,.-. BEFORE THE POSTAL RATE COMMISSION WASHINGTON, D.C. 20268-0001 POSTAL RATE AND FEE CHANGES, 1997 : DOCKE:T NO. F!97-1 ,,-. DIRECT TESTIMONY OF MICHFLEL A. NELSON ON BEHALF OF THE UNITED STATES POSTAL SERVICE .CONTENTS AUTOBIOGRAPHICAL SKETCH....................................1 PURPOSE AND SCOPE OF THIS TESTIMONY ..................... I. - Volume ....... ...3 Labor Distribution Code (LDC) 24 overlap.. ............ Express Mail collection box load time .............. Letter route driving time .......................... Special purpose routes ............................. Special delivery messengers ........................ ...5 ...5 ...6 ...7 Analytical A. B. C. D. E. II. Other A. B. C. D. E. ...2 Refinements Variable Costs Work .......................................... Express Mail interfacility service .................. Express Mail collection box access and fixed stop Allocation of Cost Segment 9 fixed costs.. ............ Expr,ess Mail rate category cost differentials........1 Cost basis for pickup fees...........................1 . ...8 ..a time8 . 0 0 ,,-,.. USPS-19A - Spreadsheet USFS-19B - Data USPS-19C - Econometric USEIS-19D - Express USFS-19E - Cost -- Refinements Collection And Analysis Procedures Analyses Mail Basis Rate For Category Pickup Fees Cost Differentials ,.r--- AUTOBIOGRAPHICAL 1 My name its Michael 2 ,,r-. 3 tra.nsportation 4 Vermont 5 1984, 6 Assiociates, 7 Mas:sachusetts. A. Research an economic 9 assiignments directed and 10 transportation. 11 applying 12 microeconomics, 13 specialized 14 Service, 15 Doc!ket 16 principal 17 carrier 18 Kolbe 19 regarding 20 railroad 21 Commission, 22 (Finance Docket 23 Industries of 24 (Finance 25 consolidation 26 No. 27 Doc:ket or research investigator for R84-1. No. Nos. of the 30800), of the 32133), portions Do&et No. control of 31522 Southern acquisition and the field to of served of provided issues 31505, respectively), extensive 1 A. Lawrence testimony in six Commerce Pacific by Rio Soo Line of in as Interstate C&NW by Union with Parcel delivery CM&W and Pacific solve United city acquisition and and Commission testimony the of developing of statistical before consulting general and also River research, R90-1, I have and/or including the this the proceedings merger Boston, numerous studies underlying competitive 32000), the February in On behalf and to econometrics before R87-1 Stamford,, Charles on operations and/or in at involves testimony No. in problems. F!Jl86-2B, Docket in typically statistics time firm projects analytical in Associate based I provided Prior participated methodologies street offices consulting My work No.'s with Massachusetts. I was ;a Senior I have I am an independent analyst systems and Wellesley, 8 Nelson. SKETCH Grande railroads DRGW (Finance MKT by Union Pacific testimony regarding the Docket (Finance the 1 anticompetitive effects 2 Pacific and 3 assisted in 4 statements 5 bodies, 6 transportation Santa the 8 Institute 9 master's the (Finance proposed of before No. other my bachelor's of verified and reports (Transportation 11 School 12 operations 13 public of degree Technology degrees in from one from In in and one with Management, sector 1977. MIT, Systems) and legal articles; in Massachusetts I received Civil two Engineering from the Alfred concentrations in systems P. Sloan economics, analysis and management. --. PURPOSE AND !XOPE OF THIS 14 testimony costs the 1978, tran~sportation research, This I have regulatory technical Southern journals. 10 15 of 30400). numerous various many merger Docket preparation authored I received 7 Fe presented and of presents the Cost Segments 6 (City 7 (City Delivery 16 the in 17 Activity), 18 (Special 19 the 20 data 21 activities, 22 driving time 23 include improvements 24 compute volume 25 other 26 that have 27 (see Exhibit Delivery results of collections of related special on motorized variable information. been This introduced the purpose as well testimony in the USPS-19A). -- These of also cost methods as the documents segment Office and 9 presents and route activities routes. accuracy costs, it refinements messenger of Activity) particular, special letter portions Carriers, Street analytical to of Delivery In delivery in development Carriers, Messengers) a series TESTIMONY new carrier and refinements used to development of changes spreadsheets -. .T-- 1 I. Analytical .2 Specific 3 implemented 4 described 5 using analytical 6 and messenger in - Volume Variable refinements the computation These below. data derived that of four have volume refinements from Costs been variab:Le have been new field costs are implemented surveys of carrier activities: 7 - Motorized a - Special 9 - Expedited Letter Route Purpose Survey Route Mail Survey Survey - LDC 24 Survey 10 ,/-- Refinemen,ts 11 These surveys 12 A. Labor 13 In general, 14 charged to 15 messenger 16 costs. 