Você está na página 1de 42

Cost of Service Standards

August 18, 2011

Sponsored by U.S. Postal Service Office of Inspector General Risk Analysis Research Center Report Number: RARC-WP-11-008

U.S. Postal Service Office of Inspector General Cost of Service Standards

August 18, 2011 RARC-WP-11-008

Cost of Service Standards


Overview
Introduction The Postal Services network was designed for the high-speed transmission of FirstClass Mail correspondence and transactions for individuals and businesses. The diversion of much of this mail to electronic alternatives raises the question of whether the high speed service that some postal products receive is still worth its cost. The overnight First-Class Mail service standard requires the Postal Service to maintain processing plants that operate through the night and on Sundays. In addition, stringent service standards for First-Class Mail and Priority Mail are the primary factors that determine if mail travels by ground or by the more expensive air transportation. Figure 1 outlines service standards in the 48 contiguous states for major mail classes.1
Figure 1: End-to-End Service Standards by Major Mail Class

FirstClassMail

1-3 Days

Stringent Service Standards Increase Cost & Dictate the Network Size

PriorityMail

2-3

StandardMail

3-10 Days

Periodicals

1-9 Days

10

Source: U.S. Postal Service Service Standards for Continental United States

Some of the Postal Services largest business mailers have stated that they value consistency over speed and they would tolerate slightly slower service to save costs.2
1 2

U.S. Postal Service, Modern Service Standards, https://ribbs.usps.gov/index.cfm?page=modernservice. These mailers have stated a preference for service that is consistent, reliable, and cost-effective with reasonable speed. (See U.S. Postal Service Office of Inspector General, A Strategy for a Future Mail Processing & Transportation Network, Report No. RARC-WP-11-006, July 6, 2011.) Further, given the diversion of letter writing and bill payments to electronic communications, the public may also be less sensitive to a relaxation in the highest speed service standards, particularly the overnight standard.

U.S. Postal Service Office of Inspector General Cost of Service Standards

August 18, 2011 RARC-WP-11-008

Given these views and the recent and forecasted declines in First-Class Mail,3 the Postal Service may want to explore a reasonable relaxation of service standards as one alternative for saving costs. To provide the Postal Service and policymakers with more information on this option, the U.S. Postal Service Office of Inspector General (OIG) Risk Analysis Research Center (RARC) contracted with Christensen Associates (Christensen) to conduct basic research to identify and measure service-related costs.4 The attached paper by Christensen examines the costs that could be avoided by relaxing service standards by 1 day. Christensen found significant service-related cost savings for preferential mail. Preferential mail products include First-Class Mail, Priority Mail, and Periodicals. They incur additional costs because of their higher speed service standards. We selected 1 day for several reasons. First, we consider it to be a reasonable relaxation that should have a low impact on customer demand given mailers stated preference for consistency over high speed.5 Second, it would still maintain the differentiation in service between preferential and non-preferential mail products.6 Scope and Methodology The paper assesses the sources of service-related costs in the Postal Services mail processing, transportation, and delivery functions. Christensen examined four areas for potential costs savings when service is relaxed by 1 day: Premium Pay is paid for work at night and on Sundays.7 Relaxing service standards by 1 day permits mail processing to take place during the following day, avoiding the bulk of night (6 p.m. to 6 a.m.) and Sunday processing cost premiums. Other Mail Processing Costs include any additional labor the Postal Service might use to meet higher service standards.8 Christensen estimates this indirectly by comparing the unit cost of the preferred product category to the unit cost of the closest comparable Standard Mail product category by presort level. Transportation Cost is the cost difference between ground and air transportation. Ground transportation is usually less costly than air, but requires longer end-to-

Notably, Boston Consulting Group forecasted that First-Class Mail volume would be about 50 billion pieces in FY 2020, about 28 billion fewer pieces than in FY 2010. Standard Mail was forecasted to remain relatively flat with total mail volume falling to 150 billion pieces. 4 Christensen has expertise on Postal Service cost models, accounting systems, and operations. Although Postal Service costing systems do not measure the direct relationship between costs and service performance, the data are necessary to develop estimates of service-related costs. 5 Christensen did not attempt to measure the demand effects of relaxing service standards. 6 Destination entry non-preferential products can receive 2-day service, so there would be some overlap for workshared mail. 7 Night shift differential is a premium paid at a specified dollar rate for all hours worked between 6 p.m. and 6 a.m. Sunday premium is paid at 25 percent extra for work scheduled on Sunday. The overnight standard for First-Class Mail means that the bulk of mail processing occurs during the night hours, including most nonpreferential mail. 8 The Postal Services In-Office Cost System (IOCS) measures the amounts of labor time associated with handling mail products. This should indirectly capture any additional work effort provided for service performance.

U.S. Postal Service Office of Inspector General Cost of Service Standards

August 18, 2011 RARC-WP-11-008

end transit times for longer haul shipments. Relaxing service standards by 1 day allows the Postal Service to shift mail that is transported longer distances from air to ground. Overtime Pay is a potential source of service-related costs. An extra day of service could allow the Postal Service to smooth peaks in labor usage during the week and avoid some overtime costs.

In addition, there are other potentially significant areas for cost savings that were outside the scope of Christensens analysis. If service standards were relaxed, processing plants could cover larger service areas, allowing the Postal Service to have fewer plants. A longer service window would give plants additional time to process the mail. This extra time would allow the Postal Service to schedule employees more efficiently and run fewer pieces of equipment for longer periods. Results in Brief Table 1 presents the service-related mail processing costs that are avoided by extending service standards for preferential mail classes by 1 day. Overall, the research identifies savings of up to $1.5 billion in premium pay and other mail processing costs.9 This is a lower bound estimate as it excludes potentially significant sources of cost savings from network restructuring and consolidation.
Table 1: Service-Related Avoided by 1 Day Relaxation of Service Standards (in Millions) Savings Opportunity Premium Pay Other Mail Processing Costs Total Amount $ 336 $ 1,139 $ 1,475

The attached Christensen paper provides the technical details of the research and analysis. We provide here some highlights: 1. Premium pay for night-shift and Sunday mail processing for First-Class Mail, Priority Mail, and Periodicals in FY 2010 had a total cost of $336 million. Relaxing service standards could allow mail processing to be scheduled largely during day shifts and away from Sunday, avoiding premium pay costs. 2. Other mail processing cost differences were more than $1 billion in FY 2010. At least some, if not all, of these differences are due to high cost levels in outgoing mail processing and elevated staffing levels needed to meet preferential mail service standards.
9

The research also identifies $120 to $197 million in transportation costs, although these are not likely achievable at the same time as the mail processing savings.

U.S. Postal Service Office of Inspector General Cost of Service Standards

August 18, 2011 RARC-WP-11-008

3. Relaxing service standards by 1 day allows the Postal Service to meet end-toend service standards for zones 3-6 Priority Mail using ground transportation. The difference between ground and air transportation for these zones was between $120 and $197 million in FY 2010.10 Similar savings may be available for First-Class Mail, but data are not available to estimate middle-zone air transportation costs. 4. In FY 2010, overtime premium pay was $811 million.11 Overtime is only slightly more expensive than straight time hours with benefits and can often provide more flexibility. Overtime can be cheaper than incurring minimum guaranteed workhours of 4 or 8 hours. The Postal Service often uses overtime to cover normal staffing requirements for a given day. Nevertheless, extending service standards should allow the Postal Service to smooth its labor usage and avoid some overtime workhours caused by Monday peaks in workload.12 The specific overtime costs related to meeting service standards, however, are not measurable using existing data. 5. Relaxing service standards provides for longer processing and transportation windows. This provides additional substantial benefits to the Postal Service including reducing the number of trips between Post Offices and plants for collection and delivery,13 improving plant operation schedules that more fully utilize 8 hour shifts,14 and lengthening machine runtime requiring fewer machines, less floor space, and less maintenance labor. Relaxing service standards also allows plants to have geographically larger service areas. This is an opportunity for the Postal Service to realign its network and capture additional, significant cost savings; however, assessing network consolidations was beyond the scope of this study and would require extensive network modeling. Our recently published paper, A Strategy for a Future Mail Processing & Transportation Network, assessed the potential savings from network consolidations given the current service standards.15 6. Current Postal Service cost models and accounting systems are not designed to measure service-related costs. Further, the Postal Service lacks important
The range is based on different assumptions of how ground transportation costs vary with respect to longer distances. The low end estimate assumes a constant cost per cubic-foot mile. The high end estimate assumes the cost per cubic-foot mile decreases the longer the distance traveled. 11 The Postal Service pays clerks, mail handlers, and city carriers overtime at 1.5 times base hourly rates for hours 8-10 of a workday and for hours exceeding 40 in a workweek. Penalty overtime at twice the base hourly rate is paid essentially for hours exceeding 10 per regularly scheduled workday. 12 Mail volume is 29 percent higher on average on Monday than the average for Tuesday through Saturday. Total work hours for mail processing clerks and mail handlers and overtime hours for city carriers exhibit some peaking associated with Monday volumes. Friday and Saturday show high overtime usage, but this is partly due to overtime early in the week leading to workhours exceeding 40 hours later in the week. 13 Currently there are multiple trips to pick up collection mail so that the plant can begin operations. Similarly, there are multiple trips to delivery units in the morning so that carriers can begin casing the mail as the plant completes its automated sortation. 14 The overnight service standard results in some operational windows that are less than a full shift. Lengthening processing windows may allow the Postal Service to better process workloads in full shifts. 15 U.S. Postal Service Office of Inspector General, A Strategy for a Future Mail Processing & Transportation Network, Report No. RARC-WP-11-006, July 6, 2011, http://www.uspsoig.gov/foia_files/RARC-WP-11-006.pdf.
10

U.S. Postal Service Office of Inspector General Cost of Service Standards

August 18, 2011 RARC-WP-11-008

service-related data including reliable origin-destination mail volumes. If widely adopted, the Intelligent Mail barcode (IMb) can provide the necessary data to compute more precise service-related costs. 7. Past concerns of requiring labor intensive manual processing to meet service standards due to insufficent machine capacity should no longer be an issue since with the declines in mail volumes, the Postal Service now has an excess of machine capacity. As the service-related costs associated with manual labor are much less than they once were, they should not be a significant source of savings. 8. Non-preferential, deferrable mail products such as Standard Mail do not have significant service-related labor or transportation costs. Conclusion As the Postal Service, the postal community, and policymakers explore how to cut postal costs, adjusting service standards should be considered. The following paper estimates the Postal Service could save up to $1.5 billion by relaxing service standards by 1 day. This is almost certainly a lower bound estimate as this paper presents preliminary results, and additional opportunities for savings exist from restructuring and consolidation.

