogramsandagencies.Thisclassificationissimilartothethreelevelstructureofbudgeting
devisedby the WorldBank that identifies the core functions as aggregate fiscal discipline, allocative
efficiency,andoperationalefficiency. Withrespecttoeachofthesefunctions,thepaperfirstdescribesand
assessesrecentinnovations,andthencontemplateshowbudgetingmightevolveinthedecadesahead.

Mysenseisthatthenexttwodecadesmaybringmorefundamentalchangethanoccurredinthetwo
previousdecades,possiblythroughbroadpoliticalandtransnationaldevelopmentsratherthanthrough
frontaleffortstoalterbudgetaryprocedures.Somechangeswillbethenaturalprogressionofdevelopments
alreadyunderway,otherswillemanatefromchangesinrelationsamonggovernments.Whilediscussionof
thefutureisinherentlyspeculative,evidenceformostofthepossibilitiesdiscussedherecanbefoundin
avantgardeideasgerminatinginthefertilemindsofbudgetscholarsandpractitioners.

Theconcludingsectionturnstothequestioninthetitleofthispaper:Doesbudgetinghaveafuture?The
question may appear trite and the answer obvious, for unless governments fade away, government
budgetingisheretostay.Thereislittlereasontobelievethatgovernmentmaybemarkedlysmallerinthe
future;itismorelikelytobesomewhatbigger.Yetthequestionwellsoutoftheanalysisinthepaper.The
trends and possibilities discerned in the paper suggest a future in which budgets may be bigger but
budgetingweaker.Thisisaprospectthatshouldconcernthosetowhomtheprocessesofbudgetinghave
beenentrusted.

1. Macrobudgeting:FromStimulustoConstraint

WhenthefirstSBOmetin1980,itwasanimatedbyasensethatoneerainpublicfinancehadendedand
anotherhadjustbegun.Duringtheprevioushalfdozenyears,OECDcountriesexperiencedoilshocks,
stagflation (low growth and high inflation), rising unemployment and escalating fiscal deficits. These
adverseconditionswereamarkeddeparturefromthebuoyantconditionsthatcharacterizedthepostwar
boom,duringwhicheconomicgrowthpropelledgovernmentexpansion.

Thepostwarexpansionissummedupinafewstatistics.In1960,OECD'sfirstyear,totalgovernment
outlaysofOECD countiesaveraged29percentofGDP;twodecadeslater,theseoutlaysaveraged40
percentofGDP,anincreaseofmorethanonehalfpercentagepointayear.Mostofthegrowthwasin
social security transfers whichdoubledfrom 7percent of GDP to14percent. The huge expansionof
government was not happenstance; it was promoted by governments which reengineered their budget
practices to boost spending. Some national governments discarded or revised longstanding balanced
budgetrules,andmanyembracedtheKeynesiandoctrinethatfiscalpolicyshouldaimtostabilizethe
economy,eveniftheshortruneffectwastodestabilizethebudget.Manygovernmentseasedthelineitem
controlstheyhadimposedtoconstrainspending,andpermittedorencouragedspendingdepartmentsto
actively bid for additional resources in the annual budget process. Virtually all developed countries
weakened budget discipline by enacting legislation mandating the expenditure of funds on various
entitlementprograms.Notonlydidgovernmentsexpandprogramsandspendmoreeachyear,butpolitical

4
 PUMA/SBO(2001)4

leaderswereexpectedtosubmitexpansivebudgets.Infact,thevigoroftheirleadershipwasoftenassessed
intermsofthenewandexpandedprogramsincludedintheirbudgets.

Economicexpansionfueledsignificantincreasesingovernmentrevenues,buttheseoftendidnotsufficeto
coverexpenditures.Whenrevenueswereinsufficient,manygovernmentseitherraisedtaxesorransmall
deficits.Neithercourseofactionentailedmuchpoliticalrisk,andneitherwasseenasfiscallyirresponsible.

Wheneconomicconditionsdeterioratedinthe1970s,publicbudgetswerestillonanexpansionarycourse.
There was an unsustainable imbalance between the momentum of the budget and the capacity of
governmentstomaintainaprudentfiscalcourse.Thiswritercommentedonthepredicamentinapaper
presentedatthe1981SBOforcemajeurerulespublicfinance.

AttheearlySBO's,themosturgenttaskfacingbudgetofficialswastocurtailrunawayspendinggrowth.
Manygovernmentssoughttoconstrainpublicspendingbyimposingfiscaltargetsinadvanceoftheannual
budgetpreparationcycle.Someexpressedthesetargetsinnominalterms,othersdidsoasashareofGDP.
Sometargetedoutlays,othersfocusedontotalrevenueorthefiscaldeficit.Theearlytargetswereset
independently of the budget process without careful consideration of revenue capacity or spending
pressure,orofthestepsneededtoimplementthem.Fewwereaccompaniedbyenforcementmechanisms
orotherchangesinbudgetrules.Thetargetswere politicalstatementswhoseprimarypurposewasto
signalthattheeraofunconstrainedgrowthwasover.Afewhadtheintendedeffectandledtosmaller
deficitsorlowerspendinggrowth.However,manyweresetatunrealisticlevelsandhadtobeabandoned
farshortoftheirstatedgoal.Insomecountries,theabsenceofeffectiveenforcementmechanismsspurred
wilypoliticians,oftenwiththeassistanceofbudgetexperts,todevisecreativemeansofsubvertingthe
targetsbyhidingthetruevolumeofpublicspendingorofthedeficit.

Secondgenerationtargetsthatemergedinthelate1980sandtheearly1990shavebeensomewhatmore
effective to the extent they are linked to budget decisions and are enforced through procedural or
substantiveconstraintsongovernmentaction.Themostimportantrecentinnovationshaveincludedfiscal
targets encased in mediumterm expenditure frameworks, targets imposed by international treaties or
organizations,accrualaccountingandbudgeting,baselineprojectionsusedtoestimatebudgetimpactsof
currentactions,codesoffiscalresponsibilitythatrequiretransparencyingovernmentfiscalpolicyand
pronouncements,andatwostepbudgetprocessthatseparatesdecisionsontheaggregatesfromthoseon
particularitemsorprograms.Eachoftheseisbrieflydescribedintheparagraphsthatfollow.

1.1 MediumtermExpenditureFramework(MTEP)

AnMTEFisanarrangement inwhichannualbudgetdecisionsaremadeintermsofaggregateorsectoral
limitsonexpendituresforeachofthenext35years.AustralialedthewayintheMTEFmovementduring
the1980sbyexpandingitsforwardestimatesintomultiyeartargetsthatrapidlygainedstandingasthe
basis on which spending departments bid for resources and the annual budget is compiled. In the
Australianmodel,theMTEFwasnotgraftedontothepreexistingbudgetprocess,itbecamethebudget
process,thatis,themeansbywhichgovernmentparcelsoutmoneytospendingportfolios.Mostofthe
expendituredetailcompiledintheannualestimatesweredevolvedtodepartments,therebyfreeingupthe
cabinetforpolicyworkontheaggregates,allocationstothemainbudgetsectors,andpolicychanges.

InAustralia,theforwardestimates(whicharemediumtermprojectionsoftherevenueorspendingthat
willensuefrom approvedpolicies) arerolledaheadeachyear andare updatedfor price changes,re
estimatesofrevenueandofprogramexpenditure,andpolicyinitiatives.Withoutthedirectinvolvementof

5
PUMA/SBO(2001)4

politicalleaders,theMTEFwouldbelittlemorethanatechnicalexercise,moreamatterofprojections
thanofpolicydecisions.Moreover,withoutafirmcommitmenttoconstrainfuturespending,theforward
estimateswouldberegardedbygovernmentasanentitlementtomoremoneyinfutureyearsandasa
weapontowrestbiggerallocationsfromthebudget.

ThereputedsuccessofAustralia'sreformshasspurredinterestintheMTEF.Butalthoughmanycountries
claimtobeapplyingamediumtermexpenditureframework,fewuseitinthemannerintendedbyits
architects.Insomecountries,itislittlemorethanamultiyearprojection,inothers,itisatechnocratic
exercisethatdoesnotinvolvestrategicdecisionsbypoliticalleaders.

1.2 TheInternationalizationofFiscalTargets

The conventional wisdom that targets are effective only when they are imposed by governments on
themselves has been challenged by the emerging role of international organizations in devising and
enforcingfiscalconstraints.TheleadingdevelopmenthasbeentheapplicationofMaastrichtsconvergence
criteriatothefiscalaggregatesoftheMembercountriesoftheEuropeanMonetaryUnion.Althoughthe
targetswereagreedin1992,theydidnotbecomefullyeffectiveuntiltheendofthedecadewhentheEuro
wasintroduced. Atlaunch, properenforcement mechanisms werenotfullyinplaceandopportunistic
politiciansmaneuvered,withsomesuccess,tooutwitthenewrules.Butwitheachyearthatthecriteria
havebeeninplace,EMUgivesevidenceofastronger,morevigilantmonitoringcapacity.Nevertheless,
thelongtermeffectivenessofexternallyimposedrulescannotbegaugeduntiltheaffectedcountrieshave
gonethroughafullfiscalcycle.TherulestookeffectduringanupturninEurope'seconomicfortunes,
therebyeasingthepoliticalburdenofmeetingthem.Thetruetestofexternallyimposedtargetswillcome
duringperiodsofeconomicweakness.

A parallel, though less structured development has been the imposition of fiscal conditionalities on
financiallytroubledcountriesdrawingassistancefromtheInternationalMonetaryFund.Typically,IMF
insists that aided countries meet specified fiscal targets and it enforces compliance by releasing each
trancheonlyifitissatisfiedwiththeaffectedcountrysperformance.Thetargetsusuallydemandfiscal
contraction, as measured bythesize of the budget deficit or by the growthof public spending. IMF
conditionalitieshavecomeunderstrongcriticismfromthosewhobelievethatitsfiscaldemandshave
mademattersworseforcountrieswhoseeconomiesarealreadycontracting.Withagrowingchorusof
critics,itremainstobeseen whetherIMFwillcontinuetoobtainfiscalconstraintfromcountriesdrawing
assistancefromit.

1.3 AccrualAccountingandBudgeting

Thepointhasalreadybeenmadethatfiscaltargetsaremeaninglessiftheyarenoteffectivelyenforced.An
essentialfeatureofenforcementisthatfinancialflowsandconditionbeaccuratelyrecorded.Tosatisfythis
condition,agovernmentmustconvertfromcashaccountingtotheaccrualbasis.Ideally,itshouldalsoput
itsbudgetontheaccrualbasis.

Cashaccountingandbudgetingprovideabundantopportunitytomisreportthegovernment'struefinancial
actions.Paymentsorreceiptscanbepushedbacktothepreviousyearorforwardtothenext;futurecosts,
suchaspensionbenefitscanberemovedfromthebudget;thegovernmentcansellassetsandbookthe
incomeascurrentrevenue;itcanhiveoffcertainexpensesintooffbudgetaccounts;andsoon.Leftto
theirownwills,budgetpressuredpoliticianstendtoengageinasmuchbudgetarylegerdenainastheycan

6
 PUMA/SBO(2001)4

get awaywith. Theyissueguarantees andtake onother contingent liabilities whichdonot appearin

conventional financial reports or budgets; they sweep idle cash from state enterprises into their own
accounts;theyshortenthefiscalyearto11monthsorexpanditto13;theyuseunrealisticassumptionsin
makingappropriationsorinprojectingfuturebudgetconditions.Theopportunitiesareasboundlessasare
theimaginationsofbudgetevaders.

Theaccrualbasisinhibitstheseopportunitiesbyrequiringthatrevenueberecordedwhenitisearned
(ratherthanwhenitisreceived)andexpenditureswhentheliabilityisincurred(ratherthanwhenpayment
ismade).Agrowingnumberofcountrieshaveadoptedtheaccrualbasisforfinancialstatements,but
barelyahandfulhavedonesoforthebudget.Asaconsequence,somegovernmentsnowreportbudget
resultsthatdiffermateriallyfromthosereportedinfinancialstatements.

1.4 BaselineProjections

OneoftheobjectivesoftheMTEFistoextendthebudget'shorizonbeyondasinglefiscalyear.Doingso
dependsonreliableprojectionsofmacroeconomicconditions,futurerevenueandspendingsifcurrent
policieswerecontinued,andtheimpactofpolicychangesonfuturebudgets.Theconventionalmethodfor
makingtheseprojectionsistoconstruct abaselinebudget andtomeasurepolicychanges against the
baseline.Budgetorganizationshavebeenmakingtheseestimatesfordecades,butwhatisnowdifferentis
thattheprojectionsarepublishedandhavebecometheauthoritativemetricforassessingthefuturebudget
effectsofproposedorapprovedpolicychanges.

In the MTEF, the baseline is used both to establish the fiscal framework and to determine whether
expenditurechangesareconsistentwiththeframework.Inasmuchasfutureconditionsarenotyetknown,
the baseline and estimates of policy change are grounded on assumptions concerning economic
performance, the behavioral responses of persons affected by policy changes, and other variables.
Countrieswhichusebaselinestoestablishandenforceexpenditureframeworksmusthaverulesforhow
the projections are made and policy changes are measured as well as procedures for dealing with
deviations from the baseline. They must also assign responsibility for maintaining the baseline and
assuringthatpolicychangesareaccuratelymeasuredagainstit.Inafewcountries,managingthebaseline
andrelatedcontrolshasbecomethefinanceministry'smostimportantbudgetresponsibility.