17 the activities 18 that perform 19 and/or 20 In 21 carrier 22 with 23 Express 24 of 25 purpose are described Distribution Code special (delivery which LDC 24, craft employees Significant of of activity in for route to LDC 24 craft activ:ity Cost for (and Cost Express Segment attributing carriers Comnonents CS 9) 6.l the street associated with USPS Devew (Fiscal Year 3 :routes Mail observations can have been has served costs of Express of 1996), CS 9 from delivery 98 that This of to Express special Mail However, result assigned reflecting Type 9. a portion of and is Segment employees Mail on :Route ,-. and 24 overlap distribution involving Mail ts account Exprsess time USPS-19B messenger only IOCS tallies past, (LDC) in interfacility street "proxy" carrisr Exhibit accrues charges delivery the in items. of be associated attributed as a type ,special Mail Costs by USPS-LR-H-1. to 1 distribution 2 activities 3 also 4 largely 5 situation 6 subject 7 caused 8 Segment and are appear has carrier to 6 and 11 been observed directly 12 24 Survey, 13 performed 14 special purpose 15 carrier street 16 the 17 respectively) 18 data 19 parameters, 20 is Cost these properly 23 they are in this matched apply, and with the the overlap which this the have Cost have been and Special in these pools develop other problem and to directly (in Route Segment 7. Survey, The information 1 and Workpaper procedures accrued is use variability needed costs eliminated. 2. are to which As a which has provided much of 2The Support Route Cost Survey, the information used to attribute the costs of special predates the proliferation of Express Mail purpose routes, distribution and delivery responsibilities among such carriers. 4 LDC is accrue Purpose Cost and activities observed analysis of LDC 24 have Survey that Workpaper cost to messenger to and special Mail activities keys the accrue of surveys USPS-C, approach, in both and Similarly, analyzed different Exhibit in Expedited carrier distribution shown that 9. Survey and of activities Mail Analysis LDC's Mail are result double-count, messenger the Segment time Expedited 22 (in route 9 (which The net delivery activities respectively). in therefore 9. collections, carrier from or such and Segment Mail). by Express Segment route LDC 24, Cost and be borne new data purpose With Elxpress an overlap, Cost to in distribution costs the costs to been However, charged as accrued 10 21 activities.' routinely distributed In 9 delivery of f-- - I. result, 2 previously 3 removed. B. 5 To date, 6 boxes 7 all a boxes 9 does proxy for used 4 Express has in the boxes 11 to their volume 12 particular, the 13 collection boxes 14 that 15 empty.3 16 always contain 17 during a day is for analysis procedures 20 each box type, 21 time factors 22 with sweeps to 24 In No. of with respect when collection swept, boxes to boxes be swept Mail and be almost repeatedly overflowing. operational park-and-loop and of Mail the accrued application volume-variable driving R90-l,, differences, separation development Express route Mail Express such to approach In that for and may have determine C. Letter Docket regular include and Express many pieces commonplace these 23 analysis contain avoid this boxes reveal of collection characteristics. data been allocation However, between time has collection Mail collection new survey 19 25 load and Mail Express boxes. ordinary mail, of between activity time a proportional and By comparison, To account load using differences reasonably unneeded Express for rarely street sweeping costs and to 6 is box collection account collection it of load ordinary not Segment estimated collection Mail-related collection cost been and Express Cost Mail 10 18 /-- the collection costs of costs new for new load associated boxes.4 time the Commission driving LR-USPS-H-153. for example, 'See, 'See Exhibit USPS-19A. 5 time. adopted That a new analysis was 1 less developed 2 (a) 3 activities other 4 limitation to 5 d~istribution 6 related) 7 account 8 other it contained (noncurbline) 11 types 12 separately. 13 points 14 estimated 15 deliveries' 16 keys 17 in 18 the 19 deliveries'. the role might regarding looping; points, it reflected for did the such occurrence be desirable of parking (b) not point even given utilize points; its a underlying (weight- and similar because (c) driving it did not activities on types. new analytical driving motorized procedures between time letter routes The different are are than routine causality the 10 that the route of than looping cost for accurate no information key Under 9 and accounted for and is in identified and volume looping which formation of routine of types of on all and studied at parking occur variability is points/dismounts', driving. the role loops mail employed, points that collection-related developed the parking activities fo?, routine I have that SSee Workpaper 1.2. 6See Workpaper 1.14. 'Deviation deliveries are nonroutine those sometimes made for expedited The effect of volume on the number estimated in Exhibit USPS-19C. 8See Workpaper 1.10. 9!;ee Workpaper 1.9. of is "deviation" New distribution the weight accounted cause f,or", of mail as is deviati,on delivery stops, such items or large parcels. of such stops is as ,-. 1 D. 2 The time 3 distribution 4 (costs 5 .updating. 