CostofService StandardsintheUnited StatesPostalService

August 4, 2011

Laurits R. Christensen Associates, Inc. 800 University Bay Drive, Suite 400 Madison, WI 53705-2299 Voice 608.231.2266 Fax 608.231.2108

ExecutiveSummary ThegoalofthePostalServiceistocollect,sort,transport,anddelivermail withintheservicestandardtargetsforeachclassofmail.Nationalservicestandards byclasswerefirstestablishedin1971.Theseservicestandardsremainmostly unchangedforStandardMail,PackageServices,andPriorityMail,whilesome changeshavebeenimplementedforFirstClassMailandPeriodicals.Underthe PostalAccountabilityandEnhancementAct,thePostalRegulatoryCommissionis chargedwithreviewingthePostalServicesqualityofserviceandrequiresservice performancereporting.However,thecostimpactoftheseservicestandardshasnot beenstudied.Thisisafirststepinansweringthequestion,whatwouldthesavings befromaonedaychangeinservicestandardsforeachclassofmail? Thisreportdescribessourcesandcausesofservicerelatedcostsinthe PostalServicesmailprocessing,transportation,anddeliveryfunctions.Whilepostal productshavevariousservicecharacteristics,wefocusonstandardsforendtoend daysfromacceptancetodelivery.Wherepossible,thisreportaimstoquantifycosts thatcouldbeavoidedwithaonedayextensionofendtoendservicestandardsfor majorproducts.Forexample,FirstClassMailcurrentlyofferingdeliveryin13 dayswouldinsteadreceive24dayservice;StandardMailwouldmovefrom310 daysto411days. Thepurposeofthisreportisnottorecommendanyspecificservicechanges orrelatedoperationalchanges.Rather,thegoalistoidentifyareasofpotentialcost savingsquantifyingifpossiblewithexistingdataforahypotheticaloneday servicestandardchange.Furtherresearchisneededifanychangesaretobe implemented.Herearethekeyfindings: 1. Thereareapproximately$2.5billioninannualcosts(basedonFY2010cost levels)inmailprocessing,transportation,anddeliveryfunctionsthatareat leastpartlyservicerelated.But,theentiretyofthesecostsarenotavailable aspotentialcostsavingsfromextendingservicestandardsbyoneday becauseoftheinterrelationoffunctions.Forexample,lengtheningtransit timesbysubstitutinggroundforairtransportationmayforeclosesavings fromdeferringmailprocessingordeliveryactivities.Also,theorganizational changesrequiredtoobtaincostsavingsincertainservicerelated componentsmaybesubstantial. ThePostalServicepayspremiumstoemployeesforworkingSundayshifts andfrom6P.M.to6A.M.,definedasnighthours.InFY2010,theservice relatedcostofpremiumpaywas$389millionunderthecurrentcost methodology.Aonedayextensionofservicestandardscouldallowpremium paycoststobelargelyavoided,sincethatwouldbesufficienttoallow processingwindowstooccurduringdayshiftsandtominimizetheneedfor Sundayprocessing.

2.

3. WhilethedirectcostsofeliminatingnightandSundaypaypremiumsare modest,shiftingprocessingwindowscouldenablesignificantlongterm savingsfrommailprocessingnetworkrealignment,sincetherecouldbe additionaltimetotransportmailbetweenpostofficesandmailprocessing plants.InFY2010,thePostalServiceincurred$4.7billioninannualcosts relatedtomailprocessingfacilityspaceandequipment.ArecentOIGstudy (RARCWP11006)concludedthatanew135plantnetworkwouldcost around$2billionlessperyeartooperatethanthecurrentnetwork, excludingtransitioncosts.Thisnewplantstructureextendsplantsservice radiibyaround50miles,andtherequiredtraveltimebetweenplantsand outlyingpostofficesmaybemorefeasibletoaccommodatewithaoneday relaxationofservicestandards. Overtimepaytotaled$811millionforclerks,mailhandlers,andcarriersin FY2010.Theservicerelatedportionofovertimecostscannotbedetermined fromavailabledata,inpartbecausethecausesofovertimeincurredinlow volumeperiodsareambiguous,andinpartsincecostsofstraighttime benefitsmayoffsetsomeoftheovertimepremium.However,thereisa substantialportionofovertimethatappearstocoincidewiththeMonday deliveryworkloadpeak,whichmaybepartlyavoidablebyadditional workloaddeferral.Overtimeusehasdeclinedsharplyasvolumehasdeclined overtherecession. Byusing24daygroundtransportationinlieuofairtransportation,the PostalServicecouldreducePriorityMailairtransportationcostby approximately$120$197millionperyear,orapproximatelyxxxxxpercent ofPriorityMailairtransportationcosts.Dataareinsufficienttoestimate correspondingsavingsforFirstClassMail,butwouldlikelybe(relatively) lowerbecauseoflowerairsharesinzones35. ThereappeartobesubstantialservicerelatedcostsofhandlingFirstClass SinglePiecemail,whichhasmailprocessingcostsapproximately$1.13 billionperyearhigherthanbulkmailbenchmarkcostsatFY2010costlevels. Theavailablecostdatacannotisolatetheservicerelatedcomponentofthe difference,butthedatadosuggestthatamajorcausemaybehighstaffing levelsusedtoclearoutgoingmailatmailprocessingplants.Incontrast,there islittleevidenceofsubstantialadditionallaborbeingprovidedforFirstClass PresortmailandPeriodicals,relativetoStandardMail,forservicereasons. MailvolumesfordeliveryexhibitastrongMondaypeak,withMondaypieces (summedoverallproducts)29percenthigherthantheaverageoftheother fivedeliverydays.StandardMailletterandflatpeakscoincidewiththe overallMondaypeak,whichmayindicateunusedopportunitiesforshaving theMondaydemandpeakwithinexistingservicestandards.FirstClass SinglePiecemailhasalateweek(ThursdayFriday)peak.

4.

5.

6.

7.

8. Issuesthathavefiguredprominentlyindiscussionsofservicerelatedcosts inthepast,suchasautomationcapacityconstraints,maybelessrelevantin lightofrecentvolumedeclinesandtheadvancedstateofthePostalServices automationprograms. Costsavingsfromservicereductionsmaybeoffsetbylossofmailvolume becausedemandforpostalproductswouldbelower,otherthingsequal, underreducedservicestandards.Aonedayextensionofservicestandards wouldnotaffectallpreferentialproductsidentically.Curtailingoneday servicewouldhavethelargesteffectsonFirstClassSinglePieceletters, wherehalfofthevolumeissubjecttotheonedaystandard.However,since serviceisnotexplicitlypricedandretailmailershavelimitedchoiceover postalservicecharacteristics,thesensitivityofpostaldemandtoservice performanceisnoteasilymeasured.Insomecases,weexpectproduct featuressuchasservicereliabilityorvisibilitymaybemorevaluablethan endtoendspeed.Analysisofthesedemandsideeffectsisbeyondthescope ofthispaper.

9.

TableofContents 1.Introduction...........................................................................................................................1
ExistingServiceStandards..............................................................................................................2 CausesofServiceRelatedCosts....................................................................................................3 VolumeLossandImplementationCosts....................................................................................3 SummaryofMainFindings.............................................................................................................4 .

2.ThePeakLoadIssueandVolumeandWorkhourProfilesforthePostal Service..........................................................................................................................................5
DefinitionofPeakLoad.................................................................................................................5 IncreasedAutomationandDeferrableMailVolumeGrowth.............................................7 PreferentialversusNonPreferentialVolumes.......................................................................7 MailVolumesbyDayofWeek........................................................................................................8 WorkhoursbyDayofWeek.........................................................................................................12

3.AnalysisofServiceRelatedCosts................................................................................16
PremiumPay.....................................................................................................................................16 PremiumPayforMailProcessing.............................................................................................16 PiggybackedCosts.......................................................................................................................17 . WorkloadShift..................................................................................................................................18 ImpactonTransportation............................................................................................................18 MailProcessingCostDifferentialsforPreferentialand NonPreferentialProducts..................................................................................................19 ImpactofTimeofDay....................................................................................................................21 ExpeditedTransportation............................................................................................................23 CostDifferentialsforExpeditedTransportation.................................................................24 Overtime.............................................................................................................................................26 OvertimeCosts.............................................................................................................................................26 OtherCausesofOvertime........................................................................................................................27

4.Summary..............................................................................................................................28 i

ListofTables Table1.FY2010Volumes,SelectedMailProductsbyServiceStandard(ODIS)............2 Table2.MailProcessingPremiumPayCostsbyProduct,FY2010...................................17 Table3.EstimationofPotentialServiceRelatedCostDifferentialsforSelected PreferentialandNonPreferentialProducts,FY2010...................................................21 Table4.AirHighwayCostDifferences,Zone36PriorityMail,FY2010Dollars........25 Table5.OvertimePremiumPay,SelectedCrafts,FY2010....................................................26 Table6.SummaryofServiceRelatedCosts................................................................................28

ii

ListofFigures Figure1AverageDailyVolumeProfileLettersandCards................................................8 Figure2AverageDailyVolumeProfileFlats..........................................................................9 Figure3AverageDailyVolumeProfileParcels.....................................................................9 Figure4AverageDailyVolumeProfileFirstClassSinglePieceLetters.................10 Figure5AverageDailyVolumeProfileFirstClassPresortLetters...........................11 Figure6AverageDailyVolumeProfileStandardRegularLetters(nonECR)......11 Figure7AverageDailyWorkhoursMailProcessingClerks..........................................12 Figure8AverageDailyWorkhoursMailHandlers............................................................13 Figure9AverageDailyWorkhoursCityDeliveryCarriers............................................13 Figure10AverageDailyOvertimeWorkhoursMailProcessingClerks..................14 . Figure11AverageDailyOvertimeWorkhoursMailHandlers....................................15 Figure12AverageDailyOvertimeWorkhoursCityDeliveryCarriers....................15 . Figure13IOCSCostsbyTimeofDay,ClerksatMailProcessingPlants.....................22 Figure14IOCSCostsbyTimeofDay,MailHandlersatMailProcessingPlants.....22 Figure15IOCSMailProcessingCostsbyTimeofDayandProduct............................23 .

