~Dn r,

OF C

aeoradue~d by

NATIONAL TECHNICALINFORMATION SERVICE

S."fedYA ?2151UNITED STATE AIR'FORCE

AIR UNIVERSITY:AIR FORCE INSTITUTE OF TECHNOLOGY

Wright-Patterson Air Force S'as*,Ohio

Best Available Copy*dw oublic w

AD-753 780

A SIMULATION ANALYSIS OF THE ECONOMICCONSEQUENCES OF ESTABLISHING MULTI-MODALTRANSPORTATION COMPANIES

Robert S. Tripp

Air Force In-stitute of TechnologyF4Wright- Patterson Air Force Base, Ohio

December 1972

DISTRIBUTED BY:

National Technical Information ServiceU. S. DEPARTMENT OF COMMERCE.5285 Port Royal Road, Springfield Va. 22151

A SIMULATION ANALYSIS OFTHE ECONOM'IC CONISEQUENCIS

OF ESTABLISHrING 'i-ULTI -I-ODALTRANSPORTATIO'1 COMPANIES

Robert S. Trilpp, Captain, USAF

Technical Report - AFIT IR 72-6

December 1972

Approved for puhlic relea-s';

*it ikto Uniiytd

-- . -n m n I I I I

• ~POTfE.'TtAL WIDE..IN4TESEST REP'ORT -3" :-- -

N~A.RNING TO EVA'LUATORS IRl',Ol INPUT-

DO IN WILL REPRODUCE PORL

S DECISION BY EVA'LUATOR S:

.4..

3.Return t~o ,;ourceW . ,-uktur'c

- . v!

IIiiO - i l i~INT' L "' Wl TL ST II ' T

.... A

DOCUMENT CONTROL DATA. R & D 71(See ... A lasorl¢ation of #lu., ?, f" Of ab~ crn ard ind.'xne atrnOtlon mnu'f he entered u'hen the overall ,!;%o Is clashified)

I. ORiGINA TING ACTIVITY((op.,e.autho) 20. REPORT SECURITY CLASSIFICATION

Air Force Institute of Technology (AFIT/ENS) UnclassifiedWright-Patterson AFB, Ohio 454332. oVP

3- REPORT TITLE

A Simulation Analysis of the Economic Consequences of EstablishingMulti-Modal Transportation Companies

4. OCSCRIPTIVE NOTES fT/pe O trport and inclusive dotes)

AFIT Technical ReportS. AU THORISO (First namne, Middle Initial, last namej

Robert S. TrippCapt USAF

0. REPORT DATE 70. TOTAL NO. OF PAGES 7b. No. OF REFS

December 1972 243 57On. CONTRACT OR GRANT NO. 90. ORIGINATOR*S REPORT NUMSER4SI

6. POJECT NO. AFIT TR 72-6

C. II. OTHER REPORI NO(S) (Any ther nuftbeti thit may be aea|inmlithit eport)

d.10. OISTRIBUTION STATEMENT

Approved for public release; distribution unlimited.

hI." qS - lf dI(b.Xi release; 1AV AFR f?9i-I ' ORING

Y Hix, "Captain, USAF ,or of Information

1.ABS TAAC T

Thi's study is concerned with the examination of the economic consequences ofestablishing multi-modal transportation companies. .More-speci-f-ical-ly,.-he purpose

) of this. research- is to ,---;examine the economic impact various combinations ofparameters or test factors have on transportation company formed from singlemodal carriers, and (-2 determine on the average'lwhich organizational form, h .'rtransportation company versus single modal carriers, is economically superior. s

In order to accomplish this goal, 4. simulation model was developed wh-ctrtifadethe comparison of a transportation company with single modal carriers.P""bleWithin this framework, the economic performance of/thLtwoorgaiZational approacheswas contrasted for both TL/CL and LTL movements. (The test factors wh.i.d.-weres---selected for analysis are: (1)rthe opqrating ratios of the-frmig-modes-l(truckand rail)-, ('rthe load factors of the forming modes, -3) the amount of availablecapacity, (4) and the level of shippers' logistics constraints.

The performance measures which were selected to describe the economic impactof operating a transportation system under the two different organizationalapproachcs are: (1) the expected contribution (to fixed and/or common costsincluding profit margin), of the carriers, (2) the actual contribution of thecarriers, (3) the total price paid for transportation by shipperz, and the amountof traffic moved by (4) truck, (5) piggyback, and (6) rail under each organizationalapproach.

DD IR, ,,,1473 UnclassifiedSi~eirtlv IIINSf~cl tcn

STrnspotaon

Multj -Xodal Tranlortation Ccw'~pani I4

Trans Porta t Ion Compan ies

CO!mOr1 Ownership

Economilc Analy-.1%

I I.

A SIMULATION A.NALYSIS OF

THE ECONOMIC CONSEQUENCES

OF ESTABLISHING MULTI-MODAL

U. TRANSPORTATION COMPANIES

F S

Robert S. Trlpp

Captain, USAr jInstructor in Management Science

Technical R .ort - AFIT IR 72-6Dece ber 1972

Apprmcd for public release;aitrlution unlimitod,

I

School of j In$9cwvlvn

Air lnor' l ir tlt , ol T,0lh10fljyAWr'lcjt I'atte.'.ui Alr" .,tt l'v., Ohio j

ACINOWLEDCX TS

The successful completion of this thesis was made possible

by the splendid cooperation given me by the members of my disser-

tation reading committee--Profsaoors Norman L. Chervany, Chairu=n,

Frederick J. Baler, and Victor L. Arnold. Their encouragement

and comm~ents have contributed substantially to this research. I

especially wish to express my sincere thanks to Proiessor

Chervany for his sujrestions and Lonstructivu rrIticlows, I also

wish to thnnk Profteors Gary M. Andrew ar.d Nicholau A. ClAtkowsk/, Jr.

for the help thoy g ve me in the initL t s whin I wan dutInIng

the objectivew of this project.

I also wish to jpreis may rat1Ludo to Alan J, tcrger and

C. Randall lyera, fallow studonLz of U-1110, for taking the time to

llsten and common on my Ideoi. I am also Cr-,tcf'..l to Mius Vla :

berube wht, expertly typed t013 dioxtr..ticn.

Acknuwlodt Lt Is made to the Univertity nf M1nn utn Coiput.,r

Ceontr for providig WoIputnur facilitico tud time witksu.u which

thin thesis could not hive been accvmplfl ad.

I WICIh LO CX .rovn m.y iratlltidu LO My paruitu (or h"Avng divo; I) 3d

my Isite rvit. Jn ticasdt;.,I~ Ac i j uld v.,:n t 'Iu ui wd , Hf ry, wii Un~t.rlI-

buted morv thon words i :eu ,xprv. o, I w 1 h -. i -rov r 1dvbtd ,

i SI

.!-

This research was supported by the United States Air Force

through the Air Force Institute of Technology, dfld reproduction

in whole or in part is permitted for any purpose of the United

States Government.

?ABLE OF CONTENTS

Page

ACKNOWLED17MNTS . . . . . . . . . . . . . . . . . . . . i

LIST OF TABLES. . . . .. . . . .. . . . . . . . . . .. vii

LIST OF FIGURES . . . . . . . . . . . . .. . . . . . . .

ABSTRACT. xi

CHAPTER

I. INTRODUCTION. . . . . . . . . . . . . . . . . .

A. Definitions

B. The Economic Is3ues

C. Purpose of the Disertition

D. Organization of the Dirsertation

II. DEVELOPMeNIT OF TIIE MOD .. . . . . . . ... .. 18

A. Scope of the Research

B. Ausumptions

C. Identificationi of the Important Variables

D. The Doclsion Enviro;ment

E. The SiMulatiol Model

II!

A

Page

III. SPECIFICATION OF THE MODEL . . . . . . . . . . . 48

A. Classification of the Test Factors

B. Other Specifics of the Model

C. Recapitulation of the Test Factors jD. Generation of the Data

E. The T:ansportation System

IV. EXPERIMENTAL DlSIG' .. 82

A. The Problem of Realism in Simulation

B. The Output of the Simulation

C. Analytical and Statistical Methodology a

V. THE RESULTS OF THE SI .JLATION ANALYSIS ..... 105!

A. Nature of the Analysis

B. Explanation of the Average Performance

Measures

C. Analysis of the Test Faztors

D. Summary

E. Economles of Scale

F. The Significance of Interaction Effects

vii- "

.IVi

. i

?a~a

VI. OTHER ASPECTS OF COMON OWNERSHIP . . . . . . . 167

A. Obstacles to Integration

B. Relationship of this Study to the

Obstacles

VII. CONCLUSIONS AND I11'LICPZ1OS OF THE STUDY .184. 18

A. Suiaary of the Research

B. Further Research Required

C. Implicatlons of the Study

BIBLTOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . 193

APPENDIX 1.

A. The t Test Involving PairedObservations of Two PopulationMeans .. .. .. ... .. .. ...... 199

B. The Analysis of Variance of theMain and First-Order InteractionEffects .. .. . . ... . . .. . .... 200

APPENDIX 2.

A. Flowchart of the Simulation Mo'del . . . 204

B. The Simulation Computer Program . . . 204

V1

7M

LIST 07 TOELLS

Table P&SO4

I K461 VariAtiOns,..N............ 64

II Othof Model Specifies ....... . 65

III FoVcontagu Irtakdown of Rail and Truck CoatA

Structurb. * ****e****... 70

IV Trucking Avwvro~o Cost% per Cwt, ..... .. . 72

V T!, Average Costh per Cwt. .... ....... 73

V!, yTrCefLigs breikdomi, of TL Costs . . . . 74

V11 LTL AverA&6 Cd*sa per Cwt, 75

Vill Not) Avore~p Coft# p'r Cwt. * .... .. 76

ix .liol(41 F.Ottor'sL Factorial Da2L. 8 a . 100

X& AvarAge IDttfaren~o ot the 3'arforzance Measures

lotwpvn thc 'k'anaortstfon Compimy and thelingle Modal Carrier%............ 110

Xb Avutago i'?ffaru:ncv as a Percentage increase()I~o * iecrtafic( of the Combiid Gingl's Modal

XI Toot YNcL, tur L sy .. .. .. .. 115

X114 IEffoclo t of is wi In Truck OperatingC RUtIU on

X1l1, I.II(tt U[ Cj1Oli4( Ni 'Iruck Upur.. Lln h-LOUU1

Illt 111 l'tio mavcu12

V11

Table Page

XlII Effects of a Change in Rail Operating Ratio on

the TL, CL Performance Measures ........ 124

XhIub Effects of a Change in Rail Operating RaLi; on

the L'L Performance Measures. . . 4 a . . 125

XIVS Effects of a Clange in Truck Load Factor on

the TL, CL Perfo-mance Measures ........ 131

XIVb Effects of a Change in Truck Load Factor on

the LTL Performance Measures . . . ...... 132

XVa Effects of a Change in Rail Load Factor on

the TL, CL Performance Measures ........ 139

XVb Eftects of A Change In RA1L Load Factor on

the LTL Performance Measures . .. .. .. . 140 IXVIi Effects of a Change in the Capacity of the

System on the TL, CL Performance Measures . . . 149

XVIb Effects of a Chanae in the Capacity of thA

System on the LTL Perform.ance Measures, .... 150

XVIIa Effects of a Change In the Level of Shippers'

Logistics Constraints on the TL, CL

Performance easurcs . ........ . , 153

XVIlb Effects of a Chance in tile Level of Silppers'

Logistic6 Constraints un tie LTL 'erformance

Heasuroes.... . . . . . . . . . .

XVIII Sumriary of the MI, Cff,.CL L on t Q Pcrfor-enceI

M asurvs . . . . . . . . . . . . . . . . . . .. 160

vLI

I t + ,;. , , . . . . . ... ..

Table Page

XIX Summary of the Interaction Effects on the i

Performance Measures ............ 161

IxI

4

.1

I

I

O2

Aix

-i

LIST OF FIGURES

Fi.ure Page

1 Structure of Equipment Allocation Decisions . . 38

Ristrictions on Carrier Pricing Decisions . . . . 43

tx

r'

ABSTRACT

Common ownership has been a subject of debate in transpor-

tation circles for years. Unfortunately, while discussion and

speculation on the subject have abounded, there has been very

little research directed at determining what the potential

of a multi-modal transportation cccpany is.

This study is concerned with the examination of the

economic consequences of establishing multi-modal transpor-

tation companies. More specifically, the purpose of th!s

dissertation is to: (1) examine the economic impact various

combinations of parameters or test factors have on a transpor-

tation co.pany iorred fron sinle mcdal carriers, and (2)

determine "on the average" which organizational form, i.e.,

transportation company versus single modal carriers, is

economically superior.

In order to accomplish thi s goal, a simulatiou model was

developed whLch made the conparlson of a transportation

company with single too,.l' carr.ors possible. Within this frame-

work, the economic pcrformanccu of the two organ zaLlorial sippronchca

was contrasted for bo~h TL/CI. and LTL mov:aonts. The te.t

factors whic'i were selectod for analynit; n-c: (1) the operating

ration of Lh formi1,,i; mode:. (truck and raJ ), (2) the load

xi

t_

T II TIT - 1 I I 111 II I I I i

factor@ of the £azming modes, (3) the amount of available

capacity, (4) and the level of shippers' logistics constraints.

The performance measureb which were selected to desrlbe the

economic impact of operating a transportation syston under

the two diffarent organizational approaches are: (1) the

expected contribution (to fixed and/or common costs including

profit margin), of the carriers, (2) the actual contribution

of the carriers, (3) the total price paid for tranoportation

by shippers, and the amounL of traffic moved by (4) truck, (5)

piggyback, and (6) rail under each organizational approach.

A fractional factorial experimental design aus utilized

to analyze the output of the sJilutlout 4udal. Appropriate

statistical costs qeru uti-iu to: (1) indicate which of the

test factors or tort factor combination* produced statistically

significant behavior in the perfornancc measures, and (2)

determine If the performance measures were significantly

different for the two orgr.nizatlonal Approaches.

It was found Lhat moot of the avvraga differencen of tha

performance measures between the traivillurtatioa company and

the aliiglu modal carriurs wero siUniffic,ir't. Xore fpecflca1ly,

for TI./CI. movuiLnta It wai,' fomd that Lha oxpccted and uctual

contribution of the tran,ljorLtAt1IC1 cCrpMP ln w.ro u1infffcar,tly

greatet thian Lh !iurn tuf hu , rum C +; r ,i for Lht ,! I O t',l - modal

)A ]I

carriers. lit the LTL category of mov"rnentol the xpected and

actual contribution of the transportation company was a&ain

mignificantly larger than the combined contributions of the

single modal carriers. The transportation company also had the

effect of significantly reduciq t&e price paid by shippern for

LTL transpor:ation services.

The prin consideration tf the resoarch was the Identifi-

cation and explanation of the manner in which the levels of thea

toot factors affect tti average perfomAnce measurs. '.ho

analysis of the offects produced by th. test fActorn reveals under

What operating .ondtions (Lest factor lovol) the tra-portation

corapany was "economically 6upirlor" tW- th; I% lo z-dal cctpsAr.e!:

and vice vorita.

The economic anpects of the common ownership controversy

are, however, but one facet of a multidimun.Junf l p1ub. m,

As such, the research hau idenLltfid the other aspects of th

controvcrsy and discucaed thOir relstlonotl- co the results

obtaLn,,d in this study. Although there re . in a lot of

unansworod questionn which must be raseirchod concernin. the

Comon ow.te.rihip qucatJon, chi tudy li a ntep in tho., d Jg':tion

Of ObU"i ni .; LhOUL, n" Iwo'i.

XI I!

OUJTER I

INTRODUCTION1

For Lu-t $out time now~ there has been a continuing contro-

versy in traraportation circles about theo desirability of establish-

Ing multi-modal transportati~on Comn~anlas In the United Status.

The arguments for and againat the owiierthIp of rr- mad of transipor-

tation by a competing mode h~ve been r~tat*J and restated, Tito

nature of thtse argumeznts Is Illuxt-rated In the following quotations

The Issues in Lhs o O-i.cehpconitroveray tirebasically Aimpl. ~Hijor or-Pror~ts in favor uf cotvioii vvner-ship arc, thAt (1) It Vuis.ltit bwt~vr sirvicu to Litc custow-r.Through the AValla.bilil-y of cooedinitid Lrtindportatiofl sorvykaunder Lt.d Coc~roll. of ; Alo~ t-C'.5-"2Y,each shi pnont can 'ja rmovud by .to:ud*o co,,-,11natfo Oimodes which oit f1lta Lite r~quiror.critj if tho cu~tone~r.Comon owncrship would not reduCo *f4ClVie CopLltlOhbecausa thura would 1istl be many sopirate mode carriersand severel "d4rtv.&nrt ores o~f Lrnnisportation" genaralcarrlert. (2) Thore are ecunols in optratlon In havitt,One s41CA force &ad tino upurotLin- oduroit 0616 eocIWhandling thoi trLinsparL'aLion oarvico. (3) Tho cuatoA.Jrconvenience would be Increstovd ,I..utia there would ba onlyone toot of arron,pnits roq(uircd, And a single carrior wouldbe responslble for any loso or dsiaage

T*he 1nwporLsint si:U1o.L ;a~IosL c~fu;iof owiio.rilp I io lude:(1) It would stifle cot-pLLloI attd r :.ult in poorer ormore expetu ~iver :cr to the shippcar 19 Lho loat run. Th4railroads would lookLtO thruir pr;:-.ry 1icill rvrcand div(vrt aill possihlu traffic to thult r~il optirationWhile O.liing the LrucV ourvic'j to dri v-2 induipt-dellLL4'.o~t of 1 u nne~i . ,!) Xaii I cjuJ or C.LIIL: r nun tru(cl I r ::,'menrts coul d not dv 311 VI it Joh ol rtinnititi- L tuCklr ;'.

hc'caunt, Lhv' d". nut h "v h l'.i : ~ .n

?Iutor-.ru rr I r tiivrv icou WUUl L o bu dtivu Juptd m4O rap idly

PI

-2-

or as yell by coon-oimterahip coapentes so It would beby nanagatents which had no othar interest, 1

The above listing of pros And tons is not exhaustive,

but should give the reader an appreciation for the types of

arguments being caot back and forth. $htl% tsuch discussion

exists, Lhre ham not beer any effective ruseerch accomplished

on the subjoct of multi-modal :rAnonorttion companioes.

The controversy has boun unresolved for sevral roAons,

such si *ntrenclied carrier .anaijorionts or it least parochial

attitudes of csrriro conct.roln their ow aud., Furthrmorei

shippers tcnd to have a short-r o perspoctiva toward the

transportisti.n plant, trying to obtoin the ximum baneJts

from thW CxLiL'4'1& i.odas. ,rmother reason V-7 there hit been no

systematic reavarch on the subject is that the present laws

which deal with Elio tranportailon company qu.tione or

In.erpr to.lono of thoo la-WO by the Intartot Cororce

Coum!psin (IC.C) have bt.en very restricttve. 7 Th6 lack of

eiiectivo r uoafrch on the topic hau reeuli.Ld n the fact that

today there eKist no Suitable dec4lon criteria for dotcuili ng

l (h yto . ,, IC ., .": , ,l i JA. Jr,, ",

Hill Book Co(., Ific,, ;sj , Jp ', *j ,

2lnfrno, Ch,4!)LvC Y1, , ,

it the formation of transportation companius would be In the

public Intertat, 50 the conitovoray continue.

An indicated in tile above orcuserits the question of whether

or not transportation compapin should be ostAblished Is a wilti-

facatnd or multi-dimonstonal problem, That is, the formatlon

of transportation companins involvas *conorAtc, leal, and social

Issuas which muit be reeearched to dotarmiic the tipact such

companie would have on the transportation system of the United

States, Theme companie would (most probably) have difLering

effects on tle various groum Iinvolvod In tho movcent of gocds.

The shippers, carriers, And reultatorv hava differing

Lobjectivos concerning the tronoportation systci which affect

their respective attitude, on tho trAnsportotion conpany isuae.

Shippers danire a low cost, efticient transportatlion systea

capAble of rateting their neoed. Carriers ,iLsh to makc n good

profit in provldls tranGportation mervica. The regulators

sit in the mtddlv buttwon thouu groups trylng to Intsura that

It might bi- pOInLed out Lhal thcre a:o oeo:a trtinsportatfloacompanies in the C'nitedr !1, .ta , rAi oxr,,pl,, would bu the routhloernPae f tIc Coi:pnjiiy, which opor;tL ut i c .'6,0W) L. ruckti rout c m ls

14,000 mile of rail lIon, and 2,300 pipellno rAl (IYr.-ilCnUip)1NrV]I. "Oil OILu lgt. TClA;: , I1rjrrvo;.n, X.VIii, ;Lotbvr 2"1,1968, 1. 9). clouthrn 1'ifci fir. ulud Il LruI'L-t o;,ur tLhatL w':rit 1) guii prior to ) u.',hr hw ' " d,".t. r uWlI2 F11 S t I:i I "d L . t n .,Ltl.LC '-.:. r'o AO' LNec'LtI un I J it iutit l-rt .1 Lit.h t -,i j•cj.l ) lt u r.i J'o, d:, ':W 1it In Lru, t nt . u-/'lal tu lit;r r.i I ro,,d:i o..i ; i,. di .,ry tru;'', :,gc IOmlpall|O , on , ' iv or,!I fo!'.f, of opi, L i ttllf,, Llw!:0 !1%;',, t,l!' ,t r , t;art,

avvu ril ,, y t' i - IdL:::: j/ I)-i' t d II:.". C 'rI f r.,;; l Cpub litc. ' I t :,L.. ;I , ':.i j ;'. . ill bc '..I II.,,:. 'l . r

At3

the needs of both groups are met simultaneously.

This paper will undertake to examine the scanomic cones-

quence of aotablinhing intarmodal transportation conpanies.

The research wll be mainly concerncd with an analysis as to

how the crearion )f such a company affects the economic well-

being of the fout:4ig carriers.

efinitiong

IB Iefore proceeding to doescriba the nature and intent of

the research, and to Insure no misunderstaudinZ In created,

the following definitions &ra presented and will be used in

the remainder of this disaueiaLion.

Integr tion of trsnspartaticn f'r.:v ivolvcs the purchafe

or acquisition of t. cransportaLion firm of one mode by a arans-

portation firm in another mode. A railroad acquiring a truck

line would be an example.

A transport4tion n im a singla !irm which owns and

operates entiticis in more than one mode of transp,rtation.

Hence, the integration of transportation firms into a

transportation co.pany involv,: ia;ue, of -hc. common crner-

!.h.ip or more Lhat ote modU of LraneporUL IaL, IN the

transiporLotti. I :tcaturc 0he terno _.o'r o:erph i a:nd

Lrnnporcrtl,, cpir y.- ar, teed ftiturchnt.:;b,Ihy and Ohlin

practice wil). be followed in thit; project.

.. .: ! ....

Coordinated transportation is defined as "a point-to-point

throuih mover ent by means of two or more modes of transportation

on the basis of regularly scheduled operations."4 Thus coordi-

nated transportation refers to a particula: type of transportation

service or product" which is independent ;f the ownership of

the mode.,. For example, a regularly scheduled piggyback move-

ment, a .r..ck trailer move; on a rail flatcar, is a coordinated

transportathon movement if the railroad ons the truck trailer

or If a trucking firm owns the trailer.

The Economic Issue

As mentioned abovu, this dissertation will undertake to

exemIne tlc cconc, c consequences of estab:ishing transportatlcn

companies. The formation of a transportation company, resulting

from combining two or mo-e carriers of dif.crent modes, could

In-olve economies from two sources. Common ownership could

leas to economies of scale and potential economies involving

the pssible reallocation of traffic from high to low cost

modes.

Economies of Scle

As alluded to In the argu:Aents for and against establish-

ing transportation compa:tics, most, If not all economic arguncnts

4Nichols A. Clislc'sky, Jr. , :n Anal"j:s and__al,_:J.on of.hc _P¢ '1 l ' . , . ,~ "' -,.., 2- • , , . .

d fI ..s~L; ic, ~4u . .itc '~J~0i ,~L,-L,;i' L rsi .y, i960), p. 10.i oduced f,orn

Lbei available copy.

- - ... .... . ... . - - - -

-6- A

on the common ownership question, are centered on the question -4

of the extent to which economies of scale can be obtained by Ithe merger of two modes (rail and truck). Co~ents such as the

following are typically made by transportation economists.-

The economies which may be realized from common ownershipare dubious, to say the least, as is evident from theearlier aaalysis of economies of scale in the variousmodes of transport ... in addition ... the competitivenature of motor ... transport would make it impossiblefor then to abscru any significant amount of the rail-road burden.

5

From the positive side,

In summnry, ... integration can lead to economies in theuse of administrative personnel, maintenance, perionnel,labor of all kinds, equipment, aad capital facilities.Many of these lead to better service thrcugh speciali-

zation of labor or capital and,,at the same time, leadto greater profit for the firm.3

As can be seen, the disagreement centers on the question

of how similar are the operating functions of the combining

4modes and is there enough similarity to aliow the transpor-

tation company to centralize functions ano eliminate duplicative jfunctions, and so forth. These kinds of questions will be

addressed in Chapter VI. As will be pointed out in Chapter VI,there really has not been enough research done in this area of

the transportation company concept to answer the question of

!5

Dudley F. Pegrum, Tra:asartaticn Econ.c-.ics and PublicPolicy, Ruvised Eidition (icz. ooc, jilinois: i'ichard D.Irwin, Inc., 1968), p. :, 63.-II. .. ,:e -nd ''irtin r. Farris, Dcm,-tc "ranc:.r-

: r~~~~~atio.. r,: .l ..r. . .. i.c , ( ::ci, .s : iuul:,9,, 326.