6 ,time 7 date. 9 ,- purpose routes proportions (by keys are used to approximately the analyze of 10 estimate 11 new econometric 12 related 13 These 14 presented drive variability in 15 E. 16 To date, aggregate 18 has 19 driving 20 time), 21 characteristics out time used", analyzing spent at of A refined is for of SPR load regarding keysI'. Exhibit need and coverage- delivery stops. USPS-19C, and analyzed in USPS-LR-H-160. Special 17 in in somewhat developed time route analyze are developed the described are variability are and are to distribution time-to-stop class purpose and used are delivery models models special variability new data and proportions1D and major old, models coverage-related my analysis, function) 17 years Similarly, and In 8 I--- Special delivery messenger as obscured time as well a function functional vs. messengers run of stop as differences been delivery-related differences delivery (e.g., has time in time accountables parameters. cost vs. related causality nondelivery to vs. the This (e.g., stop mail nonaccountables) . loSee Workpaper 1.5. "See Workpapers 1.6, 1.7 and 1.8. I'In Docket No. R90-1, t,he Commission adopted changes in the methods used to compute SPR driving time-to-stop variability Access Test (PAT)". I have from the results of the "Parcel reviewed those changes and concluded that they reasonably correct for the influences of walk and travel activitiles in the original PAT. In my analysis, I use the adjusted drive time-to-stop variability estimate of 0.6342. 7 1 Also, 2 somewhat 3 In 4 substantial 5 variable 6 and 7 development 8 coverage-related 9 stops=, the data used keys". 11 II. my analysis, new data enhancement costs of Other A. 13 To date, to 16 activities. 17 volumes 18 entirely 19 result, 20 quite 21 activities. 22 of 23 piece in developed functional that have become provide detail. driving, customer activities are identified. This new econometric models and creation of for the delivery includes analyzing time spent new delivery a volume with Mail interfacility at delivery distribution there has service been no explicit interfacility Express Mail from Such moving in these c,ther small Mail movements different from types most vehicles service, of carrier piece established characteristics carrier are do not and caused vary volume. Exhibit USPS-19C and USPS-LR-H-160. Workpapers 2.5, 2.6 and 2.7. 8 messenger requirements. cost they dedicated modest on schedules other but or involve distribution have to movements typically In particular, Mail analysis distribution movements by Express Express I'See I'See are associated the Express 15 keys Work 12 differentiate distribution variability and 14 develop outdated. collection 10 to As a that are messenger by the measurably existence with /-- /- /-- 1 In 2 associated with 3 separately in 4 B. 5 The method my analysis, Express messenger interfacility Cost Segment Mail collection 6 Commission 7 costs a proportion to the fraction 9 boxes of all collection (a) the involved form 10 the fact 11 at 12 sweeps 13 distribution 14 volume-related 15 pieces 16 properly 17 types the same points volume Express requirements of and reasons. nor the of be viewed as causing boxes swept. 18 In my analysis, 19 associated with 20 identified separately 21 C. Allocation 22 The allocation does driving sweeps in of Cost of not of clearly identify 24 legitimate 25 accounted for. The lack 26 allocation is demonstrated 27 evident purpose in the when new survey Mail data. 9 where both time 7 and Cost Segment are 9. costs costs in Cost costs, costs for functional Over Mail boxes variable by the made for collection 9 fixed a foundation (b) can stop product-specific of or, Express service fixed Segment fixed of for be located never stop Mail volume to essentially and Segment account by outbound Express Cost not tend a vehicle residual 23 did volume in collection governed Express time Mail together; the the access serviced are Neither existence are are time and Express Mail boxes stop Service to This of often fixed Postal Express types boxes and costs if ied 9. collection boxes. are Mail such ident Segment access by the that two are variable of and activities Cost box allocation that of 7 and used analyze carrier distribution previously to and 11% of Segment and have such an diversity messenger has been 9 no 1 vehicle stops, 2 are 3 movements not 4 sweeps of ordinary 5 caused by multiple 6 example, pickup 7 conducted for 8 Express 9 the associated 11 type 12 causal 13 messenger 14 identified 15 of D. sulzh 23% of to mai: entails Mail diversity in foundation for and its handled beyond there the costs 16 See Exhibit 17 E. 18 See Exhibit Cost Mail rate category USPS-19D. basis for pickup fees USPS-19E. 15See Workpapers 2.1 and 2.2. 