iii

1.Introduction Servicestandardssuchasthetimeintransitfromacceptancetodeliveryare majorfeaturesofU.S.PostalServiceproducts.Theovernightstandardforlocal FirstClassMailandthe23dayservicestandardfornonlocalPriorityMailare widelypublicized.UnderthePostalAccountabilityandEnhancementAct,thePostal RegulatoryCommissionischargedwithreviewingthePostalServicesqualityof serviceandrequiresserviceperformancereporting.Whileservicestandardsfor PostalServiceproductsmaybewellknown,measuringservicerelatedcostsin isolationfromothercostcausingfactorshasnotbeenamajorfocusofPostal Servicecostdevelopment.Issuesthathavefiguredprominentlyindiscussionsof servicerelatedcostsinthepast,suchasautomationcapacityconstraints,maybe lessrelevantinlightofrecentvolumedeclinesandtheadvancedstateofthePostal Servicesautomationprograms. Thisreportdescribessourcesandcausesofservicerelatedcostsinthe PostalServicesmailprocessing,transportation,anddeliveryfunctions.Whilepostal productshavevariousservicecharacteristics,wefocusonstandardsforendtoend daysfromacceptancetodelivery.Wherepossible,itaimstoquantifycoststhat couldbeavoidablewithaonedayextensionofendtoendservicestandardsfor majorproductswithpreferentialstandards(FirstClassMail,PriorityMail, Periodicals)andfornonpreferentialproducts(StandardMail,PackageServices).1 Thatis,preferentialmailcurrentlyofferingdeliveryin13dayswouldinstead receive24dayservice;currentlydeferrablemailcouldbedeferredanextradayas needed.Thiswouldconstituteasubstantialservicechange,butforthepurposesof thisexerciseweretainthemajordistinctionsbetweenpreferentialandnon preferentialproducts.Wearenotrecommendinganyservicechangesorrelated operationalchanges.Rather,theadditionaldayconceptisintendedtoallowcosts thatmayvarywithservicestandardstobeidentified. Intermsofreportlayout,wefirstbrieflyreviewissueswithpeakloadand similarservicerelatedcosts.Sincetheexistenceofpeakloadcostissuesdependson theconvergenceofdemandvariationsandcostinflexibility,wedescribeavailable dataonvolumeandworkhourprofilesbydayoftheweekandbyseason,andon mailprocessingcostsandworkhoursbytimeofday.Weexaminepremiumpay, overtime,andairtransportationcostsasservicerelatedcostcomponents,and discussthepotentialforcostsavingsinthosecomponentsfromrelaxingservice standards.Toinvestigatecoststhatmayresultfromassigningadditionallaborto
1Whilethepreferentialandnonpreferentialterminologyisobsolescentand

doesnotreflectfinerdetailsofservicestandardsforspecificpostalproducts,we employitasusefulshorthandtorefercollectivelytogroupsofproductswithhigher andlowerservicestandards. 1

preferentialproductstomeetservicestandards,weexaminemailprocessingcost differentialsbetweenpreferentialandcomparablenonpreferentialproducts. ExistingServiceStandards ThePostalServicesOriginDestinationInformationSystem(ODIS)provides asnapshotoftheservicestandardprofileofretailmailproducts.Asevidencedin Table1,theFY2010servicestandardforoverhalf(55%)ofallFirstClassSingle Piecelettersisoneday(overnight),asisapluralityofFirstClassSinglePieceflats; thoseproductswouldseealargeimpactfromconvertingonedaytotwodaymail. However,preferentialvolumesconstitutearelativelysmallshareofflatshapedmail overall.Onedaymailconstitutesmuchsmallersharesofparcelshapeproducts.The servicestandardformostFirstClassSinglePieceparcelsandPriorityMailis2to3 days,andtheservicestandardformostPackageServicesMailis3to4daysormore. Piecesinwhichtheservicestandardisunknown(e.g.,withnoclearoriginZIPcode onthemailpiece)havebeenexcluded.2 Table1.FY2010Volumes,SelectedMailProductsbyServiceStandard(ODIS)
Volume(MillionsofPieces) ServiceStandard Product 1Day 2Day 3Day 4Day+ Total FCMSinglePieceLetters 15,200 6,314 5,929 71 27,514 FCMSinglePieceFlats 795 461 495 6 1,757 FCMSinglePieceParcels 53 84 204 2 344 PriorityMail x x x x x PackageServicesSinglePiece 2 0 28 12 42

PercentofVolumebyServiceStandard ServiceStandard Product 1Day 2Day 3Day 4Day+ FCMSinglePieceLetters 55% 23% 22% 0% FCMSinglePieceFlats 45% 26% 28% 0% FCMSinglePieceParcels 16% 25% 59% 1% PriorityMail x x x x PackageServicesSinglePiece 5% 1% 66% 29%

Total 100% 100% 100% 100% 100%

2UnknownservicestandardvolumesaresmallfractionsofODISvolumesforFirst

ClassMailandPriorityMail,butareamajorityofSinglePiecePackageServices. 2

CausesofServiceRelatedCosts Servicerelatedcostscanbecausedbyavarietyoffactors.First,thePostal Servicemayneedtoundertakecostlyactionstomeetservicestandardsthatwould beunnecessaryifthestandardswererelaxed.Forexample,thePostalService providesnextdayserviceforlocalFirstClassMailinpartbyschedulingmail processingactivitiesfornightandSundayshiftsforwhichthePostalServicemust paywagepremiumstoclerksandmailhandlers.Also,thePostalServicemust purchaseairtransportationinsteadofslower,butlesscostly,groundtransportation tomeetservicestandardsforFirstClassMailandPriorityMailtravelinglonger distances.Second,daytodayorseasonalvariationsindemandforpreferential productscangiverisetopeakloadproblemswhencostsareinflexiblerelativeto demand.Third,thePostalServicesnetworkmaybeorganizedinpartaround preferentialproductsservicestandards,particularlynextdayserviceforlocalFirst ClassMail.Relaxingservicestandardscouldenablenetworkrealignmentwith potentiallysubstantialreductionsinfacilityspaceandequipmentcosts. Existingaccountingsystemandcostmodelmethodsprovideabasisfor measuringsomeservicerelatedcosts,suchascostsofshiftdifferentialsandair transportationcostpremiums,butareinsufficienttodevelopcomprehensive measurementsofservicerelatedcosts.Existingdataandmethodsareintendedto estimatethetotal(oraverage)costsattributabletoproductsandonlysecondarily toprovideinformationonthecausesofthosecosts.Boththecostsofproviding certainlevelsofendtoendserviceandthepotentialcostsavingsfromrelaxing servicestandardswouldbeexpectedtodependcriticallyontheoperationaldetails ofthestandardsimplementation.Notallservicerelatedcostswillnecessarilybe avoidableunderspecificapproachestorelaxingstandards.Forinstance, lengtheningenroutetimesbysubstitutinggroundforairtransportationmay foreclosesavingsfromdeferringmailprocessingordeliveryactivities. VolumeLossandImplementationCosts Demandsideconsiderationsarebeyondthescopeofthispaper,butwould beanimportantfactorfordeterminingtheneteffectsofactualservicestandard changesonPostalServicefinances.Ingeneral,servicereductionswouldreduce demandfortheaffectedproducts(otherthingsequal),solowernetrevenueswould offsetsomeofthepotentialcostsavings.Therelativemagnitudeofrevenuelosses wasamatterofcontentioninthe5daydeliveryproceedingbeforethePostal RegulatoryCommission(DocketNo.N20101),wherePostalServiceandPostal RegulatoryCommissionestimatesofprospectiverevenueeffectsvariedwidelyand wereeachsubjecttoconsiderableuncertainty.3Forservicestandardchanges,the
3See,e.g.,PostalRegulatoryCommission,DocketNo.N20101,AdvisoryOpinionon

EliminationofSaturdayDelivery,p.33. 3

centralunknownsincludetheextenttowhichcustomersspecificallydemandanX daystandard,andcustomerswillingnesstopayforXdaydeliveryinsteadof(say) X+1daydelivery.Insofarastheserviceperformancefortheproductsweconsideris notguaranteed,itwouldstandtoreasonthatpostalcustomersareatleast somewhatflexibleintheirservicerequirements. Inaddition,whilewehaveconsideredwhethersavingsincertaincost componentsmaybepossibleunderrelaxedservicestandards,wehavenot attemptedtoassessoperationaldetailsofimplementingservicechanges,including potentialnetworkadjustmentcostsorbenefits.Anysucheffects,thoughtheyare potentiallylargerelativetodirecteffectspresentedhere,wouldbepurely speculativeabsentverysubstantialadditionalresearchandanalysis.Basedonthe experienceofDocketNo.N20101,estimationofsavingsfromconsolidatingPostal Servicefacilitiesandevaluationofotherconsequencesofprospectiveservice changeswouldbehighlycontroversial. SummaryofMainFindings Hereareourmainfindingsregardingpotentialcostssavings: MailprocessingpremiumpayexpensesfornightandSundayworkhavebeen treatedaslargelyconstitutingaservicerelatedcostofpreferentialmail. FY2010mailprocessingpremiumpaytreatedasservicerelatedcostis$389 million,$336millionofwhichisassociatedwithpreferentialmailservice underthecurrentcostmethodology.Aonedayextensionofservice standardscouldallowpremiumpaycoststobelargelyavoided,sincethat wouldbesufficienttoallowprocessingwindowstooccurduringdayshifts andnonSundays.Whilethedirectcostsaremodest,shiftingprocessing windowscouldenablelargelongtermsavingsfrommailprocessingnetwork realignment,sincetherecouldbeadditionaltimetotransportmailbetween postofficesandmailprocessingplants. Byusing24daygroundtransportationinlieuofairtransportation,the PostalServicecouldreducePriorityMailairtransportationcostby approximately$120$197millionperyear,orapproximatelyxxxxxpercent ofPriorityMailairtransportationcosts.Dataareinsufficienttoestimate correspondingsavingsforFirstClassMail,butwouldlikelybe(relatively) lowerbecauseoflowerairsharesinzones35. Overtimepaytotaled$811millionforclerks,mailhandlers,andcarriersin FY2010.Theservicerelatedportionofovertimecostscannotbedetermined fromavailabledata,inpartbecausethecausesofovertimeincurredinlow volumeperiodsareambiguous,andinpartsincecostsofstraighttime benefitsmayoffsetsomeoftheovertimepremium.However,thereisa substantialportionofovertimethatappearstocoincidewiththeMonday 4