1.5 TwostageBudgetProcesses

Inconventionalbudgeting,fiscaltargetsareperenniallyatriskofbeingoverriddenbyspendingpressures
duringpreparationofthebudget.Inallbudgetsystems,thereisatensionbetweenthebudget'stotalsand
itsparts.Duringexpansionaryperiods,theparts(programs,departments,accounts,etc.)usuallywinout,
withtheresultthatbytheendoftheprocessthegovernmentagreestospendmorethanitintendedatthe
start.OneobjectiveoftheMTEFistopreventbreachofthefiscalaggregatesbyinsulatingthemfrom
upwardspendingpressurefromparticularprograms.Butwhenthebudgettotalsaredecidedtogetherwith
theparts,itishighlylikelythattotalspendingwillbemorethanthegovernmentinitiallywanted.

Tocounterthistendency,afewgovernments,suchasSweden,havesplitbudgetpreparationintotwo
distinct phases that are several months apart. In the first stage, the government updates baseline
projections,reviewscurrentmacroeconomicforecastsandthebudgetoutlook,andestablishesamedium
termexpenditureframework.Thisframeworkmaydealonlywiththeaggregatesor(asisthecasein
Sweden)withmajorspendingsectorsaswell.Atthispointintheprocess,noformalconsiderationisgiven

7
PUMA/SBO(2001)4

to particular claims on the budget. The macro budget is submitted to the legislature which votes the
spendingtotals.Severalmonthslaterthegovernmentcompilestheestimatesforthenextfiscalyear.These
spendingamountsmustbewithintheaggregatespreviouslysetbythegovernmentorthelegislature.Rules
similar to those used in maintaining the MTEF are applied to estimate the consistency of particular
spendingdecisionswiththebudgetaggregates.

1.6 HaveTheTargetsMadeADifference?

Havethefiscaltargetsandrelatedinnovationsdisciplinedpublicfinance,oraretheythelatestinalonglist
ofbudgetreformsthathavemadelittledifference?TheOECDhasaddressedthisquestionbycomparing
spendingasaproportionofGDPincountriesthatweremembersoftheOrganisationbeforethefirstSBO
convened.Thestudyshowsmarkeddecelerationinspendinggrowthandmoremodestfiscalimbalances
thaninpreviousperiods.Buttheriseinpublicspendinghasnotbeenreversed;virtuallyeverycountry
reportsthatsuchspendingisnowahighershareofGDPthanwasthecaseoneortwodecadesago.Theera
ofbigincreasesisover,butgovernmentbudgetingstillisincremental,buildingthenextbudgetonthe
resultsofthelastone.Asaconsequence,governmentspendingaccountsforahigherproportionofGDP
thanwhenSBOwasestablished.Tomeettargets,governmentshavetrimmedsomeentitlementprograms
attheedgesandhavemadedeepcutsinsubsidiestoenterprises.Mosthavetrimmedoperatingcostsby
imposingacrosstheboardcutsorbyseekingtoboostefficiency.Theyalsohavemadegreateruseoffees
andotherchargestofinanceparticularprograms.Fromthevantagepointofvoterpleasingpoliticians,
manyofthecutbackshaveentaileddifficultchoices,buttheirchoremayhavebeeneasedsomewhatbythe
newrulesandproceduresdiscussedinthissection.Yetafairassessmentoftwodecadesoftinkeringwith
themachineryofbudgetingisthattheprocessisnotallthatdifferentthanitwasbeforeSBO,andthatthe
underlyingpressuresandimbalancesofpublicfinancehavenotbeeneliminated.

2. MacrobudgetaryFutures:NeutralizingthePoliticalPressuretoSpend

ThereisastrongprobabilitythatpublicspendingwillcontinuetoriseasapercentageofGDPindeveloped
countries as their populations age. The pace of increase will dependprincipally on the cost of social
insuranceandrelatedprograms,aswellasonoveralleconomictrends.Willtherateofincreasebesimilar
tothatexperiencedduringthepreSBOdecades,orwillitresemblethesmallerriseofthe19802000
period?Ofcourse,evenamodestincreasewouldbeatopspendingwhichalreadyexceeds50percentof
GDPinmanyEuropeancountries.ThismeansthatifOECDcountriesdoaboutaswellinthenexttwo
decadesastheydidintheprevioustwo,theywillbedoingworse.Moreover,theriskthattheywilldo
muchworseisfargreaterthantheprobabilitythattheymightdoalittlebetter.

ThevulnerabilityofOECDcountriespertainsnotonlytototalexpenditurebutalsototheirfiscalbalance.
To finance the escalating costs of social schemes and the stillrising expectations of citizens, OECD
governments willhavetoextractmorerevenue from enterprises andhouseholds thantheyhave done
heretofore.Ashighasthetaxburdenalreadyis,itmayhavetobeconsiderablyhigherinthedecades
ahead.

Governmentsmaymanagetoescapefiscalmeltdownbyreversinglongstandingsocialpolicyprivatizing
socialinsuranceschemesorstringentlytargetingbenefitstotheirneediestcitizens.Thesemoveswouldroll
backtheboundariesofthewelfarestatetoafargreaterdegreethanhasbeenattemptedbyanydemocratic
leaderinrecenttimes.Pushedtothewallbytherealityorprospectofrisingtaxburdensorhugebudget
deficits,thenextgeneration'spoliticiansmayaccedetosuchdraconianmeasuresastheironlyrecourse.

8
 PUMA/SBO(2001)4

Mysense,however,isthatweareinforanotherlongspellofmuddlingthroughofchronicfiscalpressure
that,likechronicbackpain,getsattentionbutnotsomuchastocompelamarkedchangeinlifestyle.In
muddlingthrough,governmentswillstruggletoconstrainthefiscalaggregateswithinsomeacceptable,
thoughexpanding,envelope.Theirtaskwillbefacilitatedbychangesinboththeinformationcontentand
proceduresofbudgeting,someofwhicharedescribedintheparagraphsthatfollow.Butevenwithrule
changesthatarmguardiansofthepublicpursewithnewauthority,fiscaldisciplinewillbeadifficult
politicalandeconomicchore.Politicianswillhavetheirfingerprintsonthecontrols,andtheywillgarner
blameformattersoverwhichtheyhavelittlegenuinecontrol.

2.1 TheTimeFrameOfBudgeting

Ruttedinafiscalquagmirethatseemsunending,futurebudgetmakerswillregardthemediumtermas
inadequatefortheirworkjustascontemporaryMTEFarchitectsregardthesinglefiscalyearasinadequate
fortheirs.Fromtomorrowsvantagepoint,the23yearstackedontotheendoftheannualbudgetcycleby
theMTEFwillbeseenasatoomodestventurethatobscuresthelongertermoutlook.Theconventional
MTEFisacoupleofyearsshorterthanthestandard5yearplanninghorizon.

Thistimeframewasselectedtoassurethatmediumtermdecisionsarerealisticandrelevanttoannual
budgetactions.GovernmentsoptingforanMTEFhavebeenjustifiablyconcernedthatthelongertheir
budgetprojectionsareextended,thegreatertheuncertaintytheywillencounter.Theyalsoareconcerned
thatitisimpracticaltomakelongertermdecisionsbecauseaseachyearisrolledforward,seemingly
resolvedissuescanbereopened.

Despite these sound arguments for a mediumterm perspective, my sense isthat as governments face
increasedpressuretoregulatepublicfinance,theywilllengthenthebudget'stimehorizonto5ormore
years.TheUnitedStateshasmovedtoa10yearbaselineforestimatingtheimpactofcurrentpolicy
changesonfuturebudgets.Thelongerframemakesitsomewhatmoredifficultforpoliticianstoevade
fiscaldisciplinebydelayingtheeffectivedateofpolicychanges.Inmostdevelopedcountries,thelonger
horizonportraysdeterioratingfiscalcondition,astheproportionofthepopulationdrawingpensionsand
otherstatefinancedbenefitsrises.Thereissomerisk,however,thatlengtheningtheperspectivemight
introducegreaterunrealismintobudgetprojections.Ifthisweretooccur,a10yeartermorlongerhorizon
wouldunderminefiscaldiscipline.

2.2 FiscalSustainabilityAndVulnerability

Thetimeframeisnotonlyamatterofyearsbutofperspectiveaswell.Extendingtheperspectivefromone
yeartoseveralandfromseveraltoadecadetransformsthecentralfiscalissuefromthecurrentbalanceto
futuresustainabilityandtothevulnerabilityofthebudgettosurprisesandshocks.Sustainabilityrefersto
thecapacityofgovernmenttocontinueonitspresentfiscalcourseinthelightofprospectiveeconomic
conditionsanditsrevenue/expenditureposition.Sustainabilityisalwaysforthelonghaul;interestinit
derivesfromrealizationthatalthoughacountrymayappeartobeingoodfiscalhealth,futureclaimsor
conditionsmayadverselyaffectitsbudgetcapacity.Forexample,mostEMUcountriesimprovedtheir
budgetpositionsduringtherunuptotheEuro,buttheyneverthelessfacethequestionofwhetherthe
financialcommitmentstheyhaveundertakenwillbesustainablewhenthepercentageofthepopulation
dependentonpensionsrises.

9
PUMA/SBO(2001)4

Totakesustainabilityseriouslyrequiresthatgovernmentsdevisemeasuresoftheircapacitytocontinueon
thepresentbudgetcourse.Thesemeasureswouldhavetogaugebuiltinexpenditurepressure,revenue
trends,sensitivityofthebudgettodemographicandeconomicchanges,and(iffeasible)theinfluenceof
political conditions on thebudget. Economists are more skilledat analysingtrends thanin predicting
changes,especiallythosethatoccurwithlittleadvancenotice.LookingbackattheAsianfinancialcrisisof
thelate1990s,however,economistshaveconcludedthattherewasadvanceevidencethatseveralcountries
were headed into trouble. They believe it possible to convert hindsight into foresight through bold
developmentandmeasurementofrisk.

Eveniftheearlymeasuresarecrudeandunreliable, theymayinfluencethefiscalpostureofnational
governments.Mediumtermprojectionsalsoareoftenwideofthemark,buttheystillframebudgetdebate.
Inbudgeting,questionabledataarepreferabletonodata,becausegovernmentmustact.Inmostpolicy
areas,governmentmaydeferactionuntilitissatisfiedwiththeinformationathand;itdoesnothavethis
optioninbudgeting.

Vulnerability analysis and other types of risk assessment will play an increasingly prominent role in
budgeting.Onecananticipatethatgovernmentswithabudgethorizonbeyondasinglefiscalyearwillbe
themostlikelyusersofsuchanalyses.Internationalorganizationsmayalsobecomeinvolvedbyassessing
fiscalperformanceinthelightofvulnerabilitycriteria.Forexample,EMUmightassessthebudgetsof
Membercountriesintermsofsustainabilityorriskfactorsratherthanintermsofannualormediumterm
outlooks.OnecanforeseecircumstancesinwhichEMUwilldemandthatamembercountryreviseits
budgeteventhoughshorttermcriteriahavebeenmet.

2.3 UsingTheBalanceSheetToMeasureFiscalCondition

Barelyadecadeago,afarfetchedideawasproposedthatacountry'sfiscaloutcomesbemeasuredin
referencetoitsbalancesheetratherthantothebudget.Whatseemedanovelapproachsomeyearsagomay
becomecommonplaceinthefuture,asnationalgovernmentsshiftfrombudgetbasedassessmentstoones
drawnfromconventionalfinancialstatements.Ifthisweretooccur,itwouldbeaCopernicanrevolutionin
financial management, transforming the traditional relationship between budgeting and other financial
processes.

Since its emergence more than a century ago, modern budgeting has operated within a largely
selfcontainedprocess.Decisionshavebeenmade,recordedandcarriedoutalmostexclusivelyinreference
tothedocumentationproducedbythebudgetprocess.Financialconditionandperformancehavebeen
measuredintermsoftheentriesinthebudget,withoutregardtoothersourcesoffinancialinformation.
Selfcontainment has walled off the budget from financial statements suchas balance sheets, and the
results shown in the budget often have differed materially from those reported elsewhere. In many
countries,financialstatementshaveadheredtoprevailingaccountingstandards,whilethebudgethasbeen
preparedinaccordwithitsownconventions.Financialstatementshavebeensubjecttoaudit;budgetshave
not.

Severalfactorsarepressingtobreakdownthedistinctionbetweenbudgetsandfinancialstatements.Oneis
theapplicationofnewaccountingstandards,includingtheaccrualbasis,ingovernmentfinance,anotheris
thepreviouslynotedeffortbyindependentauthoritiestoassesscountrybudgetpolicies,andathirdisthe
effort to strengthen financial management in government by integrating budgeting and accoun ting
activitiesandstandards.Thenaturalevolutionofthesechangesistorelyonfinancialstatementsinlieuof
thebudget.

10
 PUMA/SBO(2001)4

Intheir1991articleurgingthisapproach,BlejerandCheastyarguedthatthebalancesheetisthebest
available measureof thefiscal surplus or deficit. They notedthat the budget isnot a comprehensive
statementoffinancialflows;itisbydefinitionlimitedtothosetransactionsincludedinthebudget,andit
countsthesetransactionsinwaysthatarepeculiartobudgeting.Budgetsdonotincludeoffbudgetor
extrabudgetaryfunds,nordotheyaccuratelyvalueassets,liabilitiesandrisks.Thebalancesheetisa
broaderstatementthatincludesallofanentity'stransactions,whetherornottheyareincludedinthe
budget.Thebalancesheetrecognisesdepreciation,deferredcosts,andreservesfordefaultsandotherrisks.
Whenthegovernmentobtainsrevenuefromthesaleofassets,thebudgetreportsagaininrevenue,the
balancesheetshowsnochangeinnetassets.