10 cost in light fixed that as by of is are differentials far pickup this no portion as incremental. Express that by messengers In are is as well show post. and For and post data to the stops, service services. activities, allocating at activities own fee, parcel messenger time These new survey or spent pickup and parcel pieces Priority street boxes.15 Mail the Mail, subclasses the time as interfacility Express collection Indeed, the activities service of are with Priority majority service almllst dedicated Mail. 10 and properly of I-. Exhibit USPS-19A SEESALISH~~T New spreadsheets time on motorized distributions street In addition, containing (W/S 7.0.5), further w/s 9. letter (W/S 9.0.3, time spreadsheet 7 and been created routes These detail as well and with cost special as in changes analysis other new spreadsheets analysis of messenger on W/S 9.0.5.1- have special changes driving cost delivery been made purpose spreadsheets and of with distributions substantial the for (W/S 7.0.4.X, on W/S 7.0.6.17) 9.0.5.5). 1. have REiLU'JFMENTS in are Cost in routes Segments described in below: 7.0.4.x Line 1 Accrued 3 Drive-to-stop previously Salesman variability cost input from W/S 7.0.4.1. variability - the figure of 0.5 has by the Conlmission, and is based on the Model". It is the best available estimate parameter. the been used "Traveling for this 4 Stop-to-activity variability - As documented 19C, econometric analysis was undertaken to to which the number c,f vehicle stops varies carrier activities. in Exhibit USPSmeasure the degree with the number of 5 Deviation delivery-to-piece variability - As documented in Exhibit USPS-1X, econometric analysis was undertaken to measure the degree to which the number of deviation delivery stops varies with the number of pieces requiring deviatj.on delivery. 6 variability of routine loops/dismounts - As documented in an estimate was developed of the degree to which Workpaper 1.14, the number of routine: loop/dismount points varies with volume (weight). 6, EXHIBIT Activity-related Rule. drive costs = L1 x L3 x L4, Activity-related Workpaper 1.2. drive costs distributed Volume-variable are distributed Workpaper 1.9. deviation delivery in W/S 7.0.6.17, w/s on the activities Chain based on costs = L9d x L5. These costs based on the data developed in Volume-variable routine loop/dismount costs are distributed in W/S 7.0.6.17 distribution of mail weight developed 2. to based USPS-:L9A PAGE 2 costs = L9g x,LS. These based on the estimated in Workpaper 1.10. 7,.0.5 l-7 Development of accrued 8 Drive-to-stop 9 Stop-to-activity variability - As documented in Exhibit USP,S19C. econometric analysis was undertaken to measure the degree to which the number of vehicle stops varies with the number of carrier activities. variability costs. - See USPS-T-19, Section I.D. ‘-. 10 14-16 17 19-25 26a Individual delivery-to-piece variability - As documented in Exhibit USPS-19C. econometric analysis was undertaken to me,lsure the degree to which the number of individual delivery stops varies with the number of pieces requiring individual deliwry. Delivery stop time variabilities - As documented in Exhibit USPS-lYC, econometric analysis was undertaken to measure ths? degree to which the time spent at individual delivery stops varies with the number of such stops made, the number of accountable pieces delivered and the number of nonaccountable pieces delivered. Activity-related Chain Rule. drive costs = L6C5 x LB x L9, Activity-related Workpaper 1.3. drive costs distributed Volume-variable individual delivery driving These costs are distributed in W/S 7.0.6.13, developed in Workpaper 1.8. to based activities on the baseei on costs = LlO x L'22C5. based on the d.ata EXHIBIT USPS-19A PAGE 3 Lines 35 Volume-variable accountable delivery costs are distributed in W/S 7.0.6.4 in Workpaper 1.7. 36 Volume-variable nonaccountable These costs are distributed developed in Workpaper 1.6. = L15 ,x L22C6. These on the data developed delivery costs = L16 x L22C6. W/S 7.0.6.3 based son the data Volume-variable fixed time at stop costs = LlO x L14 x L22C6. These costs are distributed in W/S 7.0.6.10 based on the data developed in Workpaper 1.8. 44 3. in costs based w/s 1-3 9.0.3 Develo,pment of accrued 4 Drive-to-stop 5 Stop-to-activity variability - As documented 19C. econometric analysis was undertaken to to which the number of vehicle stops varies messenger activities. 6 Customer delivery-twpiece variability - As documented in Exhibit USPS-lYC, econometric analysis was undertaken to measure the degree to which the number of customer delivery stops varies with the number of pieces requiring customer deli.rery. ,/.--- 10-12 13 14a14f 15 variability costs. - unchanged. in Exhibit USPSmeasure the degree with the number of Delivery stop time variabilities - As documented in Exhibit USPS-19C. econometric analysis was undertaken to measure the degree to which the time spent at customer delivery stops varies with the number of such stops made, the number of accountable pieces delivered and the number of nonaccountable pieces delivered. Activity-related Chain Rule. drive costs = L3C5 x L4 x L5, Activity-related Workpaper 2.1. drive costs distributed volume-variable customer delivery driving costs = L6 x (C5L14c + (C6L14c x LlO)). distributed in W/S 9.0.5.1 based on the Workpaper 2.5. to based activities on the based and fixed time at These costs are data developed in on stop EXHIBIT USPS-19A PAGE 4 Llinem 20 Volume-variable accountable delivery costs are distributed in W/S 9.0.5.4 in Workpaper 2.6. 