deliveryworkloadpeak,whichthereforemaybepartlyavoidableby additionalworkloaddeferral.Overtimeusehasdeclinedsharplyasvolume hasdeclinedovertherecession. CostsforFirstClassSinglePiecemailaresubstantiallyhigherthanmixed AADC(ormixedADC)benchmarkcostsfromStandardMailandFirstClass Presortmail.Theeffect,approximately$1.13billionperyearatFY2010 costs,istoolargetobetheresultofknownnonsamplingerrorsinmail processingcosts.Costdatasuggestatleastaportionofthedifferencereflects staffinglevelsintendedtoclearoutgoingpreferentialmail,mostlyFirstClass SinglePiecemail,tomeetservicestandards.Incontrast,thereislittle evidenceofsubstantialadditionallaborbeingprovidedforFirstClass PresortmailandPeriodicalsforservicereasons.Costdifferencesbetween thoseproductsandStandardMailbenchmarksaredrivenbypremiumpay costsandproductfeaturesotherthanendtoendservicestandards. MailvolumesfordeliveryexhibitastrongMondaypeak,withMondaypieces (summedoverallproducts)29percenthigherthantheaverageoftheother fivedeliverydays. StandardMailletterandflatpeakscoincidewiththeoverallMondaypeak, whichmayindicateunusedopportunitiesforshavingtheMondaydemand peakwithinexistingservicestandards.FirstClassSinglePiecemailhasa lateweek(ThursdayFriday)peak.

2.ThePeakLoadIssueandVolumeandWorkhourProfilesforthePostal Service DefinitionofPeakLoad Peakloadcostsareapotentialsourceofservicerelatedcostsforpostal products.AsdescribedindetailbyKleindorfer,thepeakloadprobleminvolvesthe followingcriticalelements: a) b) Theproductiseconomicallynonstorable Demandmustfluctuateperiodicallyovertimewithresulting peaksandvalleys

c) Capacitymustbeinflexiblesothattheidlecapacityimpliedby (a)and(b)isnotcostless(i.e.,oncecommittedthecapacity mustbepaidfor,whetherusedornot).4

Underthesecircumstances,efficientpricesarebasedonshortrunmarginal costswithindemandperiods;shortrunmarginalcostsinoffpeakperiodswilltend tobelow.Inthepostalcontext,thisreasoningwasusedtoarguethatcapacitycosts wereincurredinserviceofpeakpreferentialdemandsandwereexcessively assignedtononpreferentialproducts.Thisargumentwasstronglycounteredby prominentregulatoryeconomistsin1980sratecases.Thecurrentstructureofthe laborcostmethodsusedbythePostalServiceandthePostalRegulatory Commissionlargelyrejectthisview.Asaresult,thereisrelativelylittlecostdirectly treatedasservicerelatedcostsassociatedwithpreferentialmaildemandpeaks.The notableexceptionrelatestocostsassociatedwithshiftdifferentials,discussedin section3,below. KleindorferobservedthatconditionsinthePostalServicesmailprocessing systemdifferfromthetraditionalpeakloadprobleminanumberofimportant ways.Notably,postalproductsparticularlydeferrableproducts,butinsomecases alsopreferentialproductsarestorablewithinlimits.Itis,therefore,possibleboth toshavedemandpeaksandtofillindemandtroughstosomeextent.Finally,postal resourcesarenottotallyinflexible.WhilethePostalServicefacescertainlaborand capacitycostrigidities,atleastsomelaborand/orcapacitycostsareavoidableif theyareunneeded.Kleindorferconsideredanumberofscenariosfordeferrable volumeavailabilityandtechnologicalrigidityandconcludedthatmeasuredmail processingcostswereunlikelytodivergegreatlyfromactualmarginalcosts consideringtheresidualpeakloadproblem. Arelatedissueisthatcertainpostaloperationsorfunctionsserveas bottlenecksandmaybesaturatedatpeakvolumeperiods.Thestockofmail processingequipmentisrelativelyfixedintheshortterm,sothemaximum processingcapacityofaplantisafunctionofthenumberofmachinesinplace, machinethroughput,andthelengthofavailableprocessingwindows.Then,ifpeak preferentialvolumessaturateautomationcapacityandmarginalpreferentialmailis notdeferred,servicerelatedcostsarisethroughtheprovisionofalternativemanual handling(athighercost).Similarly,onpeakdeliverydays,preferentialmaildelivery volumesmayrequiremorethananeighthourshiftsworktosequenceanddeliver, andresultingpaymentsofovertimepremiumsareservicerelated.

4PostalRateCommission,DocketNo.R871,DirectTestimonyofPaulR.Kleindorfer

onbehalfoftheUnitedStatesPostalService(USPST4),p.6. 6

IncreasedAutomationandDeferrableMailVolumeGrowth Sincethepeakloadissuewaslitigatedoverseveralratecasesinthe1980s, thePostalServicehasundergonesignificantstructuralchanges,includingexpansion ofautomatedprocessingcapacityandgrowthindeferrableStandardMailvolumes relativetopreferentialproductsvolumesforlettersandflats.Wewouldexpect these,particularlyincombinationwithrecentvolumedeclines,toprovide considerablerelieffrombottlenecksandcapacityconstraintsthatmayhaveexisted atthePostalServicesvolumepeak. Forexample,automatedletterandflatequipmentdeploymentswerescaled formuchhighervolumelevelsthancurrentlyprevail,soautomatedprocessing capacityislesslikelytobeconstrainedatpeakperiods.Withinrelativelynarrow processingwindows,thePostalServicesortsnearly100billionlettersizepiecesto deliverypointsequence(DPS);outgoingprocessing,whichusesthesame equipmentforamuchsmallervolumeofFirstClassSinglePiecelettersandmixed AADClettersinFirstClassPresortandStandardMail,thereforeisunlikelytoface bindingcapacityconstraints.Automateddeliverypointsequencingoflettershas servedtoreducecarriersinofficeworkloads,whichwouldtendtobesensitiveto volumefluctuations,andtoincreasestreettime,whichincludesrelativelyless variablerouterunningcosts.However,sinceDPSinvolvescomminglingofmail classes,effectivelyupgradingStandardMailmixedwithFirstClassMail,itmaylimit theusefulnessofStandardMaildeferabilityinmanagingdeliveryworkloads. PreferentialversusNonPreferentialVolumes Theavailabilityofdeferrablenonpreferentialvolumesalsoisimportantin determiningtheeffectofpreferentialvolumesoncosts.Ifsufficientdeferrable volumeisavailableinthesystem,thePostalServiceshouldhaveconsiderable flexibilitytoavoidservicerelatedcostsarisingthroughcapacityrelatedcost premiums.Indeed,inFY2010,aggregateStandardMailvolume(82.5billionpieces) exceededFirstClassMailvolume(77.9billionpieces).Forflatshapemail,the overwhelmingmajorityofvolumeisdeferrable(includingPeriodicalsandStandard Mail)34.8billionStandardMailandPeriodicalsflatsversusfewerthan3billion FirstClassMailandPriorityflats.5Deferrableflatsvolumesarepotentially significantfordeliveryworkloadlevelingsinceflatsgenerallyneedtobesortedto deliverypointsequencebycarriers,atleastpendingwidespreadautomationofflat shapedeliverypointsequencingonFlatSequencingSorters(FSS).
5Periodicalsaretreatedaspreferentialmailforsomepurposes,butasstatedin

DMM70728.1,TheUSPSdoesnotguaranteethedeliveryofPeriodicalswithina specifiedtime.PublicationsauthorizedorpendingauthorizationforPeriodicals entryreceive,wherepracticable,expeditiousdistribution,dispatch,transithandling, anddelivery(emphasisadded). 7

MailVolumesbyDayofWeek Figures13showprofilesofaveragedailyvolumesforletters,flats,and parcelsbydayofweekandmonth,basedonsampledatafromODIS.Mondaypeaks areobservableforallthreeshapes;intotal,Mondayvolume(pieces)is29percent higherthantheTuesdaySaturdayaverage.Letterandparcelvolumesalsoshow lowersecondarypeakslaterintheweek.Onlyparcelvolumeshaveanotable Decemberpeak. Figure1AverageDailyVolumeProfileLettersandCards


FY2010 Average Daily Volume by Day All Products - Letters and Cards
700

600 Volume (millions of pieces per day)

December

500 Overall 400 August 300

200

100

0 Mon Tue Wed Thu Fri Sat Day of the Week Peak Month (Dec) Low Month (Aug) Overall Average

Figure2AverageDailyVolumeProfileFlats
FY2010 Average Daily Volume by Day All Products - Flats
250

Volume (millions of pieces per day)

200 November

150 Overall 100

June 50

0 Mon Tue Wed Thu Fri Sat Day of the Week Peak Month (Nov) Low Month (June) Overall Average

Figure3AverageDailyVolumeProfileParcels
FY2010 Average Daily Volume by Day All Products - Parcels
16

14 December Volume (millions of pieces per day) 12

10 Overall 8

6 July 4

0 Mon Tue Wed Thu Fri Sat Day of the Week Peak Month (Dec) Low Month (July) Overall Average

FirstClassSinglePiecelettershaveadistinctvolumeprofile,includinglate week(ThursdayFriday)peakandhighDecembervolumerelativetoothermonths. SeeFigure4.TheDecemberpeakforSinglePiecelettersisoffsetintheoverallletter profilebysomewhatlowerDecembervolumesforotherlettershapeproducts. Figure4AverageDailyVolumeProfileFirstClassSinglePieceLetters


FY2010 Average Daily Volume by Day First-Class Single Piece Letters and Cards
250

December 200 Volume (millions of pieces per day)

150

Overall 100

August 50

0 Mon Tue Wed Thu Fri Sat Day of the Week Peak Month (Dec) Low Month (Aug) Overall Average

FirstClassPresortletterssometimesexhibitverysharpMondayvolume peaks.SeeFigure5.FirstClassPresortlettervolumeshaveadistinctTuesday Wednesdaytrough,followedbyalateweekplateau.Standardletters(Figure6) haveanoverallpeakonMonday,butalsoshowTuesdaypeaks.Overall,Standard lettervolumesarelessvariablethanFirstClassPresortletters.Thismay,inpart, reflectsomepeakshavingusingthedeferabilityoftheproduct.