Inapplyingthebalancesheettofiscaloutcomes,adeclineinnetworth(assetsminusliabilities)wouldbe
recordedasadeficit,anincreasewouldmeanasurplus.Buttousebalancesheetsasmeasuresoffiscal
outcomes,thegovernmentwouldhavetopublishtimelyfinancialstatements.Fewgovernmentsnowdo,
butmanymorecanbeexpectedtodosoasaccountingpracticesbecomestandardizedacrosscountries.

Thepotentialdisplacementofbudgetsbyfinancialstatementshasseveralimportantimplications.First,it
willgiveimpetustoeffortstoimposeaccountingstandardsongovernmentbudgets.Ifthismovement
succeeds, it will make little difference whether financial condition and performance are measured in
referencetothebudgetorfinancialstatements.Second,itthreatensdisplacementofthecentralbudget
agencybyotherfinancialmanagementexpertsingovernment.Toretainauthority,budgetspecialistswill
havetobecomemoreknowledgeableaboutaccountingpracticesandtheywillhavetopayattentionto
balancesheetsandotherstatements.Third,variousbudgetdecisionsarelikelytobemadeintermsoftheir
impactonthebalancesheetratherthanintermsofhowtheyarerecordedinthebudget.

2.4 BudgetingForContingentLiabilities

Whetherfromthevantagepointofthebudgetorthebalancesheet,futuregovernmentswilldealmore
forthrightlywithcontingentliabilities,suchasguaranteedloansandinsuranceprograms.Conventional
cashbased budgets do not properly record government exposure to various contingencies because no
paymentismadeuntiladefaultorothereventoccurs.Thebudgetdoesrecordthesepayments,butatthis
pointtheexpenditureisbeyondeffectivecontrol.Thefailuretoproperlybudgetforcontingentliabilities
inducesgovernmentstosubstituteguaranteesforconventionalexpendituresandtotakerisksthatmay
imbalancefuturebudgets.Thistemptationisespeciallystrongwhenthegovernmentseekstoimposefiscal
disciplinebycurtailingexpenditureorreducingthereporteddeficit.Variousstudieshaveshownthatfiscal
adjustmentinvitesevasion,withtheresultthatthereportedsavingsoftenareillusory.Althoughtrenddata
arenotavailable,thereisreasontobelievethatnationalgovernmentshaveincreasedtheirexposuretorisk
through various guarantee and insurance programs. Even though risktaking by governments may be
appropriate,thelackofadequateinformationmayinducethemtounderestimateorignorefuturecosts.

Aswithothershortcomingsinfiscaldiscipline,anearlystepmustbetoobtaininformationontheproblem,
followingwhichappropriatecontrolsmaybedevised.Thisisnotasimpletaskbecauseriskscomeinmany
forms and pervade modern economies. Moreover, contingent liabilities are outside the boundaries of
conventionalaccounting.Thedefinitionandrecognitioncriteriaappliedtogovernmentfinancialaccounts
pertaintodirectliabilities,nottothosedependentonuncertainevents,suchasdefaultsonloansorthe
collapseofbankinginstitutions.Thereisnoagreementamongthecountriesthathavesoughttoregulate
fiscalrisksontheappropriateapproach,orevenonwhetherthebudgetisthebestinstrumentforthis
purpose.ThefollowingparagraphsdescribethemethodsappliedbyNewZealand,Hungary,Canada,and
theUnitedStates.

11
PUMA/SBO(2001)4

As part of its reform of the public sector, New Zealand requires all departments to prepare audited
financialstatements.TheGovernmentthenproducesconsolidatedfinancialstatements,towhichitappends
statementsofquantifiableandnonquantifiablecontingentliabilities.Theformerarecontingencieswhose
cost can be estimated. These costs are not recorded in the budget or on the balance sheet, but the
informationdoesalertgovernmenttoprospectivepaymentsarisingoutofcontingentliabilities.

InHungary,thePublicFinanceActlimitsstateguaranteetoacertainpercentageofbudgetedexpenditures
butmostguarantees,includingthoseissuedbystateinstitutionssuchastheHungarianDevelopmentBank
orforvariousstrategicpurposes,areexemptedfromthelimit.Theannualbudgetappropriatesanamount
forexpectedcallsonguarantees,anditalsolimitsthevolumeofguaranteesthateachstateinstitutionmay
haveoutstandingduringtheyear.Whenguaranteesareissued,thegovernmentpublishesinformationon
estimatedrisk,thereasonsfortenderingtheguarantees,andotherrelevantmatters.Althoughthesystem
appears to be effective for the guarantees covered by it, some major contingent liabilities have been
excluded,withtheresultthatpaymentsoftenexceedtheamountsetasideinthebudgetforguarantees.

Canada subjects loan guarantees to a budget review that is similar to the scrutiny given to direct
expenditure. Each department proposing guarantees must provide an economic analysis of projected
benefitsandrisks,includingprojectionsofcashflowsanddebtservice.Fundsmustbereservedinthe
budget for possible losses; sponsoring departments finance these reserve from fee income or annual
appropriations.NewloanguaranteeprogramsmustbeapprovedbytheFinanceMinisterandauthorizedby
Parliament.Finally,departmentsandCrownCorporationsmustreportontheircontingentliabilities;these
arepublishedasnotestotheGovernment'sannualfinancialstatement.

Sincetheearly1990s,theUnitedStateshasappliedspecialbudgetrulestoloansandloanguarantees.It
replacedthepreviouspracticeofbudgetingforthesetransactionsonacashbasiswithoneinwhichthe
estimatedsubsidycostofloansandguaranteesisbudgetedasanoutlay.Subsidycostisdefinedasall
projectedcashinflowsandoutflowsdiscountedtopresentvalue.Inthecaseofloanguarantees,theinflows
may consist of origination fees and recoveries on defaults; the outflow are the payments made by
government pursuant todefault. Thus guarantees generate budget outlays, even thoughno funds may
actuallybedisburseduntillaterfiscalyears.

Althoughitmaytakesometime,thebudgetarytreatmentofguaranteesandothercontingentliabilitiesis
likelytobecomeincreasinglystandardized.Onceauditedfinancialstatementsbecomeprevalent,theymay
evolveintoinstrumentsforreportingonthecostofcontingentliabilities.Perhapsthemostsensibleoption
wouldbetoprovisionforestimatedlossesfromguaranteesonthegovernment'sbalancesheet.Ifactual
lossesdivergefromtheestimate,theamountprovisionedwouldbeadjustedtoreflectthisexperience.
Correspondingentrieswouldbemadeonthegovernmentsbudget,whichwouldshowestimatedoractual
lossesasexpenditures.

2.5 FiscalPolicyAndEconomicManagement

Steering the economy by means of the budget is one of the premier responsibilities of modern
governments.Politicalleaderspayattentiontofiscalmattersbecausetheirperformanceoftenismeasured
in terms of the performance of the economy. In government of the future, however, major fiscal
responsibilitiesmaybeentrustedtoanindependentauthoritywhichwouldbeempoweredtoadjusttax
rates and certain expenditures in response to projected or actual changes in economic conditions. As
outlandishasthisideamayseem,itwasthesubjectofextendeddiscussionatthe2000 SBO.Drawing
uponrecommendationsbytheBusinessCouncilofAustralia,apaperpresentedatthatmeetingarguedthat

12
 PUMA/SBO(2001)4

anindependentfiscalinstitutionbeempoweredtomakeacrosstheboardadjustmentsintaxrateswithout
priorlegislativeapproval.Thecaseforfiscalindependencerestsprincipallyontwoobservations:(1)fiscal
policyisbiasedinfavorofexpansion;itfavorsincreasesincurrentexpenditure,leadingtofutureincreases
intaxation;and(2)governmentsarepronetofiscaldrift,withseriouslagsinimplementingpolicychanges.
Bothdefects,proponentsoffiscalindependenceargue,areduetotheinfluenceofpoliticsongovernment
action.Theproposedsolutionwouldcurtailfiscalexpansionandfacilitateadjustmentbytransferringfiscal
controlfrompoliticianstoindependentexperts.

Theproposalaimstoreconstructfiscalpolicysothatitisconductedinawaythatisanalogoustomonetary
policy.Justasthelatterismanagedbyanindependentcentralbank,theformerwouldbeconductedbyan
independentinstitutionthatisfreeofdirectpoliticalinfluence.Onemayarguethattheparallelisnot
justified,thatenablingindependentauthoritiestoadjusttaxrateshasgreaterpoliticalconsequencethan
empoweringthemtomanipulateinterestrates.Onealsomayarguethatindependenceintaxpolicywould
stripdemocraticgovernmentsoftheirgenericpowerandwouldcallintoquestiontheresponsivenessof
governmentstovoterpreferences.Theseandotherobjectionsaside,however,itisapparentthatalinehas
beencrossedinthinkingaboutthefutureconductoffiscalpolicy.Anideathatwashardlydiscussedafew
yearsagomayforeshadowthefutureworldoffiscaldiscipline.

2.6 FiscalDisciplineWithoutPower

Westartedthissectionbyconsideringhowfuturegovernmentsmightbolsterfiscaldiscipline;weendit
withaproposalthatconcedesthelackofgovernmentdisciplineandlooksforreliefoutsidethepolitical
process.Governmentsofthefuturemaybecompelledtoaccounttooutsideauthorities,tomaintaintheir
budgetsinaformdictatedbyothers,tohavetheirbudgetsauditedastocontentandprocess,andtohave
authorityoverkeyaspectsofbudgetpolicytransferredtoothers.

Inaglobalizedfuture,nationalgovernmentsmayhavebiggerbudgetsbutlesseffectiveinfluence
overthem.Iftheybudgetinafiscallydisciplinedmanner,itmaybebecausetheyhavebeenweakened,not
becausetheyhavebeenempowered.Totheextenttheirbudgetsimposeexternalcostsonothercountriesor
ontheinternationalcommunity,theymayberequiredtosubmittheirpoliciestoexternalreview.Eventhe
biggestcountriesmaybecompelledtoformulatebudgetpolicyinopeneconomies,inwhicheventsoutside
their borders impinge on their fiscal capacity. National governments will be increasingly sandwiched
betweenmetanationalandsubnationalgovernments,inwhichfiscaldecentralizationcoexistswithfiscal
internationalism.Pulledinoppositedirections,theirbudgetsmaybebeholdentodecisionstakenbyothers.

Thisconclusionishighlyspeculative,foritwouldmarkaretreatfromanageinwhichlegalsovereignty
andpoliticalpowerwereconcentratedinthenationstate.Thisformofgovernancemayprovetobemore
durableandadaptivethanthebadcasescenariointhepreviousparagraphindicates.Itmaybethatthe
determinationtomaintainfiscaldisciplinewillimpelcountriestorecentralisecontrolofpublicmoney,or
thatinternationalauthoritieslimittheirroletospecifyingaccountingrulesandaggregatepolicy,butdonot
playaroleinsubstantivepolicy.Anotherpossibilityisthatintheglobalisedfuture,nationalgovernments
will become the functional equivalent of contemporary subnational governments, articulating local
interests and demanding a share of international budgets for their own use. With so many plausible
scenarios, the only thing that is certain is that the fiscal future will different from todays fiscal
arrangements.

13
PUMA/SBO(2001)4

3. DistributiveBudgeting:FromAllocationtoReallocation

Thesecondsetofbudgetissuesidentifiedatthestartofthispaperpertainstotheallocationofgovernment
fundstosectorsorprograms.Thisisanomnipresentconcernbecausebudgetingisinherentlyanallocative
process.Whetheritisstructuredintermsofexpenditureitems,accounts,organisations,activities,orsome
otherclassification,themainissueinbudgetingishowmuchshouldeachentryget.Thisquestionhastwo
dimensions: the efficiency of public expenditure, and the priorities of government. The first criterion
dictates that money should be allocated so as to optimise efficiency across and within programs and
sectors;theseconddictatesthatgovernmentshouldallocatemoneyinaccordwithitsobjectives.Therehas
beenatacitbeliefinbudgetingthatthetwocriteriacanandshouldbereconciled,thatgovernmentshould
frameitsprioritiesinthelightofanalysisorevidenceontheeffectivenessofdifferentprograms.Infact,
majorreformssuchasPPBSassumethatbothpoliticalprioritiesandallocativeefficiencywillcoexistif
thebudgetispreparedonarationalbasis.

Efficiency in allocation has been pursued principally through investment in program evaluation and
outcomemeasures.Inrecentdecades,relativelylittleweighthasbeengivenindevelopedcountriesto
restructuring the budget along program lines. This approach still is tried in some developing and
transitionalcountries,sometimesasameansofconsolidatingthedetaileditemsofexpenditureintobroader
categories.Butthereisconsensusthesedaysthatchangingbudgetclassificationsdoesnotitselfchange
budgetallocations;theonlydifferenceisinthemannerexpenditurearereported,notintheactivitiesto
whichtheyareallocated.Moreover,italsoisgenerallyrecognizedthatpureprogramclassificationswhich
ignoreorsupersedeorganizationalboundariescomplicatethetaskofholdingmanagersresponsiblefor
results.