21 Volume-variable nonaccountable These cmts are distributed developed in Workpaper 2.7. 4. a. on significant The the This costs allocation of and to collection of and accrual costs is W/S 7.0.4.1, the retained L33a-L33m). transferring enhances costs following of following: letter validity route of the costs. collection the the appearing in W/S 7.0.4.1, -. common features: time at ordinary vs. Express Mail boxes; - new, Express Mail - of have - separate practice support include collection (see W/S 7.0.5, street 9.0.3 route previous The analyses 7.0.5 letter changes spreadsheets the collection a new estimate variable of route replaces time, spreadsheet analysis letter b. delivery costs = L12 x LUC6. W/S 9.0.5.5 based on the data in c: ySoreadsheet Other ccz3tS = Lll x L14C6. These based on the data developed of is separate collection of estimate coverage-related variable with estimate to the of volume boxes (see fixed time time), degree the variability Workpapers at for 1.13 stop' and (replacing which, along number of with collection time 2.4); at and, former driving stops volume. IThe time associated with sweeps of empty Express Mail collection boxes is used as a measure of the fixed time associated with sweeps of Express Mail collection boxes, and as a proxy for the fixed time associated with sweeps of ordinary collection boxes (see Workpapers 1.1 and 2.4). is EXHIBIT c. The enhanced to of by refining different d. time to the of of of e. W/S 9.0.6, analogous presented training support messenger rather treatment in carrier the street as support A description vehicle applicability use of training a pre-existing computation treated route the (W/S 7.0.4.3) vehicle use is factor components. rectify omission In letter the The analysis refined in analysis USPS-19A PAGE 5 for of than logical portion error, of office in W/S 6.0.3 which is resulted activity in the costs. in-office training mail-related, activities following are the carriers. individual USPS-LR-H-161. activities; line item spreadsheet changes is Exhibit USPS-19B LVLIIA COT,T,ECTION AND ANAT,YSIS PROCRDlm 1. General The using analyses data derived messenger the from activities. data library described from in USPS-T-19 four field have surveys> The procedures these surveys are been of used presented implemented carrier to gather in the and and analyze following references: Motorized Letter Special Route Purpose Expedited LDC 24 using Route Mail survey of these (i) log activities, relevant (ii) each forms were the course of information departure involved record or then their expedited completed associated for USPS-LR-H-157 USPS-LR-H-153 USPS-LR-H-158 USPS-LR-H-154 USPS-LR-H-159 the gathering to needed record might cause piece), and street vehicle Express in the a dedicated pieces was These log clock times stops, the Mail Study, development were stop and other recrorded. carriers/mes!sengers! the mail ID numbers classification such work. example, 'A fifth data collection, the to provide information needed rates. See USPS-LR-H-155. street that with information regarding survey by the with of out, associated regarding, USPS-LR-H-152 clocking item normal USPS-LR-H-156 worksheets Before delivery USPS-LR-H-151 information in-office information Log and to data. I a parcel (e.g. Survey surveys and to Survey Survey forms classification assigned (SPR) (EXP) (LDC) Each (MLR) during forms of provided arrival activities was unliertaken of Express Mail EXHIBIT undertaken at numbers those of any A number of data stops, delivered steps were from - careful development (previously assigned) ID undertaken this to process, of ensure including forms the the quality of the following: and procedures, including pretesting; - extensive integrity and involvement adherence - comprehensive - field After of the keypunched, the of editing data further rigor and the supervisors survey of were data from taken results. in and procedures2. the the These p,rocedures, <and, proce:s; follow-up received reporteed weighting in and other steps (ensure training; surveys w'ere to procedures; teleconference call-backs raw of to monitoring - use ..r-. the items. resulting field and USPS-19B PAGE 2 field analysis steps are and to .include described ensure data further below. 2. Data EdLl.ng Three types modifications of computerized simple other in ,/---. specific editing the procedures patterns; potential of (b) were performed original survey for correcting manual inconsistencies; review that data. These data to and, in included problems address (c) resulted that missing treatment (a) fit data of or outliers tabulations. 