10

Figure5AverageDailyVolumeProfileFirstClassPresortLetters
FY2010 Average Daily Volume by Day First-Class Presort Letters and Cards
300

January 250 Volume (millions of pieces per day)

200 Overall 150 August 100

50

0 Mon Tue Wed Thu Fri Sat Day of the Week Peak Month (Jan) Low Month (Aug) Overall Average

Figure6AverageDailyVolumeProfileStandardRegularLetters(nonECR)
FY2010 Average Daily Volume by Day Standard Regular Mail Letters (Non-ECR)
300

250 Volume (millions of pieces per day)

200 November

150 Overall

100 July

50

0 Mon Tue Wed Thu Fri Sat Day of the Week Peak Month (Nov) Low Month (July) Overall Average

11

WorkhoursbyDayofWeek WorkhoursrecordedinthePostalServicesTimeandAttendanceControl System(TACS)canindicatetheextenttowhichpeaksoflaborutilization correspondtovolumepeaks.Mailprocessingclerkandmailhandlerworkhours (Figures78)alsoexhibitsomepeakingassociatedwithMondayvolumes.While Mondayworkhoursarenotespeciallyhighforeitherclerksormailhandlers,6we expectprocessingofmailscheduledforMondaydeliverycausesmostSunday workhours. Figure7AverageDailyWorkhoursMailProcessingClerks


FY2010 TACS Average Daily Total Hours Mail Processing Clerks
1,200 December 1,000 Average Daily Workhours (thousands)

800 Overall August 600

400

200

0 Sun Mon Tue Wed Day of the Week Peak Month (Dec) Low Month (Aug) Overall Average Thu Fri Sat

Citycarrierworkhoursshowrelativelylowvariationovertheweek.Monday workhoursaverage5.5percenthigherthantheaveragefortheotherfivedelivery days.SeeFigure9.

6Thedaysshownarecalendardays,soMondayworkhoursincludelaborusedto

processincomingmailforMondaydeliveryandforprocessingoutgoingmail collectedorotherwiseenteredonMondayfordeliveryTuesdayorlater. 12

Figure8AverageDailyWorkhoursMailHandlers
FY2010 TACS Average Daily Total Hours Mail Handlers
350 November 300 Average Daily Workhours (thousands)

250 August 200

Overall

150

100

50

0 Sun Mon Tue Wed Day of the Week Peak Month (Nov) Low Month (Aug) Overall Average Thu Fri Sat

Figure9AverageDailyWorkhoursCityDeliveryCarriers
FY2010 TACS Average Daily Total Hours City Delivery Carriers
1,600 December 1,400 Average Daily Workhours (thousands)

1,200 August

Overall

1,000

800

600

400

200

0 Sun Mon Tue Wed Day of the Week Peak Month (Dec) Low Month (Aug) Overall Average Thu Fri Sat

MailprocessingovertimehoursshowDecemberandMondaypeaks.Since overtimeisearnedforhoursexceeding8perdayor40perweek,Fridayand Saturdayshowrelativelyhighovertimeusage,asovertimeearlyintheweekleadsto echoesofovertimehoursduringendofweekshifts.SeeFigures1011.Midweek 13

offpeakdaysshowsomeovertimeusage.Thismaybedueinparttoparticular facilitiesencounteringhighvolumesoffofsystempeaks,butalsolikelyindicates someamountsofovertimeusageforroutinemanagementneeds(e.g.unscheduled leavereplacements)ratherthanmeetingvolumepeaks. Citycarrierovertime(Figure12)showsaparticularlyhighmidweek overtimefloor.Thismay,inpart,reflectadditionalinflexibilitiesrelatedtocarrier routestructureaswellastheuniversalservicerequirementtoserviceroutesin theirentirety.Nevertheless,theMondayovertimepeakforcitycarrierssuggests thatdeliveryworkloadlevelingcouldreducecarrierovertime. Figure10AverageDailyOvertimeHoursMailProcessingClerks
FY2010 TACS Average Daily Overtime Hours Mail Processing Clerks
160 December 140 Average Daily OT Workhours (thousands)

120

100

80 Overall 60

40 June 20

0 Sun Mon Tue Wed Day of the Week Peak Month (Dec) Low Month (June) Overall Average Thu Fri Sat

14

Figure11AverageDailyOvertimeHoursMail Handlers
FY2010 TACS Average Daily Overtime Hours Mail Handlers
60 December Average Daily OT Workhours (thousands) 50

40

30

Overall

20

10 June

0 Sun Mon Tue Wed Day of the Week Peak Month (Dec) Low Month (June) Overall Average Thu Fri Sat

Figure12AverageDailyOvertimeHoursCityDelivery Carriers
FY2010 TACS Average Daily Overtime Hours City Delivery Carriers
200
December

Average Daily OT Workhours (thousands)

180 160 140 120 100 80 60 40 20 0 Sun Mon Tue Wed Day of the Week Peak Month (Dec) Low Month (May) Overall Average Thu Fri Sat
May Overall

15

3.AnalysisofServiceRelatedCosts PremiumPay ThePostalServicepaysshiftdifferentialstoclerks,mailhandlers,andcity carriersforworkatnight(6P.M.6A.M.)andonSundays.7InFY2010,totalcostsfor nightdifferentialandSundaypremiumforclerks,mailhandlers,andcitycarriers were$451million.Sincethereislittledeliveryorretailactivityduringtimeswhen nightdifferentialandSundaypremiumpaymentsarerequired,thevastbulkofthe associatedcost,$441million,istreatedasbeingincurredinmailprocessing activities.TheSundaypremiumrateishigherthanthenightshiftdifferential,so Sundaypremiumaccountsfor41percentofmailprocessingpremiumpaycosts, thoughnightshifthoursaccountforover80percentofthepremiumworkhours. PremiumPayforMailProcessing TheCostandRevenueAnalysis(CRA)methodologytreatsmostclerkand mailhandlerpremiumpaycostsasbeingcausedbymailproductswithpreferential servicestandards.Thekeyassumptionisthatmailprocessinglaborcouldbe scheduledondayshiftsintheabsenceofthepreferentialproductsservice standards.8ThemajorexceptionispremiumpayexpensesincurredatNetwork DistributionCenters(NDCs),whichprimarilyprocessnonpreferentialmail. IntheCRAproductionprocess,premiumpaycostsareinitiallydistributedto bothpreferentialandnonpreferentialproducts,sincetheweightsforInOfficeCost System(IOCS)talliesusedtoassignmailprocessingcoststoproductsarenot higherinperiodswherepremiumpayexpensesareincurred.Asubsequent adjustmentsubtractspremiumpaycostsatmailprocessingplantsexceptNDCs9 andpostofficesfromallproductsandreallocatesthemtopreferentialproductsin proportiontoIOCStalliesfortheproductsrecordedduringperiodswhenthe
7ThePostalServicealsoincurssmallamountsofotherpremiumpaycosts,

includingpremiumpayforChristmasDaywork.Premiumpaycostsforrural carriersarenegligible.
8Foradditionaldiscussion,seePostalRateCommissionDocketNo.R841,Rebuttal

TestimonyofJohnC.PanzaronbehalfofAmericanNewspaperPublishers Association.

9SinceNDCsprimarilyhandlenonpreferentialmail,premiumpaycostsatNDCsare

assumedtobeunrelatedtopreferentialmailserviceprovision.Theincurrenceof premiumpaycostsatNDCsisassumedtobelargelyinserviceofmaintaining efficientutilizationofNDCequipment. 16

premiumpayisincurred.ThenonNDCpremiumpayadjustmentreassigned$389 millioninFY2010,4.1percentofmailprocessingcostsexcludingpremiumpay.The differencereflectssomepremiumpaycostsinplatformactivitiesassignedtonon preferentialmail,aswellaspremiumpayexpensesrelatedtoExpressMailand InternationalMail.Table2,below,showstheassignmentofcoststothepreferential productsunderconsiderationinthisreport.Amajorityofthecost,$204million,is assignedtoFirstClassletters. Table2.MailProcessingPremiumPayCostsbyProduct,FY2010


PremiumPayCost (millionsofFY2010 dollars) x x x x x x 389

Product FirstClassSinglePieceLetters FirstClassPresortLetters FirstClassFlats FirstClassParcels PriorityMail Periodicals Total[1]

Source:DocketNo.ACR2010,USPSFY107Part5; USPSFY10NP18Part5 [1]Includesproductsnotlistedabove