3.1 Evaluation

Evaluationisanofttriedstrategywithspottyresults.Ithasprovedeasier toconductprogramevaluations
thantousetheminallocatingresources,andeasiertoignoreorexplainawayadversefindingsthantotake
toughmeasurestoimproveprogramperformance.Evaluationgivesgovernmentinformation;itdoesnot
requirethatgovernmentapplytheinformationinbudgeting.ManyOECDcountrieshavehadoccasional
successinconductingandapplyingevaluations,but(tothiswriter'sknowledge)onlytwohavehada
comprehensiveevaluationstrategy.CanadaorganizedavasteffortaroundtheComptrollerGeneralinthe
1970s,Australiaadoptedanambitiousevaluationstrategyinthelate1980s.Canada'seffortborelittlefruit,
Australia'sproducedsignificantreallocationofbudgetresources.Canadaisthoughttohavefailedbecause
itcentralizedevaluation,therebydampeningcooperationbyspendingdepartmentswhichmayhavebeen
adversely affected by the findings. Australia is thought to have succeeded because it gave affected
departmentsabigstakeindesigningandusingevaluations.

Australia'sevaluationstrategycalledforeachportfolio(whichconsistsofoneormoredepartments)to
prepareacomprehensiveplanschedulingtheevaluationstobeconductedovereachofthenextseveral
years.Whenaministerproposesanewprogram(withintheforwardestimatesprocessdescribedearlier)
theproposalhastoincludeadescriptionofhowtheinitiativewillbeevaluated.Theevaluationsarecarried
out by the affected departments, but the Department of Finance oversees the process, including
methodologyanddesign,andpublicationofresults.TheFinanceDepartmenthasmadeasustainedeffort
tofeedevaluationsintothebudgetprocessandhaspublishedanannualreportestimatingthepercentageof
thatyear'sdecisionsinfluencedbyevaluativefindings.

14
 PUMA/SBO(2001)4

3.2 OutcomeMeasurement

Arelateddevelopmenthasbeenthesystematicdefinitionandmeasurementofoutcomes.Theseindicators,
which generally measure the impact of policies on social conditions, are distinguishable from output
measureswhicharediscussedlaterinthepaper.Thereisaburgeoningliteratureonthemeasurementof
outcomes, but entrenched difficulty in applying the results to budget decisions. Outcomes usually lie
beyondthedirectcontrolofthegovernmentdepartmentcarryingouttheprogram,andusuallyderivefrom
aconfluenceoffactors,notjustpolicyinterventionbythegovernment.Itisquestionablewhetheracause
effectrelationshipcanbeattributedtochangesinpoliciesandinsocialconditions.

Governmentsthatseektomeasureoutcomesspendaninordinateamountoftimearguingoverwhethera
particularmeasureisanintermediateorendoutcome,anoutput oranoutcome,thedifferencebetween
impactsandoutcomes,andsoon.Thesesteriledebateshaveimpededtheapplicationofusefulmeasures
bygovernment.Inmyview,themostproductivewaytothinkaboutoutcomesisasdirectionalsignals,not
ascausalstatements.Theyshouldprovokegovernmenttoassesswhethersocialconditionsaremovingin
theexpecteddirections and whether policychanges should be considered. For example, regardless of
whetheritisresponsibleforthetrend,governmentshouldbeawareofwhetherinfantmortalityratesare
risingordeclining,anditshouldtakeappropriateactionsinresponsetothedata.Evenifitnotresponsible
fortheresults,governmentshouldbecognizantofthem.

In view of the weak causeeffect relationship, it generally would not be useful to base performance
budgetingsystemsonanexplicitlinkageoftheamountsspentandtheoutcomesexperienced.Itmaybe
useful,however,topublishoutcomemeasuresassupportinginformationinthebudget.

3.3 StrategicPlanning

Thesecondpathtakenbyreformofbudgetallocationhasbeentomakethebudgetmoreresponsiveto
governmentpriorities.Itmayseemobviousthatgovernmentcannotbudgetonthebasisofitsprioritiesifit
doesnothavethem.Butthelongstandingtendencyinbudgetinghasbeenforgovernmenttobeinexplicit
aboutitspriorities,tohaveprioritiesimputedfromtheallocationsactuallymaderatherthantostatethem
in advance. According to this line of reasoning, explicating priorities generates undue conflict and
complicatesthetaskofproducingthebudget.Itisbetter,therefore,forgovernmenttopreparethebudget
withoutanexplicitstatementofobjectivesorpriorities.Ifaparticularprogramisallocatedmoremoney
thananother,onecaninferthatitisdeemedtobeofhigherpriority.

Thecounterargumentisthatifgovernmentdoesnotknowitspriorities,thebudgetwillnotreflectthem.In
viewoftheentrenchedclaimsofpastdecisionsandtheactiveroleofspendingdepartmentsandinterest
groups,thebudgetwillnothavemuchmoneyfornewprioritiesunlessthegovernmentmakesadetermined
efforttosetasidefundsforthem.

Strategicplanninghasemergedinrecenttimesasthemaininnovationinenablinggovernmenttostructure
itsbudgetonthebasisofmissionsandobjectives.Incontrasttoearlierreforms,suchasprogrambudgeting
andPPBSwhichsoughttointegratepolicyplanningintothebudgetprocess,strategicplanningstands
apartfrombudgeting.Itisnotconstrainedbythetimeorfundingpressuresofbudgeting,butneitherisit
assureddirectinputintobudgetdecisions.Alsoincontrasttopastplanningefforts,strategicplanning
focusesongovernmentorganisationsratherthanonprogramsoractivities.Thekeyquestionisnotwhat
governmentshoulddobutwhatitsdepartmentsandagenciesshouldbe.Forexample,astrategicplan
mightconsiderwhetherthemissionofthecustomsagencyshouldbetoencouragetradeortointerdict

15
PUMA/SBO(2001)4

illegalimports.Onthebasisofstrategicdecisiononmissions,theagencymightrestructureoperationsand
realignitsbudgetprioritiesalthoughbudgetissuesarenotforemostduringformulationofthestrategic
plan,oncetheplanisapproved,theagencyorgovernmentwouldbeexpectedtoallocateresourcesin
accordwithit.

Itsseparationfrombudgetingenablesthestrategicplantotakeadeeperlookatpurposesandobjectives,
butitpotentiallydiminishestheprobabilitythattheplanwillinfluencethebudget.Agenciesgenerallylike
strategicplanningbecauseitportraystheminafavorablelightandbolsterstheirclaimformoreresources.
Butthemoreastrategicplantriestoinfluencethebudget,thelessstrategicitislikelytobe.

3.4 RedistributiveBudgeting

Aswiththefiscalaggregates,SBOmarkedachangeinbudgetingasadistributiveprocess.Oneofthe
earlySBO'sconsideredapaperbyD.Tarschysonbudgetingasadecrementalprocess;othermeetings
focusedoncutbackbudgeting.Thechangeinorientationhasbeenduetothereorientationofbudgeting
fromaprocesswhichannuallyallocatesincrementstofavoredprogramstoonewhichhaslittlemargin
(andsometimesnone)forspendinginitiatives.Thischangeinfortunehasbeenduetothebuiltinclaimsof
entitlementsandothermandatoryexpenditure,thelessrobusteconomicgrowthofthepasttwodecades,
and efforts to strengthen fiscal discipline. The last of these is especially important because the more
successfulagovernmentisinconstrainingthetotals,thegreatertheriskitrunsthatoldprioritieswillbe
frozenintothebudgetandnewprioritieswillbefrozenout.

When incremental resources are available, budgeting is a distributive process in which government
respondstofreshdemandsbyallocatingadditionalresourcestothem.Incrementalbudgetinghasbeen
laudedbysomeobserversonthegroundthatitstabilizestheprocessandreducesconflict,andhasbeen
criticized by others on the argument that it biases spending upward and rewards inefficiency. These
argumentsaside,thereislittledoubtthatinthepreSBOperiod,incrementalismwasthenorm.However,
incrementalism requires increments, additional resources that can be spent on some purposes without
takingmoneyawayfromotherpurposes.InthetwodecadessinceSBO,budgetincrementsoftenhave
fallenshortoftheexpectedgrowthinexpenditure.Tomakemattersmoredifficultforgovernments,they
cannolongertaxtheirwayoutofbudgetproblems.

Asaconsequence,budgetinghasbeenmoldedintoasomewhatredistributiveprocess,inwhichfundsare
recycledfromoldprogramstonewones.Theword"somewhat"reflectstherealitythatredistributionfuels
conflict,andbudgetingisaprocessthatmustdampenconflict.Redistributionisnotwholesalechange;itis
tinkeringhereandtheretofreeupmoneyforcurrentneeds.Redistributionisalwaysdifficult,butithas
beenmademoresobytheheightenedactivismofinterestgroupsontheperipheryofgovernmentlobbying
toprotecttheirbudgetstakes.Asbudgetshavegottentighter,groupshavebecomemoreassertive,and
manynowhaveinformalrolesinbudgeting.

Redistributiondoesnotdependonturningbudgetingintoazerobasedprocess.Whereverithasbeentried,
zerobase budgeting has failed. It fails because budget makers cannot ignore past decisions and
commitments,norcantheyuprootprogramswithoutregardtotheimpactsonthegovernmentagencies
carryingthemout.Nomatterwhatbudgetprocessisused,thecurrentyear'sexpenditurewillalwaysbethe
bestindicatorofthenextyear'sspending.Whenitoccurs,redistributionisonlyatthemarginsofthe
budget,notthewholesaleterminationofongoingprograms.

16
 PUMA/SBO(2001)4

3.5 UsingTheMTEFtoPromoteReallocation

Inadditiontopromotingfiscaldiscipline,amediumtermexpenditureframeworkfacilitatesreallocations
inaccordwiththegovernment'sstrategicpriorities.TheMTEFdeterminesthemargin(ifany)availablefor
spending initiatives or the net savings required to meet the fiscal targets. Proposed reallocations are
measured against the preset margin or savings to assure that policy changes are consistent with the
government'sfiscalobjectives.Becausepolicychangestypicallyhavegreaterimpactonfuturebudgets
thanontheoneimmediatelyahead,itisimportanttoassurethatclaimedsavingsinnextyear'sbudgetdo
notendupasnetincreasesinfuturebudgets.

Mediumtermspendingconstraintsarenotselfenforcing.Thedrivetoreallocatecanopenthedoorto
ploys by spending departments to substitute more costly programs for the ones they are currently
operating.WithoutvigilantmonitoringbytheFinanceMinistry,spendingdepartmentsmayoverstatethe
savingsfromprogramcutbacksandunderestimatethecostofnewprograms.Tothwartthesetactics,itis
essentialthattheMTEFincludebaselineprojectionsofauthorizedspendingineachofthenextseveral
years.Measuringproposedpolicychangesagainstthebaselinehasbecomeoneofthemostimportanttask
ofthecentralbudgetofficesinMTEFcountries.

Reallocationisfacilitatedwhenthegovernmentspecifiesaresourceenvelopeforeachsectorormajor
spending department. In the course of setting these envelopes, the government may reallocate across
sectorsbydecidingthatsomesectorsshouldbepermittedincreasesabovethebaselineandothersshould
berequiredtoproducedecreases.Theseintersectoralreallocationsshouldbemadeatthehighestlevelof
government. In parliamentary regimes, they typically involve the Prime minister and Cabinet. Much
reallocationislikelytobewithinsectors,andthismaybeencouragedbydevolvingresponsibilitytoline
ministers. Arguably, more reallocation will occur if spending ministers and managers have a role in
generatingpolicychanges.Doingthejobcentrallyatthetopofgovernmentmayresultinmuchconflict
andlittlereallocation.Thethreatoflosingresourcesandprogramsmayprovokedepartmentstoblockthe
tradeoffsdemandedofthem.Althoughtheyarenotatthecenterofpower,spendingdepartmentshave
formidable weapons at their disposal. They can withhold information needed to make costeffective
tradeoffs;theycanenterintologrollingcoalitionswithotherspenderstoprotecttheirbudgets;theycan
rallyaffectedinterestgroupstoopposereallocation.

To gain the cooperation of spending departments, it is sensible to give them a prominent voice in
reallocation.AdevolvedarrangementwouldfreeuptheCabinettofocusonmajorpolicychangesrather
thanonspendingdetails.Itwouldgivesectoralministerstheleadroleindevelopingprogramchanges
withintheirportfolios,providedthattheproposedchangesarewithintheMTEFsectoralresourceenvelope
andareconsistentwithoverallgovernmentalpriorities.InAustralia,thebirthplaceofMTEF,theforward
estimatesmaybesetatalevelwhichaccommodatesspendingincreasesormandatescutbacks.Withinthe
forwardestimates,aministermayproposespendingincreasesforsomeactivitiestobefinancedbysavings
inotheractivities.Inthisdevolvedarrangement,ministershaveunilateralauthoritytoapprovechanges
below a certain amount; changes above the threshold require Cabinet approval. This system puts the
Financeministerintheroleofrefereeingthereallocationprocess;managingthetradeoffsystemand
ensuringthatprogramchangesandbudgetreallocationsareconsistentwiththegovernment'sfiscalnorms
andpolicyobjectives.