'For example, when the initial MLR Survey revealed an unanticipated diversity in the composition of routine looping a supplemental worksheet was developed points and dismounts, sent to relevant survey participants. See Workpaper 1.14. and EXHIBIT used a. vzed to apply Procedw systematic - The edits to the - If a piece is Express Mail, services other than COD or return - If piece change checked, add a c:heck for survey change receipt critseria - Assign a (check to the street form (contains no other delivery activity - For numerical colons with decimal values points greater change value -k. were data: any checks to missing for special values. weight is greater than 11 oz. and First ,the check for First Class to a missing Priority Mail. - If piece weight is Bound Flrinted is checked, Bound Printed to a missing - Zone Rate/Other. - Interpret presence/abse,nce following USPS-19B PAGE 3 Cla:ss valua, is and than 10 lb. and Fourth Class the check for Fourth Class and add a check for Fourth Class "individual" an assigned is shown. other (e.g., or "deviation" ID number (or than time values, 9:5 is changed to time values as decimals of colons and the values or minutes observed. delivery a check) if but replace 9.5). based on the b. -1 Review - A manual review was performed to identify circumstances where missing information could be supplied or potential inconsistencies could be rectified using information available on the survey forms. For example, if a carrier reported but failed to report office hours, total hours and street hours, manual review could permit the value for office hours to be filled in (e.g., using subtraction). Outlllrs 'c - In the development of time proportions (only), a very small number of observations were excluded because they contained total elapsed amounts of driving time, travel time and/or time at stops that appeared to be excessive in comparison to normal work day limits. c. 3. &j&tina The survey Procedures sampling data procedures explicitly used provided in for the gathering different of the probabi.lities subject of --. EXHIBIT selection For for different example, large sampled with Survey, while messenger use of collection such survey and sam:pling used rates ensures that universe the from the case aro,se due to type to of the for the spent an individual working of route. To account included time terms combination observed were in to that type the for that weight inverse operate the that the the efficient appropriate weighting factors inverse of are representative to have the This of the the each carrier/route because activities, will of consideration identify Basically, the be observed route is by works SPR Survey carrier/route to amount sampled time probability IOCS to period. be tc on that weighting sampled the IOCS tallies proportional actually proportion in with SPR Survey. an additional employee this, EXP sampled observations. different employee the drawn. sampling. in in survey IOCS tallies on a particular amount the study. intenti.onally facilitates reflect estimates of the general, SPR Survey, use time intentionally In develop they the were necessitates that resulting combinations ref1lect that applied which In but be applied. bee:n develtoped probability sampling under were high in stratified universes units probability resources, factors the delivery routes low of weighting of disproportionately disproportionately 'The use portions USPS-19B PAGE 4 the type of factors type of time it was Exhibit The econometric carrier/messenger analyses surveys USPS-19C performed are on summarized data from the below: Model 3: at Delivery Stops = f(# stops, # Accountables, # llx!aEmTime Model stops Theory The number of vehicle stops is caused by the number of activities that must be performed Model Specification/ Rationale Quadratic/ Plot data, simplicity, flexibility Dependent Independent Variables 'Jariable Model Deliveries fo 1: = STOPS = # vehicle stops on sampled run ACTS = # activities undertaken on sampled run of 2: = The number of stops required to perform individual delivery is caused by the number of pieces requiring individual delivery; ‘high volume" stops are fixed, and do not vary with piece ~OlUlW The time spent performing individual delivery is caused by the number of stops that must be made for individual delivery, as we:11 as the numbers of accountable and nonaccountable pieces; "high volume" stops differ from other stops in these relationships Quadratic/ Plot of data, simplicity, flexibility, alternative model Quadratic/ Simplicity, flexibility, alternative DELS = $4 delivery points receiving 1-4 individual delivery pieces sampled run = ~Cumulative time at individual delivery stops c,n sampled run VOL = # pieces delivered to points receiving 1-4 individual delivery pieces sampled run model TIME on Ni = # individual deliver-y stops sampled run at location i <XI on --. EXHIBIT USPS-19C PAGE ,- Model stops 1: = fo ACTS**2 Model Deliveries f Model 3: Time at Delivery Stops = f(# stops, # Accountables, # Nonaccountab& 2: = voL'*2 VA = # aocountable pieces tc individual delivery stops on sampled run vo = # nonaccountable pieces t0 individual delivery stops sampled run VA+'2: on VOff2 Separate models 'estimated for 1-4 :piece and 5+ piece ,stops ,(each estimate separate model) MLR, SPR, EXP (LDC 24 portion), EXP (LDC 22, 23, 27 portion), LDC MLR, SPR, EXP (LDC 24 portion), EXP (LDC 22, 23, 27 portion), LDC SPR, EXP (LDC 24 Iportion) Alternative None' Same, but without limitation t0 1-4 piece stops' :Ssme, but with :single N variable (and N**2), with snd without3 :separate estimation for 1-4 snd 5+ pc. stops Datasets used to Model 'The preferred model was al.so estimated on a preliminary, incomplete dataset. 