PiggybackedCosts Piggybackedmailprocessingcostsincludemailprocessingrelatedcosts suchassupervision,mailprocessingequipmentcapitalandmaintenancecosts,and costsofmailprocessingfacilityspace;theserepresentalargecomponentoftotal mailprocessingcosts.Asamechanicalmatter,piggybackcostsareassignedtothe premiumpayportionofmailprocessinglaborcosts,andpremiumpaycosts includingpiggybackedcoststotaled$587millionforthepreferentialproductsin Table2.Whileitmaybereasonabletoassignalargershareofmailprocessingnon laborcoststopreferentialproductsbypiggybackingpremiumpaycosts,ingeneral thesizeoftheclerkandmailhandlershiftdifferentialswillnotservebroadlyasa logicalbasisforpreferentialproductsnonlaborcostpremiums(ifany).Some piggybackedcosts,suchasmailprocessingsupervision,alsoincurshiftdifferentials thatwouldbeavoidableifthesupervisedmailprocessinglaboroccurredduringday shifts.However,someotherpiggybackedcostcomponents,suchasequipment 17

maintenanceandcustodiallabor,wouldtendtoincurhighershiftdifferentialcosts astheycurrentlytendtobecarriedoutduringdaytimeperiodsofrelativelylow mailprocessingactivity. WorkloadShift AvoidingpremiumpaycostsrequiresnightandSundayworktobescheduled attimeswherepremiumpayisnotrequiredi.e.,duringdayshiftsandawayfrom Sundays.Currently,theTour2shift(approximately7A.M.3P.M.)isalowactivity periodinmailprocessingplants.Atleastintheory,aonedayextensionof preferentialmailservicestandardscouldpermitmailprocessingtobescheduled largelyduringdayshifts.Thatis,theextensionwouldallowmailprocessing currentlyperformedbetween6P.M.and6A.M.,largelytoprovideserviceforlocal FirstClassMailduetobedeliveredthenextday,tobedeferredfor12hourstothe subsequent6A.M.6P.M.dayshift.Notably,overnightFirstClassMailwould converttotwodaymailundersuchaprocessingschedule;asdiscussedbelow,the effectoncurrent23daymailwoulddependontheresultingnetworkalignment. Shortofawholesalereorganizationofmailprocessingintodayshifts,there maybecostsavingopportunitiesfromreducingSundayworkload,especiallyfor nonpreferentialmail,sincetheSundaypremiumisrelativelylarge.IOCSdatashow thatalargeshareofSundaydirecttalliesatplants(tallieswithanassociated product)isfornonpreferentialproductsincludingStandardMailandPackage Services(41percentinFY2010).Rescheduling41percentofFY2010Sundayplant workintoSaturdayorotherdayshiftperiodswouldreduceSundaypremiumcosts by$65millionmostofwhichisbornebypreferentialproducts. ImpactonTransportation Inadditiontodirectsavingsofshiftdifferentialcosts,dayshiftprocessingas describedabovecouldextendthewindowsfortransportingcollectionmailto outgoingprocessingsitesandfortransportingprocessedincomingmailtopost officesfordelivery.Currently,narrowlocaltransportationwindowsresultingfrom 1dayservicestandardslimitthePostalServicesabilitytoconsolidatecertainmail processingactivities,includinginitialoutgoingprocessingandfinalizationofmailin automateddeliverypointsequencingoperations.ThePostalServicetakeslimited advantageoflongerwindowsundercurrentstandards,suchasforSaturday collectionmail,byconsolidatingSaturdayoutgoingprocessinginanumberofareas. ProvidingservicewithinnarrowprocessingwindowsmayalsolimitthePostal Servicesabilitytoalignstaffinglevelswithdaytodayvariationsinworkloads. Longerprocessingandtransportationwindowsduetoextendedendtoend servicestandardsmayallowplantstohavegeographicallymuchlargerservice territories,possiblyforbothoutgoingandincomingprocessing,whilemeeting dispatchrequirementsfortwoandthreedaymail(currentlyoneortwodaymail). 18

Withlargerplantserviceterritories,somecurrenttwodaymail(e.g.mailfor neighboringSCFsunderthecurrentnetwork)couldremaintwodaymailwith plantsexpandedserviceterritoriesafternetworkrealignment,mitigatingthe effectsoftheservicestandardchange.10Themagnitudeofpotentialcostsavings woulddependontheextentoffacilityconsolidation.ArecentOIGstudy(RARCWP 11006)concludedthatanew135plantnetworkwouldcostaround$2billionless peryeartooperatethanthecurrentnetwork,excludingtransitioncosts.Thisnew plantstructuremaybemorefeasiblewithaonedayrelaxationofservicestandards. MailProcessingCostDifferentialsforPreferentialandNonPreferentialProducts InadditiontousinghighercostresourcessuchasnightshiftandSunday labor,thePostalServicemightuseadditionallabortomeetservicestandardsfor preferentialmail.Forexample,alongstandingclaimofPeriodicalsmailersin regulatoryproceedingsisthatthePostalServiceinefficientlyprovidesmanual handlingofPeriodicals,ostensiblyforservicereasons.11SincePostalServicemail processingcostsarebasedinpartonIOCSsampledatawhichmeasuretheamounts oflabortimeassociatedwithhandlingvariousproducts,anyadditionalworkeffort providedforservicepurposesshouldbereflectedinmeasuredproductcosts. However,theservicerelatedportionofworkeffortisnotdirectlyobservablein IOCS. Inanumberofcases,themailprocessingcostsforpreferentialproductsare substantiallyhigherthancorrespondingnonpreferentialproducts(e.g.Standard Mail).However,themeasuredcostdifferencesbetweenpreferentialandnon preferentialproductsmayreflectnotonlytheeffectsofendtoendservice standardsbutalsodifferencesinmailmixwithinproductse.g.,presortprofile anddifferencesinbundledancillaryservicessuchasforwardingandreturnto senderprocessingforundeliverableasaddressed(UAA)mail.Mailprocessingunit costsusedtodevelopworksharingdiscountsalsoreflecttheCRAredistributionof premiumpaycosts.IdentifyingappropriatebenchmarkcostsforFirstClassSingle PiecemailisparticularlychallengingsincethepopulationofSinglePiecemailvaries widelyincostcausingcharacteristicsincludingautomationcompatibilityand implicitpresortand/ordestinationentrycharacteristics(fromlocalmail).
10Inthismodel,mailthatcanbetransportedbetweenplantswithin24hourswould

havethreedayservice;thismayincludesomemailcurrentlysubjecttothreeday standards.
11PostalRegulatoryCommissionDocketNo.ACR2010,InitialCommentsofTime,

Inc.,pp.79.WeunderstandthatthePostalServicehasrespondedbyeliminating Hot2Cprocessing. 19

Toisolatecostsassociatedwithendtoendservicestandardsfromother sourcesofproductcostdifferences,wecontrolforthreemajorsourcesofcost differencesnotrelatedtoreallaborinput:(1)differencesinpresortlevel,(2) provisionofservicesforUAAmail,and(3)premiumpaycostallocations.Wecontrol forpresortdifferencesbycalculatingcostsusingthepreferentialproductsmailmix combinedwiththeclosestcorrespondingStandardMailcosts.ForSinglePiecemail, weusecostsforthemixedAADCpresorttierasthebenchmark(mixedADCfor flats).12TheresultisahypotheticalcostbasedonStandardMailproduct characteristics,butusingthepreferentialproductsvolumelevelandpresortmix. TocontrolforcostdifferencesduetobundledUAAservices,weuseIOCSsample datatoidentifytheportionsofproductcostsassociatedwithreturnedand forwardedpiecesandotherhandlingofpiecesidentifiedbydatacollectorsasUAA. WesubtractthedifferenceinUAAunitcostsbetweenthepreferentialproductand thecorrespondingStandardMailproducttoremovethecostofthepreferential productsbundledUAAservicesfromthecostdifferenceafterpresortlevel adjustment.Finally,wesubtractthepremiumpaycostdistributedtothe preferentialproduct.Weconsideranyremainingcostdifferenceaspotentially servicestandardrelated. TheresultsoftheanalysisarepresentedinTable3,below.Theanalysis suggeststhatpotentialservicerelatedmailprocessingcostpremiums,apartfrom premiumpaycostallocations,arelargeforFirstClassSinglePiecemailandsmall forFirstClassPresortmailandPeriodicals.Sincetheestimatesaresubjectto samplingvariabilityaswellastheinfluenceofcostcausingfactorswecannot readilyincorporateintheanalysis,wejudgethesmallpositiveandnegative differencesforFirstClassPresortmailasprovidinglittleevidenceofadditional realmailprocessinglaborinputforthepreferentialproducts.TheFY2010cost premiumforFirstClassSinglePiecelettersandflatsoverStandardMail,adjusted forpremiumpayandbundledUAAservice,is$1.13billion,or25.6percentofthe associatedproductcosts.WeadditionallycomparedFirstClassSinglePieceletter andflatmailprocessingcoststomixedAADCFirstClassPresortbenchmarks (mixedADCforflats)tohelpisolatethecostsofservicefeaturesofretailFirstClass MailfromthoseofservicefeaturesapplicabletoallFirstClassMail.Wefoundthat FirstClassSinglePieceletterandflatcostswere$973millionhigherthantheFirst ClassmixedAADCbenchmarkcost,suggestingthatthebulkofthecostdifferenceis
12Wealsoexcludecancellationandremotebarcodesystem(LDC15)costsfromthe

SinglePieceFirstClassproducts,sincethoseactivitiesareprimarilyrelatedtothe depositandentryfeaturesofFirstClassSinglePieceandarenotrequiredforbulk enteredautomationcompatiblemail.FY2010RPWdatashowthattheFirstClass SinglePieceproductis0.8percentnonmachinable,butweallowforadditional manualhandlingofSinglePiecelettersonthemarginsofmachinabilitybyassigning anonmachinablemixedAADCbenchmarkcostto2percentofletters. 20

specifictoFirstClassSinglePieceproductcharacteristics. Table3.EstimationofPotentialServiceRelatedCostDifferentialsforSelected Products,FY2010.