17
PUMA/SBO(2001)4

4. DistributiveBudgetFutures:WhoWillGetWhat?

Despitetheattentionpaidit,redistributionisasideshowinbudgeting;themaineventstillisdistributing
money,usuallyonanincrementalbasistopowerfulclaimants.Notwithstandingthebleakscenarioswhich
showbudgetsanddemographicsonacollisioncourseearlyinthe21 stcentury,budgetingwillcontinueasa
process for distributing spending increases to agencies, households, and other beneficiaries of the
governmentslargess.Onthebasisofpoliticaltrends,onecanexpectbudgetingtobecomemoreopenand
sensitivetodemandsbyclaimantsmobilisedtoprotectorenlargetheirshares.Voters,interestgroups,
NGOs,andotherswillbecomemoreassertiveandthebudgetwillbecomemoretransparentaboutwhois
benefitingandwhoislosingfromthegovernmentsfinancialdecisions.Thus,evenasbudgetingbecomes
moretechnocratic,itwillbemorepoliticised.Theupshotwillbemoreconflictinbudgetingasrival
interests vie for advantage. Yet because of the compelling need to resolve each years budget, new
methodswillbedevisedtoregulatefrictionoverresourceallocations.

4.1 EBudgeting

Informationtechnologywillopenthebudgettogreaterpressurefromaffectedinterests.Itwillgiveinterest
groups,whichhavemultipliedinalldemocraticcountries,bothinformationandaccess,moretimelydata
onpastorpendingbudgetactionandgreateropportunitytoinfluenceoutcomes.

Manygovernmentsnowpreparethebudgetinelectronicformandmakekeydocumentsavailableviathe
internet.Mostofthisinformationhaslongbeenpublic;whatisnewistheeasewithwhichitcanbe
accessedandmanipulated.Thoughbrowsersandsearchengines,interestgroupswillkeepinformedon
budgetoptions,thelineupofthosefavouringoropposingparticularcoursesofaction,thepersonsor
offices responsible for making the decision, and other details that enable them to intervene more
effectivelyinthebudgetprocess.Moreover,asinternetusebecomescommonplace,matterswhichonce
wereheldinconfidencewillbepostedontheweb.Governmentswillbeimpelledtorelaxtheirsecrecyand
thebudgetwillbecomemorepermeable.

Akeyissueiswhetheranyinformationshouldbepostedbeforethebudgethasbeendecided.Itwould
requireonlyafewkeystrokestomakeagencyrequests,baselineprojections,proposedpolicychanges,
spendingoptions,budgetassumptions,andotherrelevantinformationavailableontheinternet,andonlya
fewmoretoenablereaderstosubmitcommentsonamessageboard.

4.2 PlebiscitaryBudgeting

Once budget information and options are widely available, government will be able to use new
communicationtechnologiestopollcitizensonpendingissues.Thetypesofpollsareasvariedasarethe
formsinwhichgovernmentsmakebudgetdecisions.Onepossibilitywouldbetosurveypublicopinionon
broadquestions,suchaswhetherthegovernmentshouldseektoreducethedeficitbyraisingtaxesor
cutting expenditures; another would be to present citizens with specific questions, such as whether
transportspendingshouldbeincreasedby3percentor5percent.Otheroptionswouldbetoaskvotersto
rankvariousprogramsorsectorsinorderofpriority,ortoindicatewhetherspendingineachprogramor
sectorshouldbeincreased,decreased,ormaintainedatthecurrentlevel.Governmentcantrytoreplicate
actualbudgetdecisionsbyaskingvotershowtheywouldallocateafixedamountofmoneyamongvarious
claimsonthebudget.Pollingcanbeinteractive,withsuccessiveroundsofquestioning,beginningwith
broadissuesandthennarrowingtothespecificissuesfacedbybudgetmakers.

18
 PUMA/SBO(2001)4

Dependingonhow it isconducted, a plebiscite canbe either advisoryor a means of makingbudget

decisions.Intheformersituation,governmentwouldhavetimelydataonvoterpreferences,butitwouldbe
free to take whatever budget decisions it wants; in the latter case, government would be bound to
implement voter preferences in thebudget. Turning budget decisions over to aplebiscite wouldhave
profoundimplicationsfordemocracy.Voterswouldhaveadirectchanneltoexpresstheirpreferences,but
theywouldalsohavetobemoreconsistentonbudgetmatters.Iftheballotwereproperlyworded,they
wouldnotbeabletovoteforbothsmallerprogramsandbiggergovernment.Plebiscitarybudgetingwould
provokeinterestgroupstobemoreactiveandtolobbyvotersforsupportonbudgetmatters.Government
would cometo resemble a permanent campaign, in whichthe period betweenelections is filledwith
governmentsponsored public opinion polls, yearround ad blitzes, and other efforts to sway voter
sentimentandbudgetdecisions.Insomecountries,theextensiveuseofpollsbygovernmentleadershas
alreadyblurredthedistinctionbetweencampaigningandgoverning.Plebiscitarybudgetingmaybenearer
thanwethink.

4.3 FormalGroupParticipation

Onmostissues,votersinteractwithgovernmentthroughtheintermediationofinterestgroups.Inall
democratic countries, many more such groups now exist than did a generation ago, and the number
interested in the budget is vastly greater than it once was. The proliferation of groups has been
accompanied by a breakdown in corporatist institutions, in which government regularly consults with
socialpartners(suchaslabourunionandbusinessleaders)beforeadoptingmajorpolicychanges.Rather
thannegotiatingwithahandfulofleaders,governmentnowisexposedtoaphalanxofgroupswhichhave
conflictingagendasanddemandmorefromthebudgetthanitcanbear.

Budgetingwouldbreakdownifgovernmentgavetoomanygroupsaseatatthetable.Nevertheless,one
canenvisionafuture inwhichpolitical leaders formallydiscuss budget issues with interestedgroups
beforetheyact.Whileitisunlikelythatgroupswillbegivenaformalvetoovergovernmentaction,an
informalvetomaybejustasgoodforthem.

Asthenumberofgroupsmultiplies,theirspanofinterestnarrows.Ratherthanbeinginterestedinwhole
sectors,groupspayattentiontoparticularprojectsoractivities.Accordingly,onecanseeafutureinwhich
moregroupsgetsomethingfromthebudget,butformosttheslicetheygetisquitethin.Slicingthebudget
canbedonebyearmarkingfundstoparticulargeographicareasorprojects.Withthinnerslices,more
groupscomeawaywithsomething,butthelessthetotalclaimonbudgetresourcesisless.

4.4 ClassbasedBudgeting

Buyingoffinterestsbygivingthemsmallamountsofmoneymaybeasensibleaccommodationtothe
hyperpluralismofcontemporarydemocracy.Buttherestillwillbepowerfulbroadbasedgroupsmore
interestedinoverallallocationpolicythaningettingalittleextramoneyforafewprojects.Inthepast,
thesegroupsconcentratedontheamountsallocatedtoparticularsectors;inthefuture,theymaybemore
concernedaboutwhetherparticularsocioeconomicclassesaregettingafairshareofthebudget.This
interestisreflectedincompilationofawomensbudgetorachildrensbudgetbyNGOsinvarious
countries.Thesetypesofbudgetsarenotanalyticexercises;theyareadvocacystatementswhichmarshal
data on the allocation of budget resources to argue that particular classes in society have been
disadvantaged.

19
PUMA/SBO(2001)4

Classbasedbudgetswillbepreparedinthefuturecoveringmajorfissuresinsociety;menversuswomen,
richversuspoor,youngversusold,oneethnicorracialgroupversusothers;oneregionagainstanother,
andsoon.Whatwillbenewabouttheminthefutureisthatratherthanbeingissuedbyoutsiders,theywill
bepublishedundertheimprimaturofthegovernment,firstassupplementaryschedulestotheregular
budgetdocuments,butovertimeasauthoritativestatements.Oncethisoccurs,classbasedbudgetswill
becomedecisionalclassifications;inthecourseofproducingitsbudget,governmentwilldecidehowmuch
tospendonrichorpoor,menorwomen,youngandold,etc.

Whenthiscomestopass,nationalbudgetswillbecomeflashpointsforsocialconflict,sharpeningthe
adversarial,tradeoffcharacteristicsinherentinbudgetingwithscarceresources.Aspartofthiscombat,
theprotagonistswillfightoverclassificationsandanalysisofexpenditure.Forexample,onecananticipate
protractedargumentsoverwhoaretheendbeneficiariesofgovernmentexpenditureonhighereducation,
andoverwhetherthisexpenditurebenefitstherichbecauseahigherproportionofthemattenduniversities,
orthepoorbecauseuniversityeducationgivesthemopportunitytoimprovetheireconomicposition.These
debateswillbetrulydivisiveifpublicmoneyisallocatedonaclassbasis.

Implementingclassbasedbudgetswillrequirenewaccountingrulesandprocedures,possiblysimilarto
thosedevisedinrecentyearsforgenerationalaccounting.Ashasbeenthecasewithrespecttoeffortsto
account for the incidence of spending and taxes by age cohorts, classbased budgeting will stir up
interminableconflictovertheassumptionsusedinestimatingbudgetallocations.Itremainstobeseen,
however,whetherthisformofbudgetwarfarewillleadtomoreredistributionortohigherspending.The
firstpossibilitywillmaterialiseifthevariousclassescompeteforbudgetallocations;thelatterwilloccurif
theclassescolludetogetmoreforeach.

4.5 AllocatingNationalIncome

Whetherbysectororclass,thebudgetbattlesofthefuturewillbeoverhowmuchshouldbeallocatedto
eachsetofclaimants.Thesharescanbeexpressedaspercentagesoftotalexpenditureorastheproportion
ofincrementalresourcesallocatedtoeach.Thesetypesofcalculationsarepervasiveinbudgeting;themain
differenceisthatinthefuture,thetradeoffsmaybemoreexplicit.

Butthereisanotherpossibilitythatisanythingbutconventional.Ratherthanallocatingbudgetshares,the
budgetwillallocatesharesinnationalincome.Thisnovelideaisrootedina1989proposalbythelate
AmericaneconomistHerbertSteinwhourgedthattheUnitedStatesshouldbudgetnotthe$1 trillionin
thegovernmentsbudgetbutthenearly$5trillioninthenationalincome.Andwhathastobecomparedis
thevalueofalternativeusesofthenationalincome,notofthebudget.Theproblemisusuallydiscussedin
termsofatradeoffbetweendefenseexpendituresandothergovernmentexpenditures.Infact,therelevant
andrealistictradeoffisbetweendefenseexpendituresandallotherusesofthenationaloutput.

Steinsproposalillustratestheinclinationofeconomiststoturnanalyticaltoolsintodecisionalrules.Itis
analyticallyusefultoexaminethesharesofGDPspentonhealth,defense,educationandotherpurposes.
Butdoesthismeanthatthebudgetshouldbedecidedonthisbasis?Operationalisingthenotionthatboth
publicandprivateusesofnationalincomeshouldbebudgetedwouldlikelyrequireacentralplanning
apparatusthatwouldgeneratetherigiditiesandinefficienciesrifeinplannedeconomies.Itispossible,
however,toenvisiontheallocationofthegovernmentsexpenditureofbudgetresourcesratherthantheall
nationaloutputalongthelinessuggestedbyStein.

20
 PUMA/SBO(2001)4

OneargumentinfavoroftheSteinmodelisthatitwouldforceexplicittradeoffsamongcompetingclaims
onnationalincomeorbudgetresources.Incontrasttoconventionalbudgeting,inwhichbidsforresources
aredecidedserially,ofteninisolationfromoneanother,thisapproachwouldcompelgovernmenttoweigh
therelativemeritsofrequests.Inasmuchasthetotalspentcannotexceed100percent,when(asislikely)
budgetrequestsexceedthatamount,somewillhavetobecutforotherstogettheirshares.

Arguably,makingbudgettradeoffsmoreexplicitandtransparentwilldomoretopoliticizeallocationsthan
toimprovetheirefficiency.Lookingtoanuncertainbudgetfuture,onecanforeseepoliticalpressureto
allocateportionsofthebudgetinpercentageterms.Forexample,parliamentmaydecreethatnotlessthan
aspecifiedpercentageofGDPoroftotalexpendituremustbeallocatedtohealth,education,orsomeother
program.Quitelikely,claimsentitledtoafixedsharewouldhaveapreferredpositioninthebudget,while
thosenotreceivingsuchallocationswouldhavetocompeteforanyremainingresources.Obviously,this
double standard will motivate strong groups to seek legal entitlement to fixed budget shares. If they
succeed,budgetallocationswillbebothpoliticizedandrigidified.

4.6 InternationalNorms

Entitlingcertainsectorstosharesinthebudgetorinnationalincomewouldbeconsistentwiththetrendto
baseallocationsonnormativerulesratherthanonthediscretionofpoliticalleaders.Atpresent,most
normativeallocationsareconcentratedinmandatorytransferstohouseholdsandgrantstoregionalorlocal
governments.Thesetwocategoriesnowaccountforwellover50percentofnationalgovernmentoutlays
inindustrializeddemocracies.Theywillaccountforanevenhigherpercentageinthefuture.

Nowadays,nationalgovernmentsoperatetwoparallelbudgetsystems.Oneisorganizedaroundrecurring
(typically, annual) budget decisions, the other is driven by normative rules which specify eligibility
standardsandpaymentformulas.Theformerislegallydiscretionary,thelatterismandatedbystanding
law.Althoughannuallyappropriatedfundsgarnermostoftheattentionduringpreparationofthebudget,
they cover a shrinking portion of total expenditure. Ruledriven spending usually is prescribed in
permanentlawthatcontinuesineffectunlessitismodifiedbynewlegislation.Theseexpendituresare
controllable, but most carry over from one year to the next without substantive change. One may
confidentlypredictthatmanygovernmentswillimpelledbyfiscalanddemographicpressurestocutback
normativeexpendituresinthedecadesahead.Iftheysucceed,theywillmerelyslow,notreverse,the
growthoftheseexpenditures.