'This alternative model was only estimated on a preliminary, incomplete dataset. 'This alternative model was also estimated on a preliminary, incomplete product eliminated the dataset with a specification that included The cross-produmzt terms (VA*VO, VA*N, VO*N). (due to lack of an operational foundatimx) estimation of the alternative model on the full cross- term,s prior dataset. were to 2 EXHIBIT Model Econometric technique from the results preliminary presented below. weighted sum of output to Similarly, is output the It is ts3 Cost programs in and/or noted SPR and special the sum of Segment 9. used in USPS-LR-H-160. that for EXP (LDC 22, purpose = route by preferred alternative models, in econometric as are 27 portion) analysis by as well Mode.1 1 and EXP (LDC 24 portion) the Data weighted inverse of sampling rate models, both 23, Model 3: Time at Delivery Stops = f(# stops, # Accountables, # 2: Data weighted inverse of sampling rate by from da,ta a weighted Computer presented = Data weighted inverse of sampling rate The econometric Model Deliveries f I.: stclps miviries) USPS-19C PAGE 3 Cost and analysis Model 2, results is Segment 7. LDC results are a EXHIBIT SUMMARYOF ..---, w 1: WONOMETRIC stoos = USPS-19C PAGE 4 ANAT~XSES (Activitid Complete Data/ Preliminary H M!d.el i-'- Data EaJL Preferred Adi. R Model N EaJL % Total EC& sraes' - MLR 0.9831 206 0.99538 0.9831 213 0.99440 [I - EXP SPR 27 0.9583 0.9647 571 52 0.96460 1.00909 0.9376 0.9606 684 65 0.92280 1.05 0.00628 0.99372 - EXP 24 24 c -- LDC 0.8967 0.9634 112.9 220 0.94985 1.01018 0.9196 0.9049 1399 261 0.94962 0.96750 0.07899 0.92101 12zzl Output output to to w/s w/s 7.0.4.x, 7.0.5, output to w/s 9.0.3, C3L4 p3Gq = 0.9944 CPLP = 0.9995 C9L5 = 0.9661 (BY value (BY value = 0.9996) = 0.9657) 'Estimated from activity and activity/stop data contained USPS-LR-H-157, USPS-LR-H-158 and USPS-LR-H-159. 'Model value of 1.007 truncated to 1.0 due to absence of operational foundation for diseconomies of scale. in EXHIBIT OF ECONOHmC 2: Preliminary Alternative .N Adi. R- td!d.d - MLR r- - EXP LDC 24: 24 output output output to to to Deliveries Data/ Model YaJL USPS-19C PAGE 5 ANAT,YSES (cont'd) = f(VoM Complete Data/ Preferred Model Adi. RH YaL-0 % DELS Rt.Type EaL 0.7261 116 0.94421 0.8799 117 0.995'75 0.98904 0.5912 334 0.831.71 0.9102 373 1.06 0.95472 0.95472 0.3064 35 0.41286 0.9465 44 0.70963 0.9099 0.5638 172 979 0.91965 0.66254 0.9550 0.9461 1197 201 0. 96206 0.96091 0.97685 0.93866 W/S 7.0.4.x, w/s 7.0.5, w/s 9.0.3, C3L5 = 0.9848 (BY value = 0.9852) CPLlO = Cl.9519 C9L6 = 0.9396 (BY value = 0.9395) <Model ,value of 1.015 truncated to 1.0 due to absence of operational foundation for diseconomies of scale. 'Weighted mean of SPR and EXP :27; weight = # DELS (from WP 1.3). OWeighted mean of EXP 24 and LDC 24; weight = # DELS (:Erom WP 2.1). ,-iGzq EXHIBIT C ANALYSRS Model T1'm e = 3. IStows. Adi. 1. Preliminary N/Cross-products Stops) N VA vo 2. Complete (All Stops) N VA "0 3. Accountables.wntables) R2 N 281 0.5025 Data/Single N 0.2772 317 0.0485 0.1277 0.4547 N 4. Complete Data/Single (5+ PC. Stops) N VA vo N 5. Complete PC. stops1 N (CQnt'd.) 0.0689 0.1907 Complete Data/Single PC. Stops) N VA VO (l-4 r-. 0.3092 Data/Single (All USPS-19C PAGE 6 0.1721 321 -0.0518 0.1398 0.4915 0.8612 % TINE 70 Total m 0.5300 0.1251 0.2624 Data/Ni (l-4 Data/Ni (5+ PC. VA VO 6. Complete Stops) 70 VA 0.6846 0.0520 VO 0.3002 N r-7 0.9527 'Model values of 1.0148 operational foundation "Model value of -0.0907 operational foundation of the VO variables in and 0.0503 prorated due to lack of for diseconomies of scale. treated as 0 due to absence of for negative volume variability. Neither this model is statistically significant. EXHIBIT output out.put output to to to w/s w/s w/s C9L14 C9L15 C9L16 7.0.5, 7.0.5, 7.0.5, 2 adi. kfQd& 7. Preliminary N/Cross-products stops) N "A vo 8. complete (All Stops) N Data/Single (All R 0,.5720 N (BY value (BY value (BY value = 0.8963) = 0.0592) = 0.0445) XaL 857 0.5570 0.2369 0.0403 Data/Single N 0.6058 1052 0.7982 VA 0.0980 vo -0.0145 9. Complete Data/Single (l-4 PC. Stops) N VA vo N i?- Complete Data/Ni PC. stops1 N VA 0.6324 1049 0.7048 0.1550 0.0237 10. Complete Data/Single (53. PC. Stops) N VA vo - = 0.9117 = 0.0479 = 0.0461 N 0.2034 127 % Total YaL 0.7411 0.1168 -0.0555 (l-4 vo OL!tput output output to to to w/s 9.0.3, w/s 9.0.3, W/S 9.0.3, C9LlO C9Lll C9L12 = 0.6860 = 0.1342 = 0.0219 (BY value (BY value (BY value = 0.6865) = 0.1334) = 0.0237) USPS-19C PAGE 7 Exhibit USPS-19D L RATE CATRGORY COST DIFFRRFIYTIALS The Express source rate for categories. each of Express the Mail of intervening This it makes observation. the in this the those significant related transportation lT,his analysis differentials. order to and based to this, it on differences to in causal costsI, omits any to has during the the 1980 Mail in reliably categories are study rate scarce. meas.ure throutgh been direct necessary to the field develop current cost differentials analyzed with respect cost There unit factors. the