(CostsinThousandsofFY2010Dollars)
MPCostw/ piggybacks 2,482,214 255,057 1,231,048 3,436,232 978,649 MPCost@ Standard UnitCosts Difference 2,253,193 229,021 211,548 43,508 1,038,671 192,378 2,120,254 1,315,979 704,684 273,966 Premium PayCost 91,636 7,581 37,534 112,104 26,991 Difference afterUAA and Premium 90,044 8,390 9,890 953,582 178,107

Product FCMPresortLetters FCMPresortFlats PeriodicalsOutsideCounty FCMSinglePieceLetters[2] FCMSinglePieceFlats

UAA[1] 227,429 27,538 144,953 250,292 68,869

Notes: [1]UAA=(RPWvol*(UAAunitcostStandardUAAunitcost))/100 [2]Assumes98%ofvolumeautomationand2%nonautomation,nonmachinable

Ouranalysisisnotcapableofisolatingtheportionoftheapparentcost premiumforFirstClassSinglePiecelettersandflatsoverpresortedmixedAADC (ADC)benchmarksthatisspecificallyduetoservicestandards,sothecost premiumsforFirstClassSinglePiecemailshowninTable3cannotbeassumedto beentirelyrelatedtoendtoendservicestandards.However,weconsidereda numberofpotentialconfoundingfactorsandconsideritlikelythatasubstantial portionofthecostdifferentialisservicerelated.Forinstance,mostFirstClass SinglePiecemailisautomationcompatible,andalargefractionislocalmail,which shouldreducecostdifferenceswiththemixedAADCbenchmarkvolumes.Also,the magnitudeoftheeffectistoolargetobeanartifactofknownnonsamplingerrorsin mailprocessingcosts,suchassystematicmisidentificationofproductsinIOCS. ImpactofTimeofDay ServiceperformanceforFirstClassSinglePiecemailparticularlypieces subjecttothe1daystandardis,ofcourse,amajorservicefocusofthePostal Service.Auxiliaryevidencepointstoatleastpartofthecostdifferenceasbeingdue tohighstaffinglevelsinoutgoing(Tour3)mailprocessingoperationsaimedat meetingtheservicestandardfor1dayFirstClassSinglePiecemail.MODS productivitiestendtoberelativelylowinsomeoperationsthatprimarilyserve collectionmail,includingISSoperations(imageliftforbarcodeapplication)and outgoingmanualprocessing.Intradayprofilesofmailprocessingcostssuggestthat Tour3costsarehighrelativetoTour1,thoughincomingdistributionworkloadsare muchlarger.Weobserveacostpeakformailhandlersandclerksbetween10P.M. andmidnight.SeeFigures1314,muchofwhichappearstoinvolveworkeffortto clearoutgoingmail.Figure15,whichshowscostsbyproductandtimeofdayfor lettersandincludesmailprocessingatpostoffices,showsthatFirstClassSingle Piecelettershavecostpeakscoincidingwithlateeveningdispatchtimesandin 21

morninghourswhenDPSrejectsandresidualmanualmailaresortedtocarrier routes.TheabsenceofanetcostdifferentialbetweenFirstClassPresortmailand StandardMailsuggeststhatthelateeveningpeakdoesnotrepresentextralabor usedtoinductincomingFirstClassPresortmailvolumesfornextdaydelivery. Figure13IOCSCostsbyTimeofDay,ClerksatMailProcessingPlants


FY2010 IOCS Tally Dollar Weights ($000) by Hour of the Day Plants - Mail Processing Clerks
450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 1:00:00 AM 2:00:00 AM 3:00:00 AM 4:00:00 AM 5:00:00 AM 6:00:00 AM 7:00:00 AM 8:00:00 AM 9:00:00 AM 1:00:00 PM 2:00:00 PM 3:00:00 PM 4:00:00 PM 5:00:00 PM 6:00:00 PM 7:00:00 PM 8:00:00 PM 11:00:00 PM 12:00:00 AM 10:00:00 AM 11:00:00 AM 12:00:00 PM 9:00:00 PM 10:00:00 PM

Tour 1 (11:00 P.M. - 7:00 A.M.)

Tour 2 (7:00 A.M. - 3:00 P.M.)

Tour 3 (3:00 P.M. - 11:00 P.M.)

IOCS Tally Dollar Weights ($000)

Hour of the Day

Figure14IOCSCostsbyTimeofDay,MailHandlersatMailProcessingPlants
FY2010 IOCS Tally Dollar Weights ($000) by Hour of the Day Plants - Mail Handlers
250,000

Tour 1 (11:00 P.M. - 7:00 A.M.)


200,000 IOCS Tally Dollar Weights ($000)

Tour 2 (7:00 A.M. - 3:00 P.M.)

Tour 3 (3:00 P.M. - 11:00 P.M.)

150,000

100,000

50,000

0 1:00:00 AM 2:00:00 AM 3:00:00 AM 4:00:00 AM 5:00:00 AM 6:00:00 AM 7:00:00 AM 8:00:00 AM 9:00:00 AM 1:00:00 PM 2:00:00 PM 3:00:00 PM 4:00:00 PM 5:00:00 PM 6:00:00 PM 7:00:00 PM 8:00:00 PM 11:00:00 PM 12:00:00 AM 10:00:00 AM 11:00:00 AM 12:00:00 PM 9:00:00 PM 10:00:00 PM

Hour of the Day

22

Figure15IOCSMailProcessingCostsbyTimeofDayandProduct
FY2010 IOCS Mail Processing Direct Tallies Letter Mail Only
100,000 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 1:00 AM 2:00 AM 3:00 AM 4:00 AM 5:00 AM 6:00 AM 7:00 AM 8:00 AM 9:00 AM 1:00 PM 2:00 PM 3:00 PM 4:00 PM 5:00 PM 6:00 PM 7:00 PM 8:00 PM 11:00 PM 12:00 AM 10:00 AM 11:00 AM 12:00 PM 9:00 PM 10:00 PM

Tour 1 (11:00 P.M. - 7:00 A.M.)

Tour 2 (7:00 A.M. - 3:00 P.M.)

Tour 3 (3:00 P.M. - 11:00 P.M.)

IOCS Mail Processing Direct Tally Dollar Weights ($000)

Hour of the Day First-Class Single Piece First-Class Presort Standard Mail Non-ECR

ExpeditedTransportation Thecostofpurchasedtransportationvarieswidelybymode.Ground transportationisusuallylesscostlythanairtransportation,butitrequireslonger endtoendtransittimesforlongerhaulshipments. InFY2010,purchasedairtransportationexpensesforthePostalService totaled$1.98billion,includingPriorityMailcostsof$xxxxxxxxxxandFirstClass Mailcostsof$515million.TheFedExdayturnnetworkaccountsfornearlyall PriorityMailairtransportationcostandamajorityofFirstClassMailcost. RelativelylittlePriorityMailiscarriedbypassengerairlinesduetoairlinesecurity regulationsbarringpackagesoveronepoundfromflyingaspassengerairliners bellycargo.FedExdayturncostsaretreatedascuberelatedandnondistance relatedinthePostalServicestransportationcostanalyses.13Highway transportationcostsaretreatedascuberelatedanddistancerelated.
13Thatis,thePostalServiceeffectivelypaysapricepercubicfootforFedExday

turntransportationthatdoesnotvarywithzoneoractuallengthofhaul.The unobservedunderlyingeconomiccoststoFedExforprovidingthedayturnservice presumablydodependondistance. 23

Basedontransittimesforcompetitorsgroundservices,itispossibletomeet endtoendservicestandardsoftwodaysuptoapproximatelyzone4,threedaysto zone5,andfourdaystozone6.Groundservicetozone7and8destinationswould requiremorethanfourdaysintransit. Providing23dayserviceforFirstClassMailandPriorityMailtosomezone 5andmostorallzone6(orhigher)destinationsrequiresairtransportation.With2 4dayservicestandards,itispossibletoservezone5and6destinationsusing groundtransportation.Somezone5PriorityMailcurrentlyhasathreeday standard.Forzone6mail,fourdayservicetypicallywouldrepresentaoneday increaseintransittimeforFirstClassMailbutmayconstituteatwodayincrease forPriorityMailcomparedtocurrentstandards. CostDifferentialsforExpeditedTransportation Computingcostdifferentialsbetweenairandgroundtransportationis complicatedbylimitationsofthePostalRegulatoryCommissionscurrently acceptedvolumevariablecostanalysisforhighwaytransportation.Afeatureof sometransportationcostsisthatthecostpertonmileorcubicfootmiledecreases withthelengthofhaul;thisiscalledadistancetaper.Largerdistancetapereffects willtendtoreducetransportationcosts(perpoundorpercubicfoot)forhighzone shipmentsandraisethemforlowzoneshipments,relativetoacasewherethereis nodistancetaper.Asapracticalmatter,alargerdistancetaperforhighway transportationtendstoincreasethecostdifferentialbetweenairandgroundmodes inmiddlezones.Unfortunately,theexistinghighwaytransportationmodelsdonot provideempiricalestimatesofdistancetapereffectsforPostalServicehighway transportation.Asaresult,weprovidearangeofairgroundcostdifferentialsbased onthreesizesofadistancetapereffect. Weobtainedaircubicfeet,aircosts,anddistancerelatedgroundcostsfor PriorityMailfromcostdatafiledwiththePostalServices2010AnnualCompliance Report.14Thedistancerelatedgroundcostspercubicfootassumeaconstantcost percubicfootmile(i.e.nodistancetaper).Wealsocomputedthegroundcostsper cubicfootimpliedbytwosizesofdistancetapereffects,correspondingtolengthof haulelasticitiesof0.2and0.3.15Theseareintendedtorepresentmodestdistance tapereffects.

14PostalRegulatoryCommissionDocketNo.ACR2010,folderUSPSFY10NP27. 15Specifically,thesearetheelasticitiesofcostpercubicfootmilewithrespectto

lengthofhaul.Ineachscenario,theaveragecostpercubicfootmileisthesame. 24

PostalServicedataindicatethatxxxxxxxxxxxxwasspentinFY2010to provideairtransportationforzone36PriorityMail.16Thisincludesxxxxxxxxxxx forzone35mail,forwhich23daygroundtransittimesmaybefeasible.The estimatedsavingsfromsubstitutinghighwaytransportationforairtransportation rangefrom$120millioninthenodistancetapercaseto$197millioninthe scenariowithalengthofhaulelasticityof0.3.Thepotentialzone35savings, whichentailrelativelylittlechangeinexistingservicestandards,rangefromxxx xxxxxxinthenodistancetaperscenariotoxxxxxxxxxxxinthe0.3elasticity scenario.Theremainingsavingsresultfromextendingzone6servicestandardsto fourdays.SeeTable4,below. Table4.AirHighwayCostDifferences,Zone36PriorityMail,FY2010Dollars
GroundSavings,DistanceTaperScenarios NoDistance Taper Elasticity0.2 Elasticity0.3 x x x x x x $120,495,788 $164,282,528 $196,747,462

Zone AirCubicFt. 35 x 6 x 36 x

AirCost x x x

SomesimilarsavingsmaybeavailableforFirstClassMail,thoughdataare notavailabletoindicatetheextentofmiddlezoneairtransportationcosts.We understandthatthemiddlezonetransportationmixforFirstClassMailhashigher utilizationofhighwaytransportation,sotherelativemagnitudeofhighway transportationsavingsrelativetoairislikelytobeconsiderablysmallerthanfor PriorityMail. Withintheframeworkofaonedayextensionofendtoendservice standards,savingsfromsubstitutinggroundtransportationandfromdeferringmail processingwillnotbeavailableatthesametime.Sincelettermail,andtoalesser extentflatshapemail,isnottransportationcostintensive,theresultsfromsection 3,above,suggestthattransportationcostreductionsarelikelytobelessfruitful thansavingsfromdeferringand/orconsolidatingmailprocessing.Sinceparcels (includingPriorityMail)areprocessedindistinctmailstreamsfromlettersand flats,itmaybepossibletoobtaintransportationcostsavingsselectivelyforparcel shapeproducts.