Mostcurrentnormativeallocationsarecountryspecific;eachgovernmentadoptsitsownrules.Inthe
future, however, many will be determined by norms that cross national boundaries. International and
regionalorganizationswillprescribeminimumstandardsofexpendituresforvarioussocialprograms
such as health, education andenvironmental protection. They will promulgate convergence criteria in
eithermoneyorprogramterms.Forexample,theymaydictatestudentteacherratiosorrequirethata
minimumpercentageofGDPbeallocatedtopubliceducation.Theymayestablishstandardswhichcompel
governments toallocatecertainamountstoparticularprograms. Forexample, theymaydecreethat a
governmentmusthaveatleast500childcareplacesforeach1000childrenbetweentheagesof15,orthat
allschoolsoffercertaincourses.Thevarietyofnormsisasboundlessastheinterestsofaffectedgroups.

At the outset, many criteria will be advisory or indicative; over time they will become binding, and
governmentswillberequiredtoaccountforexpendituresinwaysthatenableinternationalmonitorsto
auditcompliance.

21
PUMA/SBO(2001)4

Oneparticularlyburdensomesetofnormswillpertaintorevenuesharingbetweenthenationalandlocalor
regionalgovernments.Theformerwillhaveresponsibilityforextractingtaxesfromcitizens,thelatterwill
betherealspenders.Underthebanneroffiscaldecentralization,nationalgovernmentswillbeputinthe
politicallydifficultpositionofbeingblamedwhenthingsgowrong,butwillhaveinadequateauthorityto
putthemaright.

4.7 BudgetingWithoutAllocating

Sinceitsinception,budgetinghasbeendepictedasaprocessforallocatingfundsamongalternativeuses.
Many of the decisions have pertained to the allocation of incremental resources, but the increments
normallyhavebeenofsufficientconsequencetoinvestbudgetingwithpoliticalandfinancialimportance.
Budgetdecisionshavematteredbecausetheyhavedeterminedhowtheseresourceswerespent.

Over time, however, the decisional capacity of budgeting has been chipped away by statutory
requirements, international treaties and obligations, changing fiscal relations between the central
government and subnational governments, the shift in fiscal risk from enterprises and households to
governments,andthedemandsofinterestgroups.Thecumulativeeffectofthesedevelopmentshasbeento
transformmuchofbudgetingintoameansofaccountingforpastdecisions.Thescenariodrawninthis
sectionindicatesthatevenagreaterportionofexpenditureswillbebudgetedinthismannerinthefuture.
Asunsettlingasthisconclusionmaybe,itmaybeanappropriateconditionformaturedemocracieswhose
voterswantneitherbigexpansionnorbigretrenchmentofgovernment,andwherethemostimportant
issues have been settled (at least for the time being) by decades of policy adjustment and program
accretion.Inthesecountries,mostvotersareclusterednearthecenter,andthepoliticalpartiesfightloudly
butoversmallmatters.Democraticstabilityisnominoraccomplishment,anditisdueinparttothedulling
ofpoliticalconflictbybudgetarystasis.

ThelossofbudgetaryflexibilitywasarecurringconcernoftheearlySBOsandhasappearedonthe
agendainvariousformsoverthepasttwodecades.Entitlementsareagrowingpartofnationalbudgets,
andtheyareheretostay.Thedamagetheydotothebudgetiscompensatedbythesecuritytheybringto
households.Anyonewhobemoansthespreadoftheentitlementstatemustmarvelattheeconomicwell
being it has brought to its citizens. The two developments are interlinked, making the budget the
dependentvariableinthepoliticalrelationshipbetweencitizensandstate.Butwhatisnewandsomewhat
alarming, if the future envisioned here is credible, is the rigidification of heretofore discretionary
expenditures.Thebigallocativequestionsstillinthehandsofnationalgovernmentshowmuchtospend
onserviceswillbedecidedbynormsandformulasinfuturebudgets.Thelastpreserveofbudgetingis
threatenedbythesamelogicthathasmadeentitlementssopopularandpervasive.Ensureinterestsand
classestheirshareinthebudgetbymakingallocationamatteroflegalrightratherthanofdiscretion.Give
citiesandcommunitiestheirshares,theoldandtheyoung,theschoolsandthehealthclinics.Budgetary
choicemaybeweakened,butallforthegoodofsociety.

Perhaps this scenario will turn out to be another false alarm and budgeting will persist as a robust
allocativeprocess.Butifnormativebudgetingmaterializes,itwillbenecessarytostopthinkingofitasan
allocativeprocessandtorecognizethatithasmetamorphosedintosomethingthatappearstobethesame,
butisnot.

22
 PUMA/SBO(2001)4

5. OperationalBudgeting:FromControltoPerformance

Fiscaldisciplineandfinancialallocationareprincipalconcernsduringpreparationofthebudget.Thethird
objectiveofbudgetingpromotingefficiencyintheprovisionofpublicservicesfocuseslargelyon
implementationofthebudget.Whilethefirsttwoobjectivesgetmostoftheattentioninbigbangreforms,
operationalmattersconsumemostoftheworktime ofbudgetstaff.Theflowofcommunicationsduring
budget execution and the sheer number of actions that are reported and controlled dwarf other work
demandsofbudgeting.

Despite its sometimes lowly status, operational budgeting is important because it affects the cost and
qualityofservices,thevolumeofgovernmentexpenditure,thesizeofthecivilservice,andrelations
betweencitizensandgovernment.Citizensknowtheirgovernmentthroughtheservicestheyreceivefrom
it.

OperationalissueswerenotprominentattheearlySBOsbecausetheagendawasdominatedbythebig
issuesdiscussedintheprevioussections.Buttheygainedattentioninthesecondhalfofthe1980sas
innovativegovernmentsinAustralia,theUnitedKingdom,NewZealand,Sweden,andelsewheresoughtto
improvepublicperformance.Eachofthesecountriesmovedboldlytoshiftbudgetingfromcomplianceto
performancebygivingmanagersfreedomtoruntheiroperationsastheydeemedappropriate.Several
forcesstimulatedthistransformation,includingtheimportationofnovelmanagerialpracticesfromthe
businesssector,recognitionthatthepersistentriseinoperatingcostswascompellinggovernmenttospend
ahighershareofnationaloutput,theabilityofpoliticianstotakecreditforcuttingexpenditureswithout
drawingblameforcuttingprograms,andrealizationbybudgetofficialsthattheywereovercontrollingthe
mostcontrolledportionofthebudget.

The operational budget agenda parades under a number of monikers: new public management,
managerialism,performancebasedbudgeting,andmarketisation.Thenamesreflectdifferentapproaches
tothesameissue.Someusemarkettypemechanisms,otherrelyonmanagerialskillsandjudgement,
others on strong accountability arrangements. All recognize that transforming operational budgeting
requiresmajorchangesinthemanagerialsystemswithinwhichbudgetingisembedded.Incontrasttopast
reformswhichsoughttochangebudgetingwithoutregardtopublicmanagement,recentinnovationshave
beengroundedonthepresumptionthatbudgetingisasubsetofmanagementandcannotbereshapedin
isolationfromotherprocessestowhichitislinked.

MostOECDcountrieshavemadelittleprogressindismantlingmanagerialcontrolmechanisms,though
mostappearinterestedinimprovingoperationalperformancebyeasingbudgetcontrols.Thereisnosingle
modelformovinginthisdirection,norasingleinnovationthatwillaccomplishallthesoughtchangesin
governmentoperations.Thehandfulofinnovationsdiscussedbelowhavebeenselectedbecauseoftheir
breadthandwideapplication.

5.1 ManagerialDiscretion

Anessentialstepinshiftingoperationalbudgetingtoperformancehasbeentodismantlemanyofthe
operationalcontrolsthathavebeenappliedforgenerations.Thetypicalcontrolspertaintoinputsand
operate through the detailed itemization of different objects of expenditure, such as travel, supplies,
utilities,andpersonnel.Eachtypeofcontrolhasitsownrulesandprocedureswhichareenforcedby
controllersatthecenterofthegovernmentorindepartmentalheadquarters.

23
PUMA/SBO(2001)4

Inresponsetothevastexpansionofgovernment,manycountrieshavepartlydecontrolledpublicspending
byconsolidatingtheitemsintobroadcategoriesandbygivingspendingunitssomediscretioninshifting
fundsamongtheitems.Manyalsohaveplacedgreaterrelianceoninternalcontrols,inwhichthespending
agencyisresponsibleforassuringthelegalityandproprietyofexpenditure,incontracttoexternalcontrol
whichveststhisresponsibilityincentralagencies.Afewhavegonemuchfurtherandhaveeliminated
virtuallyallcentrallymaintainedcontrolsoveroperatingexpenditure.Insteadofsplittingtheoperating
budgetintonumerouspocketsofmoney,theynowgiveagenciesalumpsumforallrunningcosts.Within
this operating budget, managers decide how much to spend on travel, salaries, and other items. The
operatingbudgetiscashlimited,barringagenciesfromspendingorrequestingmorethanwasprovided,
evenifinflationoutrunstheestimatesonwhichthebudgetisbased.Moreover,somecountriessubtractan
"efficiencydividend"fromtheoperatingbudgettoreflectexpectedgainsinproductivity.Thisdividendis
typicallyinthe12percentayearrange,and pressuresagenciestobemoreefficientinusingoperating
resources.

Wheredecontrolistakenseriously,itisnotconfinedtothecenterofgovernment,butisdevolveddownto
operating andfieldunits. Each field unit or local office gets its operatingbudget and, within certain
restrictions, has freedom in spending these resources. Thus, managerial flexibility redefines both the
relationshipbetweencentralagenciesanddepartments,andbetweendepartmentheadquartersandfield
units.Thelogicofthisstrongcommitmenttodecontrolisthatmanagerialimprovementcanoccuronlyif
managersarefreetousetheirskillsandprofessionaljudgementinrunningtheiroperations.Theyareeither
freetomanageornot.Iftheyare,servicesarelikelytobemoreefficientanddeliveredinamoreflexible,
responsivemanner;iftheyarent,managerswillstresscompliance,evenifsomecontrolshavebeeneased.

Butthislogichasnottriumphedeverywhere.InsomeinfluentialOECDcountries,apremiumstillis
placedonuniformityinservicesandcentralcontrolstillpredominates,thoughnottothefullextentitdida
generationortwoago.Insome,thereisconcernthatmanagerialdiscretionwillunderminethecivilservice
systemandleadtothereplacementofcareerofficialswithshorttermappointees,ashasoccurredinNew
Zealand and the United Kingdom. In fact, attitudes toward the civil service sum up key differences
betweenthereformmindedandstatusquocountries.Thelattersensethelossofacivilserviceethic,the
formerseerigiditiesinthetraditionalpublicservice.Thisisanissueforwhichtherecanbenofinal
resolution,noranysinglebestway.

Centralagencieshavebeendeeplyaffectedbythemarchofmanagerialism.Asrecentlyasageneration
ago,theirmainfunctionwasmaintainadministrativecontrols,especiallythoseregulatingpersonneland
finance. In some countries, various control agencies have been abolished. for example, the personnel
agencyinAustraliaandtheaccommodationboardinNewZealand.Butgovernmentsdonothavethe
optionofdoingawaywiththeirfinanceministriesortheirbudgetdepartments.Somefinanceministries
haveledthedriveformanagerialreform,othershavegonealongreluctantly.Allhavetogothrougha
periodofadjustment,unlearningoldrolesandrelationshipsanddefiningnewones.Theymustdecidehow
todrawthelinebetweenlettingoperatingmanagersmanageandinterveningwithadviceorrestrictions.
Theymustdecidehowmuchcontrolshouldbethrownoverboardandexactlywhatshouldberetained.And
theymustfigureouthowtoretainpoweratthecenterofgovernmentwhenmuchoftheirpowerbasehas
beensurrendered.

5.2 PerformanceTargetsandReports

Therearetwosidestothenewmanagerialism.Oneissummedupinthephrase"letmanagersmanage"and
revolves around the divestiture of input controls; the other is impliedby the phrase "make managers

24
 PUMA/SBO(2001)4

manage"andhastodowiththeimpositionofstrongaccountabilitymeasurestoassurethatmanagersare
not abusing their discretionand areproducingthe intended results. In everyreformist country, ithas
proven easier to move ahead on the first agenda than on the second. But the countries which have
progressedthemostinfreeingupmanagershavealsobeenthemostcreativeanddemandinginintroducing
newaccountabilityregimes.

Accountability deals both with money and performance, and in the guise of performance budgeting
(discussedinthenextsession)triestomoldthetwotogether.Onthespendingside,accountabilityentails
theaccrualbasiswhichpurportstoholdmanagersaccountableforthefullcostofoperationsandaudited
financialstatements.Afewcountrieshavebrokennewgroundinthecostingofservices.NewZealandhas
beenthemostventuresome,chargingoperatingbudgetsbothfordepreciationandfortheuseofcapital
(computed as net worth on an entitys balance sheet). Managerial minded countries also require
departmentsoragenciestopublishannualreportsdiscussingtheiroperationsandcomparingplannedand
actualresults.Ontheperformancesideoftheledger,managerialistgovernmentsgenerallyrequirethat
performancetargetsbespecifiedinadvanceandthatactualresultsbemeasuredagainstthetargets.These
countrieshaveinvestedheavilyindevelopingperformancemeasuresandcompilingrelevantdata,though
someoftheefforthasbeenwastedinsteriledebatesonthedifferencesbetweenoutputsandoutcomesor
betweenintermediateoutcomesandendoutcomes,andsoon.Theapproachestakenbyvariouscountries
appeartodiffer,butallemphasizecomparison.