with now extremely differentials between factors. differentials in Express is category cost-causing associated categories. costs rate Mail involved have employed determine rate Express funsztions that impossible rate of primary etc.). mailstream or the volumes different Basically, methodology, factors, for as the now outdated, and piece methodology Mail served among PO-to-Addressee with Express determined ,,-. the Because Under costs in has is distribution, impractical new methodology respect category than associated ca~usal rate now be repeated. other among times changes categories costs which labor period, not (1960) differentials stuldy, outgoing Because :Study coist This (acceptance, could Cost determining tabulation ,- Mail four rate is Therefore, Express (quantification Mail of are categories no evidence acceptance with of or distribution aside from rate to categories transportation cost a EXHIBIT differ from related one another primarily with respect associated with USPS-19D PAGE 2 t'o delivery- costs. The delivery-related categories are different Costs determined delivery methods, f Unit CQst through Express ($000) 2. Delivery/Total 3. Delivery i$OOO) 4. Piggyback 5. Delivery-Related costs ($000) 6. Piece Volume (ooo)8 7. Delivery-Related Cost/Piece ($1 Mail Cost costs Cost 27,488 98,098 associated by Mf=thod 6+7 24.747 of 9. 51,217 712 863 23,401 51,074 5,380 1.41854 1.410:24 1.41823 1.49538 1.19693 2.227 33,188 76.375 6,439 0.1.15 19,017.852 1.745 99725 21,882.762 Deli verv Ff.QL3i 5,479 ,945s’ 19,307.%2 with lQ 025g3 .ooss' Cost Window service in Lines 1-4, 'Source: FY96 'Source: Cost ?Source: Cost %ource: Cost 'Source: FY96 *Total distributed unit rate as follows: Deleted segment: 1. the different 9s206 4,171.824 3.490 and platform operations are reported and consoli.dated in Lines S-7. IOCS tally ,analysis. Segment 7, W/S 7.0.3. Segment 9, W/S 9.0.4. Segment 10 spreadsheets. IOCS tally analysis. based on results of Express Mail 1.543 separately Study. 118,229 64,380 1.836 EXHIBIT ,- sast se- USPS-19D PAGE 3 6+7 Ct.9 *. Pcs. A. PO--to-PO B. PO-to-Addressee C. Same Day Airport D. Custom 9. Fracfion A. PO-to-PO 1.0000 B. PO-to-Addressee 0.2906 C. Same Day Airport 1.0000 D. Custom 0.8761 287.914 18,432.792 Designed and provide the differentials between the basis the 4,171.824 15.874 67.114 0.2996 0.3440 0.0658 0.0237 0.1002 0.0000 delivery-related mean delivery-related for across establishing rate the 0 cost cost per "per piece" for 669.802 each piece Cost Per of rate $1.836 cost Piece Neal2 ($1.721) PO-to-PO $0.115 PO-to-Addressee $I..858 Same Day Airport $CI.115 ($1.721; Custmn 5c1.492 ($1.836)" Designed 63.422.242 categories: Delivery-Related -'er $0.022 service and platform operations are reported separately and consolidated in Lines 5-7. in Lines 1-4, lOReflects omission of delivery-related costs from per-piece charge. Cost basis for per delivery charge = $0.492. PWindow /--. 21,815.648 0.042 586.814 Designed category 19.001.978 0.042 The differences .- 287.914 Exhibit USPS-19E COST RASIS The general approach scheduled pick-ups (by for type) route is based special carrier/messenger survey:; previous 1. SPR Indiv. Time/Stop' Del. 2. SPR Drive Time/Stop (Indiv. Dels.) calculate delivery Information the to on driving carriers. replaces FOR PICKUP is use FFWS costs time/stop fo.r on-call plus time messengers and spatial drawn from the rescent this approalch, to support of messenger delivery Stop and at stop purpose whi'ch (costs as a proxy. 301.69 i 3. SPR Drive Time Time/Stop o.71732 (Ll x 216.40 216.40 288.28 315.57 504.68 531.97 1.33102 1.33102 L2) 4. SPR Pickup (seC.)3 5. Total Time SPR Pickup at Stop Time (L3 + L4) 6. Street Support Factor4 7. Productive Hourly Rate City Delivery Carriers 8. 9. - 26.083 26.083 City Delivery Carrier Piggyback Ratio 1.31468 1.31468 SPR Pickup Cost L7 x LS)/36001 $6.3985 $6.7445 'Source: lSource: 3Source: 'Source: [CL5 x L6 x USPS-LR-H-157; Cost Segment USPS-LR-H-157; Cost Segment USPS-LR-H-158. 7, W/S 7.0.5, CSL22/C6L22. USPS-LR-H-158. 7, W/S 7.0.1. Includes related volumetime EXHIBIT 10. SDM Cwt. Time/Stop' Del.. 11. SDM Drive Time/Stop (cust. Dels.) Time 12. SDM Drive L2) (Ll 13. SDM Stop 14. Total L:L31 15. support 16. Productive Hc~urly Rate Special Delivery Messengers 17. Special Delivery Piggyback Ratio 18. SDM Pickup C0e.t [CL14 X L15 x L16 x L171/36001 Time/Stop Pickup (sec. ) ' Time SDM Pickup at 0.6828’ x Stop Time CL12 + Eactclra Messenger 170.41 170.41 515.69 505.60 686.10 676.01 1.57627 1.57627 24.411 24.411 1.49520 1.49520 $10.9648 $10.8035 % SPR' 0.5974 0.7902 % Messcngerl’ 0.4026 0.2098 $8.2369 $7.5961 Wtd. Ave. 5Sourc'e: 6Sourc'e: 'Sourc'e: %ourc'e: associated In/Out) 9Sourc'e: 10Sourc:e: Cost USPS-19E PAGE USPS-LR-H-158; Cost Segment 9, USPS-LR-H-158; Cost Segment 9, with activity and 6430 (Obtaining USPS-LR-H-157; USPS-LR-H-157; USPS-LR-H-159. W/S 9.0.3, USPS-LR-H-159. W/S 9.0.4. Includes related volumetime C5L14c/C6L14c. Includes portion codes 6519 (Training), 6522 Mail, etc.). USPS-LR-H-158; USPS-LR-H-159. USPS-LR-H-158; USPS-LR-H-159. of C6 (Clocicing 2
© Copyright 2026 Paperzz