I

-7-

whether or not there are substantial economies of scale to be

obtained from combining different modes. Therefore economies

of scale will not be rtly considered in this study. The

effects of not directly considering economies of scale on the

analysis of the results of this project will be explained in

detail in Chapter V.

Economies Resulting from the Reallocation of Traffic

The major focus of this research project will be on the

analysis of the potential economies of common ownership

involving the possible reallocations of traffic from high to

low cost modes. An explanation of why it may be possible for

a transportation company to obtain economies frcm reallocating

traffic from high to low cost modes deserves some attention.

Currently, carriers have the leeway to price services

be.ween out-cf-pocket and fully distributed costs 7 under

8current regulatory policy. Assuming firms are profit

7/

7Out-of-pocket costs is a term used in transportationreferring to the added costs incurred in performing anadditional service. Fully distri'uted cos:s are the total ofvariable costs and a prorated portic;n of applicable fixedcosts including a profit allcw'ance for a transportation move-ment.

8The Passage of the Transportsion Act of 1958 added para-graph 3 to the Rule of Latc ..a.i"-, (Section 15a) o' the Inter-state Co:m:jercc Act. In partic u.lr, the Congrcss declared that."rates of a carrier shall not be held up to a particul:fr 'ACal _

to protect the traffic of .- othr node of tr.nspor:ticn,giving due cnnsidrL'.io: 0 tI ccC'ivs ' the 'iLo.al0

transportICai polic,. lccharL i:' th's act.' n Lfcct, this

gt.We@v@ off th eet #are "d $two thave is axess capacity,

is tkh toil4roa thAt oinga modal carriers will carry trof tic

of 1"S is it cftrfl~uto# .o..'thinu to mrgin. Iscauss eachad. %a" d1Eferet toot characteristics (differing amounts of

ftiadl aftd vorl1Iiio costs) and :Oiven that ctarters are profit

(Notmsior., there omutopo in~ t;vanspowt~tibnt today a situation

imW ~t1& to e tho1.0A. dirocted the Intorotata commYrc Con-4411t (14e.C.) tG bam rates f -r ivorvicto providod by carrikirmI 01~ 4~.V qf V(M~A of Oto sittitiont.

ortior tou 11ie , .cc diLe0 01.1L rAL.Lc hould b6rlowit Ih.Mi tv Wor~cd Lhio cari.'rN * fair o4por-Ii IIt t6~ i oUt*-t. in i f taWoro juat :A d

tho Otfcii a nv# ~i w~til Iive or -;o wtS("o 04~ fvaic;kri Of 0'QV to th~of eLIr c;%rricru,

lib. V turi ' fo~vr t 90Rl of Wtt.0u wHoi- a to10LAul fi.. 0,Mrui. NOW~ YOV' U01 HAW A11.4 Harford~ R,.Co.,4~

eiI~eI~l~ ~IW4ik4' WIA 00vJc1w tho WA:?. (t ) thr Suri Co-.rL

dv ile Imi d o irunvIais A-U'tv4o (;UuL ac~tiwa ful ij ditri-

diaMiuiiiij Ac.H~ Ai i~tic ;n,"r fi 10 ilfi .aOfid litLifl :.tz~tt

IOUP ~ ~ ~ ~ ~ 9~a9 Iii GYFug.4igt ro %I)V- ii .. L o tlit 0Cv~4'lC

forllq -I "4 14i ,--& 0 yr wila Y he(2) 11141 C 41

911 ttif bifi 1R rtg i 4ttV.l#'~ ~ LU rt" 141r l41141buiA eletN pij."61dtli-fiaAre no ok,. to ullydii~vi

i"i

-9-

where carrieus of one mca are compstinS for traffic which can

be moved mor profitably at the sam or competitive rates by

carriers of snother mode (with dLf!aront coot characteristics).

This point is substantiated by past chsirmin of the I.C.C.

Howard Freas when he pointed out in 1958'

However, when competitive traffic is hauled at a minimsm 4of profit by carriers whose costs are relatively high, theAlow cost carrier who at the same or lowor rates couldprovide the service at a revonable profit is deprived ofthe business, By having the hi h cost carrier perform theservice, the overall charges to the public are not reducedas other trafflc must bear a disproportionate share of thetotal trAnsp. rtaion burden. Thus, t'he public is preve.tedfrom receiving the benefit of the mor economical se.rvice.. 0

Thus it in evideut that thar. exists the potontial for a transpor-

tation company to ach.eve economies by realIocating traffic from

high to low cost modes. Whether this is p~ssible or not deperd Ion many factors which must be analyzed.

To date there h n been no quartitativw microcconotic analybip

of the transportation company concept. Peter S. Douglas has

supported this findint and states in a recent article, "No pro- Iponent of 'c.cwnon ownership', however, han yet undertaken pubi.icly

to identify the economic forces that tight make coordination of Isoparate mod es luss Coitly wider a single : eagL-mcnt than undcr

separate mL3LIgmnton."I

US. C'ofl:vjHu * Th itu ,ce[a.tti on Cc-.'.erce,, " -ic'n lTrn~nn ort.- I .~' I ':' ' (Tha Uoyic ',ort), S. )Pt. 445, t671'

Conn.rutm, Jat u;.IU r,, 19,I, p. 217.

'€ Lcr ,S. Dou'L11g, "[he Lc .nuc r I c I rrul a: 'C .O of:Ownvt,1 ,I{ " [,(C,C i'. l:i.,. ! J u :.l .?' J . -, ..:1'') . ''' .-

-10-

Purpose of the Dissertation .

In the .'tght of the previous disoussion on the economic

issues of thb common ownership controversy, the purpose of this

dissertation may be stated. The purpose o this research project

is to examine the economic impact various cocbinations of paraeters

or test factors have on a .:ansportation company formed from single

modal carriers. In the accomplishment of ;his goal, the research

project will attempt to provide Information on the cauie and

effect nature the parameters have on the -,:onomicC performance

of the transportation company as cotparod to Lhe economic per-

formance of the single modal carriero of which the transportation

company Is comprised. A secondar!, but ir.portant, goal of the

research is to provide a basis for future research efforts.

The Research Approach ,

The reeearch approach used in this project will be to identify

the exogencous controllable and uncontrollable variables, the

ondogeneous variables, the constraints, and the relatlionships

rbetween them. Taking these factors into consideration, a

mathematical simulationi model will be constructe:d. Once the

Mathematilca odel ha'a been devulopd, a hypothetical tra:nspor- j

tatLion systuc° will be analyzed In w'h'ch tl,. parametera and con-

atraints will be allowed to syst-c:I.a1.cally vary to dctermi.e1u

F what effect .ach has on the solution of the trodel.

It Hhou!.d 1e polt tJ cit that hIS Ly,)u o all apprach tc

the probl.,,i rg niw-casrv ior the fol.o-,J, rL,, s; : (1) h : ,'!1

kI

. *. . ' : - - .; " 'l z : . : : - ' -- - '-- .. . . . - - " ' - ~ - - - ~ - - - - .

it I not necessarily easy to abstract the real world into s)mbolic

models, it Ii the only practical way in which the operations of an

integrated transportation company can be compared with the operations

of competitive single modal carriers made up of its corposite com-panies.

There is no data on which one can draw which compare the operations of

a truly integrated transportation company with competitive single

modal companies. That is, while there are some transportatior.

companies in the United States, they do not compete directly with

single modal carriers with exactly the same route structures, loadiI

factors, management skills, and so forth. (2) As one might guess,

the carriers, shippers, und regulatory authorities as a group are

not willing, for their own parochial reasons, to experiment with one

segment of our transportatlon system to see If the concept really

"holds water". Furthermore, even If one experiment did work, this fact

would offer no proof that the concept would be valid under all ccn-

ditions. Thus, one of the benefits of utilizing mathematical simu-

lotion model]; is that It is possible to manipulate the pe.rameters

Involved which make two situations differe.t to deterirdne the outcote

on the solution at a rr'asonablc e:xpense. (3) The formulation and use

of a simulator to analyze this probl, area has never been attcpLed

before and hopCfully this approach will be u.,,d as a building block tc

extend quantitative ,.2thod.i to aid carrier n anagezients.

The data used In thils tu y 1: ll be hypothietical. y(t Will

reflect aCLtuOI diff!rcrice, In L opera.l:g cli %rlct.cri.tc- between

I.4

~ ~ Ii 7 - ____ ____ _'_

= - .. ...... ..... - . ... .. .. "---- . - . .... .. -

nodes. Realistic hypothetical data serve :h,, purpose of this Idissertation as well as real data, because they will be used as

Scomaon denominator for comparinS the same transportation modes

opetating under different organizational structures.

With respect to simulation Perse, it has becn definedas follcws:

A simulation of a system or an organism is the operationof a model or simulator w:hich is a ropresentation of thesystem or organism. The model is amenable to manipu-lations which would be impossible, too expeisive orimpractical to perforn on the entity it portrays. Theoperation of the model can he studied and, from it,

properties concerning the behavior of the actual systemor subsystem can be inferred. -

In essence, simulation can be viewed as an experimentalmeans fcr generating an artificial history of a system forpurposes of analysis,

12

Thus one of the distin uishin; characteristics of simulation

studies is their reliance on the model-buildLng approach to examine

problems. H1illier and Lieberman have put forth this proposition

as follows:

...simulation typically is nothing more or less than thetechnique of performl-n;; saz-plin& experiments On the modelor system. The experimen-.s are done ou rhe model rethirthan on the real system itself only bLcaAi the letterwculd be too in'onvcnient, e:pnsive, and time consuming.Otherwise, sim-ulated e:pcrJ.rtents should be vewed asvirtual.y indIiFLinuisihbi c from ordi.-ary StOList.icalexperlmnts..

12 Nori:r, L, Chervany, A S.::. 1 .at.:o; .. l'._ of C ..sh Flc.P tt L :!r n. ,.! :.' .I L[ : : ... 1;-,- Ar. .. .. " " .' i:', . D., ; A .

1968), p. 10, quot!; , fr; Mj n SuLf,., -i.;ulation of the li'"Aim-ric :, C ; i:9 )bir .9$D), p. 5C9,

1 F}ri,. d r i,- 3 '.; ' u L : C-.-..J I -- .<: c~ .,. n, I' 'r. !- : .t o.p'r o ...... " ( *., Y..nc co, CA i :n' a: I:cl.: 1-;;.:,

_____,1; ,7 I .,,'}

• ?i- : - i ._ -- - 7 ;/ / ]- --- = . . .......... ... ........

-13-

It must be recognized, however, that while there are certain

advantages In utilizing this model-buildir.g approach to analyze

a problem area, there are corresponding disadvantages. As Hillier

and Lieberman put it:

Hathematical models have many advantages over a verbaldescription of the problem. One obvious advantage Is thata mathematical model describes a problem much more concisely.This tends to make the over-all structure of the problem torecomprehanaible, and it helps to reveal Important cause-and-effect relationships. It also facilitates dealing with theproblem In its entirety and considoring all of its inter-

relationships simultaneously ....On the other hand, therb are pitfalls to be avrided whe:i

using mithematical models. Such a model is necessarily anabstract Idealization of the problem, and apprc,:-iationsand simplifying asuumptions generally are requlred if themodel ia to be tractable. Therefore, carp must be taken toinjure that the model remains a valid representation ofthe p1ol m.14

With th,.se advantages and disadvantates In mind this model-

building approach will be utilized to achieve the objectives of

this dfssertatlon. The study will be undartaken in three distinct

phases. Fir.L, the model used to analyze the economic consequencesA

of estnblishlng transportation companies wIll be presented.

Secondly. the test factors and porfonnancc mesures utilized In

the st.id,; will be presented. In this phase of the study, the

relationship of tie rode! to the real world will be critically

diucussed. Thirdly, the evaluaLloua of the results of LhL sIzu-

latiors will be presOiied as well as the irplicatlons the study haf

on the rc-.]J world and future re:arch.

14Ibid., p, 15.

-14-

Contribution of the Dissertation

Speculation and intuitive belief hav. guided the Arguments

concerning the establishment o lormation of transportat!cn compa-

niem. Investigation, not speculation, Is required if objective

decision criteria are to be established, as to when or under what

circumstances transportation companies would be in the public

interest. As indicated above, developing criteria for this purpose

is A multi-dfmCIisIonl undertaking. This dissertation is con-

cerned with one part of the economic feasibility of common ownership.

The dissertation will answer such questions as under what operating

conditions, :raffic conditions, and cost conditions will the transpor-

tatlon compa-y result In lowar costs for transportaition users and/or

greater profits 15 for the industry than by kceplng the individuil

modes separate.

The resilts and methodology of thiE study could hopefully be

incorporated in a rIgorous u-;temS analysis of til coiacept which

should be performed by the I.C.C. or Department of Transportation.

I

The major thrust of the research could then be thought of as a

necessary part of a vsysteta analysis on thc co u omiership contro-

versy which identifies sone, of the economic irpicts on the carriers

involved in forring transpor~atiOn companies, The Infornation-_I

1 5Prof I tability w:il be rcJ9urcd In Lerm!i of the Lotal contri-bution W,:Iu to 'in ;d/or cc-:.'Al cc1 LS tclud t;prof it a.,:r, -For v deta lcd uxl);1-t1I ,1 of vity hilh r.c.i,;urLcn, LWaS u-cd .ce

Infra., ch.'ptr v, p'. ;,0-l62 II

___ A

I ' -;15-1

generated and mnttodology used i t this dissertation could be

incorporated as part of the analysis nucesiary to determine if such

companies would be in the public interest. The I.C.C., if it were

to attempt such an analysis, would also have to consider not only

the economic aspects of the problem, but also such factors as to how

such companias would affect shipper convenience and service, effect

on carrier enployeas, legality, desired leels of cozpetition,

regulatory problems, and whether or noL such companies could ba

effectively organized and mananed. Only after all these questions

have been reziearched can appropriate decision criteria be for.ulatc.

as to when the formation of a transportation company would be sucially

desirable. IThe major focus of this research will hopefully provide a

now pearpective and a foundation for research on the question of

the transportation company concept. The research should, however,

provide guidelines as to under what circuwr.stances a transportation

company will be more profitable than single modal competing carreris.

The methodology utilized in this study could also be used, is a

foundation, to develop a ratioral basis for determinin when carriers I

should provide coordinated or sIncle modal tranrpoitatlon sur'Icu.I

Finally it is hoped that thiu disertat~on wi .1 be used i. a

building block to ctvnd the applications of quentiL1LJW( IOetl,1

in carrier n.igU ,(:it. -

A

VAOrganization of the Dissertation

The rem.Inder of this diusortatlou consists of six chapters.

Chapter 11 Involves the formulation of the model. The chapter

deals with the definition of the relevant decision variables,

explanation of the form of the objective functions, and statement

of the set of constraints within vhtI:h the simulation will operate,

The chapter alsm presents a diacuision of the scope of the reisa.rch

as well as a description of the pertinent assumptions.

The first part of Chapter III identifies the parateters as

well as the random components of the model which will be varied

in the analycic of the model to determinc what effect each haR on

the solution of the model. The re alrdar of the third chanter

explanu hou the cost data will be genrat.d. The CoSLirL ci ori.es

for the modem are established and an explanation of the manner in

which the tent factors effect costs is given. Following this

explanation, the Li,.sportaticn systae which was modeled is presorLed.

Chapter IV focuses attention on the nature of the experl-

mental deeign atid the analytical und statistical methodolog7, used

in the dissertation. The chapter discusses the problez of realism

in the simulatlon and prcscnts the pcrfor:ance reaeurLli or sudmary

statistics which are used to describc thc resulLs or the nlnu-

latioi. Thu chapter also discu'ojcu the nature of the particular

exp'rirontal def;1 i utilized In the study,

The fifLth chapt,-r .L and ana)yzc,; ijh rcrult.s of thu

a I m' at I en". T ! . . " 's ICaLe-:'A which tcst faCr.ors ni r 'fIl fI-

A.,

-17- "

cantly afiect the economic impact of a transportation company

via-S-via single modal carriers.

Chapter V1 Identifies the other aspects of the comzon owner-

ship controversy that were not considered in this project and Ni

discusses their relationship to the resear:h accomplished in thir,

study.

Chapter VII discusses the results and Implications of the

study in the broadur frarawork of the controversy.

B1

I

I

I

.i

III

A

-18

CHAPTER 11

LI DEVELOPMENT OF THE MODEL

Very broadly stated, the purpose of this chapter is to:

(1) develop the foundation for the simulation model utilized in

this project; and (2) present and discuss the simulation model.

The chapter begins by identifyi g the scope of this research

effort. Next the important assumptions which are made during

the conduct of the research ate discussed. Following this

discussion, the key variables of the simulation mudel ar. ;re-

sented with the purpose of introducing the model in an overview

fashion.

The chapter will then present a description, in mathematical

and verbal terms, of the nature of the decision environment

within which the shippers and carriers operate. Attention will

be given to the interactive nature of the process by which carriers

mak , quipmerit allocation decisions and shippers choose t.ethols of

trat,portation for movvments. The nature of the manner in which

carriers determine rates is then presented. The simulation tiodel

which will be used for analysis in the project is then presented

and described.

I Sc_po of the l'c5carch

The American transpurLation systtn, which Is coancened w;ith

the movement of f r o lfght, is s;o pervaiwv that it i: a1r.o. t b(:'Y'nd

Ii

-19-

imagination. .the system is composed of thousands of origins

and destinations connected by millious of miles of roads,

railroad tracks, canals, rivers, ppelines, and air routes.

There are fivC modes of transportation concerned with the

movement of freight. Thesc modes involve tiotor, rail, water,

pipeline, and air transportation. These modes literally move

billions of c-n-miles of freight each year.

To simplify the analy.is and to limit the scope of the

research tz, a mdnageable size, th3 research will restrict the

tumber of origins and destinations, the nui.ber of foundlng modes,

and the number of conodicies considered.

Because there are many possible cczbinations of modes whi:h

could be furned into a transportation comp.ny, the research w%-Ll

consider the two modA corbinarion which would most probably

have the most significant economic. i=pact on both the shippers

and carriers coicerned. Air carriers and freight forwai'ders are

involved with a relatively minor a-.ount of the total freight

movements. Domestic water carriers are limited to their geo-

graphically controlled route structures. ?ipelines carry n very

limited product line. The analysis will therefore limit the

system considern-4 to the two modal case involving a railroad

and a trucking cornpany, a.nd a traisrortation company which operates

Ii....filiiilq-ii+Ti'iiNMI",1 e

both the rail tAn highway modes. These two nodes are by far

the most important modes involved in the movement of,,freight.

1

In simplifying the analysis by considering only two modes,

the results which will bu obtained are restricted to the modes

considered. This is so because each mode has different economic

or cost characteristics which must be considered in the research. -

The extension of the analysis to transportation companies con-

sisting of more than two modes is conceptually straight forward,

although not necessarily easy to accomplish. To consider other

two modal transpurtation companies composed of different modes

should be a simple process once the methodology is established.

This study will address itself to only: a portion of the

total rail-truck market. The study will focus attention on

ccmmon carrier truck and rail movements including Plan I and

2Plan II piggyback mnovenents. Other forms of piggyback move-

ments will not be considered.2 Furthermore, the analysis wjll -i

also consider only one product. This product may be considered to

be the class of cormuditics which are subj. ,ct to intermodal

competitioi for movement.

I1n 1969 the total estimated revenue of all regulated ire 4 6rntcarriers was approxiCrately 26.4 billion dollars. Of this " alregulated railroads and .ator carri. r. ace unted for 92."V, cfthis total ci 24.3 billion dollars.. "America. Tric!king Trcnds 1970-71", Departments of Pecarch and Tra:nspor Ecao:crjcs and Pub -i. .Relations, Ar.crican Truc*i .. socia.tins, Inc. ('-.azhingto., D.C.), p. 16.

2Plan I piggyback rovements uti2i;e co.-.n carrier truck

trailers on rail o,-d f!'Lcars. Plan I pM;2.ac, .ne-.e.-ts 'involve rail o r:',! il r:iler. and flatc r.. 'hor, are other cr.sof pigg:'bazk ;:o.'.- -. iii )";o!'. shKi:.r -:::d truck tr . r ,and stii2. ot t .-rs "..'hicli c,, r Lo f, 'i;;h or.'a rr.

.I

-21-

Restricting the analysis to one product with average commodity

characteristics does present some difficulcLies. The commodity

charactaristics of goods does affect the cost of movement. Once

a system is solved for the average commodity, it may prove worth-

while to adjust the costs in the analysis by the amount these

characteristics cause the costs to deviate from the average. If

this is done, the transportatior firms would have exact commodity

break points for the modes over a cartain segment.

The size of the transportation market which is subject to the

intermodal conpetition is an important factor ior determining if a

transportation company has a greater profi: potential than two single

modal carriers. If all traffic were subJe.:t to intermodai competitlon,

there couid exist many possibilities where a transportation company

could generate a greater profit contribution than the competitive

single modal carrier by allowing each node Lo be used in its most

profitable market segments. If there were no traffic subject tc inter-

modal competition, there would be no increased profit potential for

a transportation company by allocating each mode to its proper

economic rcle,. There may however still be some economies of scale

involved in this situation.

To the extent that service differs between the modes, the

effective size of the amount of traffic which is subject to inter-

modal competi tion is reduced. For ivstainc, if speed between

nodes were substantia.ly d'fcre;t bct*.:en two pointE cn" this

was an irpcrLant dccision v,.rlable for a s'.iip&r, this facitr

• A

6ftU redia &Me amout of traffic subject to Intermodol com-

1*01ioo. Visa offoct this would have would 'be to redu(.e theI

Afta IN Alt *ft64ttva savinigs are likely to result by the

1*elitomfl of traffic to the most profitsble nodes, This

tyf# &f oituttio mild result also 1k a shipper'a logistics

*yatea vie tat up only for on. made of transportation, regard-

I##of wtioti,or or not it is ths low cost aarrier by which the

IrenOPenrition tconpeny would liku to m~ove 'ie s hipmtunts. To the

usoit tib typ of vituaLton prtvontA the movomont of goces

Ily OF IW C cuil carrier, 010 lads will be thle economic Impact

csitod Uy moiingt eaiportation companioi. Theme types of

Oitififo~ 665 10 ir6crpcrAtsd lNO the antldi. by conbtratrning

the *A'.ner iii wich1 movieetnto con br nedo Voitwt-an oigins and I

kstesifing the flut-6g of origin and Jestinaion pairs to

thai which Its mi~n o~I doos not really affect the generality

of tile reauha., For insitakne once (lic broeA points where modal

aIlji olv"B'I 'j"'cIt arc- Novid on~ a partictilar 11CLWork sag-nnt,

oil Iic. e#IIn and d-astinmiown paid bnuweori those pointis should

U*sivoed 10-1, ho ae' i.xdum %r miJt1 oda~l .L6U.-

011d i~~. (OL I, JIMI L1101 WI .1h I peced o!i the anelys1:j

r,fil1 Ili OIb fl~ncta "Li1u. Thew HIIL10bif Will be. a short-

li IndL, 0;4! 911iltioil tkuU14 wlllcJ will

1I II ;- Wi~j~ l lt 11U :.(h L 4h, Wci. a c ',ptirl~ g th10 oco nl J C

,______ . ,--.- , -..- . t I I I i

It fl - - -. .- .

-23- 1performance of a transportation company wich a single modal rail-

road and trucking company for a single timg period. This time

period may be thought of as being one day. The analysis will

not suffer due to this limitation because, as will be shown in

a later section of this chapter, the decision processes which

are being simulated remain the same from time period to time

period.

Asumptions

Now that the scope of the research has been described, this

section will identify the pertinent assumptions which will be

utilized in the formulation of the simulation model.

The study will assume that the appropriate regulatory azency,

the Interstatc Commerce Cor.'.ission, will continue to safeguard

the public interest. In other words, it will be assumed that

the activities of a transportation company, as well as the single

modal carriers, will not be allowed to run contrary to the public

Interest by &uch actions as charging exhociritant rates or dis-

regulatory supervision allows the author to assumc that the rates

charged by the singi. modal c.rriers and the transportation

company will be deter,,Ined by a con sstenL policy which will be

cxplaled in a later section. Further.tore, it will be asisuned

that the I.C.C. through it.; YaLe Ma!'1n? policy will contrcl the

average opcrating ratios of thc modc . The I.C.C. w:ill thuS

indirectly c ,, trol the /!-,,Lr.a:c rrIC of r t r for ! r. h i. ;

I

• -

-24-

Since the time frame necessary for'tha analysis is a short-

run analysis, one can assume that the physical plant-of the

companies is fixed. ThiL is, the number of: tractors, trailers,

flatcars, locomotives, terminals, line miles of track, road milt.s, I

and so forth, will be considered as constaats in the analys's.

Furthermore, the technological state will be litited to the

current state of the art.

For certain movements, the physical characteristics of

products restrict certain modes from carrling them. For Instance,

the shock resistance, or size or weight of some co=modities could

limit them from being moved by any one mode. This factur reduces

the number of combinations of coordinated .ovezents to those

Involving the xodes which are physically capable of carrying

the goods. The moda.l developed below will consider the co~odity

to be capable of being transported in a standard railroad box:ar

and/or a standard truck trailer. Piggyback movements will be

limited to the standard configuration of one or two standard

truck trailers on a flatcar.

In addition the assumption will be mado that when the

service characterlstics of the modes are equal, shippers have

no real preference bcLwceu the modos If they are physicall:

capable of dealing with cach r.

It will also be ass%:cd that shipnents subje't to logistlcs

constraints will be movN'.1 before the ca'rior s ::l,., cerzt itrvc Itratiic. 1hi . ri:,t.alc. eca rc. arr I ,s ti ht be e:.:jcL.d

JII

-,: - --- - - - --

-25-

-* to satisfy a guaranteed market before they would enter into

the competitive market place.