TheUnitedKingdomfavorsasmallnumberoftargetsforeachagency;thesearecompiledinanannual
reportissuedbythegovernmentinasingledocumentthatcomparestargetsandresultsforthepastyear
andprovidesnewtargetsfortheyearahead.Theadvantageoftargetsisthattheyaretightlyfocused;the
disadvantageisthattheymayexcludeimportantelementsofperformance.TheUnitedStatesprovidesfor
eachdepartmentandeachmajoragencytoproduceanannualperformanceplandiscussingwhatithopesto
accomplishintheyearaheadandanannualperformancereportdiscussingwhatwasactuallyaccomplished
inthelightofitsplans.NewZealandandSwedengenerallyrelyonannualreportstoassessresults.

Asperformancemeasuresandreportsbecomemorecommon,thesewilllikelybesubjecttoreviewby
auditorswhomaycommentontheaccuracyofthedataorontheinterpretationsdrawnfromthem.Italso
isprobablythatsomestandardswillemergeoncompilingandanalyzingperformanceinformation.Itis
unlikely,however,thatthesestandardswillbeasauthoritativeasthosewhich governthereportingof
financialresults.Thefutureevolutionofperformancemeasureswilldependasmuchonhowthedataare
use as on how they are collected and presented. Performance measurement rests on the questionable
assumption that measuring and publishing results will make a difference in the results. The marger
managerial movement rests, however, on a different proposition: that unless managers have strong
incentive to be efficient and productive, generating data will make little difference. This is why
performancemeasurementisseenasasubsetofmanagerialreformratherthanasthemainevent.

5.3 PerformanceContracts

Onewaytoincreasetheuseofperformanceinformationistoincorporateitintocontractswhichspecify
theresultstobeachieved.Whilethetermperformancecontractissometimesappliedtorelationsbetween
governmentasapurchaserofservicesandoutsidevendors,recentdevelopmentsincontractingpertainto
relationshipswithingovernment.Contractsarenowformalizedinsomecountriesspecifyingthepayand
responsibilitiesofseniorandmiddlemanagers.ThechiefexecutivesofNewZealanddepartmentsare
employedundertermcontracts(forupto5years)whichspecifythekeyresultsexpectedofthem.Similar
employmentcontractsarenegotiatedbetweenseniorandmiddlemanagers,andondowntheadministrative

25
PUMA/SBO(2001)4

chainofcommand.Thesecontractsenablemanagerstonegotiatetheirownpayandtermsofemployment
withoutregardtostandardcivilservicerules.AsimilararrangementprevailsintheUnitedKingdomfor
thechiefexecutivesofgovernmentagencies.

Performancecontractshavealsobeenintroducedtospecifytheoutputsorservicestobeprovidedby
government entities. In the United States, department heads (who are political appointees) negotiate
performanceagreementswhichindicatethestepstheywilltaketoimproveoperations,introduceservices,
ormakeotherchanges.Ratherthatbeingcomprehensivestatementsofalltheyresultstobeachieved,these
agreementsconcentrateonmattersthatthegovernmentanddepartmentheadsdeemtobeofparticular
importance.InNewZealand,ministersnegotiateannualperformanceagreementswiththechiefexecutives
oftheirdepartmentsspecifyingtheservicesandotheroutputstobeprovidedduringthefiscalyear,andthe
resourcesthatwillbemadeavailableforthesepurposes.

These and similar contractlike documents are not genuine contracts. For one thing, the relationship
between the parties to the agreement is not truly arms length; for another, the government does not
normallyhaveeffectiveredressifmanagersfailtoperform.Iftheschoolprincipaldoesapoorjob,sheor
hecanbedismissed,butthegovernmentrarelyhastheoptionofclosingtheschoolorcuttingitsbudget.
Moreover,thecontractsrarelyspecifyalltheservicesoroutputstobeprovided;rather,theyidentifythose
matters of particular interest to one or both parties. It may be fruitful to think of the contracts as
establishinganongoingrelationshipandabasisfordiscussionbetweentherespectiveparties.Theycanuse
thetermsoftheagreementascheckliststoreviewprogressinachievingparticularmilestonesorastalking
pointsinperiodicmeetingsbetweenthetwosides.Theyaresignalingdeviceswhichspurmanagersto
focusoncertainaspectsofperformanceandputthemonnoticeastowhatisexpectedofthem.

5.4 AutonomousAgencies

Fordecades,effortstoimprovepublicmanagement involvedtheintegrationofactivitiesandseriesinto
broaddepartments, each withabroadswathofresponsibilities.Therewasadepartmentfortransport,
anotherforeducation,anotherforhealth,ondownthestandardlistofgovernmentresponsibilities.In
creatingfunctionallyintegrateddepartments,governmentstrovetoeliminatefreestandingagencies,andto
placethosethatsurvivedwithinthedepartmentalstructure.

Managerialismhasgivenrise,however,tothedismantlingofcohesivedepartmentsbyremovingoperating
units and giving them broad independence to carry out their assigned responsibilities. While this
developmenthasoccurredinonlyafewcountries,othersmayjointheagencybandwagonin theyears
ahead. Typically, when an agency is established, the government (sometimes through the parent
departmentorministry)definesthemattersoverwhichithasoperatingdiscretion,itsdutiesandfunctions,
andthemannerinwhichitwillbeaccountableforfinancialandsubstantiveperformance.Thegovernment
appointsachiefexecutivewhohasfullauthorityto'runtheagency.

Britain and New Zealand provide alternative models of agencification. Britain created independent
agencies to free service providers from central control; New Zealand created agencies to free policy
makers from undue influence by service providers. In Britain, approximately 75percent of the civil
servantsemployedbycentralgovernmentworkinthe130NextStepsagencieswhichhavebeenformed
duringthepastdecade.TheseagencieswereestablishedpursuanttoaGovernmentstudywhichfoundthat
previousbudgetandmanagementreforms,suchasthefinancialmanagementinitiative launched bythe
ThatcherGovernmentin1982,hadfailedtoimprovetheprovisionofservices.Thereportrecommending

26
 PUMA/SBO(2001)4

independence for service agencies was titled Next Steps; hence the entities created pursuant to its
recommendationsarecommonlyreferredtoasNextStepsagencies.

Beforeitlaunchesanagency,theGovernmentdefinesitsresponsibilitiesandassessesitscapacityforself
management.Itthenrecruitsachiefexecutiveanddrawsupaframeworkdocumentwhichspellsoutwhat
the agency is to do, its relations with the parent department, and the discretion it will have. The
Governmentpublishesannualperformancetargetsforeachagencyand,asnotedearlier,comparesresults
tothetargets.Approximatelyevery5years,theGovernmentcommissionsanindependentreviewofthe
agency's operations and performance, and decides whether it should be continued, terminated, or
restructured.TheNextStepsprocessiswidelyregardedashavingbroughtsignificantimprovementinthe
efficiencyandqualityofservices.Ithasbothreducedcompliancecostsandenabledagenciestobemore
flexible and responsive in delivering services. Despite the origin of Next Steps in the Conservative
ThatcherMajorGovernments,ithasbeencontinuedbytheLabourGovernmentheadedbyTonyBlair.

IncontrasttoBritainwhichwas motivatedtofreeupserviceprovidersfromtheclutchesofWhitehall,
NewZealandseparatedserviceprovisionfrompolicyadvicetogiveMinistersfreedomindefiningpolicies
andinpurchasingservices.InNewZealand,appropriationsaremadebyclassesofoutputsandarevotedto
Ministers who may purchaseservices from their own departments or from alternative providers. This
formal distance between them enables ministers and departmental chief executives to negotiate the
purchaseagreementsmentionedearlier.Eachdepartment'schiefexecutivehasdiscretioninhiringstaff,
using available resources organizing work, and producing the required services. Accountability is
maintainedthroughannual reportandaudits,output measures,andevaluationofthechief executive's
performance.

Thereputedsuccesses of theBritishandNewZealandmodelshavespurredothercountriestocreate
independent agencies. But as agencies have become entrenched, both Britain and New Zealand have
become concerned that the number and independence of these entities impedes efforts to coordinate
policiesandprogramsthatcrossorganizationallines.Bothcountrieshavesoughttodevelopa"wholeof
governmentperspectivewhichmaydiminishtheoperatingindependencepreviouslygrantedagencies.

5.5 MarkettypeArrangements

Manageralism relies on the skill, judgement, and professionalism of those who produce government
servicestoimproveoperatingresults.Withoutthispremise,itwouldmakenosensetodecontrolpublic
managementandempowercivilservantstoactastheyseefit.Butnotallreformersagreethatentrusting
managers with operating discretion will improve results. Some hold that nothing short of customer
sovereigntythroughcompetitionandchoicewillmakegovernmentsresponsivetocitizensindesigningand
providingservices.Incontrasttomanagerialismwherechoiceisvestedinserviceproviders,amarkettype
approachgiveschoice,throughvouchers,pricemechanisms,usercharges,andotherdevicestothosewho
receiveservices.Andincontrast tomanagerialismwhichintroducescontracts withingovernment, the
marketapproachusesprivatizationandcontractingouttopromotegenuinecompetition.

Atitscore,themarketapproachispredicatedontheargumentthataslongasgovernmenthasmonopoly
powerintheprovisionofservices,itwillbe wastefulandindifferenttotheinterestsofthoseitserves.
Accordingly,governmentshouldeitherprivatizeservicesoropenthemuptocompetition.

Themarketstrategyhaslittleappealtothosewhobelievethereisinherentadvantageinhavingpublic
servicesprovideddirectlybygovernment.Argumentsandevidencethatcitizenscangetabetterdealif

27
PUMA/SBO(2001)4

they are free to purchase services are irrelevant to those who believe it wrong to privatize public
responsibilities,suchasschools,ortosubjectthemtomarkettypecompetition.Itislikelythereforethat
marketizationwillbeconfinedinmostcountriestocommercialactivities,suchastelecommunicationsand
othertradingactivities,andwillnotsignificantlypenetratecorepublicservices.

6. OperationalFutures:GovernmentasaProducer

Inviewoftheclashingpositiononpublicversusmarketprovisionofservices, itisdifficulttoenvision
howgovernmentoperationsmaybecarriedoutinthefuture.Butontheassumptionthateffortstoprivatize
orcontractoutbasicserviceswillcontinuetoberesisted,andthatmarketpenetrationwillbemarginal,this
sectionconsidersanumberofinnovationsthatmayextendtheboundariesofmanagerialismandbolster
markettypearrangementswithingovernment.Inotherwords,governmentwillcontinuetobudgetforand
provideservices,butwilldosoinwaysthatheightensensitivitytothecostandqualityofservices.

6.1 PerformanceBudgeting

Allocatingresourcesonthebasisofservicestobeprovidedisanold,appealing,andelementaryideathat
hasmadesuprisinglylittleheadway.Manygovernmentsincludeperformanceinformationintheirbudgets,
butdoingsodoesnotmeanthattheysystematicallymakespendingdecisionsonthisbasis.Theconceptof
performancebudgetingintendedhereisoneinwhicheachincrementinexpenditureisexpressly linked
withanincrementin outputorperformance.Implementingthisconceptrequires thatgovernmenthave
reliabledataontheunitcostofservicesandthatbidsforresourcesbestructuredinamannerthatfacilitates
themarginalanalysisofcostsandoutputs.Fewgovernmentscurrentlyhavethiscapability,thoughmany
compileperformanceinformation.

Inconventionalbudgetingasumofmoneyisexchangedforthetotaloutputorworkofaspendingunit,,in
performancebudgeting,theoutputsaredesegregatedintounits,costsfactorsareattachedtoeachunit;and
thetotalexchangeddependsonthevolumeofunitsandtheamountspaidforthem.Implementingthis
strict version of performance budgeting requires governments to have cost accounting systems which
distinguishbetweenfixedandvariablecostsofgovernmentservices,andmeasuresthemarginalcostof
changesinthevolumeorqualityofservices.Whilethistechnicalfeatispresentlybeyondthecapacityof
mostgovernment,themainimpedimentstoperformancebudgetingarebureaupolitical.Spendingagencies
areuncomfortablewithbreakingdowntheirworkintostandardunits, andalsouncomfortablewith the
notionthattheamounttheygetshoulddependexpresslyontheamounttheyproduce.Thismorethanthe
technicalproblems,isthereasonwhyperformancebudgeting,whichwasaleadingbudgetreformhalfa
centuryago,hasmadesolittleheadway.

Nevertheless,performancebudgetingisanideawhosetimewillcome.Progressinmeasuringandcosting
outputsanddemandsforbothefficiencyandqualityintheprovisionofserviceswillimpelgovernmentsto
allocateonthisbasis.Ofcourse,oncebudgetsexplicitlylinkincrementsinresourcesandservices,itwill
beasimpletasktooutsourcegovernmentworkandtoexchangeresourcesandoutputonacommercial
basisthroughmarketmechanisms.Thus,performancebudgetingwhichseekstoimplantmanagerialismin
government,hasthepotentialtouprootit.