16Weexcludesomecostforlowerzoneairtransportation,whichweunderstand

includesAlaskaandHawaiimailwhichreceivesairtransportationinpartdueto thosestatesuniquegeographiesandotherpolicyconsiderations. 25

Overtime ThePostalServicemayprovidestaffingforpeaksofnondeferrableworkload inpartbyextendingemployeesshifts;thismaytriggerovertimewagepremiums. Overtimecostsmaybeviewedasaformofservicerelatedcostsfornondeferrable workload,sincetherewouldnormallybelittlejustificationforincurringovertime costsfordeferrableworkload.Incontrasttothetreatmentofpremiumpaycostsfor shiftdifferentials,overtimecostsarenottreatedasacostofpreferentialmail serviceinthemailprocessingordeliverycostmethodologies. OvertimeCosts ThePostalServicepaysclerks,mailhandlers,andcitycarriersovertimeat 1.5timesemployeesbasehourlyratesforhours810ofaworkdayorhours exceeding40inaworkweek.Penaltyovertimeattwicethebasehourlyrateispaid forhoursexceeding10perregularlyscheduledworkdayormeetingother contractuallyspecifiedcriteria,butlaborcontracttermsspecifythatpenalty overtimepaymentsarenotrequiredinDecember.Notethatovertimepaidtofull timeemployeesforworkinglongdaysearlyintheworkweekforinstance,on Mondaysorhighervolumedaysafterholidayswilltendtotriggerovertime paymentslaterintheweekasportionsofendofweekshiftsexceed40hours.The secondaryendofweekovertimepeaksareclearlyvisibleinFigures1012,above. AsshowninTable5,wagecostsofovertimepay(thepremiumoverthe straighttimewage)forclerks,mailhandlers,andcitycarrierswere$811millionin FY2010.Citydeliverycarriersaccountforamajorityofovertimepayexpenses,and carrierovertimeoccursatanaboveaveragerate. Table5.OvertimePremiumPay,SelectedCrafts,FY2010

Craft Clerks MailHandlers CityCarriers Total

OvertimeHours 17,946,805 8,219,239 36,713,489 62,879,533

OvertimePremium PayCost 232,967,141 99,385,304 479,082,653 811,435,098

OT%of Workhours 6.40% 9.30% 9.70% 8.40%

Source:FY2010NationalPayrollHoursSummaryReport

Therawovertimewagepremiumsmayoverstatetheactualcostof employingovertimehours,dependingonthetypeofstraighttimelaborthatcould beemployedinthealternative.Theeconomiccostofovertimedependsinparton thenetdifferenceinbenefitscostsbetweenovertimeandstraighttimehours.For instance,overtimeusedoesnotincreasethePostalServicescontributionsto employeehealthcareandretirementbenefits.Thealternativestoovertimeusage 26

maybehiringmoreemployeesearningstraighttimewagesplusbenefits,or incurringwageguaranteesforemployeescalledtoworkoutsidetheirnormalwork schedules.Iffringebenefitcosts(apartfromsocialinsurancecontributions)are allocatedtostraighttimeworkhours,thenthenetcostofovertimeoverstraight timeworkhourscanbeaslowas$12perhour.Sinceovertimeisflexible,itcanbe cheaperthanincurringworkguaranteesfornonscheduledemployees.17 OtherCausesofOvertime Someportionsofovertimemaynotbecausedbymailproductsendtoend servicestandards.Aparticularchallengeisidentifyingthecauseofovertimeusage attroughperiods.Figures1012showthatinthemidweek(TuesdayWednesday) trough,thePostalServicerecordsapproximately120,000hoursofdailyovertime formailprocessingclerks,mailhandlers,andcitycarriers.Annualized,thisimpliesa baseof36millionovertimehours.Eventhoughthebaseovertimehaslittleclear connectiontosystemwidevolumepeaks,localizedvolumepeaksthatarenot evidentfromnationaldataneverthelessmaycausesomeofit.However,other routinemanagementissuessuchasprovidingunscheduledleavereplacementsmay alsodriveaportionofovertimeuse.Particularlyforcitycarriers,someovertime incurrencemaybeduetouniversalserviceconsiderationscitycarrierscannot effectivelycurtailportionsoftheirroutesaftertheyhaveleftforthestreet.Existing datacannotidentifytherelativecontributionsofthesecauses. Inall,asubstantialportionofovertime,slightlylessthan27million workhours,exceedsthebaselevelobservedattroughperiodsandmaybereduced byshavingsystemwideworkloadpeaks.Anindeterminateportionofthebase overtimemayalsobeduetolocalizedpeaks.Limitedadditionaldeferralofendto endservicestandardshouldenablesomeavoidanceofovertimeworkhours.Indeed, sincethepeakforlargelynonpreferentialflatvolumecoincideswiththeMonday peakforpreferentialletters,theremaybesomeasyetuncapturedopportunitiesfor peakshavingwithinexistingservicestandards. ThePostalServicehassharplyreducedovertimeusageratessincetheonset oftherecession,from12.9percentofcitycarrier,clerk,andmailhandlerhoursin FY2007to8.4percentinFY2010.Continuedpreferentialvolumedeclinesshould reducetheneedtoincurovertime.Additionally,newlaborcontracttermsproviding additionalstaffingflexibilitymaycontributetofutureovertimereductions.

17Laborcontractprovisionscallforminimumguaranteedworkhours(typically4or

8hours),orpayinlieuofwork,foremployeesworkingoutsidetheirnormal schedules;thesedonotapplytocaseswheretheemployeecontinuesworkingintoa regularshift.See,e.g.,NALCUSPSNationalAgreement,20062011,Article8.8. 27

4.Summary ThispaperdescribescomponentsofservicerelatedcostsintheU.S.Postal Service,definedascoststhathypotheticallycouldbeavoidableifexistingendto endservicestandardswereextendedbyonedeliveryday.18Comprehensive measuresofservicerelatedcostsarenotpossiblegivenexistingPostalServicedata andcostingmethods,whicharegearedtowardmeasuringtotalcostofproductsand notthecostimplicationsofproductsservicefeatures.WhilethePostal AccountabilityandEnhancementActchargesthePostalRegulatoryCommission withmonitoringserviceperformanceforPostalServiceproducts,thoseeffortsdo notencompassmeasurementofthecostsofserviceperformance.Nevertheless,we canidentifyapproximately$2.5billioninannualcosts(basedonFY2010cost levels)inmailprocessing,transportation,anddeliveryfunctionsthatareatleast partlyservicerelated.Table6,below,summarizesthesecosts. Table6.SummaryofServiceRelatedCosts
CostsinMillionsofFY2010Dollars
MailProcessing PremiumPay (Plants) Product FirstClassSinglePieceLetters x FirstClassPresortLetters x FirstClassFlats x FirstClassParcels x PriorityMail x Periodicals x Total[4] 389 OtherMail Processing Costs 954 [1] 186 [2] [2] [1] 1,139 Overtimefor Clerks,Mail Zone36Air Handlers,and Transportation CityCarriers [2] [2] [2] [2] [2] [2] [2] [2] 120197[3] [2] [1] [2] 120197[3] 811

Notes [1]Potentialservicerelatedcostsdonotappeartobestatisticallyand/oreconomicallysignificant [2]Potentialservicerelatedcostsarenotmeasurableusingexistingdata [3]Servicerelatedcostsdependon"distancetaper"forhighwaytransportation;costscannotbe savedincombinationwithplantpremiumpay [4]Mayincludeotherproductsnotlistedaboveandcostswithnonservicerelatedcauses

Viewedaspotentialcostsavingsfromhypotheticalservicereductions,the costsinTable6cannotnecessarilybesummed.Addingenroutetransittimeby shiftingmailfromairtogroundtransportationmayforecloseprocessingordelivery costreductionsthatwouldinvolvestorageofmailinthesametimewindows.Also,


18Wearenotadvocatingforanychangeinservicestandards.

28

theorganizationalchangesrequiredtoobtaincostsavingsincertainservicerelated componentsmaybedramatic.AvoidingpremiumpayfornightandSundayshifts, forinstance,wouldrequirereorganizingthevastbulkofcurrentmailprocessing activitiesintodayshiftwork.Thedirectlaborcostsavingsfromdoingsomaybe modest,thoughfacilityconsolidationopportunitiesfromextendingprocessing and/ortransportationwindowscouldleadtolargeindirectcostreductions,insofar asthePostalServiceincurssome$4.7billioninannualcostsrelatedtomail processingfacilityspaceandequipment. Finally,demandsideeffectsofservicereductionswouldtendtopartlyoffset costsavings.Aonedayextensionofservicestandardswouldnotaffectall preferentialproductsidentically.Curtailingonedayservicewouldhavethelargest effectsonFirstClassSinglePieceletters,wherehalfofthevolumeissubjecttothe onedaystandard.However,sinceserviceisnotexplicitlypricedandretailmailers havelimitedchoiceoverpostalservicecharacteristics,thesensitivityofpostal demandtoserviceperformanceisnoteasilymeasured.InsofarasStandardMail volumesexceedFirstClassMailvolumes,andsomeexistingFirstClassMailis preventedfrommigratingtoStandardMailduetocontentorotherrestrictionson StandardMailmatter,webelievethatthereissomepossiblyconsiderablelatent demandforserviceatlowerlevelsthancurrentpreferentialmailstandards.

29

Você também pode gostar