Furthernore it will be assumed that the transportation

:; company faces the same demand pattern av the two single modal

companies of which it Is composed.

The final assumption which will be made in this project is

that the transportation company can be effectively organized in

such a manner as to achieve its objectives. The objective

function: of the transportation company and the single modal

carriers will be presented in this chipter.

The assumptions utilized in this study will be critically

analyzed thruughaut this thesis to detarmine the impact they liave

on rcleting the results of this study to the real world. In

addition, Chapter VI devotes special attention to the discusslon

of these dssumptions and their relationship to the real world.

Identification of the Ircortnt Variables

Now that the pertinent assumptions which will be utilized

in the formulation of the model have been stated, this section

will identify the important variables -With which the nodel -All

be concerned. Vic purposu ouf this secLoae is to acquaint the

reader. on i. very general bi s,vIt, th1( "input and output"

varlables used in the jJr.ulaL1v:t analy:ls., This rater.al J5

presented at this time to facilitate tl- dcr2..din, of "he

matliematical ,"of t.:c .... fsJo.i t.

.,....u . o ie 1t; o .,v r ':t t l t

-26-

simulation model which will be presented In the next two

sections of this chapter.

There are two classes of variables which the models, to be

developed in the next section, manipulate to arrive at a solution.

These variables are exogenous and endogenous variables. "Exogenous

variables are the independent or input variables of the model

and are assumed to have been predetermined and given independently

of the systen modeled."3 Exogenous variables may be subdivided

into controllable and uncontrollable variables. "Endogenous

variables are the dependent or output variables of the system

and are generated from the interaction of the systers (model's)

operating characteristics." 4 Endogenous verlbles are synonomous

with the decision variables in this study. The parameters in

the moL9s formulated below may be thought of as specific values

of the exogenous variables. In other words, the models will be

solved with different parametric values. These parametric values

may be thought of as "snapshots" or a specific value of an

exogenous variable.

3Thomas It. Naylor, Joseph L. Ballntfy, Donald S. Burdick,Kong Chu, .no.!!t__rSimulationhchni uo,, (New York: John Wiley& Sons, Inc., 1967), p. 10.

4Ibid., p. 11.

I

-27-

Exogenous Variables .

Attention will be focused first, on th. exogenous variables

and In particular on the uncontrollable exogenous variables.

The cost of movement between any two given points may be cou-

sidered an exogenous uncontrollable variable which is a

function of uany other uncontrollable variables.

There are many variables which affect the cost of movement.

These variables may be divided into two categorits--those which

are concerned with the characteristics of the commodity being

moved and those which are concerned with the route over uhich

the movement will take place. The following cormodity charac-

teristics all have an effect on the cost a3SoCiatOd with the

movement of c particular commodity: loading characteristics,

susceptibility to loss and damage, volume of traffic, regularity

of traff~c, Lnd the nature of equipment required. The folloving

route characteristics also directly affect the cost of move-

ment: distance, operating conditions such as geographical and

weather factors, and traffic density.

Another exogenous, uncontrollable variable Is the demand

for transportation between points. While it may be true that

Individual carriers can affect the demand for their services over

See for instancc, Cernane, Glagkowsky and llcskett, Oyc_.,Chapter 4.

I

I • ! i I - i • ml• n i 'a III • ' i~ -m I I m 'l l g i,..= ......- ... .....

-28-

time, the short-run nature of the study has the effect of fixing

the demand for the carriers invclved. The same is true for the

amount of traffic which must be allocated to one mode or the

other due to shippers' logistics constraints.

It sh..uld be pointed jut that the nature of this analysis,

being a (very) short-run economic analysis, has the effect of

adding certain variables to the list of uncontrollable exogenous

variables which over a longer time period ,Mould be controllable

variables. The short-run analysis also ha3 the effect of more

or less "fixing" these variables at the values they have assuced

at the "moment" of observation.

The controllable exogenous variable ia the anal:ysis is

the schedule of carrier operations. This variable is a function

of both cost and demand. For instance, rail management may

institute a policy of moving freight only when the train reached

100 cars (approximately). This may have the effect of limiting

service for a certain comunity to once-per-day service. In

essence the scheduling activities of the carriers create the

amount of capacity which Is availabla at each point for move-

ments, although total capacity is fixed.

Endoge',ous '. rlahies

17he model3 for the single mcdal co.pa{c',tz and the transpcr-

tation co:.pany which will be prvcsentcd ir the next tecticn, prcduce

a number of cndo:c..:, or d,';ic, virl I!,:, Th -

arc IV :;!. -it 'A Iuc...,.d ,.' .'h C .) '.,J.L will .! :v.c d bc(t'cc

-29-

origins and destinations by truck, rail, or piggyback, and the

amount of capacity the carriers will allocate to the various

origins. The rate at which carriers price their services to

shippers is also an mndogenous variable, although the levels

rates can assume are bounded by reguiatory restrictions as will

be explained in the next section.

The rate shippers are charged is dependent upon the cost

functions of the carriers. The rate between two pointc is

dependent not only upon the co..odity and route characteristics

of the movement but is also dependent upon the competitive nature

of the movement. As will be discussed below, carrl.-q have n

certain amount of leeway in pricing their jer-Aces when con-

fronted with competition.

The amount of Cwt. moved between origins and destinations

by each of the methods of movetant is dependent upon: (1) the

demand for transportation; (2) the amount of traffic which is

constrained to move by each mode due to shippers' loIstics

constraints; (3) the amount of capacity carriers allocate to each

origin; (4) and the manner in which shippers select the rn.des

for traffic which is not constrained, i.e., competitive traffic.

The equipment allocation dccisions of the carricrs is dcpe:-dent

upon the amount of contrlbutlion to fixed wd/or cov::on costs

including profit allowance each novcI.eCt makes. That is the

carriers will. allocate thuli' equIpr-nt in uucih a..:u~cr as tcc

maaximix( the'.r cxpt:t Cd co t l i L h i : tt COt.Lr"l Ut .. I

4

( -30--:

U2

is dependent upon: (1) the forecasted demand between origin and

destination pairs; (2) the cost characteristics of the move- A

mnto; (3) the aount of traffic constrained for each mode by

shippers' logistics systems; and (4) the competitive nature of the

movements

The Decision Environment

From the discussion in the preceding section it is apparent

that the decision process of the carriers and shippers are inter-

active to a large degree, The nature of this decision environ-

ment will be explained in detail in this section.

The decAgion environnent within which the shippers and

carriers operate can be viewed as a constrained minimization

problem which interacts with a constrained maximization problem.

In other words, shippers will choose the least cost, as measured

by rate, method of movement between two points given that this

method does not violate the shippers' logistics constraints.

Operating within this decision fra.ework, the carriers will

attempt to mtximize their contribution to fixed and/or common

costs including irifit margin by allocating their equirnent to

the most "profitable" route segments. The an.ount of capacity

available tU shippers at oach oriin then is dpend,'nt upon the

equipnnt allocation decisiuns of the carriers.I.

-31- jIdentification of the Symbolic Terminology

The nature of these interactive decision processes will

be examined in detail following the identification of the neces-

sary symbolic terminology.

Let X be the number of hundred weight (Cwt.) moved fromij k

point j to point k by mode i. The designator i equals 1 when

the movement is by common :arrier truck, 2 when the movement Is

by Plan I piggyback, 3 when the movement is by Plan II piggyback,

and 4 when the movement is by rail. The designator j represents

origins; k destinations. Thus the variable X would indicate

the number ot rfwt. P-ved from A to B by truck. In other words,

the X arc the decision variables which in the sclution wllijk

indicate how much product will be moved by what modes between

given points to satisfy demand.

Let rijk be the rate per Cwt. charged by a carrier for

moving the product between j and k by mode i. Similarly, let

c 6Cik be the out-of-pocket cost to the carrier for moving the

product between j and k by mode i. The i's, J's, and k's are

defined as above.

Let represent the demand in Cwt. for Lransportation

service bet'een the two points j and k.

Y ijp. is the C3paciLy in cwt. allocated by mode i for move-

ments between j and k. Yljk and Y is the a.ount of single

6 Supra, rhapter I, p. 7.

p

-32-

modal truckirg capacity available for over-the-road and Plan I

piggyback mo~ements, respectively. Y3Jk im the amount of rail

Eowned trucking capacity available for Plan II plggyback move-

merts. is the amount of rail capacity allocated for

movements between j and k. F is the imount of flatcar capacityR

which the railroad has available for both types of piggyback

movements. FT the amount available for Plan I moves is equal

to P - LEY That is, the railroad does not have to provide7,, 3jk

flatcars for Plan I moves. The railroad will only do so if it

is not using all flatcar capacity for Plan II. FR and F willR I

also be expressed in Cwt. capacity available by multiplying che

number of flatcars available by the capacity of two standard

truck trailers.

Wik will be the amount of traffic which must be loaded

ito truck trailers, eitler rail or truck company trailers, at

- movement to k. This is the quantity of product for which

shippers' lcgistics systems are set up for trucking operations

only. In other words, some shippers only have terminal facilities

for truck trailers. Similarly U is the amount of Cwt. -hich

must be shiped by rail due to Gore shippers' logistics constraints,

between j and k,

Z is the amount of product which must be moved betWeen ijk

points j and k by the truck over-the-road. This quanLity .ay

be thought, c as tie az.cunt of traffic ,hih .t bc "aovcd froter

than what, Lh... r.. or pi - oi,.s can offer. ,.aiu this is

a logtiL[c.al cor;tr.aint of the 9tnIp :rF, h-h -1',t be satif[.A.

J

-33-

The Shippers' Dilenuna

The nature of the problem which shipp.,rs face, that of

choosing a method of movement, is a constrained minimization

problem which fits within the framework of linear programming.

In other words, after the carriers have male their equipment

allocation decisions, the shippers must choose how to move their

products. The linear programming formulation for the shippers'

selection amcng alternative modes when faced with a system

of single modal carriers is presented belo,,.

Shippers' Modal SelecLiuii ModelWhen Faced With Single Modal Carriers

Mmn Z "krlJk ljk + r2JkX2 jk +r3jk'3jk+ r 4jkX4 jkljk

Subject to.

Demand satisfaction

(1) Xlj k + X2j k + X3j k + X4j k - d Jk for all J and k

Capacity constraints

(2) X ljk +X 2 jk Y2k for all j and k

(3) rx S FR3jk R

(4) kX^. sF - TEXj2jk R jk3jkjk jk

(5) X 3jk -Y 3Jk for i J and k

(6) X Y for all J and k4jk -4Jk

,*1

..34-

Logistics systems constraints

(7) xljk + X2jk + X3jk Wjk for ail J and k

(8) X4jk I Ujk for all J and k

(9) Xlj k z zJk for all j and k

Xijk > 0

The different modes of transportation have different cost

characteristics which are reflected in rates in the objective

function. Faced with these sets of rates, shippers will allocate

their traffic to the low cost mode, given that their logistics

system does not constrain their choice.

The demand satisfaction constraints--equation (1)--insure

that demand is met providing the carriers have provided adeq%, fe

transport capacitv at the various points--equations (2), (3),

(4), (5) and (6). If the carriers do not provide enough capacity'

at each origin, there may be no feasible solution to the above S

problem. In other words, the demand for total transportation

services may not be satisfied between all origins and dastinations, 4

if the carriers do not allocate their equipment in such a manner

to make demand satisfaction possible. The manner in which carriers

make t:eir c,uipmcnt allocation decisions will be discussed in

the follotwing section of this chapter.

*1.i

I

[ - . - - - ... . ... .. " ~ 4C.s.$&~. - .~......*. .......... ..-...- -- -

-35-

Equations (7), (8) and (9) constrain the solution to be

compatible with logistics systems of the shippers. Equation

(7) indicates that certain shippers have logiptics systems which

are capable of handling only truck trailers, although a decision

must be made as to how to move the goods--by truck, Plan I or

Plan II piggyback. Equation (9) indicates some shippers must

have their product moved by truck -or speed of delivery. Equation

(8) indicates some product must be moved in rail boxcars again

because of shipper logistics systems demands.

Similarly, the linear programning fornulation for the

shippers' selcction among alternative modes when faced with a

transportation company consisting of the same two modes is

presented below.Shippers' Modal Selection Model When

Faced With Transportation Company

Min Z UirX + rr + j + r 4X kjk rljk 3J k 4JI

Subject to.

Demand satisfaction

(1) Xljk + X3J + X d for all J and kljk 3jk+ 4jk Jk

Capacity constraints

(2) X +X -Y +Y for all J and kljk 3jk - uk 3Jk

(3) ZEZX jjk

(4) X4jk Y for all J and k

Fk= a

(6 4J j for all J and k

41 ~ j j for allii and kj

Ijjk "0

This torstiibaloI% to very airtilor to tOat of the two zo(.a1

#oVbtIIMY~s Poal On*C uiajot difference, howelitr, is thict the

#41of 4oftoiatt variablims X *;.4 Y are absent In the model.2jk 2jkI This arie iiwcuow Oie vraii~yortatioi coupany owne both modes

50thjt I'etiL I rpg"yt'#A 16 fdtontica to Pl.an 11 piggyback.

Whapio it tiw oowiorn c'uil Lin developed 'or this type of pig-y-

j beck wvpIent, but to kvv~y the intarpretau'on of symbols as easy

as 0oaeblo X~kwill bo ubud to IndicAto 4 pigeyback movement

Ily Ohe t r~biaopr teti ll c'Tpaly.

V~jatins 8.(2). (3) anid (4) survo the sAme purpose o

In the ktinle uodal mumpeotivio model, that is to inoure that

doo,4 to satisfied It therv ik adoquet(A capacity available.

j4li ii (5) , (6) taid ( 7) con!3 tra~ins 010 3olutiOzn to be cumn- 4

PAttblv Wi tit 5111Pp.firl' JrjI w~titn dUAandfl,

Tha1_ giv pr, 1 tornn buciut' o.' 1.heP - tructure have trivial

801o)t w *lhe, '11if 110L11il tU te prugr..1 will alwayn Invo1,C.

0hf PP"'" ' C11604~ Il- OR 10hu I/Cori L Inode Of trA11mpOrt fit ton given1

I\-37-

this is not ine-cmpat-ble with their respective logistics systems

aind providing that earriara' allocate their equipment in such a

tanner that t.is is possible. This fact will simplif; the for M.-

lation of the simulation model.

The Carriers' Equipment Allocation Dilemma

The above discussion of the shippers' dilemma is but one

part of the total riicLion environment in which the carriers and

shippers operate. This section will -ocus attention on the nature

of the problez the ci riers face.

If a transportation company is to create any economic benefits,

it must be able to offer transportation services at the same or

lower rate to chippers while maintaining or Improving upon the

profitability which single modal firms could obtain. As

mentioned previously, this would be possible if the transpor-

ta.ton company could achieve economics of scale and/or if the

allocation of equipment by the transportation company results

in greater total profits than the manner in which the single

modal companies allocate, their equipment.

If the allocation of equip-(cnt to origins Is different

between the two organizatiunal forma, i.e., single modal carriers

vis--Is a transportotion com ,any, his w.ll have an effect on

the profiLia'i.iLy of the two forr.s as u-cll as; on the price paid

for tranoipo tation servlccs by usterJ. It the single modal

Fiftuat oi: Lhc I .r, c.. r m'.U the , iL!.r.:.,d tvLi .:,Lc ic:cp..dent

l._!_t. , , . h

equ ]l,..', cuA

U -38- -trucking capacity to allocate each origin .qr truckinfmove-,

ments and PLn I piggyback movement.s. Similarly, the railroad

independently decides the amount of boxcar, rail owned truck

trailers, and flatcar capacity to allocate to each oeigin. The

transportation company, on the other hand, makes capacity allocations .1

as a single "profit center" which may result in different capacity

allocations to the origins than those that the single modal Larriers

make. This situation is displayed graphically in Figure 1.

Figure 1. Structure of Equipment Allocation Decisions

A ...... : Truckers

ru a

A *~ Rail

carrier'sshipper choice allocationof service of equipment

Single Modal Case

- IratspcrLa-tonA --- )Company

Ashippcr choP-c carrier's .*1location

of servJcc of equip.unt

Trd!nvport2Lion Ceonp,.nny

CA -39- "

The allocation of equipment to specifte origins by the

carriers det.rmines the amount of capacity available to shippers

at each origin. In other words, the solution of the above problem

places the capacity constraints, Yijk' FT and FR In the cost

minimization linear program of the shippers. Operating within

the framework where shippers are trying to minimize their transpor-

tation costs, the carriers wish Cn maximize their expected contri-

butions to fixed and/or comon costs and profit margin by the

determination of price and the allocation of equipment.

Mathematically, the trucker wishes to maximize

r ClJk)Y Jk + (r~Jk CZjk)Y2jkl

wher Y ' r And C are definetd above. The Plan IIljk' Y2jk' ljk9 ljk

piggyback movements are coordinated movements which r.ecessitates

the sharing or splitting of the contributions to zixed and/or

common costs and :.,rofit margin between the two firms. This

contribution we-ild have to be split between the firms on the

basis of the cost contribution made by :ach firm.

(r' - C ]k)Y2jk represents the truckers share of the expected2jk

conribution.

Similarly, the railroad wishes to maximize

11r2, - C2Jik)y2jk + (r 3 jk - 3jk)Y Jjk + :r 4 jk -c4jk~yq'jkl'jk J

r - C2jK"2jk Is the railroad share of expected contribution

2' C2 dy~j

ZZq

-40- A

resulting from Plan I piggyback movements. (r3jk - C3jk)Y3jk

and (r4jk - C4jk)y4jk are the rail expected contributions from

Plan II piggyback and boxcar movements.

Operating withip the framework of the second linear program,

the transportation company also wishes to naximize the contri-

bution to fixed and/or common costs and profit margin. Mathe-

matically, the transportation ccrpany wishes to maximize

EE(r'jk Cijk)Yijk

ijk

To accomplish these objectives the carriers must deternine

where to allocate their equipment. The carriers are constrained

by the total amount of equipment (capacity:, they ow-n and by the

logistics cor.straint3 of the shippers. ConcomLitant with their

resource allocation decisions the carriers must also determine

what rate they should charge shippers to u,e their services.

The ,arriers' Pricing Dilce.a

This section will outline the nature of the carrier pricing

dilemma and indicate what leeway carriers have in pric'r.g their

services.

In general, becau.qe the tw:o modes have different cost charac-

teristics, i.e., before differing anounts of fixed and variable costs, *he

out-of-pockct costs and fully distributcd :osts for the modes7

will differ for a given origin and destilnaon pai:. Because

A full c::planitlcn of the ( ffcrcrc-- f r. t!,e cos'L charactcr-i~t CS (.; ,e tr l:- ....! rai ,; ! , ".'Ii hi :,L

III (ti ,; , 1. 63 ). Tih 'c .:i : o':i 2' a d ,:c i t ,,f

the pf;.t:t which afY' . Lh-Li cost 0, , "'..",.

S-~ --... ~.....

-41-

the costs of the modes will, in general, be different for the

movement of the average product between tvo points, one needs a

pricing or rate making scheme to determine the price of transpor-

tation services between vario, s points. To be consistent wtth

8current transportation policy in this matter, the following

price setting mechanism will be utilized.

When the out-of-pocket costs of the high cost carrier are

less than the full costs of the low cost carrier, the full costs

of the low cost carrier will be the allowable floor for the rate..

This statement is illustrated in Figure 2(a). In other words,

the high cost carrier maZ, if he wishes, price his service at

the full costs of the low cost carrier but no lowL.r.

The high cost carrier may also price his services at hic

full costs. It may appear that there is no choice for the high

cost carrier but to price his services at the level of the full

costs of the low cost carrier. This is not the case however.

Between any two given points the high cost carrier ray have sold

his services to some shippers on the basis of lower toLal

logistics costs for the shippers even though their transportation

costs may be htthc.r than need be the case. ifn other words,

transportat 4 on co.:ts are but one of the costs involved in

accoMplishing; the 1oi:tic fuHction. Soze shippcrs -ay elect

for intaurc to trole-off hiIxr LransortaLl(. costs for I.e-c r

itwventory conts, and -o fort. In tI .Cdcis JWvulc,."' in Ge

. 9 .r.i, ch-tkr ', p. 7.

-42-

previous section, these situations were incorporated in the

logistics cot,straints which force a cert3in amount of traffic to

move by some modes regardless of which is :he low cost mode.

In light of this, the high cost carrier will price his

services at either his full costs or at th. low cost carriers'

full costs depending upon which price cont:Ibutes the Lost to

profit. If services are priced at the same level the traffic

botween those points will be arbitrarily split equally between

the modes. The high cost carrier will thei examina how much

traffic the logistics constraint requires be moved by his mode

and the amount he estimates he could obtain by pricing at the

low cost carrier's full costs. The high cost carrier vill then

determine under which price his contribution to profit and fixed

costs will be largest and will accordingly price his services at

that rate.

When the out-of-pocket costs of the hfgh cost carrier between

two points, is greater than the full costs of the low cost carrier,

as depicted In Figure 2(b) the full costs of the respective modes

will serve as their rates between two points. This is so because

it has been assumed thaL the I.C.C. will allow a specific rate of

return on each movement between all points.9 The I.C.C. would

9 In practice, the I.C.C. duc.; not atL:pt to control the rateof return on each c.o-.0dity bctw,. .n every - ri;)fn and .' , ;tInain.The ".C.C. dJ:es atLc1-.'L Lo co'tiz"ol Itnc raLv-c - r..u,.n ,,:r cirrierSin the aggro3.,te, co..hkrig all. .o.vnt of aai ccr.:.,o. ,: L.

-43- I

Figure 2. Restrictions on Carrier Pri:ing Decisions

(a) Allowable Floor for Ratzs

full costs

'wable floor for -- full coltsrates

out-of-pocket costs

out-of-pocketcosts

high cost low cos: carriercarrier

(b) Allowable Ceiling for Rites

full cost including profitallowance

out-.of-pozkct cost

full costsincluding profit

allov'ance

cut-of-pocketcos :u I

high cost low cost carriercarrier

control the rate of return by spc-cifylng o: controlling the avera;c

operating ratios of the wcdcs.

I2; h

- - ---.-.. .- . . . - . - .... . . ..-

U' -44-

A

The Simulation Model

The ecornic evaluation of a transportation company versus

sin independent trucker and railroad which i1 be ur, derraken in

this study is based upon the simulation of the interactive -J

decision environment just described. In the preceding sections

of this chapter, all of the components needed to "construct"

the simulation model have been presented. This section will bring

the component parts of the simulator together and explain how

the simulatior was undertaken.

The simulator cz the single modal carriers and the transpor-

tation company wlich was developed indicates how the firms should

price their service and how they should alLocate their equipment

based on the. criterion that each f~r- wisht:s to maximize its

contribut.on to fixed and/or common costs including a profit

margin. Solutions of the simulation model were generated for

differing levels of the parameters which will be described In

Chapter III. The solutions to the simulator may be thought of

as a sensitivity analysis on the cost coef.icients In the

objective functions of the carriers, the level of logistics

constraints, and on carrier capacity. More detail on Uhe type

of analysis that was performed is also presur.ed in Chapter III.

This approach rcquireq that the carriers kno;w the full costs

(rate) of the .ow cost n.cdL and their c-'n rvIporcLiVe Costs. his

is reasortable , that L.c rate of the luw cos. carrie: "'" be

publish d A:l ,,,.I.C.C. rrc dar- .. ,.

-45-

also requires that each mode know's how much Lraffic is committed

to each mode for movement due to shippers' logistics constraints

and alternatively how much traffic in each route segment is subject

to intermodal competition. This is not unreasonable providing

the carriers have alert marketing research departments.

The same simulator can be utilized to determine the equip-

ment allocation and pricing decisions of both the single modal

carriers and the transportation company. The manner in which

the algorithm is applied varies slightly for the two organi-

zatlonal approaches, however as will be explained. To determine

equipment allocation and pricing decisions of the single modal

carriers, the algorithm is first solved for a given set of

parameters for the rail mode then proceeds to the trucking mcde.

The simulator will now be explained int detail. The Inde-

pendent railroad first determines the full costs of making rail

boxcar, Plan 1, and Plan II piggyback movcnents for each route

segment in the transportation system. The rail carrier then

examines each route se;neut to determine i., one of the methods

of movement he controls (rail, Plan I or II) Is the low cost

method on each scgment. For tho.e route se 'ents where he does

not control thc lo,. cost rvthod, he iternis if he can price

the services of one of hi mtWods at the full cost of the

trucker,

Once the rndlroa.J ,L r:.nc: whcr.,c he can c:'pcte for uncon-

straiIcdJ or c,-.M:. L itic tL.:f C' T.1 I Ch Of IIit

. . . he .~l .'t.. Uj .u h f l~

A

-46-

market he cart expect to obtain. Using theue predictions, he

determines which methcd of movement makes the greatest contri-

bution for each route segment. In other words, on the segments

where he has the low cost mode, say boxcar, he would compare the

estimated contribution of his other methods priced at the boxcar

rate and determine whicl. method made the greatest contribution.

Once the estimated contributions for each route segment have

been completed, the railroad ranks the expected contributions.

This ranking determines the most profizable manner of pricing and

equipment allocation for the carrier. The carrier will illocateI

his equipment beginning with the most profitable expected contri-

bution and proceed to allocate his equipment down the rank until

his capacity is exhausted. It should be remembered at this plint

that it has been assumed that shipments subject to logistics

constraints will be satisfied first.

Since the algorithm uas accomplished for the :ail carrier

first, the routine will now Le run for the trucker, for a coratible

set of parameters. Before the trucking routine is accomplished,

however, the amount of flatcar capacity the railroad will offer to

the trucker n.ust be determined. If the rail carrier has used all

its flatcar capacity, the trucker will be pre-empte'd from particl-

pating in Plan I moves, If the rail carritr has some flatcar

capacity left, this amount vilt be made available for the trucker

in case he should find it dr rable to use Plan I piggyback inve-

ments. The ,'i:.ulalin Of the trucl.-Lr dc;2Isiunn are arLILIou,

to these of the ra rd; Lhcrcforz n,.d not le repeated.