28
 PUMA/SBO(2001)4

6.2 PriceBasedBudgeting

Onceagovernmenthasthecostandoutputdataneededtoformulateaperformancebudget,itmaytake
bolderstepstoimprovepublicmanagementbybudgetingonthebasisofpriceratherthancost.Aprice
basedbudgetisoneinwhichgovernmentauthorizesexpenditureintermsoftheamountagreedtobepaid
ratherthanonthebasisofthecostofproducingtheservices.

Forgenerations,governmentshavepreparedbudgetsbyaddingupthecostoftheinputspurchasedby
agencies.Butgovernmentappropriatesmore,agencieshavemoretospendoninputs;ifitgivesthemless,
theymustspendless.Ifcivilservantswinpayincreases,orifmorestaffareaddedtoanagency'spayroll,
thegovernmentvotesmoremoney,regardlessofthevolumeofoutputproduced.Thisbehaviormeansthat
inbudgetingthecostofinputsalwaysequalsthepriceofoutputs.Ifthecostofinputsrises,sotoodoesthe
priceofoutputs;ifgovernmentreducesthepriceofoutputs,ittherebyalsoreduceswhatagenciesspendon
inputs.

Asconventionalasthispatternisingovernmentbudgeting,itisalientomosteconomicexchange.A
customertypicallyisindifferenttothecostbornebythesupplier.Asaconsequence,anefficientsupplier
mayprofitbyeconomizingoninputs;aninefficientsuppliermaylosemoneybecausethecostofinputs
exceeds the price charged for outputs. Significant gains in operating efficiency can be reaped in
governmentbyseveringtherelationshipbetweencostandprice.Governmentwouldnegotiateaprice
without regard to the cost borne by affected agencies in producing the agreed services. Assuming,
however, that services are supplied inhouse by government agencies and are not outsourced to
commercialvendors,therewouldhavetobestrongassurancethatthepriceisright,thatitisreasonable,
andthatanefficientlyrunagencycansupplytheservicesfortheamountthatitwillbepaid.Withoutthis
assurance,thereissubstantialriskthatthepricewillbetoolow,resultinginhiddenreductionsinoutput(or
inquality)orinabilityoftheagencytopayitsbills.Itwouldbequitedifferentforacommercialvendorto
closeitsdoorsbecauseitunderchargesforservicesthanforagovernmentagencytobeinsolventbecauseit
cannotoperateattheagreedprice.Itisthedifferencebetweenfirmsgoingbankruptandschoolsshutting
down.Oneisacommonoccurrence,theotheristrulyrare.

Exceptinareaswhereoutputiscontractedout,governmentsdonotpresentlyhavetheknowhowtobudget
on the basis of price. Governments lack cost accounting systems to measure what particular services
shouldcost,noraremostoftheirbasicservicessubjecttocompetitivetenders.Yet,significantprogress
hasbeenmadeinreparatingpricefromcost.Whengovernmentallocatesarunningcostbudgettoan
agencywithouttallyingupthecostofinputs,itmovesinthedirectionofthepricebasis.Similarly,whenit
cashlimitsthisbudgetorsubtractsanefficiencydividendfromtheamountprovided,governmentweakens
thelinkbetweenpriceandcost.Pricebasedbudgetingwillbeadifficultfeat,butitisfeasible.

6.3 VariableBudgets

Performanceandpricebasedbudgetingareassociatedwithanotherpracticalinnovationashiftfrom
fixedbudgetsinwhichtotalspendingdoesnotdependontheamountproducedtovariablebudgetsin
whichtheamountpaidbygovernmentisdeterminedbythevolumeandqualityofoutput.Asalienasit
may seem to government budgeters, variable budgeting is standard business practice. Although some
portions of their budgets may be fixed, firms allocation production costs on the basis of the volume
produced,andmanytieexpendituresonresearchanddevelopments orothervariables.Governmentsdo
not,however,becausetheyregardappropriationsasfixedlimitsontheamountsthatmaybespent.In

29
PUMA/SBO(2001)4

effect,governmentbuysalltheoutputofitsspendingagenciesatafixedprice;itdoesnotspendmoreor
lessiftheamountproduceddeviatesfromthebudgetedlevel.

Arguably,theamountpaidshoulddependonvolume;ifitdoesn't,thegovernmentwilloverspendwhen
outputfallsshortofthebudget,anditwillriskhiddencutsinserviceswhenoutputexceedsthebudget.

A critical difference between fixed and variable budgets is that the latter are inherently performance
budgets and the former arenonperformance budgets. After all, ina variable budget, spending varies
becauseperformancevaries.Butwhenexpendituresarefixed,thebudgetisindifferenttoperformance
evenifthegovernmentprofessestohaveaperformancebudget.Devisingavariableperformancebudget
entailsthreeadjustments inbudgeting. First, expenditureshavetobe basedonprice rather thancost.
Ideally,apriceshouldbesetforeachunitofoutputratherthanfortotaloutput.Second,theconceptsand
methodsofcostaccountinghavetobeutilizedtodistinguishbetweenfixedandvariablecostsaswellas
betweenmarginalandaveragecosts.Withoutadequatecostaccounting,thegovernmentwillnothave
sufficient information to set a fair price for services. Third, it is necessary to distinguish between
expenditures whichshouldvarywithvolumeandthosewhichshouldnot.Asageneralrule,variable
budgetingshouldbeappliedonlytothoseserviceswhenthedemandforservicesisexogenous;itisnot
directlyinfluencedbytheentityprovidingtheservices.Thepassportagencyshouldhaveavariablebudget
becauseithaslittleimmediatecontroloverthenumberofapplications.Butthepoliceagencyshouldhave
afixedbudgetbecauseitdecideshowmanypatrolstoundertake.Ifitsbudgetvariedwithvolume,itwould
beabletoboostitsbudgetbyprovidingmoreservices.Variablebudgetingwouldrequireacasebycase
determinationoffundingarrangements.Withinanagency,someservicesmayhavefixedappropriations
whileothersmaywarrantvariableexpenditures.

6.4 Citizen(Customer)Rights

Inthefuture,governmentperformancewillbearightofcitizens.Inthesameway,thatcitizensareentitled
tohealthcareorincomesupport,theywillbeentitledtoperformingschools,ontimetransport,courteous
civilservants,prompthandlingofapplications,andotherpublicservices.Whenthisoccurs,performance
willhavebeentransformedfromameansofmeasuringandassessingservicesintoassurancethatservices
willbeprovidedassetforthinthebudget.

Servicesasrightsofcitizenshipwillnothappenallatonce.Theserightswillbetriedinsomeareas,then
extendedtoothers.Atfirst,therightsmaybepoorlydefined,withinadequatespecificationofqualitative
features;overtime,therightswillbeelaborated,astheyhavebeeninsomanyareasofsocialconduct.The
early rights may have no or weak redress; they will be rights without remedy. But remedies will be
introduced gradually, includingcompensation for subpar services. One practical remedy wouldbe to
empower citizens to obtain services from alternative providers when government fails to meet the
standards. Another would be to give citizens vouchers enabling them to purchase services from any
eligibleprovider.Ifthisweretooccur,citizenswouldbeempoweredascustomers,withtherighttotake
theirbusinesselsewherewhentheyaredissatisfied.

Prototypecitizensrightsmodelsareemergentinthecitizen'scharterideapioneeredinBritainandimitated
elsewhere.Althoughthecitizen'scharterwasallegedtohavestartedasapoliticalstunt,itmaturedintoa
seriousefforttospurqualitativeimprovementsinservices.Entitlingcitizenstocertainservicesmaybean
effectiveattractiveoptionforthosewhobelievethatpublicservicesshouldnotbeprivatizedorcontracted
out.Ofcourse,entitlingcitizenswillbeefficaciousonlytotheextenttheygetthepromisedservices.Asin

30
 PUMA/SBO(2001)4

otherentitlements,theobjectiveisnottocompensatetofailurebuttoassureperformance.Ifperformance
isnotforthcoming,definingservicesasentitlementswillbeahollowgesture.

6.5 Receivership

Underthebestofcircumstances,notallagencieswillprovidecitizenswiththeservicestowhichtheyare
entitled.Shortfallsinperformancemaybeduetoimpropermanagementofagencyfinances,lackofskills
neededtocarryoutassignedresponsibilities,anuncaringattitude,orcitizenswhoareindifferenttothe
quality of services or do not know how to obtain improvement. In these circumstances, the most
appropriateremedymaybetoplaceoperationsinthecustodyofareceiverwhoisauthorizedtoreplace
staff,managethebudget,andtakeotheractionsneededtoraiseperformancetoacceptablelevels.An
alternativewouldbetogivecitizenstheoptionofobtainingservicesfromotherproviders,butthismaybe
oflittlevalueforcitizenstrappedinpoorcommunities.

Receivership is not a new idea, though it is rarely applied to public programs. It typically refers to
situationswheretheentityisinsolventandunabletopayitsexpenses.Buttheconceptcanbereadily
adaptedtogovernmentalprogramswheremoneyisamplebutperformanceisbelowminimumlevels.

Thefutureshapeofgovernmentoperationsishighlyuncertain,moresothanforthtwootherbudgetary
functionsdiscussedinthispaper.Withequalplausibilityonecanenvisiongovernmentyieldingtomarkets
orgovernmentsdigginginandprotectingtheirtraditionalways.Theprobabilityisthatgovernmentswill
moveindifferentdirections,withsomemakingbroaduseofcontractsandothersrelyingonrulesand
controls.Almostallwillpursueimprovedperformance,buttheywilltakedifferentpaths.

Doing nothing will not be an option, however, because citizen expectations will continue to rise.
governments will be confronted with a choice: improve performance, spend more on operations, or
degradeservices.Mostwilloptforperformance,buttheywillsucceedonlyiftheyarewillingtodismantle
manyofthecontrolandcompliancesystemswhichshapethestructureofgovernmentandtheprovisionof
services.

7. DoesBudgetingHaveaFuture?

Ofcourse,theansweris"yes"ifthequestionreferstoproceduresforpreparingandimplementingbudgets.
Nolessthannow,futuregovernmentswillcompilebudgetsthataccountforrevenueandexpenditure.
Nationallegislatureswillappropriatefunds,spendingagencieswillfilereportsandcarryoutauthorized
programs,auditorswillreviewfinancialreports,andonandon.Procedurally,budgetingwillbeinrobust
health.

Theanswermaybe"no",however,ifwhatisimpliedisthatbudgetingwillbelesscapacitatedtodecide
thefinancesanddirectionofgovernment.Manyof'thedevelopmentsalreadyunderwayaswellasthose
thatmaylieinthefuturewillremakebudgetingintothedependentvariableingovernmentfinanceand
policy.Ratherthandrivingdecisionsonmoneyandprograms,budgetingwillsweptalongbypowerful
tides.Thebudgetwilldulyregisterwhathasbeendecidedalreadyorelsewherewhetherbyformulaorby
others,butitwillnotbetheforumformakingmanyofthedecisions.

31
PUMA/SBO(2001)4

The following paragraphs sum up major developments underwayin budgeting's fiscal, allocative, and
operationalroles.Thelistisnotcomplete,butitneverthelessaddsuptoastrongcasethatbudgetingwill
bedisplacedfromitsfavoredposition.
Sandwiched Budgeting. National budgets will be influenced by international rules and
requirementswhichprescribehowtheymanagetheirfinancesandwhattheyspendmoneyon,as
well as by local or regional governments which will have claim to much of the nation's tax
revenue.
NormativeBudgeting.Anincreasingportionofnationalbudgetswillbeallocatedbyfixednorms,
which will govern transfers to households, grants to subnational governments, and program
allocations.
Exogenous Budgeting. Revenues and expenditures will be increasingly dependent on outside
influenceseconomicandsocialconditions,capitalflows,exchangerates,theageandincome
structure of the population, medical technology, childrearing practices, and service levels.
Changesinthesefactorswillcompelautomaticadjustmentsinnationalbudgets.
GovernmentasFinancier.Thenationalgovernmentsmainbudgetresponsibilitywillbetolevy
taxesthatfinancespendingbyothersinternationalorganizations,households,andsubnational
governments.Theservicesitdirectlyprovideswillaccountforadecliningshareofthebudget.The
separationoftaxingandspendingauthoritywillgeneratebothpoliticalandbudgetarystress.
Balkanized Government. National governments will be fractured into numerous independent
agencies,eachwithitsownbudgetandeachempoweredtooperateandmanageitsresourcesasit
deems fit. Ministries or departments will still be responsible for developing policies and
coordinatingprograms,buttheywillhaveweakleveragevisavistheindependentagencies.
Higher Expectation, Less Trust. Budgets will be made in a political environment in which
citizensdemandmuchmorefromgovernmentbuthavelowconfidencethatwillperformwell.
Highexpectationswillpressuregovernmenttospendmoreandimproveservices,lowtrustwill
makeitdifficultforgovernmenttoextractmorerevenuefromitscitizens.

Budgeting has a future, but what it will be determined by what government becomes. Budgets and
governmentsareintertwinedtogether;itisnotpossibletoenvisionrobustbudgetingunlessgovernmentis
strongandcapacitatedanditisnotpossibletohavestronggovernmentifitsbudgetlacksthecapacityto
regulatepublicfinance.Thekeyquestionthereforeishowgovernmentwillevolveinthedecadeahead.
Answerthatquestionandthequestioninthetitleofthispaperalsowillbeanswered.

32