0 -47-

The algcrithm for the transportation company differs slightly

from the one presented for the single modal carriers. First,

the transportation company would not have to estimate how much of

the competitive market it could obtain for each method of movement.

The company vould only have to estimate the total size of the

market, as do the single modal carriers, but would then determini

how to most "profitably" move that traffic since it controls all

methods of movement. Also the additional bookkeeping of determining

how much flatcar capacity is available for Plan I piggyback mov:

is eliminated.

This section does not complete the discussion on the simulatior

' model. In fr.ct this section has presented just the skeletal

framework of the simulator upon which the following two chapters

will expand. Chapter III presents a detailed discussion which

focuses on the test factors which will be analyzed and describes

how the cost data was generated. Chapter IV describes the per-

formance measures used to evaluate the output of the simulati-

and discusses the nature )f the experimental design utilized in

the study. It is only after these chapters have been read that

the reader w-ll have a full understanding of the nature of the

simulator utilized in Lhis project.

I" I

-48- I

CHAPTER III

SPECIFICATION OF THE MODEL

The preceding chapter presented and described the nature of

the simulation model which will be utilized to determine the effects

selected test factors have on the economic perforance of a transpor-

tation company in contrast with that of two single modal carriers.

The description of the simulation model was very general in nature,

however. The parameters which the model manipulates to arrive at a

solution were identified but only in an overview fashion.

The purpose of this chapter is fourfold. First, the chapter

will fC:Us attention upon the specification of the test factor

sel *ite to- analysis and will explain how these factors effect the

parameL.c values in the simulator. The discussion of this pcrtlon

of the chapter will include the reasoning behind the selection of the

test factors as well as the Identification of the specific values of

the factors. The second objective of this chapter is to identify the

other cpecifics of the model. In this section of the charter the

random components of the model will be specified and the values of

these raadom variables will, be Identified. This section wIll also

describe the shipcru' modal selection policy that wIll be uscd by

Athe shippers fn the rodel. Thirdly, rhe manner in which the carinZ

data will be generated for use in the simujation model will be idntified.

The te;t factors which affect the cost of movenont 'J l also be

idcntlficd, and the --stuve of Lh functionial relatin;hi bCL-twen

WI~Ii2

cost of movement and test factor values wll be explained. Finally,

the trarnsport.tion system which was modeled will be presented.

Classification of the Test Factors

The parameters which affect a carrier':; economic posture may

be classified as environmental, commodity, and route factors. There

are many environmental factors which could affect the economic well

being of a carrier. For the purpose of diseussion, the environ-

mental factors may be divided into two groups--managerial and

geographical factors.

Ianagerial factors, such as the competnce of personnel, organi-

zational effectiveness, and management-labor relations, are concerned

with the managerial effectiveness of a carrier. These environmental Ifactors will not be directly considered in the inalysis because the

transportation company and the single modal carriers which will be

compared have been assumed to be able to achieve their objectives

of maximizing their respective contributionq to fixed and/or cornion

costs including profit margin. Thus the competence of personnel,

organizational effectiveness, and so forth ol both organizational

forms, i.e., a transportation company vis-5-vis single modl carriers,

have been implicitly assuzed to be of equal effectiveness. ':he

validity of this assumption will be discus . :d at length in Chapter VI.

Geographical factcrs such as weather or climate conditions anr.d

the geographi.-al terrain over which a carrier upcrates are cor,cerncd,

as the na:c Implies, wjth the gcc;;raphfcal environ-cnt of a carrier's

route strucLu:-L. Since the ra.lIrcad amd . llf:i cc: :.. My Pnd tho

!g

-750-

transportatlon company trill operate over the same route system these

factors will iot be directly considered in the ceparison between

these organ55ttiott alternatives. That is, the operations of thea

two single modl firms w1l1 be compared with the operations of a

transportation company which I-As the same route structure and hence

would face tho same climatic and geographical conditions. These

enwronmental factors will be considered In]irectly, however, as such

factors would affert the cost of movement over the route syster of

the cArriers,

Commodity factors are characteristics of the goods, which c-~-te

differences in the custs of movement, For instance, one clasb

goodc tAy have loading characteristics which would require special

hxndlin. Jinquos which would roquire additiorkal labor nr capital

outlays by the carrier thus affecting the cost of movement. As mentioned

in Chapter 1I. commodity factors will not be addressed in the study.

This is because the rail and truck modes have literally thousands of

comaodities many which have characterlstics which affect the cost

of movement. If one were to consider a number of commodities, the

number of decisio, variables developed in the previous chapter would

have to be mu Ilplied by the number of comoities considered. Thus

ftr the sake of simplicity the study will c ncentratce on the "average

commodity su:)ject to Intermodal conpeLiton for movenent." While

thin nirtliIJ(:atiL.. zay effect the Zcuerality of the results of

the study to scome eXtent, the object of the study is to obtain a broad Ipicture Of the effecL!; of the factorg consJdL-r- ra.ther than to cc,.-

cttU111l' tE, ol !;;Ic( c . .. ,"':,nts of cr 1-c L;j in I ct : .1,.,0 .'.., ,q " ...

,-51-j

Furthermore this simplification may no: be as serious as it first

appear . For Instance it is not likely tha: the commodity charac-

teristics of a shipment of canned corn differs substantially from a

shipment of canned motor oil. Nor is it likely that the characteristics

of stoves differ substantially from those o! washing machines. InN1

other words, there are classes of goods for which the commodity

characteristics are substantially the same. In fact, the classi-

fication of goods into a relatively small number of groups is the

starting point for the carriers rate determination process.

That is, each of the hundreds of thousands ofshippable commodities, ranging from aardvarksto zymometers, is placed (classed or classified)

In some one of a relatively small number of "classes".Then instead of shipping a co=odity, In effectone ships a certain quantity of a certain "class".

This fact would enable the solution of the model for "the average

commodity subject to internodal competition" to remain valid for

a number of comnodities. A solution for each specific class of goods

could be obtained by using the average costs applicable to that class.

While this ende:avor should be of interest tD the firms which may be

inteLested in forming a transportation company, this is not necessary

to obtain a broad picture of the effects specific factors have on

the operating characteristics of a transportation company, which is

the object of this dissertation.

Route facLrs may be conveniently classified Into two groups

for purposes of discussion. Internal route factors arc those factors

]Sampson and Farris, 9pE.ft., p. 162.

I

--. . . .

. -52-

which the carriers can contol'to someexten:. Examples of internal

route factors are carrier operating ratios, load factors, and carrier

capacity. External route factors are those factors over which the

carriers can exct little or no influence. One: external route factor

would be the level of shippers' logistics constraints. Thus,

examination of the route factors lies at the heart of the purpose

of this project. That is the effect each of these factors has on

the operations of a transportation company compared with the operations

of the single modal carriers should indicate whether it. is possible

for a transportation company to achieve economies from the reallocation

of traffic from high to low cost mo4es. These route factors are

the test factors which will be analyzed in the study to determine

the effect they have on the economic performance of the transportation

company and the single modal carriers.

Internal Route Factors

The intertial route factors which will be studied in this thesis

are: (1) the operating ratios of the composite single modal companies;

(2) alternate levels of load factors of these companies; (3) the

capacity of the carriers; and (4) the size of shipments.

The levels of the factors may be considered to be an upper

bound a:d a 1~wer bound of the factots, Ii other words, the valuc

of a particular factor for most companies should lie bet:4een

these bounds. Thus the CvaluatJ ,: of the changes In The level of

these factors should prouvdu ni -Ihti as to how the!;, factors would

___ ___ __"__ __ . ......... .......

affect the economic consequences of a tranwportation company formed i

by carriers vith factors which lie between these bounds.

Carrier Operating Ratios

As stated in Chapter 1, one of the potential sources of

economies resulting from the formation of a transportation company

is the possible reallocation of traffic from the high cost mode to

the low cost mode on a particular route segment. Examination of

carrier operating ratios, both truck and rail, should give insights

as to whether or not this is possible. Thn operating .atio is the

ratio of operating expenses t; operating r.tvenues. Thus the operating

ratio of a carrier directly affects the contribution to fixed and/or

common costs including profit margin which may influence the manner

in which the carriers make equipment allocation decisions. One would

be interested in determining if a change in the level of the rail

and truck operating ratios causes the equipment allocations of the

transportation company to be significantly different o am the manner

in which the singlc modal carriers allocate their eq, ment. Another

question of interest is whether a change in the operating ratios

causes the contribution of the organizational alternatives to be

significantly different. As will be explained in the next clapter,

variour pcrfiranc measures will bc calcu'ated from the smu] tlln

runs which will be helpful In ans,,,erir.g the2se questions.

The specific opera3lng ratios which will be considered are 99.'; 'I

and 91% for the truckln firm r.prcenhLn a below avec'3g- or :nef-

ficient cirrir a;d a I raa carri, r, re:.pecti vc. T, opcrav,",n

7

U -54-

ratios for the rail mode will be 85% and 63% representing an inef-

ficiently anu an efficiently uanaged carrier. Within these spectrums

one would find the average carrier.

Carrier Load Factor

The load factor of a carrier shall be defined to be the average

payload transported over the carrier's route system. In other words,

the load factor is the average weight of shipments transported between

origins and destinations. The load factor of a carrier directly

affects the cost of movement, as will be explained In detail in a

later section of this chapter. Thus, this factor should have a direct

bearing on the manner in which carriers allocate their equipment and

on the profitability of the firm. Also of interest would be the Imanner in which operating ratios and load factors interact to affect

carrier allocations and resulting profits and cost to users.

The specific load factors which will be analyzed for the trucker are

100 Cwt. and 300 Cw.;t. representing a low load factor and a high load

factor, respectively. The load factors for the railroad will be 400 Cw-.

and 800 Cw,. representing a low load factor and a high load factor. 3

The average operaLing ratio for Class I and II motor carrlerswas 96.2% in 1970, ("Amerjc,.n Trucking frctnds 1970-71", op.cit., p. 19).The approximate avurage cperaVn;, raL'u' of Class I railroads was 79. I. 3-1969. ("Ycarbook of Railroad !'act3 , ]97L Edlitio., Econcrlcs & r:'ranc -

Departm 'n t, Assci ation of A-'crican Railroads (Chicago, Illinols: April19 70), p. 9.)

3 TDe avera),o, load factor 'or Class I :totor carriera of cneralfrelght was 264 't in !L9.9, (' rc ruc in; ir.n.. 19Y ,-7l,

oP.cIt., p. 34.) "ihe averaOge w,- ;;t of a carloaid cf ra 1 frojgdhtwas Lppro.Iiauc ' 28 C',t. - i ]"69, ("Ycar-ao of .::Licad .- 'ts' ,

r Pi.,Jt., P. ~5o.)

-55-

91I

Size of Shipments

The stuay will segregate movements into two categories truckload

(TL)/carload (CL) movements and less-than-truckload (LTI,) movements.

This distincAon 4s necessary because the'cost characteristics in

the TL/CL movements are different from those of LTL movements. The

reasons for these differences as well as the nature of these differ-

ences will be explained in a later section of this chaper.

TL and CL quantities shall be defined to be shipments which are

capable of beinig moved in either a truck trailer or a boxcar. In

other words, the TL/CL shipments received from the shippers are large

enough to be transported by a number of truck trailers or a (different)

number of boxcars. Thus, TL/CL shipmcnts .:an be moved by truck,

piggyback, or by rail, subject to shippers' logistics constraints.

LTL shipments are those which are not large enough to fill a

truck trailer. Since the railroads have phased cut there less-than-

carload (LCL) business, these types of shipments are subject to move-

ment by truck or piggyback only.

Capacity of the Carricrs

The capacity of the carriers should also affect the manner in

which the carriers make their capacity allocations and thus affect

the pr.ofALaility of fir.hs. This factor tay also affect the price uselr

pay for tra.-;porLation survice. Like the other factors disci:',sed, the

capacity of the system will also assumc two levels. 'I,,Z:t is, the

system wide ca pacity of the carriers will bo ;tu icd for .ffc.'Cs at

-56-

F two levels. The lower level of the system will be defined to be that

combination of carrier capacities which is capable of meeting all the

expected demand requirements for movement within the system. Whether

or not the carriers allocate their equipment in such a manner as to

meet that demand Is subject to analysis. The upper level 0i -vstem

capacity shall be defined to be that amount of capacity owned by each

mode such that no one would add additional capacity. In an operational

sense, the upper level of system capacity will be that amount of

capacity such that each carrier is capable of meeting at least all

the expected demand rcquirements within the system.

The capacities of the carriers have been segregated into two

groups--a certain amount of capacity has been reserved for TL/CL

movements and a certain amount has been reserved for LTL movements.

In the real world it is very unlikely that a carrier would segregate

its equipment into such groups. This artificiality was introduced

so that the two classes of novements--TL/CL versus LTL--could be

studied separately to determine if the different organizational approaches

(transportation company versus single modal carriers) have significanti)

different economic consequences in so far as the two market segments

are concerned. Although the sJr.ulation could have dealt with the

total market seZP;'nt by using average cozts of movement including

both CL/TL and LTL in the sa:.:z cost figure, this would have lost the

information as to ho% the or;;aniztiont1l for&.s perform under the FI ze

of shipment para.'tcr.

,________ , M_1 ---- - I

-57-

The amount of capacity set aside for each type of movement by

the modes is also somewhat arbitrary. For instance, the specific

percentage of capacity reserved for TL movements by the trucker may

be relatively large as compared to what one may find in the real

world. That is the trucking mode In the simulation has reserved an

amount of capacity equal to that which the railroad has reserved

for CL movements. This may indeed not be the usual case in the real

world. It must be stressed, however, that although this capacity

split is arbit:ary, for comparison purposes it is consistent. The

performance of a transportation company will be compared with the

performance of single modal carriers operating with exactly the same

set of parametric and random values. As will be discussed in great

detail in Chapter V, the analysis will focus attention on whether or

not there are significant differences in the performance between the

two organizational approaches. llcnce it is of greater importance

in this study that the parametric and random components be internally

consistent between the approaches than is the absolute values of the

parameters. More will be said about this matter of internal consistency

in the next .;hapter.

Before proceeding further, it should be pointed out that in practic.e

the rail mode would alsc not allocate a certain aMount of capacity

for plggybaLk LTL move:3u;ts as this :;tudy has doac. Agaiin this 'vas

done to dotcr-.ine if the two tra'tsportatlcn approaches vwould resuIt

in a diffcrent equip ent allocation p..,trern and icn goucIv rcsult in

signIfica'itlv diffo:rcr t cfL (2 Onic ccn;('urnc,..q

WO¥=-WWTI

~-58-

External Route Factors

The external route factor which will oe studied in this thesis

Is the level of the shippers' logistics constraints. The size

of the shippers' logistics constraints will also assume two alternate

levels.

Shippers' Logistics Constraints

Since shippers' logistics constraints directly affect the manner

in which carriers allocate their equipment, this factor will definitely

affect the economic situation of carriers. It has been assumed that

the logistics constraints of the shippers must be satisfied. This is

not an unreasonable assumption given that certain shippers' logisticsI<

systems dictate the movement of goods must be allocated to specific

modes. The carriers should also be expected to satisfy a guaranteed

market before they would enter the competitive traffic market.

As was mentioned earlier, the size of these lfgistics constraints

determine the size of the market which is subject to movement by more

than one mode. The second leval corresponds to the situation where the

logistics systems of shippers doces not constrain the allocation of

equipment by the carriers to a large extent. In other words, the

second level of this factor provides for most of the market being

subjec. to i.tercodal con-petition. The first level of this factot

constrains te solution alo:q.g no--al lines to a ri'ach greater e:tent.

Eximination of this factor should provide guidelines as to under

what compe titive circur5:5Lrcc, If cny, a tranvpo-tat-cn ccmpany

would out perform ci,,I, 2c'dal c---:'p:ies.i

Other Specifics of the Model

This section will identify the random components of the model

and will present the shipper's modal selection policy. There are

two random variables incorporated into the simulation model. These

are: (1) The forecast of the market share of competitive inter-

modal traffic the carriers predict they can obtain; and (2) The

demand for transportation services between specific points.

Carrier Market Share Predictions

The market share of the traffic subject to intermodal competition,

each carrier predicts he can obtain, will be treated as a rando.

variable in the study. This is in keeping with the difficult real

world problen of forecasting market shares. The market share a firm

obtains is determined by a number of factors and complex interactions.

For example, the market share a transportation firm obtains is dependent

upon, among other things, the general economic conditions, the service

reputation of the firm, the effectiveness of the firm's advertising/

sale3 progran, and the logistics systems of shippers. Thus the =arket

share a firm predicts it can obtain is likely not to be 100% accurate.

The firm may over estimate its market share or under estimate its

market share as well as possibly having an accurate prediction. It

Is because ' the large number of incractions and uncertaintiAes

involved in the csti:.,;tjon of market sh.arcs that he prcdict.c7! of

markete shares a carrier will obtain will be Lreated as a rand-.n

variable.

J3

t-60-

The values the variable may assume fo- the trucker and railroad

will be 40%, 50%, and 60% of the market subject to intermodal conpe-

tition. In other words, when these carrlers can compete for traffic, '

they will estimate they can obtain either 40%, 50%, or 60% of that

market for a particular route segment. The particular estimate for

each route segment will be determined by a table of random numbers.

Wien a carrier cannot compete on a given segment, the other carrier

will attempt to estimate the size of this market and if profitable

allocate equipment to it. The values in this case will be 80%, 90%,

100%, 110%, 120"% of the market. Thus for any given route segment,

the carriers may over estimate the market which would result in

excess capacity at some nodes or they may under estimate the market

resulting in unsatisfied demand between so'ne points or they may

exactly estimate the market which would satisfy all demand and leave

a minitaum of excess capacity.

The transportation company, likewise, is confronted with the

same market prediction problem. The values of the estimated market

share the transportation company expects to obtain will be 80%, 90%,

100%, 110%, and 120% of the market subject Lo intermodal competition.

The particular estiat1e for each route segment on a particular run

will be determined by suring Ithe si " mDal carrier's estimates

in order to Freserve internal consistency. In other words, if on a

particular run the trucker estiniates lie can obtain 60' of the inter-

modal compoeltiva n:irket cn one scg.cnt and the railroad estm=teOS it

can obtain 5Q:, the tran:sp:rtaticn cCMpa1ny ":oud es':iate the sze of

LA

-61-

the market at 110% of its true size. In essence, this allows one

to make a cross comparison between the performance of the single

modal companies vis-g-vis the transportation company in which notA

only are the parametric values the same but also in which the random

components are consistent across the organ. ations.

The Demand for Transportatlon Services

The demand for transportation services between any two points is

determined to a large extent by the demand for shippers' products.

If the product is manufactured at point A and there is demand for

the product at point B, the product must be transported between these

points. A more succinct way of stating this proposition is thate

transportation provides time and place utility J:, goods. Viewed

from the carriers perspective, the demand for transporation services

can be thought of as a random variable. The variability of the size

of the transport market is of course directly related to the problem

the carriers have in predicting their market share.

The size of the transportation market subject to intermodal

competition ,ill be allowed to assume three value.;, a high volunC of

traffic, an average volume of traffic, and a low volume of traffic

for each size of shipmient level. The specific level of denand butwoen

specif'c poi is will be dete .zined -y a L)le of random numbers.

Shippers' Sel-ta1 S.-n Policy !

In thu cc-..pet.iLtvc -ingle -oda2l c=.pany situation, the carric.rs

mus t dctir-n.inc ii the y can c-'.cu e in a markcL 5Cflnt Sub A = to

-62-

intermodal ccmpetition. If they can, they must allocate equipment

to that segmtent consistent with their forecast of the amount of

traffic they can obtain in that seSment. If the carriers allocate

more equipmert to a segment than there is demand and if this allocation

results in more than one method of movement being offered to shippers,

the shippers must chose the manner in which to move their goods.

These methods have no service differences (or these differences are

unimportant as perceived by the shippers), or else the shippers specific

needs would h.ave been included in the logistics systemr constraint.

If the carriers have collectively uader estimated the amount of

traffic subject to intermodal competition, the shippers have no

choice but to accept what is offered, up to 100% of the demand being

met. If the carriers over estimate this market, the shippers must

choose what methods of movement they will use. Under these circum-

stances, the follov!ng shippers' modal selection policy will be adopted:

The shippers will allocate their traffic equally between the carriers.

This is in keeping with the assumption that when the service of

competing carriers is equal the shippers have no real preferences

between the codes.

In the case of the transpprtatJon company, operating under the

assumption that it will face the same demand pattern as the sif.;le

modal cor.mpanies of which It is coprised, there may be no need for a

shippers' modal selection policy. This is so because the transpor-

-63-

tation company would most probably allocate only one method of

movement to any competitive market segment.

Recapitulation of the Test Factors

Due to the length of the preceding discussion, a recapitulation

of the test factors, shippers' modal selection policy, and the random

components of the model is in order. The test factors and their

appropriate levels are outlined in Table I. Table II displays the

random comporents of the model and describes the shippers' modal

selection policy.

Generation of the Data

As stated in Chapter I, the data which will be used In this study

will be hypothetical. This section will explain In detail how the

data will be generated for use in the simulation model. The discussion

will also focus on an explanation of which test factors affect the

cost of movement and the manner in which they effect these costs.

The costing figures used in this study will be on a per hundred

weight (Cwt.) basis. These costs will be ascertained by ailocating

the total costs between specific origins and destinations on the

basis of the amount of Cwt.'s transported between them. Once the

full costs of each :.ode have been detervined, the pricing t.echansm

explained in the last chapter will be utilized to dctcriine the

appropraLc rates be:,teen- t,:o points. A computer program has been

develo[c! to calculate the full costs of the modes. "The costs will

differ each time a parameter Is variud.

'- : i- .. .-..

-64-

14 P-4- 00

4)U C%.. 00 U .joo Vor 0 co C:00 -. Cs 0~) O-.0 wl 1 0 c-Q

CA -T 4- -

*4v 5*J -t:

U u I"C

I-4:

C . ; . 0 U -

..'. o

.. .. '"

"-

-.- -,_-.

-65-

TABLE II

OTHER MODEL SPECIFICS

RANDOM CO,.PON.,TS: VALUES OF RANDOM VARIABLESA. Carrier Market Share

Estimates

1. Truck,r 40%, 50%, 60%2. Railroad 40%, 50%, 60%3. Transportation Company 80%, 90%, 100%, 110%, 120%

B. Demand for TransportationSarvices on Each RouteSegment1. CL, TL Shipments 30,000 0.'t; 60,000 'WL;

90,O0C' 04t.2. LTL Shipments 15,000 C.':t; 30,000 (.Cwt;45,000 Cwt.

SHIPPERS' MODAL SELECILON POLICY:If the carriers have colloctivol, under astimiated tl~a amount (.ftraffic GubJc2t to Inter-odal co:,potitlon, the shipIcri have riochoice but to 8CCepL Whirt If- offor(d, up tu 100, of the cer;anIbeing met. If the carrior3 collectively over eL'mato thiimarket, the shippers will allocatc thuIr traffic. equally betw c nthe carrIers.

The CosLTnl C!oS7t4rThvraor, Pt'. iny cuvtttng C t,:vl,-.,i Oo~ (..4d liao for atwily""1ir,4

thti COAL of ruvillo tiod4, Trh foOibovjll ! C ttRLt IAdY b'..i1

chobtEI Iieo u-u thliy rev il Inf ttiiiL Aia oi t I ....' uff c t 0

IIIt u, - I" . I IuI I 1 f 14 ill ) It u l . ., ,-, I , 1i --1. I I i .

ati-l I t , 7 .-1., I,h, all-, t;I. Ur l . It ,: I I, . l I, II t ,- .I LI

I--. .. -- .. ..

-66-

or over-the-road coits,.pick up and delivery or car spotting costs,

terminal costs, and billing costs.

Line-haul costs are definid as the movement over theroad (or road bedi, either from one terminal to anotheror, in scme cases, fron a shipper's location to destination.Terminal costs are defin'ed as the handling and rewtorking offreight to maLch origins and destinatl.)ns of outgoing andIncoming freight. Pickup and de,'verv,, ccsts are defined

as the function of pickin& up and deli/ering freight withina specified terminal area. Billinp ... costs are thoserelated to paper work coLts for each shipment.

Each of the out-of-pocket costing categories is measuredin terms of its relevant service unit. A service unit isthe apprcpriate varial)lc assiSned to a specific costingcategory. This concept Is e.actly the same a-, that used ingeneral accuunting in cost allocatioi. The accountant alsomust find some method of allocating indirect costs such asront.

The service units for line-haul c).ta are vehicle hoursand vehicle mllefi, becauze bcme coats In the line-haulcategory vary with tfi" (oe., drivers' wages) and othersvary with diotancr (e.'., fuol)."

Line-haul cotts alo0 vAry with tho anount f product carried.

TWirs, an appropriaLt mnaaurc for line-hAul costw would bo the

5W4 of tho total Mi4eage coLt (dantan.,_) nd' total hourly costs

(wage-) 11vidod by thu tutal t utber of Cvt.-tlle_ ttnhportod

by the torrir II k gIven pyriud,

I1,

UYI1. 61 it of lltl lij l~t 1

-67-

would: simply multiply the line-haul costlCwt. -cile by the average

length of hat.!. For example, if the average line-haul cost/Cwt.-

m Iile was .35 cents/Cwt. mile and the avera~e length of haul was

300 miles; the average line-haul cost/Cwt. would be 105 cents. To -

convert the average line-haul cost/Cwt. to specific line-haul costs

between two points with a specific load factor, one would multiply

the average figure by the ratio of the actual length of haul/system

average length of haul and dIvid2 by the ratio of the actual load

factor/averaje load factor. For Instance, if the average line-

haul cost wall 105 cents/Cwt, bazed on an aveorage load factor of

300 CWt. And average lhiath of haul of 300 miles, the specific

lisie-hatil COAt for A rnc-,c,;acnt of 150 Ocut. .)vcr 900 miles w!ould be:

Aavaroaa ata dintinic actual. liad fatctor

cost a~~~ ~t~a aoqe~idf~~*actual lilsi-houl coot

( 105 900/300 1 50/300 )

F ~As OtI ivi iaul co~rtA, j'.ckuI A~id doliveryj ccomfi~

Ave dlLr';i':n L ii:U ~ u till, '.11,4 Pkiyu1u iuld

delJvLtY CUaL~b lj'/ LhU tAii (A, (ti4ndb.J ii II .t by>3tL'.

V'ol i u ral I J.1th UjjLLf Cull', i I AL I'll', tIurIr 0~ L~; F-. V j I;u dm

As L.a1 91 1 til ui I:I I lit Iu,' Utiii Ii t 11Cnj 11~ V )V"'tId Ii it. . I I i .~

#,H 111 l t ll l t h oil 1-, % .lj ... it , il, fil -I 11.1 1 1.I W 1 V. 1. L IP 1 -11 Illi r I I li

-68-

PUD cost/Cwt. for a movement with a specific load factor can be

obtained fron the system average figures by dividing by the ratio

of the specific load factor/system average load factor.

Terminal costs will be measured in the same manner as Pt'D

costs, i.e., on a per Cwt. basis. The discussion in the previous

paragraph of how to convert to specific average costs/Cwt. for

given load factors are applicable to this category of costs also,

Billing costs are the expenses associated with the paperwork

functLions per movement. The average billing cost/Cwt. is obtained

by dividing this somcvhat constant cost by the system average load

factor in Cwt. To convert thls average coat to a cost/Cwt, for

particular movements, one would divido by the ratio of the par-

ttculpr load factor/systen average load factor.

Thu above out-ot-pockut costing catagories are derived (ren

the recognix,:d I.C.C, standard Accouintig >,go:i for wiotor carrior.,

PaIl carriers have different costln& catogorlog ouch as iltching

and yard exim hne b4t iN ordhir t,-; have a tasi for coiiparfl.5ii the

ahoW CALuL'-i'ic will bo U§ed for both rnods1,

Aw alludiO to otd.ovoi th lfii 4 Il .!."o(k'. d fOl'

tIj- 1 l i d,-I i t I I h1 4 t 1 lit-.h V I t l , t 1-3 1I , 11 -

li"11i1 0 fit %ji-Y# Alid -1 !'l1 t~l l 1"11J 41 G3,il l la l l 6411 6'4 1..In i1: ( V 1.11i rutil <,h -,, h

firpi l IrA! , i : l i 1,1 , - I t i ., lt ! : l'l'lv 1,11 1 114 1 1--1 l.1. l , i -11

-69-

associated with yard operations such as car spotting, switching,

and classification of freight. Blllng costs are comparable as

they stand.

Once the out-of-pocket costs have been calculated, one must

allocate a portion of the total systems fLxed, and/or ccmmon costs

to a particular movement to determine tho fully distributed costs

for that movement. Fully distributed costs are found by adding

a percentage of each out-of-pocket cost category to cover co=..on

costs that do not vary with traffic. These would consist of

officer's salaries, depreciation, and the Mie. The Cost Finding

Section of tho Bureau of Accounts of the interstato Cormerce

Comni.lon hAv.w found commaon conts to be about 10% and 20% of

fully distributed costs fur rotor cartier, and railroads, respectively. 6

Thum, to fin4 th fully distrbutad cott fir a mov omnt for the notor

carritr and railroad, onl would tAke thu ovraV4a out-o -poikut co.ts

anJ dividn them by 90% ind 8U.: repactevol.,

Th@ fln-4J. adJu~ttvilt fur CoAL Ing 14 a prof It allovkumco, Int

thin atudly tho pr f IL al uwaocu wIll be .iadv thrugi tho upurat.11n

t o. 'romt, at iddrd ac uut lt g prutlu, thr opratitig ratio

S(14 1 tit-(l , ..l ....1 ..l..-.....! ' - ,, b.y1V- Ig W . .

; I. I A 4 I - I 1 I

h I f . 1-.

-70-

Average Industry Cost Data

As mentioned in the introduction to tils section, the research

will utilize approximate industry iata in the various costing

categories sc that the study may deal with carriers which reflect

the differences in operating characterlsti.:s between modes. Table

III presents the average cost structures for the rail and truck

8modes. The cost structure of the trucking mode Is an average

which includes both TL and LTL movements. The cost structure of

the rail mode is based on CL movements,

TABIE III

PFRCENTAGE B"REUAKDC U OF RAIL ANfD TRUCKCOST STRUC'UIU

Truck Rail

% varlable 90% 80%

Out-of-pocket co. ts

1,111e-h14ul MA:Pt 64% 67%

rUD (car spottL g) coAL 18% 107

Tortnhiai (6'aitching 6 ctie!lfi|cat|vi) cott 20%

hillilr. cooL %

rj L t ry fur them ui.-t;r t ,lv, lI tfir t11A)v d.,u':a u' ,btuIll d 1 j" t;iIV; !,. . * 'l I'll, I i l "

ia ~Lr * fu) I i : i tlIL' t II , Lti 2 i. ohil ,' ii I .il '

,l , i1 fi j i, f , i -MA

I - I -ltj I

1 I .i f ,

!

-71-

The study will segregate movements into two categories TL/CL

movements and LTL movements. This distinczion is necessary because

the cost characteristics of the TL/CL movements are different from

LTL movements.

Truck Cost Data

For TL movements, the cost structure for truckers given in

the preceding table is not appropriate. This Is due to the fact

that a substantia' portion of trucking business Is LTL and necessitates

larger expenditures on PUD and terminal costs than do TL moverents.

The average cost structure for truckload movements will be derived

in the following par-praphs.

The average revenue for truckers hauling general cor.odities

Is approxinately 7 cants/ton-mile; and the average lenglth of haul

9 iis about 300 tndles. !.ultiplyfiig the revanuo figure by the average

length haul yield. the average revunue per ton of $21.00 or $1.05/

CwL. Thoi nverage load factor of the truckor will. be takcn to

ba 200 Cwt./vohlclv novc .tit.

Am rxii|t,,od AhovU onu¢ co the parametvra thaL will be varied

J6 the1 11,,roi h!M I'dt b of tilu uIrriti l tO ,It Uhf 0l1 Wh1 t o rfOcth0( hn nit I trhc p ft tht I It' of th,' I rrit l - i , I ;. -,,wa io Aid ill

tlt! jq' (if *I tit I.,1, It.J l Iycf i I fL, t,.

l i. t .I i VI?

a I I t I

-72-

ratios will te 99% and 91% for the trucking modi~ representing

abelow average or Inefficient carrier and a well managed carrier,

respectively. Within this spectrum one would fInd the average

carrier.

Multiplying the average revenue per Cwt., just calculated,

by the operating ratios and then utilizing the breakdown of the

average cost structure of motor carriers in Table III yields the

average trucking costs per Cwt. displayed In Table IV.

TABLE IV

TRUCKING AVERA.GE COSTS PER Cwt.

Load factor 200 Cwt. 200 Cwt.

Operating ratio 91% 99%

Total revenue 1.050 1.050

Profit riargin .0945 .0105

comv'On C')Srta .0955 .1040

Total. OuL-of-pockot costs .8600 .9355

Unabot coats .5504 .5987

I'UD COOL .1548 .1684

Terrnflita. C-Out .1204 .1310

1111 hii; copt .0344 .0374

f it 'I 11-''l . i'- II~I ~1 Y u ii t th t df

--; :. E ... .- -: . . , ::_:.-. . .- .. :. . ... .. . .. . ._. -. J

-73-

when small shipments could be added to the TL at the terminal,

since the trailer could be moved direct from the customer's door

to destinaticn. Billing costs should drop also, perhaps as much 14as one-half, since there will usually be only one billing per

vehicle with TL moverzents. Deleting these costs fror out-of-

pocket costs and recomputing coizxon cost and profit margin by

dividing the out-of-pocket costs by the % variable then dividing

this figure by the operating ratio yields the TL average costs per

Cwt. displayed in Table V.

TABLE V

TL AVERAGE COSTS PER Cwt.

Operating ratio 91% 99%

Load factor 200 Cwt. 200 Cwt.

Total revenue .790 .790

ProfIt margln .071 .008

Common costs .072 ,0783

T utnl out-of-Vo'et co tfi .647 .704

Line-houl cot .551 .599

1'J1l) (.(j;I .077 ,08/

It ' :1I ), :n. t ,r)j9 I021

lit I / ,,,I21 1

i

-.74-

The percentage breakdown by costin~g category for TL move-

rents Is given in Table V1.

TAB LE VI

PERCENTAGE BREAKDOWlN OF TL COSTS4

% variable 10

Out-of-pocket costs -

Line-haul cost 86%

PUD cost 12%

Terminal cost A-A

B1lling cost 2%

Using the ovcrall average trucking costs per Cwt. and the

derived TL average costs pcr Cwt., one can determine the LTL

average costs per Cwt. in each costing category. This is so because

the overall average trucking costs per Cwt. are a weighted average

W 2(M Cost) w Average Cwt., where tile weitlits, W I and W 2 are tkle

percn~ge of Cwt, whiich are LTL and TL, r-!rpect1vel',.

The percentage of LTL traffic cartried by regulated &vrncr:.

cutwmiodfty cnrri.-jrs is apiproxhdnate1V '0'i# so tlic pereentage of 11,

10 LantI yoiir Cla)I I I n.otor c:ardir or. Ct mui.r.iI fri ~h

F. 3 1 l LIM ',~' If LI L. Il190 1'~ 0 1 . 1 U f I I'( IILII - 1 1. L'!- i

I lit___ --11-0 -, -1

-75-

traffic is approximately 50%. Using these weights and the data in

the other twc tables the LTL cost data was calculated and is

presented in Table VII.H4i -

TABLE VII

LTL AVERAGE COSTS PrR Cwt.

Load factor 200 Cwt. 200 Cwt. 2

Operating ratio 91% 99%

Total Revenue 1.310 1.310 :2

Profit margin .118 .013

Co=oon costs .119 .130

Total out-of-pocket costs 1.073 1.167

Ii Line-haul costs .551 .599

PUD costs .232 .252

Terminal costs .240 .262

Billing costs .050 .054

Rail Cost Data

The svora', revenue for rallroaJni In 1969 was approxiraoly

1.35 cents/toni-nile; and the avr,'a~ lonOih of haul was about

600 rilefl. The ,vurn,' rev.:uo N!;ure wi ll bo a-.U itrd up;'ard

to 4 evn I/(', I ~-HI v I)ctitiv Ci fon..u'r fi .,uto Q Iic Iu,, a lnrqc3

tiiiich'i elf 11111. ci t . fi IU1- I %iI iI 10 C jIlc c ,It I :It W itil I rioL

I l t. . i ' ' . ."iI ,, ' , , , ' .

-76-

co=odities sibject to competition for movements by rail and

truck would havo higher costs of handling and demand greater rates.

Multiplying tae ave.'lze length of haul and the average revenue

figure yields the average revenue figure of 120 cents/Cwt. Using

an average load factor of 600 Cwt. and the average cost structure

of railroads presented above, the cost data in the following

table were calculated. The operating ratios for the rail mode

will be 85% aid 65% representing an inefficiently and efficiently

managed carrier.

TABLE VIII

RAIL AVERAGE COSTS PER &dt.

Load factor 600 Cwt. 600 Cwt.

Opezating ratio 65% 85%

Total revenue 1.20 1.20

Profit: margin .42 .18

Co~mon costs .156 .234

Total out-Of-poCkot costs .624 .816

LIne-haul costs .418 .547

PUV coar': .062 .082

Terminal costv .125 .163

b l lg "oa~s t4nA19 .024

(..) -77-

Piggyback Cost Data

Once the average cost categories ha%? been developed for

the single modal carriers, these costs cn . uied in computing

the cost of piggyback transportation. A piggyback movement ntaals

using a highway tractor and trailer for pi kup and delivery opeiations

and transferring the trailer to a railroad flatcar for line-aaul

transportation between two points. Hence, piggyback covements are

coordinated rail-truck operations and as such, the cost of these

movements e.mbody selected cost charecceristics from each mode.

[A] ...simplification of the transportation process inpiggyback Is the by-passing of railway yard classificationat the origin and destinaticn. Piggyhack operations needas an initial classification yard no',ing mre than a trailerramp site with a parking Jt. Classification of cars bydestination usually can be ac!omplished in the process ofloading tlh trailer onto flatc.rs at -'he origin. Alrjclassificatio-n yards enroute can be by-passed to a largeextent. Switching enroute under most circu.stances nightbe done by diesel road-switcher locomotives simply droppingcars at t:all." ramps located outside major cities.

..................... . . . . . .

Like trucking costs, pigyback costs usefully aredivided into terminal and line-haul e:;Fenses. Terminalexpenses, in turn, can be scparitctd Into those con-ventionally asEociated ViLhi truck tra:sportaticn and thepeculiarly p gg;back c,sts of loading and unlcading thehighway tiallcr onto nd offi the rall':ay flatcar and thecost of rjaLkl up Li.,! p0 jgjy)ac: traln.

ligSybac" pcl.up and dltlvvur' coiLRt are cusrntiallytile Fnam . '1; tlc' , .coclc(,1 nIg tvL'- tia.rotd truckop( r tl-:'s. 1 2

I,.)i hp,L c:].,y, I I -I 11tu l r. 0:., 1 or 1 g1 ;.'rtl;" m ov.?-

m.nts wIil bu thuruo (i Lh rallto,id adJu twk for ditf a.'viiua Ill

' aq I l 1 I i i. "1 " ill

-78-

load factors. The piggyback PUD costs will be those associated

with two highway trailers plus a cost of S4-or $2 for =oving

each flatcar into loading andunloadin; position and building and

breaking the train,, for the railroads with operiting ratios of

85% and 65%, respectively. The piggyback ter=mnal costs vill be

those of the highway mode plus a $9 or $5 charge per trailer for

loading and unloading the trailers on and off a flatcar for -he

inefficient and efficient -ailroad respectively. 13 The piggyback

billing cost3 will be the sum of the costs for r'-o trailers and the

average rail billing cost.

The piggyback common costs will be obtained by dividing the

ral. and tru:k cost contrIbutions for this type of =ove.ent by

the corresponding percent variable--Q% and 90s, respectively.

The piggyback profit contribution will be determined by dividing

the full costs contributed by each mode by the appropriate operating

ratios. For instance, If the out-of-pocket trucking costs were

$27 and the o-it-of-pocket rail costs were $32, the out-of-pocket

plus comnon costs for thil movement would be $27/.9 + $32/.8

$30 + $40 = $70. The full cost of this covo would be

13tr. G. W. Tolfer, kan.,cr of TOC, Burllntcn :orth.rnRailroad lr,'icaLed that the pro:.:.c.tw c0 t of a n" a'dunlocAllng t,-i r olio lI ...... - ; "I . /tri ..r. LA Ler

also Sta;ted L;': th'e SWI cc, s;s. for '7 n l J at ars I:s-ol o a d I i i i n a d u --. c ' ' ' 1- , :: a s 7,1. . . -" : " ') r C n : C .

and Lhat a s'iLch!r o' v avcrae ccui h.und1e 2 ctrs/huur,Thu.i the sal I i~i'* and lo.iin,; cost:, u-e. in t.; stud>' are i, t.e

ranj,'V that I5 rL;rc-:::..,: v . : f thy r ! al'...rid,

-79-

(30/.91 + 40/.65) $33 + $61 - $94, for a trucker with an,-

operating ratio of 91% and a railroad with an operating ratio

of 65Z.

The cost data which will be used in this study are average

costs, There are ceitain dangers involved when one uses average

.:L.. Average costs are dependent on the volume of "production".

If one is dealing with a well-eos tablshed on-Zoing £rti in which

the average "production" costs have rcmaincd relatively stable

over the past several years, then one can feel relatively at easeI

In utilizing past average cost data for determining future courses

of action, at lcast within so-e rangc. For mo-t large transportation

firms, the average cost per Cwt. moved are relatLively stable since

outputs are large and e:pected incr v.: ;r In output small by

comparison.

The TranH)ortatlon Syqtc.m,

The prev'ous secton has f the costing categories

and presented the average costs In each caLegory that will be

used a.; the foundation for analysis. The t'ection also indicated Ihiw the various para:uLvr wuuld impact on each of the costing

categori es . TVIJ Rectcioi will IdCitify the transportation systen

t1,1 l te '- be investigat.ed,

The tranniorL ',t :ti y!,cm v.iti cl ..'lll be .tudJL,,' conslsts of

the five poinU dL.p; Ict.:d bv]ow, Eiih1 point Is oerved by a railroadA B C DI;

75 13!) "- J . - -" Our)

-80-

and a motor carrier and a tr'ansportation company composed of the

two carriers. Point B is located 75 miles fro-. A, C is 150 miles

from B, D is 300 miles from C, and E is 600 miles from D.

The average costs presented in the previous section will

provide the basis for dc'.ermnng ,:hat the average costs are between

each origin and destination above. In other words, one would multiply

the average line-hnul costs by the ratio o.4 the actual distance/

system averate length of haul for the todes, then readjusE cornon

cot s and the profit allowance. For Inst nce, If the average line-

haul cost was 105 ccnts/C4t. based on an average load factor of

300 Cvt. and average length of haul of 300 t, lce, the specific

line-haul cost for a movement of 300 Cwt. for 900 miles would

be (105 .00) 315 cen.s/Cwt. This line--haul cost would be added300

to the other out-of-pocket costs, PUD, tor.itnal, and billing.

The out-of-pocket costs would be divided by the % var.able for the

partIculnr mode to obtain the co.Mon plus Out-of-pocket costs.

This figure would be dlv~dLd by the opertlng ratlo of the modt

to deterinine th0 fulll co'nL of thc LMVe:.MicL.

Cotsidoring MovMOnts both to and fro'i specific points in

the above tr.i:iqporLatio n syst m, ticir ere 20 dlff,-rent route

fegmtits, ' nhrv run;*.ruvc:;ncnLs, fro:n left to rlght, A to B

(detilgiviod A3). , AD, A, I:C, BD, l;, L:), C, and D.. Thle

corruspnidlin,; :;!ovc:; In th- uLhvr dl rt tion are ED, IC, l, I..A,

DC, DB, I)A, CB, CA, ani ,. v.::;Cntn in both 21 rutlnli rI hotld

bM S cON;l rL d .:. :,: in ,. l. :MeL r,:.-.-rt..' ,y,;t :

--------------

[: -81-

not have balanced traffic patterns or even the sie geographical

conditions in both directions which cause the costs of movement to

differ vith the direction of movemnt. This thesis will, hcweve:,

limit the route segments considered to those unidirectional move-

ments from left to riGht.

This simplification described in the previous paragraph will

reduce the computationnIl requirements by cr.oe half and will not

affect the integrity of the results. The reason why attention

need be focused only upon unidirecticnal movemonts In this th.siq

is that each factor which would n.akc the dirr-ctional moveents

have differing conts will be ox.m,-nud for effects at both levels

(in dIfferent runs) during thL SinUrlLicn. So in actuality, both

directional movements are being consldcred in different runs, even

though the lettcring of the rmove-ments is unidirectional.

i

A

-82-

41

CHAPTER IV

£XPERIU.N'AL DESIGN

This chapter will focus attention on the nature of the

experimentanl decuCn as well as the analytical and statistical

methodology used in the dissartation. Your opecific arcas of the

experimental design aspects of the disnecrtalon will be discussed.

First the problem of realis in suimulation will be addressed.

Second, attention will focus nt the nature of the output of the

aiMulu.lon. Next, tho perforrnn.ce ticaourea used to describe the

output of the simulation will be prosnted. Finally, the specific

experinc'ntal deigr which wat utilizcd will be discuQuod. -a

The Problem of Rciallsm In Simulation

One of the taout imporLantt prublu js ont. faces when uti11zI1n&

simulntion to inveutigate a now or proposed system is that of

insuring thit thu model is an accuritto repreo|nnLatlon of thu

syste being annalyzed. Chervany haq succinctly suirainrized Lho

problemn of using finul.ation todis to atai.yzo real world probleAn-

at follc:j

II llthe' d,_'Ivt, op.'1 ,.0'jjinI d a ly.i l of t," ' r:rdi.] Ithe, riojit vr i;I:;t .:t : I. L ,I.!.I coiic4,n rl it, rall :.:i of LCm"lel - 9'8 r lgap 'p -t'... A hvi .,:at .. , tud.l ILl LH A

pcrfecL oI u t.'.tty fur ''Ic Iitrv:.I', v., Ai .-. . hbvVnrilmI'i f.Ct(irn u:;dL'r c , , :.' , vri.d i;. ' (or ui.cv

0141 onc) (I. a t.117.0 '~ W!, lit - 1 W-LI :10 1LI l4 i v.,c"cowjt itl, . Thl ij ::l t ic~i •COI~j imt , 111:3 I~::~ yIL . : yi.,h Lifly, hu'-.'. ,r , I in I-.

A hi.vkd r ,i 0 L(,:L, i ' .,' t.'.LOA . I I.II..,p r :. t .:1 p r, , .':I of I.,ull t .l it ; h it l , : .

r--, I ' ttI . ,, I 1,,I : ., i I, . i'',ILn r'

____ ____ ___

* -83-

reollsm arises In all model-building attempts, it ICCei! tobe even more troub4eao:i'3 thtan iinual in siv~uliition studies.

Although this realim problum assurios many forms, Itseems that there iare two basic h~crarcial levels that gcnerallyappear. Vie first level concerns the structural form--rmodulsubsystemui and oubayctem' interrelatlonvihlps--of the modelbeing simulaited. The second level forumas upon the determl-nation (;f the spucif ic datnt I npotA that, are requl red LO mzkethe miodel Analytically operationa~l. 1

StLCurAl Realinn

The problem of structural reallt: focum on theseleiction of clomcntri to be Includnd In tho vtadel ind thomanner In which they aire connected. The following qunoutonetae typical of thorle Lil-1 MUSt 1)0 anWWrCd At thill lt!v0l: DOthec variablin inClut!d 11. Lite mudel ai':curaiti'y rid luct OldaImportant connstruct.- iound In ~vlt7Aro the vatrInblooomitted frou tho modul Aufficiliily ui~irpornitt that falluroto conni1dvr then will not bins the. ctmiclu14lon erawn irti n Elaysipof the ffodql?

Sinco thil firm (a) boini, modeled fin this project..,Larelhypothetical, thLe -AHly to teat thia modul for Structuralreallt J isomewhat lini toil .

oleo. C. P. *bonii hi mtj 1'Jg" Lvd ain ap-pronich to vAlidatlonIn skru) ntion of hyjpoi.lcti cal oz L~ini zaitt., , Concorohl his~model of dcl cion mtid Iift 1maii.Ion 11y!!1Wns wi111 Ahi u a Yput 01CIalfirm, he atatum:

The firstL quitsLon to bti aloked :ihout, ourmodel would prtipcrly bV, "L~j;ri the irode] cor-rcr-poiid to tho en1i wurld?'' In uthev wurd:110"DJo thv ilfur'.f I I~ i it: o'd J I oi lo yiltteVh;11 roIit il nhlIy* i!jJrCtI(At r IM wol.1 U i I i i CCjiII?"

WC Woulid nr) I v xp',. wI 't e*I, I t't iiod(

all 1'oil I. Oi .j ('~ t: *0i'vC

We do tl .', i'.V t h i t Li hu lii.''v

Jr. if t1I'i1pioil I A i~u iL J til of r- c i

behavior. We cannot, of course, completelyvalidato this beliof, but 'what wo can and willdo ii. to set forth the major Ingr:dcnts ofour decision rules for separate exanIKlation.We will attmpt to Justify thane rules by

relating them to existing theory In thescientific litersare,of economics, accounting, orthe behavioral sciences, or to the lit, ratureonl bu ;Itieas Vriict,'e, 2

The structural forn Of thiq model was devluped along the

lineu suggcsted by Biunifl. It shuud be strOesed at this point,

hwever, that the model is a simplificatioti of the real world. Tha

hypothetical flrf,,o modeled In Chapter It are auuined to ract txi all

5a tuAtlons to mnaxiti ize their profit CotLrlhutLoln. The model has

not considored anly tocial welfare rola that miaiy modern buoiness

concerns hav' odopted. To the. extont. that transportation firns do

not attempt to max1Lmize prc.'It. whother duo to the fact that they

are limited goal aatnfiv,' jr due to the faCt that they hava some

social wlfaru role to ul ill., tho rvoulLt of thit project are

limittd to Ljit degrv,.

Furtiertore, an mcntioned in the previous; chaptcr, the Maiilna-

ment: and oporationla p rsniinl an w:lt an the org'ni nt1A~iJ atructure

of these firi i are conaidard to b( capablu of roachtii thutir goal

of profit xiaexmlation itt, Joct to the uncertairitien of predicLing

the muarknt u:r, tnlu1c.ibl, The dJ vI!irLdxL vn ha i 1no fltiL

addr.'oi, d O t ] l '(J i li !:,.i of how10 .t. i-duj.l IN op"r;I J fll!" I:.i'';L ih!

. hid. , p 1 ) /,,.,7, 'Iu,) L 1 e.i, (;h r I'i 11 W MI ti , :. foitlI

11.,1 1 , JiL'*, ! 'Ji' 9t * ,' "

-Sz

-85-

accomplished to have the equipment needed at the right place at the

right time in order to maximize profits. !'hile these are definite

limItations of the analysis, all firms considered have the benefit

of these slmplifications. Thus the end result In that they will

hopefully cmncel each other out in the study.

What happcns when the results arc attempted to be applied in

the real world? To the extent that irporflect scheduling cporations,

Ft and 4iffora,t levels or quality of ptrsoitnel are found in the real

world, the object of profit ma'imization will suffer. The author

u would argue, however, th.a, Cven ith these quallfctationf the rkulsa

of the study should atill be a valid indicator of which combinnations

of facLoru or fli'w pruc. sauin ipocifIc level; of factori thould

lead to iIgnificait economle as a reoult of foriwtng Lranspor-

tation eompanies from two ulntlo modal carrlorr,. At least the

factors which aro found to be sajnffi c lnt in the atialy.1ss of th,1

results of this simulatlon should prc,:ide a bsIs for predictin

which cotnhlnotlons of factors should lend to signiflcaitn econo.-leq,

There is oien other nrea Wn whic c thu inodul differs from the

real world, Thii modu whilch hav bot-i devalopcid considori a

trcn portt lI, systeIn oe'Cr h I ci or, r;1k ,, o n Y o r a AIIron A Ic) p. f- '.

and erti, r tic'I rrv cu::p.i:ry :iiid ), :1'' Al tho pu. Ow ri'tItlg reaults of

thi n w.yn r i'Lbh the v t'cralt hfli w..t.1 ihlt Ar , .. ' obtalwL.d wArell

ii tr:lflm 1O t1*ntJui Coi p: , opi'r t , vc / T l! U l :,1 flyf '. 1, 1r S iiOL

Ilikely in Lh, rt-l wej Id LthtL n tr,, porr:'rtf.'ri, 1;v,'terl, or a cu. -

i t , It/I i ) p r t I -:I of n r, I , t i. 1, . ' . , :1, o . .' It I t. ,d

-86- 1

and one trucking company. It is far more likely that on a given

system, or segment of a system, that there will be more than one

firm in each mode competing for the available traffic. Thur. the

element of intramodal competition has not been considered in the

model.

What effuct will this simplification have on the interpretation

of the results of tne analysis? In essence, this simplification

of the real world aseumes that Ce individual railroad a:d trucking

company can maintain a relatively fixed portion of the trunspor-

tation market, when confronted with compctition from other rail-

roads and trucking firms. This ancumption may be reasonable if

the two comptiniev are wall stablinthod fivi s e i plificaLtcn

also assumes, however, thtat the trunfportation company, formed

from these two firms, faces the same relative denand patteni an

the two firms when they were operating sliuly. This may be a

tenuous s:;umption. There is, however, no mort-ts of tWsLInU its

validity short of actually for mt| the Lransportation company In the

real world, The reader IN left to his own co:icluAion on the validity

of this assuniplion. Th. effect of the non-validity of this

sbaumptluII wtul]d b to Inicaro. (or do.ro:a-e) the ec.no::.c I rpact

trniinporLati on coEipat:ic would hav dup,;,d ni ui,0i o it ' LI l- the

forma i O;i of Lithe c:qpIL:y Incret ncd (or dncrce;.;d) the r,:tr'.et Ah:,re'M

Of the crU:;,b li 1 firvm,.

;=I-

-87-

Data Realism

The specification of values for he various parametriccharacteristics, the second hierarchial level of realism, isalso a more difficult problem in simulatiors of hypotheticalcompanies. ... The currently suggested techniques for gener-sting input data are bared upon the assu.'ption that thevalues can be determincl from the analysis of real-worlddata.3

This has been the approach u.cd in thir study. Care was taken to

use approximate industry average date for the costing of move.n.-.,

As pointed out by Chervany, howisver, another criterion that should

be used tor data generation is that the numerical values use-d be

JnterilqnlSy consistent. "This means that the absolute level of the

nuaerIcal data, whiile Important, is not as important; as the inter-

relationships atrong the covponeuts of thIs numnrial data sot".

Thus care was takeii in this study to use data which was internally

consistent, that it the relatlonshipa of rail operating ratios,

load factors, and so forth, to truck operating ratios, load factors,

and so forth, are in the relative proportirIs that they would be

found in the real world,

The OutpIut of t'o .IMUlnttoR

The simulation noodal being n tudled -wno coutructed to doter.iine

the result.mt econoni C IMpact a tran.portaLion counp,,ny would havo

and if this .'1p7CL Would be au1,qtai 'Ja!ly Iffcrk.:1t than th M :o

iiiny, ,s n d, ] co:'j:'II : of -. ! ' ci IL 1:. c '..i' d, I ich run of th

I1bid., p. 14,).Ed .°J

-88-

simulation generates a number of statistics for each carrier--

the transportation company, the railroad, and the trucker--that

are a function of how each carrier prices his services and allocates

his equipment. Each run Indicates, for instance, the rate at which

each carrier decides to price his services consistent with the

rate makhiig schc.e outlined in Chapter for each route segment

a well as the specific capacity allocations each carrier makes

on each route segment. Although such detailed information may be

neccasary to explain why a certain factor has a significant bearing

on the results, it is much too detailed for any type of logical

prenentation. A numbcr of su .nary statistics have therefore been

developed to aid ii the presentation of tha results of the study.

The remaindar of this section shall present these surnary statistics

or performance measures.

There are probably an Infinite number of perfor-ance measurf.F

one could devs-1 o oeaoure such a nebulous a thing as the c,'! -.-

Iuiprct or cc,:nc ic co:.-equencc of operating a transportat1e,,, .-

under two diff.rnL organizaLioaal approaches. There are t.... .-.

dimensions t.o the probl.e, of measuring the economic iract........-

tation .c:ipaxy tr ie sInglu t:odal carriers create. Oe', sl-

azte'pt to n::.d rn.ca.urep which Indicatc hw the Opcrot i,:.V; o"

,-.i-palen iffct t1.cmj,.Lv, o iSt %%n!.. as ,.:c . .1, t !'. ,.. .(c2

their u.,crr . Vit this i t, nu the Ililo 'nd M1:- !. O:W , .. ,:..

mea sute i'll ,e tj'ta't tr a y . th e 7C!U !e t' uf tiI l v,, .. .. :l

ru't; ; (' r!.. ,,ad c 0',t L, b tL J:!; '. : " f c ' ori ", t . ,. .-.,; :

th: ".arr!. r+,. .,,t Lhe tot . pu 1J : paid !or L',.. tr , :.ij-r':.' '.: : .. -.....

-89-

Tha nature of the simulation provides a unique opportunity

to determine the effect the creation of a transportation company Iwould have on the manner in which movements would take place vis-

A-vis single modal carriers. Therefore three statistics have beenA

compiled to indicate whcther o. :CL a transportation company would

make movements which were significantly different than the manner

in which single modal zo-pa-.Jies ,cace movements in concert. These

three statistics are the number of Cwts. moved by truck, amount

o1 Cwts. oved by pfg',sck, and the ;iount of C..4ts. moved by rail

.joxcar. Thc..e ptre:.-t . ,ill b2 kept for the tra..srcz-

tatlon con:panv, '_.e rJT! oa .:z' 0-? trucking cowpry.

.a Ccntrib, ". ,

There i:. an additional elerent of cometltlon present when

t. o single modal companies atternpt to obtain a portion of the com-

petitIve traffic which a transportation company eliminates. The

sinle modal companies must not only cope uith the problem of

estimating the size of the market subject to competition but must

also estimate what prcportion of this market they car. obtain.

'nr. - LSstjmLates of the proportion of the market they can obtain

for h -articular segnent is directly related to the amount of

c.:pacl v the, will allocate to that rer.. This is so, because

-is will allocate their equipment Lc maXImize their

expet,: d contribution to fi/ed and/or common costs and profit

warg !ns. 71hus th2 sikqjlc 2:.Yal carricrs h.ive an additional t

:" u: .. L: niy to r,..j.- with th t tIrar,. po t. iun co-,1.p:i:'y Ic...: nut.

u: ui

77-

-90-

A transportation company must only estimate the size of the market,

then attemprt to allocate to that market tl.e most ?rofitable oethod

of movement. It therefore appears that the expected contribution

of the carriers shoul4 be examined for effects of the factors since

this has a direct bearing on the allocatiortL of equipnent and

correspondingly on the actual contribution of the carriers.

Actual Contribution

Actual contribution or total contribution of the carriers is

probably the most important perfoinance measure which reflects the

internal valu c the organizational form. This measure is of

extreme inpor c if one is trying to examiine the merit of

establishing transportation companies. Thus the effects the

factors have on actual contribution is at the heart of the study.

The actual contribution the s Ingle modal carriers enjoy Is affected

not only by shippers' logistics constraints, as is the transportation

company contribution, but; in addition is affected by the shippers'

modal selection policy. Since the competitive market could be

allocated equipment from both single modal companies, the sh1ppers'

must choose the modes to use which of cource affects carrier contri-

butions.

Price Pnid for S'....-rices bIy_ Users

The price paid for trinsportatio nervice 3 by users should

indicat' to 5s.c u:,I ::t, th. c ,omic coL oi the traz.1pertatfon

oc. ; t. c rrlyi r:-r o'ic r c. "J o : ippr; r ,.

-91-

This performance measure is probably the most important external

indicator of tie economic value a transportation company would have

for Its users. Thus the study hopes to Indicate which combination

of factors, if any, would significantly affect the price paid for

transportation services under each of the organizational patterns.

The total price paid for transportation services is a function

of the demand for transportation services and the rate charged

by carriers. More specifically, the total price paid for transpor-tation by shippers - r X r

ijkXJjk ijk'

where,

X - the amount of Cwt. moved by mode I between j and k

r -k = the rate/Cwt, for mode i between j and k.

It is Impor'tant to stress the relatlonship of the price paid

to both the amount of Cwt. moved and the rate/Cwt. In other words,

the total price paid may increase if one organizational form satisfies

more demand than the other. An increase in tot&. price paid could

also occur If the rate (price./unit) increases. Conversely, a

decrease in total price paid by one organlzatanial form, can

occur if less demand is satisfied or ii the rate (price/unit)

decreases more thin the other organizaLIonal approach. Thus one

cannot tell by inspectIAon whether an Icrcase in price paid resulted

from an increase on the rate or because ure dciand was satisfied

eiLhur by the transpor'iLon cQQipany or .,! singlc modal cc::panies,

The conversc !s true for a dccrcce i:, total price paid. tecaus-:

of this rc lati,: hip i;c.-.1 care t:j 11 be L kcn i r the di,; u;

-92- '

of the results to indic~te which of the two variables created

the Increase' (decredso) in the price paid ;erformance measue. -

Amount df Trucking Movenents

One of the arguments put forth against forming transportation

companies is that a transportation company would be dominated by

the rail mode with a tendency therefore to drive independent

truckers out of business. This study offers the opportunity to

observe how a transportation company would allocate movements to

its composite modes. The criterion for modal selection within

the company when not constrained by shippers' logistics systems

is profit (contributi6n) maximization. It will, therefore, be

very Interesting to determine if a transportation company moves

a significantly different amount of goods by truck than the

aingle modal trucker,

Amount o Pi $-hac: 'ovements1

This statistic is being tested for significance for the

same reason ad the las.t measure. In particular, It Is directed

at determining if the or;anizational approach has a significant

effect on the amount of goods moved by piggyback. It should also

be pointed out that this analysis carried to its logical con-

clusion should provide the ans,'ers to what Is the optimal equip-

ment mix for a trcispo-LUttun co;' pjny as %,ell as for single

modal carriers.

I:

, ,I-93-

Amount of Rail Boxcar Movements

The statistic is being tested for signi icance for the same

reasons as the last two measures. -

. An alytical and Statistical Methodology

In the preceding sections of this chapter, the problems of

utilizing a simulation experiment were discussed and the performance

measures to be used were presented. Chaptcr III discussed the

test factors which are to be analyzed in the study. This section will

Introduce the analytical procedure and the statistical tests which

will be employed to ascertain the effects the various test factors

have on the selected performance measures.

Hypotheses to be Evaluated

This section will present the hypotheses that will be exa.:ined

in this study, but before doing so a rec3pitulation of the experi-

mental design developed thus far will be given. The economic

performance of each single modal company r:'d the transportation

company will be described by six statistict--expected contribution,

actual contribution, price paid by users, tind the amount of Cwts.

moved by truck, pigyback, and by rail. The principle focus of

the rescarch will be placed on how the two orgainzatfcnal approaches

differ with respect to the performance noasures. That Is, the

analysi.,; of the ouLput Uf the simulatilon will be will be conce.::%O

with the diffcrc;.c. bct,,Con the transportaL lon company pcrfor:.-nce

rea,;urc! Stid th2, Su:LI of ti l)crforl.:anc, r. eisures of tLh sing,.Ie

I' ... -"- I[" is - r¢ i " ri , - ' i . .. ' - • . 7 ' i i ' - . . .... . .

S. .-:a

-94-

modal carrie:a. For each simulation run which inrorpo-

rates u difflrent combination of the inter.ial and external

route factors, these descriptive statistic3 will be calcu-

lated. The hypotheses which will be tested are concerned

with: (1) the, mean differences between the performance

measures of the tranhport~tlon company and the single

modal carriers, and (2) the effects the c€zabinattons

of the teat factors have on the economic performance

of the carriers.

The analysis of the test factors is concerned

with whether the test factors "cause" significant changces

in the values of the perforuaance meu-,urvu irom their

average value. It is also of interest to deter.ine

if the mean differences between the organizational

approaches are statistically significant. In other

words, one in interested in determining if the per-

formance measures of the transportation conpany are

signifi.:antly different, on the average, from the s=

of the measures of the single modal carricrs. The null

hypothcses are for each pt'friance - aIr Lhat the

mean of the tranup;ortation company and thC Mean OF the

sum of the single modal carrier pcrfor.i::ic-.. z,-,a:aurva

are the same. The altt.riam.Ive hypoticsF' .irv thaL the

mean:; are dlffernt. The nLurc uf the output, bolIn

that the rand"a componentr and parametric values were constant A

across the o-ganizational alternatives, inlicated that a t test

involving paired differences was appropriate.

Since the analysis of the result Is l rgely con,-rned with

the effects brought about by char-es in the levels of the test

factors, a much more detailed description of the statistical

procedure utilized for this purpose will be presented below,

The analysis will be mainly concerned with: (1) determining

if a change in the level of any single test factor has a signifi-

cant overall effect on the perforuance measures over all combi-

nations of the other test factors; (2) determining if any combi-

nations of any two tezt factors has a significant effect on the

performance measures. 6 The first type of affect Is called a main

effect, while the second is referred to as a firut-ordcr Inter-7

action effect.

They hypotheser which will be tested are concerned with the

nature of a11 of the a.-ifn effects and first-order InteractionI T h pedx1 a cuso nteapr ~acc~o

The nature of the t te.qt which was u.ed Is described in detal1itk Appendix 1. For a diccussion on the a'ppropriateness- of this par-

ticular tnot, the reader is referred Lo Gu-_.nlher, o,.,, p. 24.

61'he word coc..bltaLion ag u;td . bove, refers Lo the :umber of cxpcr-

mental co:nditi.ons or ni.is-..c nu:7,cr of arran:.:ientj of the leveir of Ith- tClit fattors that are possItI.e In Lhi th j.j,-ri twtta1 vULLi11G.

7Thr,er are also r; ccnJ-o-Jar, Lt!krd-.orJL'r, fourth-order, andfif I~-rdu Itt' rt: .: * ffcc~.*icI ar wLi LLie teoi t factr,

which will 1: 't I( . L!;- I .i' i'. , : c * natLIr(, O LheIspec fi1c c..:p r ,., ' . ;i ;'$ ut ., Ic t']il; ': dc . : not pro.'d!.for Lhu:1,.;A:, of v c L . "iik f-,tu:' av~d r,oic :; fur t.

&LeleCCU.,fl 'I Lhe e..ic :...::il d,:.;,;., u -.1 in "ii:; prujc" ci " be

Si ]L .

effects associated with six of the seven test factors selected

for analysis, The test factor which will tiot be tested for effects

is the size of shipments factor. Instead of testing thin factor

for effects, the entire expe-iment will be repeated for this

factor at each of Its two levels. This Is because, as explained

earlier, , the costs of moving CL/IT shipments differ from those

Involving LiT movements. Also the choice of how the movements

are to take place differ for the two sizes of shipments. In the

IT/CL shIiptucits, movements may take place by truck, both of the

piggyback miethods, or by rail, wheread LTL shlpments may take place

only by truck or by both piggyb~ack methods. In capsule form, the

test factors which will be analyzed at each level of the size of

ahipments factor are (cc& Table 1, Chapter III):

1. !tigtjivatrcun loiw truk0c rating rittion-An Analysis of

the effects of a 99% operating ratio vie-1-vis a 91%

operating ratio.

2. 111eh vervus L_j:Lrail pcra n ~aor--An analysis of

the effecte of an 85% opur~ting rntio V1S-A-V ' A 65%

operating ratio. -3. 1^ow versus hl.), truck lr~ad factor-An analysis of the

effects of mi nvc'raju truck load factor of 100 0C4t.I

vis-6-vi s an avkrage truck load factor of 300 Cati.A

4. Low versus hi ;i roi l.oaid faictor-An analysis. of the

~~1 ef (Uct; of an nvver-igr r.-i I load fnectur of 4e00 CwL.

vi --- vJq -:~:i .veur;u; r-. I Jc'd f.: tar of P3OO (

-97-

5, Hih versusw _sUstom capAt--An analysis of theF4

effects of excess system capacity Via-h-vis adequate

system capacity to met system w'ie demand.

6. HLh versus low lpvo of shin' loc.t!.s constraints--

An analy~tsaof the effects of a s)ste constrained to

a let$e e::int by shippers' loilstics systems demands

via-h-vis a system constratised very little by shippers'

logistlcs systems.

The null hypotheses which will be tested are that none of the

main effects or Interaction effects differ from vero, That IF, the

null hypotheses state that none of the tact factors (or combinations

of test factors being considered) produce significant differences

between the parformancc moeaures of the trr.tsportatioka company

And the single modal carriers, R(ejection of any of these null

hypotheses Indgcate that the specific te t factor (or €ombitation

of test factors) Lin. hAve important Cause ittd effect relationships

which affect the econoxic performance of tLe transportation corpany

vie-Ai-vis the mingle modal carro.ta Conw, tsely, non-|'oJection of -

any of the null hypotherea implies that thV Specl c tCet filc.Lor

(or combination of test factors) 'njy .to. hv:ve Important caus'u nnd

effect telaiionAilfia which affect the ecow:omIc perfor-:zttce of the

trannportatlicn corpany as compared to the :iitjlc tod~al carrlci~t.

In the precedin6 %Ltvm'-n L thu ord T r.iv has beou strer;ed.

[ ThIn Js because It. mlut i be (r,,hnslI d Lh13t Lh(, u tF.1Ls Which arc

obtalor-d prt.ll~ to Ow. t-OIill. of OW;~~t~r ;u 1A" .C,.

[.

That is, the results are obtalfod from the .aipulatfon of a

mathematical model which was formulated to resemble the real world

phenomenon. Before the results may be scatcd as truths research

and analysis In the real world is required. The results may stand,

however, as working hypotheses until vorifi.d or disproven by real

world research.

frsctional Factorial Iworimets

In light of the preceding dlc,;usuion, the experimental design

chosen for this projuct muot allow for the analysis of the main

effacto and first-order interuction ffecti. Factorial experi-

ments allow for such analyala. 8 "A foctorAal e,pcrIm=nt It one

in which all levels of a given factor are co'ibined with all levels

of evary other factor In the. experiment." 9

In thts project six toot factorp (oacli with tuo levels) have

been selected for analysis for cach lovel of the size ofr ohipent

10 6parameter. This may bo describud au tvo 2 factorial exporl-

mentm, ona for each Mihiumc:it Lil a. lluior -ino for ono corn-

plato replication of the experinc:It there are 128 possIble combi-

nations of factore fur each of Lhu tLhrcu ..irricrs that muill be

For an eXcIl]i:nit dli cuaslon u n tho a ivwutu(!. cn"d usefulnu.::i

of the facto'lal tu:: n for . i. 1nx; r!:Ie lt nOf eOW., 1ature 'A

thin pL,) l('L: , P'IV C1 1 'VW 1'.Y i. 'j P " . 5(1

9 Lh itr( I. 1hick<,:, Vudru:.t.'ni ( I n -c :' In L-: , Af

p. 78.

0 s v 1 ,_t .jr., c , ,cr , .

simulated if this complete factorial experiment was run. If

additional riplications were added to increase the reliability

of the results, the number of simulation runs required becomes

?von larger.

In order to reduce the n=ber of runs required and yet to

obtain as much information as poosible concerning the main and

first-order InteractLon effect*, a one-half frect:3nal ractorial

design was utilized in th1s utudy. A fractlo' . .t.rial desagn

uses, as the name Im plies, a fract . 'rIp>C*Le 0.. A rectorial

experiment. The one-half fract~onal faCtorIr -"h1dih jas

used In this proJect is identified In Table IX.

In this poartiulir djI|gn Lhit n lin eiecti 4.iid Lbt|, firit-

order Interaction offects aro confoundod with the hIgher order

interaction effccto, The hlrhct ordcr intcractlon effects with

which the min, and first-order Interaction offectn are confounded

are refrred t. an the allasos of the ncio a;&tI first-order interaction

effects.12 Thus, In etfect, if one of the null hypotheses is

'ffccLISa -u confoulld.d IL.'th One unoLhur .'hon thu camo €,ta

and enttIIaLor of tli' il','. ,,* ' u 0 'o ' rmilo those (both)effect". (Wil11 1 : ' 2denh, 1, T I.:n v': sd. I n of ,x er.:.t(Belmont, Czill foin Ia: Wao.,worLh Puo ,th: :i ' :'!, Inc., 1968), p. 3J.2,

In e 1 gn uSCed II this study oach naln effect and first-

order iILt: iLl :, t'fiecr: 1!1 colfOUrdJ(d WI'Lh oIC hi;'.;vr order I nter-action o ff,.ct., Th,, doffltii; coitl. L. " il:h I (ll,:iLo' Which LifF' :L

are confuJnJL.d, Cr L1I:' dI.:,lg In .. ,., lo dc (';ine tn till 41.)Ut anl (e ff . 0 2,c . : 'u11 :2, ] i Y J: :,:' t, I : '.'i C;> ,': . , L by' .I,.' 'L fi:tc.of tilli t it, (CMi : l8 t , ,-,L ,r ,. i "liil CIJ Lii .. i bu f of

ithe All iflucL 15, tic., ( .i ') '".i c CL.

III

'1iLi

g- .

-100-

TABUE IX

ONY'-IHALF FRACTIONIAL FACTORIAL DESIGN

S(I) ab ac bc !.

abe d cd bd M

beef acef abef of

ade bdef cdaf abcdef

abco ce be e

J6 abde acde bcde

af bf cf abef

Lcdf acdf abdf df

rejected one would not know which of the effeets (Lte alias or

the effect of interost) caused the rejection. In order to

circumvent this difficulty, the higher ordcr effecto will be

msrumod to be zero. If the render has some doubt about thcb validity

of thJs assumption, the results of this stuJy will have (for hint)

at least, narrowed thu causative factor to the proper alina group. 1

'3 T e appaarnco of a s-m.ll letter in the table moans a factoris prvsvoIL aL Lh, HUCOi, (ilt:or[ -,Lv) lvlT1, 1he absenco of n letter -moans A fActor I at tLi1 fNtSt (ILaudard' level. (1) siuifies-ill facturr ura at ti, £li':JL luVJL,

For a muci more dut. A11ud ']l;CU!;Jon of the c;.)nc!ptu dlncugE :,din thin sn c ttl, , thi r&-;dvr J:, r r d t.o Che rvn y , _J )p. 2,8-

257.

..:= . .- .. 7.. . . . .

Sttistical Analysi'

The 3ath2uatlicl model of the fr-ictlooal faLtorial design

ued in this study i precanted belowv Is

(1) XIjklmn + Ai + I +Ck + DI +Em + T +

ABJt+ AC + AD + i +A +

BC~ + BDjl + BEjM + BF + CDkl +

cc + Crk + DE + DFn +-

+ CIjkmn

where

X the observition of the parforrince

veasure for a 3iven run

U " the iean oboervation of th e perform~ance

measure from all runs

AI . ., t " effects of the factors

Al tJ , , F, Y n - firat-order ititrefction effects of

the test !4ctor-

tJjklm " unexplained variation of the obser-

vation of a particular run (error

teom) jI, J, k, 1, m, n - I If th obervatlon var. Ccerated

from a rujn .n which rho test factor

was at level I or 2 if the tCst factor

wa at ).ev.l 2.

FA

1 clku , 9p.: I., UP "2'"D.231.I*

. . ...

-102-

The major f~cus of this research project will deal with the

statistical tests (and c€rresponding explanation) of the null

hypotheses presented in a previous section of this chapter. The

tests will be conducted to determine if the difference between

the performance measures of the organizational approaches are

significantly different from zero. The analysis of variance

(ANOVA) technique will be employed to dete'-mine if the differences

in the measures are significant.

Since the test factors and their corresponding levels were

preselected by the author for analysis, the fixed-effects analysis

of variance model is the apprpriate stati;tical terKviique which

will be used to analyze the effects of Inturest.1 6 To utilize

this technique the £ollowing assumptions are necessary:

1. Ti n error terms, Cijklmn are Independently normally

distributed with a mean of zero and a e.onstant variance

217

2. The effects are additlve. 18 I

3. The aliases of the maln and first-order interactiou

effects are zero.

1 6Gucnther, op. ct., p. 37. There is another A:OVA modelwhich is appropriate when the facLors are "cIected at random. Itis referred to a t,,i random Cefect . mooel.

1 Ibid., p. 36.

1 8 Chcrvny, op.ct, p. 260.

( 19hisn assu;-'pLJon w, discurssd In the previcu5 section oftils C1 , I te'r.

I

-103-

What impact would the non-validity of these assumptions

have on the analysis? If the "true" raodel of the experiment is

not additive the size of the error variance will increase which

would lead to less significant results being reported than is~20

actually the case. The author has no reason to believe that

the "true" model would not be additive however.

The remaining discussion will focus on the assumptions con-

cerning the error terms.

...studies have indicated that inequality of variances do.'snot seriously affect the F-test [the test statistic utilizedin the analysis of variance] ... if the sapLe sizes arcequal (which is the case in this project].

A further assumption we made in discussing the analysisof variance techniques was normality of the observations.Although the test is derived under this assumption, investi-gation has shovn chat failure to satisfy this condition haslittle effect upon the F-test.

We should make every possible effort to obtain independentrandom samples. Nonrandomness can veryIeriouGly affect theconclusions we draw from an experiment.

Thus the most critical assumption in the analysis of variance tec -

nique centers on the independence of the samples (runs).

Because the sunary statistics used in the study are differences

between the erganizational forms, the same set random numbers were

used to compare the tra::sportation company with the single modal

compani-os In - given run. This should acf. to sharpen the contrast

between the two approaches. Yet, to insure the independence of

20Chervany, o ., p. 263.

21 Guenther, __. ciL,, p. 63.

-104-

observations betveen runs, a different sat of random numbers was

utilized for .,ach run to determine the random components of the

model. 2 2

(~~ a mure dr-tal Iodl di.sci!:ion on thev nto~ crfa .1 rUe'nted

In th9 !vi ct I n, Ow caL is rc fcti-ed LU Clicrva.y , till. p. 257-266.

-105-

CHAPTER V

RESULTS OF THE SIMULATION ANALYSIS

The nature and intent of this project has been described in

the previous four chapters. Chapter I presented the background

and described the nature of the research effort. The simulation

model was presented in Chapter II. The test factors and costing

data were described in Chapter III. The last chapter outlined

the experimental design which was used t; analy-e the results of

the simulaticn.

This chapter discusses the results of the simulation.

Attention wil; be focused on the effects upon the performance

measures produced by the various test factors. The analytis will

be concerned with describing the nature of the effects and dis- -

cussing the r!odel operatlons which caused them. The implications

of the results for the real world will be reserved for the final

chapter.

Nature of the Anaj.is

As previously stated, the purpose of Lhis dissertation Is

to deterrine the effects the selected test factors have on the

econon'c con-!quences resultfng from the .- rMItio1 of transpcr-

tation co.pnlvs vlb-A'i-v~s siniJe tioial carriers. Thereforc, the

principal focus of the, results o: the research will be placed on

how the two org.')ii~.tttonal ap.proach,,.: differ with respect to the

1:

)-106-

performance measures. That is, the analysis of variance that is

being used to statisticallyanalyze the output of the simulation

will be concerned with the difference between the transportation

company performance measures and the sum of the performance

measures of the single modal carriers. More specifically, thi I

sammary statisiics which will be analyzed are the performance

measures of'the transportation company rminus the performance

measures of each single modal company. An example should help to

clariiy this measurement approach. SuppOse that on a given simu-

lation run the actual contribution of the transportation company A

was $224,000.00 as compared to $163,000.00 for the railroad and

$52,000.00 for the trucker. The statistic which would be entered

into analysis of variance calculations would be $224,000.00 -

($163,000.00 + $52,000.00) - $9,000.00. I"., order to determine if these differences are statistically

significant, a standard of comparison is required. The analysis of I

variance technique, that is being utilized to analyze the results,

measures the effects of the test factors in terms of deviations

from the mean of the performance measure in question. In other

words, the standard of comparison 'scd in this study is the

average of the differe:;ces of the perforrance measures between

the traosportation co-pany and the single modal carriers. The

discu;:;ion of Lhe risults will dcscribC the iechalSnM which causc:d

the effects.

I

-107-

As mentioned in Chapter IV, 64 different simulation runs were

required to .easure the effects. The analysis concentrates upon

the differences, between the organ!zational approaches, of the

six performance measures: (1) expected contribution, (2) actual

contribution, (3) price paid, (4) amount of truck movements, (5)

mount of piggyback movements; and (6) the amount of rail move-

ments. For each of the simulation runs these statistics were

gathered.

It was also mentioned in the last chapter that the mean

difference between the perfor.ance measures of the transportation

company and the single modal carriers would be tested for signi-

ficance. In other words, the results of this type of a test

would indicate if there is an appreciable difference between

the performance measures "on the average". In order to obtain

the average difference In perforance measures beti:een the organi-

zational alternatives the above six measures were averaged over

th simulation runs. The average differences for each level of

the size of shipment parameter are presented In Table Xa.

* Explanation of the Acrac Performance Mcasurcs

Some explanation of how tu interpret the averaCe performance

tmeasurts Is ,-cesSary, but before bL .this discussion the

relationshi p. bct".Len the purfon,.ance r.easurcs will be recaI iLu-

lated. ThCe XpCcted ccntributlon of the carricrs is a projection

made by the carrier,: baied upn10 their forcca!;t of the total

F-r-108-

market for transportation for each route segment and their fore-

cast of their market share. The actual contribution of the

carriers is based upon how they allocate equipment to the various

segments, the accuracy of their forecasts, the shippers' modal

selection policy as well as carrier pricing decisions. The price

paid by shippers for transportation services is a function of the

shippers' modal selection policy, the demand for transportation,

the carriers pricing and equipment allocation decisions. The

amount of truck, piggyback, and rail movements are dependent upon

the pricing and equipment allocation decisions of the carriers

as well as the shippers' modal selection policy and market demand.

The first three performance measures--expected contribution,

actual contribution, and price paid--are directly concerned with

tha economic impact the two organizational alternatives have on

the carriers and users of transportation. The latter three

measures--amount of Cwt. moved by truck, piggyback, and rail--are

determinants of the first three measures. In other words, the

expected and actual amount of Cwts. moved by each method of move-

ment deternines the expected and actual contribution of the

carriers and the price paid for transportation by users. For

instance, t.;e actual ccntributlo, obtained by the carriers is

d -.adcr.t upon the aount of p!F.gybac., truck, and/or .; movements

that take place. Thesu intcrdepcndvnclc3 between 0c performance

measures :Illl aid in the cxp,, atLion of U.a cause and effect relation-

ships b, L c., : I ; t f;;clors a:. s i ni [1 ,c! prfcr:.:.;c: c !; ur .

-109-

As Indicated in the last chapter a t test was utilized to

determine If the mean differences in the prformance measures

between the organizational approaches were statistically signifi-

1cant. The results of the test are indicated by the use of

asterisks in Table Xa. One asterisk (*) indicates the hypothesis

is significant at the 10% level. That is, one asterisk means the

null hypothesis is not accepted (rejected) with the probability of

10% that the null hypothesis is in fact true (Type I statistical

error). Two asterisks (**) indicate slgnificance at the 5% level.

Three asterisks (***) indicate the results are significant at

the 1% level. 2

As can be seen in Table X, the differen.cs between the reans

of the transportation company's performance measures and those

of the combircd cingle modal carriers are in most cases signifl-

cant. Table Xb presents the relative differences In performance

measures betoieen the transportatlon company and the single modal

carriers. For instance In the TL/CI. market, the Lransportatlon

company made 44% fewer piggyback moventr than did the single

modal carriers.

1Supra, chapter iv, p. 98.2Tiere sgonlf ca:e lt,.Lcls a ;proi r c for a !;In.le

test, when ole cou.; Lc-;L , hc pr ,,: 1, t offal.Lly rcJc.-t;:g aLt !(:;J ;L u. true null. IS:ote: I. [1 - p(not r.jccting 'u all the Ltit.S)]. !;cO, for example,Menderhill1, cp..f, t., p. 175.

-110- A1

TABLE X

81a" AVERAGE DIFFERENCES OF THE PERFOR..NCEHFASURES BETWEEN4 THE TR.,ASPORTATION COMPANY

AND THE SINGLE MODAL CARRIERS

TL/CL Movements LTL Movements

Expected Contribution $24,221.53 $ 18,158.30

Actual Contribution $23,143.72*** $ 17,513.62" *Price Paid 7,662.38 - 92,475.94

Truck Movements - 7,481 Cwt. -101,650 Cwt.

Piggyback Movements -49,331 Cwt. 91,306 Cwt.

Rail Movemen.s 70,388 Cwt. ---

Significant at the 10% level a

Significant at the 5% level

*Svnlficait at the 1% level

4

.1- 1

I

TABLE X

b THIE VERAGE DIFFERENCES AS A PERCE;TACEINCREASE (+) OR DECREASE (-) OF THE COCMINED

SINGLE MODAL PLaOr,-WNCE MEASUlES

TL/CL Novements LTL Movements

Expected Contribution +112 +17%

Actual Contribution +11% +16%

Price Paid + 1% -12%

Truck Movements - 42 -35%

Piggyback Movements -44% ----

Rail Movements +24%

As indicated in Table X, one may conclude the expected contri-

bution and the actual contribution of the transportation company

is larger th.n the single modal carriers, although at different

levels of ailnificance. The price paid by shippers is significantly

less for LTL movements when transported by the transportation company.

The price paid for services was not significuntly different between

the two approclies for TL/CL mover..nts. In the LTL category, the

transportation company rmovcd lese goods by truck and more. by piggyback,

which contrluted both to the higher contrIbutlon and reduction In

3TC fJ 8urC, In this table ri.present thc ratio of the averagcdifferonce, -:i perforance mcasurLs abh Xa) to the avcrage

combined si,'ge modal i erfor.-ance t.r asures, 4-, ' 411%.

4Therc yore no ,j; 2 c. :udal 1 ggyback tovnnt in .'fLfreight . The 1ncred:;c in C:t , dL. by pif;yha-k in thy tran por-ta lien cc:npi;ty sL LL, 'n ;, ; 91.306 CWL.

ill ii

-112-

price paid by shippers. In the TL/CL cateSory of movement,

the transpor:ation company made more movemjnts by rail and fewer

by piggyback than did the single modal carriers. In general,

this occurred because the single modal trucker participated

in Plan I moves when he was not the low cost carrier. In a

large nimber of cases when this occurred, the rail mode was low

cost but Plan I was competitive. Hence the transportation company

shifted Plan I moves to rail in those cases, which of course,

effected contribution and price. Other factors played an important

role on the performance measurcs which will be explaired in great

- detail in the next section of this chaptor.

To reiterate, the figures in Table Xa represent the averAgC

dIffercncuts between the transportation co.pany performance measures

and those of the railroad and trucker. For instance, in the TL/CL

category of movements, the trancportation company had Uenerated

$23,143.72 on the average more contrIbutJCn than the railroad and

trucking companies. In the some manner, in the LTI, category of

movements, the transportation company moved 101,650 Cwt. of

product less by truck than did the single modal carriers (the trucker).

It should also bu p.Ji Lcd out that thc differences between

the a ount of truck movcmcnLs, In the LTI. caL. c;ory, if not co:.:-

pletely mad, tip by p1g6 yback ,o tnt;, In uther word.;, the

transporUt..ii corp:ilan:, r ov, J0.1 ,650 U t'. :;,; by truck than the

V; i;Inglo 1.Ol1 carriers, bur. ;Ilftfi only 91,306 Cwt. to plg,,ybacL.

moV-;(:nL!;, C* n th, oL!.i r d - " ./ C!,':t.g , , thu ft.wcr U

.-- " Ij -... .. _

(.) -113-

truck load ard piCgyback movements are mora than compensated for

by rail moverints.

Why is this so? In the TL/CL situatio. many factors may

contribute to this effect. For instance, there were situations

on some rout(* segments whore the trucker was not the low cost

carrier but could compete for traffic, If the trucker exhausted

his equip:3ent capacity on other route segments, in the single mudal

setting, dcmand for service was unmet on some route sccents.

The transportation company, on the other hand, If it had rail

capacity could substitute this scrvice if not constraInted. Thus,

the differences in the amount of novcwat by each mode need not

sum to zero. Similar conciderationu come into play In the LTL

category of movement.

The LTL category of movement, however, doos have a unique

situation which must be discussed. In the mudel, the single .odal

trucker really has no competition for LT1. movements. This ia the "

same as sayirg the model is considering only that LTL portion of

the market which mover by comNTon Larricr truck. At any taLe, the

trucker and the railroad do not have to participate in PlAn I

piggyback to,.'ov.ni.nts. If thcy do participate, they will split the

profit cootr!butloti on the basis of cost conLrlbuticns a!;

explained III Chvpter 1.1. S1imcO the tru kur has t,, real compne-

tition for LTL traffic, he ,:II act his LTL rate at his full coi;La,

consistent wiLi the pr'i c :; pollc; or;t; bli ;!,L d in Chnpt,.r i.,

Thur:, In' I.T . :,L; ti,, : i]t ;'.c;!,1 i'uT'.. I1!; :tOi o)..r o L

part i i,~n an I :- '-,,P

-114-

In the case of TL/CL movementn, the full cost of tne lw cost

mode will govern the action of the uingle Aotail trucker. If Plan

II piggyback or the rail mode In low cost, the trucker may wish to

participate In Plan I movements to obtain some share of the

Acompetitive narket.

Regardless, the transportation company will establish the rate

at the full cost of the low cost mode. In the case of LTL move-

montp either at the trucking full costs or the piggyback full costs

whichever is lq.et for a given route segment. Thus in the single

modal .TL markct se,;me:nt, the product will be moved at the full costs

of the trucker, while the transportatior. company will price its

services at the full cost of the low cost method. 5 In a large

number of canes, the full cost of piggyback movc".ents were lower

than the full costs of truck movements. This is what creates the

large difference in price paid for traasporuation services in the

LTL category between the transportatlon company and the single

modal carriers, as shown in Table X. Furthermore, the rail

flatcar capacity was the limiting resource of the traisportation

company. Because the transportatlon company was pricing services

at the full cost of the low cost mode, some trucking ,ove-ments

Ln pra.'tLcO it Play be r-nore likvly that the transport;,tion

company would price It!; ,;.irvJcc,; 1i, thu Li. 1 . r1 at the trucker 'n.full CO!:ts. "i o prii '.r1l po!Icy w'vu ' dopelti u.,oai icw thl, I.C.c.would react LO tle tri.orLa-caU ct;,ya preini, tL -urvicc(,; Inthe LTIL mark.-c, at It.,; fill coC t:L;. if t.he Lri.:p-ortation c c:.,.y;..'did price at !.;c! ti'ucc '- ; EulI cob:t , thc1 di tfrcec h.,t'.',,un the tEL.p0Ctvd ;'.:"i actull coli t ib . t[uw,,; of L1; i :i. u i i; Lu-wutlld 1, cCoM,: k've.i lair;;k,. tLh.!ll th ill 101) V X.I

~-115-

wegre not attractive even after all the flatcar capacity had been

exhausted. This accounts for the differencea: between the amount

of trucking zoveents not being fully compunsated for by piggyback

Umovents for LTL movements in Table X.

Analysis of the Test Vr ctor

Before proceedin& with the analyb!,i of the test factors, the

symbolic representations of the test factors will be recapitulated.

The symbolu which are associated with each test factor are displayed

6in Table XI,

TABLE XI

TEST FACTOR SY'MBOLS

A Truck Operating Ratio

B Rail Operating Ratio A

C Truck Load Factor

D Rail Load Factor

E SytteM Capacity

F Shippers' Logitics Consrraints

lhui sectlon prese:mts the analy!,; of the effects produced

by the six tLet factors, for .ach love! of the size of shiIp'.:ets

parameter, on iach o4 the perior a-;.ce mcasurcu. The analy';J ,r ill

be compoi,cd of two parts;. First the effucts which are found to be

statitftC all.y r:t; .1flcant . 1) ., !d.t .cd, St'ce:'d', a:1

6o detai cct r LiHhe v t ( t L,-; t £act o; n, the

rei ¢,r I, rvf -rL-d Lu j i,,bl. I , C,:, cr ]. I, W.m",, p. 1,',,

-116- -

explanation of what "caused" the factor to be significant in terns..

of the manner In which the model operated will be given. One cannot

be absolutely sure, hawever, that the explanations will be complete

or accurate. This is because of the complexity o! the interactions

of the sysre which required simulation to be used as a methodology

to study the transport;:-on company concept.

As indicated in the last chapter, the main effects and Inter- Iaction effects will be tested' ,or significance. The main effects

and interaction effects are numerical measures of each type of effect.

"In general, an effect is the average difference between observations

(of a particular performance measure) generated from simulation runs

where a test factor is at its alterilaLe lezel, and the observations OR

where a test factor is at its standard level". 7 The effects are

defined mathematically as follows:8

1. MAIN EFFECT:

ME _ -21 (r Q 1, 6)

where: il

:HE main effect of the jth test factor

Xij a Ith observed measure when the jth test factor is

at its altcrnate level,

XI a ith observed neasure when the jth test factor

is at Its standard level,

7Chcrvwny, pP_ _ *', p. 272.

8Lld., 1. 273.

--' - " : -'= - " .' --- .. .. . , -" .. ... - ...... ="I ... ..... -,,.: -, .

-117- -

the number of observations when the jth test

factor is at its alternatQ level, and

n n the number of observations when the jth test

factor is at Its standard level.

2. FIRST-ORDER INTERACTION EFFECTS:

nj X n n

1 2 11 j ~ I.-IIE~~~ 2 IX- 2)2

J n 1=

(j1 M 1, ... , 6)

02 = 1, 6)01 0 J2)

where:

IEI = first-urder interaction effect of the jlth

and j2th test factors,

X ith observed measure whcn the two test factors

1 21(1 and j2)are both at their alternate levels,

X ith observed reasure when test factor j is at _i.)I

its standard levcl and test factor j2 is at its

Ii alternate level,

1 ith observed measure when test factor is at its

alternate level and tast factor j2 Is at its

stanlard level,

XC ith owserved measure when both trst factors

(j 1 and j2) are at "heir standard lcvels,

[IL

I.-8

JlJ2 * the number of observations when the two

teat factors (j1 and J2 are both at their

alternate levels, .a - the number of observations when test factor

Jl is at Its standard level and test factor

J is at its alternate level,

n * the number of observations when test factnr

is at its alternate level and test factor

J2 is at its standard level, and

n * the number of obser%, atlons when both test

factors (j1 and j2) are at their standard

2evels.

An example should help to clarify how the measurement calcu-

latIono are made. From Table X the difference between the expected

contribution of the transportation company and the expected contri-

bution of the single modal carriers is $24,221.53 for TL/CL move-

monts. 1he average difference between the expecced contribution 4when the truck operaring ratio was at its alternative level, i.e.,

91 was 24,169.67, Similarly, when the truck operatitig rdtio was

at Itr, stnndard level., i.e., 99%, the avervge difference in . -

eKpected con.ributlon bctwccn th'2. organiza.l nal alte rnaives was I

?4 273.339. The dlffer nce bet'..ei t',ese a.'rages ($103.72) is the

main (ttect cf the low opnrating ratio (91;;) over the high operating

ratio (99").

I

.. .. __...... , -... . . ... .7 --- r- .: -- T - ..-- . n.. --

-73

The operational interpretation of the interaction effects

is much the Eame. For example, the interaction effect of truck

L operating ratio and system capacity (AE interaction, see Table

X11a) resulted in an increase in the difference between the i

expected contribution of the transportation company as compared

to the single modal carriers of $8883.32. This represents the

average of the expected contribution when both factors, A and

E, were at their standard levels plus the average when both

factors were at their alternate levels minus the averages when

one factor was high and the other factoc low,

Before proceeding with the analysis, the direction of the

change in the economic performance measures brought about by changes

in the levels of test factors deserves some discussion. An increase

in the expected and actual contribution indicates that the transpo:-

tation company is more "profitable" than the single modal carriers.

A decrease in thses performance measures indicates under what

circumstances (test factor levels) the single modal carriers may

econocically out perform the transportation company. Similarly,

an increase in price paid would mean tho single modal carriers

were more economical to users than a transportation company.

Cqnversely, a decrease in the price paid f)r transportation, Indi-

cates the transportation co:;mny ray bc more economical to users

than single xodal competitive carriers.

-1;-0----

The effects of the test factors are summarized in Tables XMIa

through XVIIb. The results of the analysis of varjance are identified

by the use of asterisks as described earlitr in this chapter.

Effects of a Change in Truck Operatini Ratio (99%-91,) on the

Performance Measures A

The effects of the change in truck operating ratios are pre-

sented in Tables XIa and XIMb. These results indicate that no

significant change in the performance measures occurred due to

changes in this factor. The lone exception to the lack of signifl-

cant effects took place in the amount of rail movement measure in

the TL/CL category of movements. The interaction of a decrease

in truck operating ratio with a reduction in system capacity caused

more movements to take place by rail boxcar. The reason for this

result is that 'the change in truck operating ratio did cause more

movements to take place by truck, although not a significant amount,

by shifting piggyback movements to the trucking mode. When the

trucking mode capacity was restricted, the transportation company

would in general, when not constrained, shift trucking moves to

rail moves, especially on the loingcr route segnents where the rail

mode even if not lot cost made a suhstantial contribution to profit

and fixed ar" 'or co-'no.n conts.

AI=I

-I

4'1

.- C £ %1(1 in P N'0.4 0n

.C " ~ I'% N

I. w f!W 0 6 t

a: 0

0~4 %n &M ?4. 0

I-C,

%0 a, e 0

9. L I -4 - %0

(i)C

r'rJ cc'.

*r4

4.4 " 0 r% .

.4f4

'I

OD co n0 0 i4 PV. r- 0 N

'4i

03 * P 000 ~

V 0N

~~V A~

w4 -4 0 I0. N IN 4

N~I %n N .

0: A

~-. 4~~ 31

-123-

Effect's ofa Change In Rail Operating Ratio 8.5%-65% on the

Performance M.asures

The effects of the change in rail operating ratios are pre-

sented in Table XIIIa and XIIIb. There are a number of statistically

significant results in all the performance measures except on the

mount of truck movements.4-

Exected Contribution--In both the TL/CL and LTL categories

of movement, the reduction in rail operating ratio caused a

significant Increase in the expected contribution of the transpor-

tation company as compared to the single modal companies. There

was also a significant interaction of a decrease in rail operating

ratio with an increase in rail load factor in the TL/CL movevent

category. This interaction (BD) had the effect of lowering the

expected contribution of the transportation company vis-a-vis the

single mo'lal co.pani es.

Actual Contribution--Again for both TL/CL and LTL movements,

the reduction cf rail operatiJng ratio caused a significant increase

in Lhe actual contribution of the transportation co.pany as com-

pared Lu the .jnile modal carriers. There was also a significant

int.raction of a decreat;u it, rPu opLralnL ratio with an Increse

in rail load factor (BD) In the TL/CL MoveC-,rnt catVCry. As one

Mig;ht CXpect, U;MS I ntcrac'. . had ti.c sar. cff,:ct on actual c||tr t-

bution ag IL d d on v-pvcu'.-d contrILut Icn; that I , It cauicd a

lu 'icbrJ . of Liu. actu.i co trLributlon (.f thi! rniprt:1,|un c,:.;..': v

vl'.-;4-vi; thc :,;igin . i.' . . r .

I

i:: | i '- , " ' . . . . . . ... .. . .- ...

l .

P4 WN

C~ 0% 0% OD 0ne 4-44

N0

II

u I., C

et :4 "3

cnI I

0A

t~' ., , . , - "

0

1. .7- l _ c

U 4 9"IC4 0 07 fn

Us-.~L Li r- nU 0 0 .

IP -I0 ~ V

A A

UJ0

-

V.

ho 1

'. 1 ~ 0) V0 LA

oft . -4 ~ ,~ fI '

-. 4. - .

1% .4 N ^q N1

,A p44 I I I p

9- jc 0

t-I

.4'13 ,,t, IF., .

4 , LA- co/) NC'4

i',~~a ""t0.%"' "" '

l

InI

." ~ 0 .* ' t

l k- -4~

-~ f~4 49 N

-126-

Price Paid--The change in rail operating ratios caused a

significant reduction In the price paid fir transportation services 7

for userx of the transportation company vis-h-vis the single modal

carri-er.- - .he TL/CL category of movement:. This factor had no

effect ou price paid for LTL movements. There was one signifi-

cant intera n with the truck load factor (BC) in this performance

measure for .'L movements. This interaction, a reduction in

rail operating : tio and an increase in truck 1wad factor had the

effect of loweri,: the price of service for the users of the

transportation tor. -,q as compared to the single modal carriers.

Amount of P.'','. Movements--The rcduction in rail operating 01

ra.io caused a sigr f .-h ly fewer amount of piggyback movements

to ta~ke place in the tre.:. .ortation company setting than in the

single nodal carrier e',lr.-.nt for TL/CL movements. There was

no significant different ,..ei the systemsq as far as LTL move-

ments are concerned. jAmount of R.il MLv.-.ntser piggyback movements were

compensated by a iar.r .. wount .: - ,ovemcnts in the TL/CL

category. There aie, of c..'ur.e, I " ovements being considered 1in the ITL category. "',c re o.. o' operaLlnl ratio con-

bined with an ineroazcn '.. rail I.-.311 :..,uced .' nig~iifl

cant irteracLion (b!) " had tie t ,, dluciig the nurber

of rail rnov-:rints.

" ' t'" ' ---' -" ±-71/ "- -, w 7- "' ' " 'Z ; -- .--i .u- , _ ... : ..:':I

-127-

Explanations--What interrelationships within the sim-.!:

system causel these changes in the performance measures? To !.: r

the discussion, it was pointed out that the transportation compsny

experienced larger expected and actual contributions in both

categories of movement than did the single modal companies with !

a reduction in the rail operating ratio. The reasons for thej

increased profit contribution of the transportation company over

the single modal cotipanies are the same in both cases. The

transportatioi company shifted traffic frcm truck and piggyback

to rail movements. It did this to the extent possible. That is,

the transportation company shifted as much traffic to the rail mode,which the single modal comnpancs were moving by rail and piggyback,

as it could a ithout violating shippers' lcgistics constraints and

capacity constraints.I

The significant interaction effects under expected contri-

bution and actual contribution, BD, indicates that the combination

of a low rail operating ratio and a high rail load factor signifi-

cantly reduced the expected and actual contribution of a transpor-

tation company as compared to Its single modal counterparts. It

must be remembered, houever, that these results must be interpreted Iin conjunctlon -'fth the re:nults displayed In Table X. It can be

seen that the BD interaction is ncgativc In each case. Thus, the

BD Int:raction effect is to lower the figure of the expected and

acLu-1 contrIut !on !n Table X. The reason the B! interaction Is

negat~v I.; "hat thl. cob wir. en is thle mrost favorable rail po;ition

-128-

for single zodal carriers. Thus the formation of a transportation _

company with these rail characteristics whi.le still improving

the profit situation of the transportation company over the single I-Ul

model carriers does not result in as large an increase as the

average presented in Table X.

The effect of a reduction in rall operating ratio was ti.

reduce the difference in total price paid for transportation by

users between the organizational approaches. The reduction in rail

operating ratio made the rail mode in both organizational settings

the most "profitable" method of movement o:. a large number of

route segments. Also, the reduction in ral opcrating ratio made

* the rail mode much more competitive with the independent trucker.

In other words, on some route se~ments when the rail mode had a

high operating ratio the only traffic it could obtain was that

reserved for the mode due to shippers' logistics constraints. On

other route segments the rail carrier would attempt to compete

with Plan I1 movements. Under these circumstances the single modal Itrucker, in a number of situations, would exhaust his capacity

before demand was satisfied. The transportation company, on the

other hand, because it controlled both modes made different allo-

cation dccisions (based upon total system wide contribution

("profit") ma:.iti zat ion) ,--hich resultid in more demand being

satLisfh !. Lecause :ore dc:mand was sa*&Isficd, the corrcsponJing

total price p;:id for servirce was greatcr for thCe transporLatlon

Ct

,:' o:'a%"tha ti: :i, lc e~l] ca rJrsvl~n t,. r fl ~d. ha] ai

-129-

bigh operating ratio. Thus the reductiot. in rail operating ratio,

since it enrbled the single modal rail ca:rier to be more com-

petitive, reduced the difference in the price paid between the

transportation company and the single modal carriers. In fact,

on a large number of runs, where the rail operating ratio was at

its low level, the differe..e in price paid between the two

alternative organizational approaches was zero. Thus, the overalli=

effect of a reduction in rail operating ratio was to increase the

contribution of the transportation company over that of the single

modal carricrs at substantially the same price to users as that

paid to the single modal companies.

The interaction of a reduction in rail operating ratio and

an increase in truck load factor (BC) had much the same effect as

that described in the last paragraph. The increase in cruck load

factor coupled with a reduction in rail operating ratio made piggy-

back movements, in general, the dominant method of movement in

both organizational settings. Hence, the difference in price paid

for transportation seivices betw'een the organIzational alternatives

was very much closer together than when the truck load factor

was at its low level. In other words, the difference in price

paid for services bet;w .en the organizatio;'al alternatives under

these circtt::.-tancvs (factor levels) vas zero or nearly zero in

all cases, as opposed, to s ,uations when those factors vcre at

their standard leve_.

4

-130-

Effects of a Change in Truck Load Factor (100 Cwt.-300 Cwt.)

on the Perfc.,nance Measures

The effects of the change in truck load factor from 100 Cwt.

to 300 Cwt. are presented in Table XIVa and XI11b. The change in

this factor produced several st:itistically significant results.

Expected Contribution--The increase in the trucl- load factor

significantly reduced the expected contribution of the transpor-

tation company vis-f-vis the single modal companies in the LTL

category of movements. This factor had no significant effect on

expected contributions for CL/TL movements. There were two signlf'-

cant interaction effects for this measure. The CD interaction I

effect, increased truck load factor with low system capacity,

was to significantly increace the expected contribution of the

transportation company vis-A-vis the single modal carriers. The

same kind of effect was caused by the CD interaction effect, high

truck load factor with high rail load factor.

Actual Contribution--The change in truck load factor produced

no significant main effects in this performance measure. There

were two significant InLeraction effects however. Again for the jTL/CL move-mcnts, the interaction effect with system capacity (CE)

was to Increase the actual contribution. This same kind of reaction

was produced by the iateraction effect with rail load factor (CD)

in LTL r.oveo.cnts.

.. .. ..... ..... .I

94C N ?It~t~r4 rs.%-4 N 4-

4j -4 4 .- 4

%4i. 0

0. u. ~

Ev IA

00 -4 - 'r'0 '% 0. N- 7 0o

i- -. 4 4~ n I

44

VL o

Ca,

Go N 0 611 00C.0 t'-s 0

2 m 04 0 N 4

449

LI P. N %D 1-4 N1? 0 on a, N .1

ci :1-4 LN C- N

wd 41C

F-I 04

-C. di 0% P- c' 0~~9U C,0) . @4 4 i-

v' r- , I1

1.41..

-133-

Price Paid--The main effects of a change in truck load factor

was significant in both categories of movm.ment. The effect in

both cases was to Increase the price of services of the transpor-

tation company compared the single modal carriers. Several

interaction effects were significant as veil. The Interactions of

truck load factor with (1) rail operating ratio (CB), (2) system

capacity (CL), and (3) system logistics constraints (CF) were all

significant in TL/CL movements. The last two interaction effects

acted to increase the price paid while the first acted to decrease

the price paid.

Amount of Truck overents--The change in truck load factor

had no significant effect on the amount of product moved by truck

in the TL/CL movement classification. This test factor change

did, however, cause a significcnt increase in the amount of truck

movements of the transportation company vIs-A-vis the single modal

carriers for the LTL movements. Therc were also two significant

interaction effects in the LTL classification. The interaction

with rail load factor served to decrease the nuwber of truck move-

ments significantly. On the other hand, the Interaction of truck

load factor with shippvrs' logistics constraints served to increase

the ar-ot ,! Oat. moved by truck.

"M--TI Of arnount of Cwt . rov'd by

piggyback wa. rslJinif lcaitly I crca::cd in TIL/CL, sovcments a:nd waq

stgnl fic-m:ity dLcc.:mD in l.'L';. b.:)w:.A2;t. Dy n tnrret:c in truck

-134-

load factor. There were three significant interacton effects.

For TL/CL mcvements the interaction with ,. low system capacity

(CE) significantly decreased the number of piggyback movements.

For LTL movements, the interaction of high truck load factor _1

with high rail load fa,'tor served to Increasethe amount of Cwt.

moved by piggyback. On the other hand, the ititeraction with .1

low shippers' logistics constraints acted to reduce the number of

piggyback movements, jExplenations--Because of the large number of significant

interactions associated with a chauge in the truck load factor,

dispussion of the effect of a change in this test factor will be

centered on the main effects. Interaction effects will be exa.ined

for trends rather than discussing each significant interaction

effect Individually.

In the LTI. movements, the increase in truck load factor caused

the expected contribution of the transportation company compared

with its single modal counterparts to decrease. This Is due to the

fact that with a high truck lo3d factor less craffic was allocated

to piggyback movements by the transportatlon Company. This is

because an increase in the load factor of the trucker reduces the

cost :,cr C'. , of moving goods. SinICe cos Is directly related

to the rate (the lu]] cost), the truckcr beca.,o more compctJtLvc

with pljggybark MoVemcnts. Hence the tran.portatLon cCr"M1- y 1ad

I.Itth, opportunity to fihtft tr:,fric f.ro'i truck Lu pigyhback L~rv,:.c:lts

wit ~Li I 'it -, i t r',..t(.r profiL corLrlbo,. ion. T ' ff, t of t

-135-

change in truck load factor did decrease the difference between

the two systems actual contributions but uas not quite statisti-

cally significant at the 10% level.

That the above reasoning is correct is borne. out by the

fact that the transportation coMpany moved significantly many

more Cwt. by truck than when the truck load factor was at its low

level in the LTL category. Correspondingly, the transportation

company moved much less traffic by piggyback than when rhe truck

load factor was at its low level.

A peculiarity in the main effects trought about by the change

in the truck load factor is that the price paid for LTL transpor-

tation by tratsportaLlon coSn-any users Is cignificantly hfghur

than whei the truck load factor was at itn low level. This is an

interpretation problem which again deals with the nature of the

perfor. ;nce measures being a difference between the two approaches

to trilisportation. The incrcace in truck load factor did signifi-

cantly raise the price paid from the average found in Table X.

Even :ith thi.r rise in price paid betwecn the two forns, the

transportation conpany wot'id still have negative effect on the

total price paid by users. That Is, if the deviation from the

average cavted ly an Increase In truclk load factor was added to

the average diffrcnc, thu figure t;ould still be negatlve by a

l.arge aioun t t hur indicati::. the tra.nsiportation coi:.pony still creates

Iecolol;J ed i for A; pk, r. *

:1=

-136-

Before considering the TL/CL movements, the effect of changing

the truck lead factor on the LTL interact;''on effects will be

examined, Four of the six significant interaction effects were

the interaction with a change in rail load factor (CD). The other

two significant interactions dealt with the interaction with a

change in shippers' logistics constraints (CF).

The CD interactions had the effect of significantly raising

expected and actual contributions and raising the amount of Cwt.

moved by rail while reducing the amount of truck movements. This

is as one would expect. The combination of high truck and rail

load factors made piggyback movements the least cost (rate) maximum

contribution method of movement. The piggyback movements had a

larger contribution associated with them than did the truck move-

nents. Hence the transportation company substituted piggyback

movements for truck movements subject to logistics and capacity

constraints. This had the effect of increasing expected and

actual contributions over the single modal trucker.

The significns CF interactions had the effect of raising

the price paid by users over the average and reducing the number

of piggyback movements below the average difference as expressed

in Table X. Again this is so because, in general, an increase

In the load factor of the trucking mode made this method of move-

amnt dominant. Most moves took place under these circumstances

by truck, which In general had a larger price tag than when coordi-

nated transportation was used under different factor combinations.

-137-

The change In truck load factor significantly Increased the

TL/CL price ?aid and the amount of Cwt. moved by piggyback. The

increase in truck load factor made piggyback movements an

attractive latermecdiate length of haul method of movement. In o

the singlc .. ,3dal sltuatie:i, however, the trucker tended to make

truck movements as much as possible because of the necessity of

splitting the contribution in Plan I moves. The t'anspartation

company utilized piggyback movements to a greater extent when

the total contribution was greater.

Thie effect of this factor on thu price pald by users is more

complicated Lo analyze. Ccrtilnly, tLhe Interaction with low system

capacity (CE) Is understandablc. The price pald Ognlffcantly

Increased wh-n the truckicn, capacity was exceeded. This is

because alternaLive more expensive mcthod- of movement must take

place to satisfy der.nnd. As far no the rain effect Is concerned,

the analysI.; of the printouts of the simulation Indicated that the

increase in truck load factor made trucking woves more attractive

to the trucker aR well as the transportation company. The sJmu-

lation printouts revcaled that In a large number of casr,, due

to chance, the 141ngle inocli. carrier would run out of capacity

before dunanJ was s;atJsfJcd. Thc trn, po~-tatJ(:, ccr;,a' whiltl

sti l ,l cra ti.8 u:ide r lc .. ! Lu:: cz,:'(:,tv had x, odd L I afrI 1 uuL0ft

of truckliq c.yaciLy aL 8 di ,'C:,J . Thi, extra capacJLy :a.

what Lhe ,':, 1',...J h idJ av J ial(' for l'Ii- If 1:0W- ;. Ihu; LhC

trw ~~'~'cr~iilw. t Ii' :i uf thu d;i ~~

-138-

as a result the price paid for services correspond ngly rose.

Effect of a Chan~e In Rail Load Factor (400 Cwt.-800 Cwt.) on

the Performance Measures

The effects of the change in rail locd factor from 400 Cwt.

to 800 Cwt. are presented in Table XVa and XVb. The change in

this test factor produced many significant results.

Expected Contribution--The increase in rail load factor had

the effect of significantly reducing the cxpccted profit contri-

bution for both TL/CL and LTL movements. The cbange in this

fcctor also produced three significant interaction effects in the

expected con:ribution perfor.ance measure. In TL/CL rovements

the interact.Lo, of increased rail load factor with (1) a teduction

in rail operating ratio (DB) and (2) a reduction In the size of

shippers' logistics constraints both servcd to significantly reduce

* t1-e expected contribution cf the transportation cocpany as com-

pared to the single modal companies. In the LTL category of

Smovements, the interaction effect of a change In rail load factor

with a change in truc- load factor (DC) significanLy IncreIsed-

the expected contribution .easure.

Actual Cm:,r1 ut1':--T:e nain cffects of I:creasln , the rail

load f'tor 'n both r,a'cj,-.. t Ca 'o"c'; 3 s to 'J,JA'icantly decrease

" the actual co:-:tributin C4 the t:ran:,;ortatb)n cc:c-.any as oppoicd

to tsr in.]: ;:Ow,.!:. carr t'r;. Ti,c s.:. r, terct c; ef :c that

were ; J: Li ;pra otecd Loio.rihut ion rii'.vU are : it-

cant for the otun': c,: ,.:2 I f ... -,)c ,';. , ";'.- .;t' .:

gn

,% m

to aL9p

V4 N

~~41Aj 49 4

Aj ~ ~ ' W4 'a ChW4PI-fl 0 .4 4 U

r% 4 q fn o

ioiP4

4l W4 .W4

.0 It V.,I VIOaa u~o

!016

4

u ~ Go0 0 0%

U4

4 4

tDi ? 14F 4

v.4 N N~ .2 .4 ft

0 4944 a'9 In a

%0 In %D n

1%T EW4 I I .

494

Ine I ~

rn 4n 9 cc en4 o % - C" 4

Ij a I

- . -141-

effects-DC for LTL movements and DB and DF for TL/CL movements--

produced the same kind of reaction uf the system In this measureme2nt

as they did for expected contribution.

Price Paid--The change in rail load factor had no signifl-

cant effect on the price paid for transportation in TL/CL move-

ment category. The change in this test factor did, however,

significantly reduce the price transportation conpany customers

paid for transportation services as compared to the single modal

counterparts in the LTIL category of moveme2nt. The interaction

with shippers' logistics cor"strai-ts (DF) also significantly reduced

the price paid for services in the LTL category.

Amount of Iruc. ovo.,e.ts--The increase in rail load factor

significantly reduced the number of Cwt. moved by truck in the

transportation compzny situation vis-S-vis the single modal carrier

in the LTL category of movements. Also in LTL movements, the

interactions with increased truck load factor (DC) and with reduced

shippers' ic-gistics constraints (DF) sig-nificantly reduced the

e.mount of truck movements.

Amouit of P1;': , ,.t-.h_ change in rail load factor

signif czi:-.t incr-:.tscd the nu:n.cr cf C- *. noved by piggyback fcl

1,.1 ov..:cntS. This iacLor c.a:;u ,had thC e:act Cpc JLe effcct

o the :1 c.t. of pj'g::hac',- :c:'nts in tie TL/CL movc:.cnt ca, itory,

ThcrL ,c ,c t:o -ii.t i:.t.. :' ;.: c . c ls lro. 1;1 the TL/CL

Catiory, LI-C ,iit a of

/ i

-142-

constraints (DF) also served to significantly reduce the amount

of Cwt. moved by piggyback methods. In tl~e LTL category, the

interaction with increased truck load factor (DC) served to

significantly increased the number of Cwt. moved by piggyback.

Amount of Rail Movements--The change in rail load factor

led to a significant Increase in the amount of rail boxcar move-

ments by the transportation company over what the single modal

companies moved by rail, as one might expect. The interaction

with a low rail operating ratio (DB; served to significantly

reduce the number of rail movements of the transportation company

as compared to the single modal companies.

Explanations--The ex q.anation of how a change in rail load

factor produced significant effects on the TL/CL performance

measures will be presented first, after whIch attention will be

given to the LTL performance measures. The Increase in rail load

factor significantly reduced the expected and actual contributions

of the transportation company as cozpared to its single modal

counterpar:s. This is due to the fact Liat the chan.e in this

factor made the single modal railroad, in most cascs, the do--

nant carrier. In other worcds this tst factor cha:;je, caused for

the tost part, most of the singe :.od,-.l traffic to Lcve by rail

except where constrained. Thus the for:;ation of a transpor-

tation co.:.:mv had li'tt2. effect on ;e .a:;,cr i ; ., ch -:, .2.ts

too- placu u;i2Lr thi, , ci u' cc , :.:Id hc-., e Cid ;not incrcase

profit contr!bution as In other siLur',iu::;.

.. _ .. . . ... .. .. . . .. ... ... . " -"= -- "'-' - ,Z .,- " '- --

.--- - T -" ... "- --.. ... ..L.-,-, Z- .. .. .-. -rs-Z "'t-- .. . ...- s rr,- . ... .r .,...

-143-

In most situations the singlr modal trucker would attempt

to compete with the single modal rail carrier in short hauls by

using Plan I moves. Thus the amount of Cwt. moved by the transpor-

tation company by piggyback was significantly reduced as compared

to the amount the single modal carriers mcved by this method.

This is because, under most circumstances, the rail mode was low

cost but piggyback Plan I was within the price range of competition.

Thus the transportation company shifted movements which the single

modal carriers moved by piggyback to rail. This correspondingly

explains why the change in rail load factcr led to a significant

increase in the amount of boxcar movements.

The significant interaction effects with rail operating' ratio

(DB) were explained for the various perfozmance measures earlier

in this chapter, when the change in rail cr.qating ratio was

examined for effects. The interaction with a lw level of shippers'

logistics constraints (DF) signif!,-intly reduced the expected

and actual contribuLion of the transportation cotpany as well as

the nuber of Cwt. moved by piggyback. The lower level of shippers'

logistics constraints allows the carriers more freedom in selecting

the methods of movement to salsfy shippers' demand than does a

high level of logistics ccnstraints. As such, the interaction of

high rail load factor with a lo, level of log1itLcs ccnstraints

produced the sane kind c ffect as the main effects of changing

rail load factor, only to a larger degree.

-144-

Attention will now be turned to examine how a change in rail

load factor affected the LTL performance teasures. The effects

of increasing the rail load factor was to significantly reduce

the expected and actual contribution as wFll as the price paid

by shippers and the amount of Cwt. moved by truck, of the

transportation company vis-S-vis .he single modal companies. The

reason for the significar.. decline in expected and actual contri-

bution and price paid is that the high rail load factor has the

effect of lowering the out-of-pocke: costs per Cwt. which is

directly related to the rate charge shippers (the fully distributed

cost), hence the price paid by shippers is less than in the single

modal situation. The full costs of the LTL piggyback movements iwere substantially less in most cases than the fully distributed '

truck costs. In the .single modal situation, the trucker would

set his rate at his fully distributed cost since he does not have

to participate in Plan I moves. This had the effect of raising

the profit contribution of trucker relative to the transportation

company because thL transportation company priced its services

at the full cost of the low cost mode. Even with this difference

in pricing between the systems, the transtortation zo.pacoy cut-

earn-d Cie sin~e mal CM-a- nC . The effect of increasing the

rail load factor war to decrease the amnount of expected and actual

profit contribution dif rtntial bet'1ccn the systcn-.s. Even so, if

one loohs at Table X, onc can -tc that even after subtracting the

effect!; oi this, factor fri'O Lit, aVeraVU. in the table the

... ... ] . ... . . ...... I1l7 i ' ". .. . . .. . .... .. . . ... , ,ig - .

' -145--

transportation company is still more "profitable" than the single A

modal companies.

It should be pointed out that the pricing actions of the trucker

differ substantially betwee!, the two classes of movement--LTL and

TL/CL. In both cases, the trucker does not have to participate in

Plan I moves if he does not wish to. The same is true for the

railroad. In the TL/CL market, the trucker has an incentive to

participate in Flan I moves that he does not in the LTL market.

In the TL/CL market, the railroad is competitive for traffic that

the trucker wants to carry and vice versa. Thus if the trucker

is not low cost he will want to participate in Plan I moves if

this is the only manner in which he car. obtain some contribution.

This is not the case in the LTL market. The railroad does not

actively compete .r LTL movements In the portion of the market

9considered In this study. Thus the trucker will price his LTL

services at his full costs ii all cases.

The result of increasing the rail load factor was to signifi-

cantly increase the nubher of piggyback movements and correspondingly

decrease the number of truck movecnts. This had the effect of s-gnifi-

cantly reducing the price paid by shippers which is due mainly to the

lower priced rail related pi&-yback substitution for truck move-ents.!9

9 1t must be rc:,mbcrcd that this study, In examining o;ly co_..ncarrier rail and tru:k :.ovc'aent.;, andi associated 1lan I aid Plan 1piggyback t:''c mrnts in both traffi' caLt. rics. To lend scmc crcc'(:to the above .tatcMr.&t t ' it railroads do r.ot actively see' L'fL ::C'e-ments in Pan i1 _O.C: "ic auLlbor Contacted L.e 3urlin 'I. a r :"C.rn-.Railroad to Jcter-':, . type of "raffic Lei;.- carried !n La' I!moves. Yr. Robert Dro;o~u in the "iCO-C ithparL:zc;t jt dcatcd ;'o::i-Ulatly 992. of Plen'. II :s arc IL lots. To do.,I'e chec C; i 7 y ',M rs. Ro,:::ary IIurJ 1- thi -rei zht r.AcntL S O' ; , ':as c ..n.. L. ., .1; S".0

plzic'd th:t' C!';t i::.:tC .f 95;; of t1e i ,lun 11 U ':rrnt ' ,c.re

Ii .W~' C~ Li,

The interaction effects with a change in truck load factor

(DC) which bad signific:ant effects on the performance measures were

discussed in the secion explaining the effects of a change in truck

load factor, There were two significant interaction effects of

an Increase in rail load factor with a reduced level of shippers'

logistics constraints (DY). The DF interactions served to

significantly reduce the price paid and the amount of truck move-

ments. This is due to the fact thaL reduced levels of logistics

conhtraints allow the transportatio:. company to substitute piggy-

back service for truck sorvice to a -,ter degree than a high level

of logistics constraints.

Efraetu or a Chsn in the ap of the ys tor on the

Pcrformance keamures

The effects which resulted from a change in the level of

the capaclty in the Pystem are displayed in Table XVia and XVIb.

The chonge in the level of this fu:;tor also produced several

aignifitai1t risults.

t. p(.tcd Cont.ribut tcn--Tlh rw:tucton in the level of system

Sca j'd,, Ly had 110 0iptc i;ih 1 fcct On. tht, e.xpced contribution

of Ow Lranv i, ortnai.ion cut .-;)nny ir. ci ; d to .hi ; n&'3 ( r~ndal

cohiptcn ln In elth r tih.' TI/Cl. or 111 cat.,dory of urjvvmol -. The

I it.. r it c wI t ! ,: w tL!v of tr .i. L, l i I fac:tor (iC) 1gn 1IfI it mt y

-1'titift t1V V4i-' II t'od con'.r but ii el L'th: t.rn.n , or tat. I :i copanmy

,i#i11,- v 1; Lt- ~ ~~l . ! l :r l !

* I I

.40

N. N 9* *W

;5 I~~ 0 00 to

4 ICU

4;e

~.'hI %

.0 p.4 w~ N% m N

N~~0 4.44P4. to P* e

~ .4

V% to% A m LONN N V% co 0N

t DMNN N 0%

* II

u m N ' ,

I I

90 r4 00('

S4 N N I Ico ' P % m t

.-0

& j 0 0 &I o

u4 0'

." , . .

'o H

S4I 0 0 V N

IN r%-4 o 40-

. q- * r4 1 - 94 c

t, 49 1 I .I , I

U I

,, xP...& .. 4 0 4

. , SI I

U- F4 5,41* I I I * * S l

0.1.4 ~ ''A ~ ..

KS~~~4 OA -A' 4 -

,.' .. 4 0 04 W . 9- . ,

,4~ 1% 9 Il4 u Ir .

U, IIII' , 41 r ' I .,

44 tn 44141

UsU

I U U

U L

I S

... . . . ... . . . .. . .. .. I .. . . . 4'" " I.-. . ../ .. .. .. . . . .. .. .. I ' " . .. .. . :' .. . .. . .. . . . . . . . .

-149-

Actual Contribution--As in the case with expected contri-

bution the change in the level of system capacity did not signifl-

cantly effect this performance measure in either category of movement.

The EC interaction significantly increased the TIfICL actual contri-

bution, the swe reaction it caused in the expected contribution

measure.

Price Paid--The change In this tert factor had no appreciable

effect on the price paid for LTL moveients. The reduction in

the level of system capacity did significantly raise the price

paid for transportation services for TL/CL movements. The

interaction with a change in truck load factor (EC) acted to

significantly increase the price paid by users in the TL/CL

category of movement.

Amount of Truck Mov-.ent.--The decrease in the level of system

capacity significantly incr led the number of LTL truck movements

the transportation company ade compared to its average usage of

truck movcmcnts. This test factor change had no effect on the

amount of truck moverents in the TL/CL raket.

Amount of Pi:,,;vlack :e.c..L--Thc change in the level of

capacity sniLiflcantly increased tht nunb, r of pJggyback move-

ments in the TL/CL ,arket. Lwhf].e it actcd to sig:1iftca'tly

decrea!we the amount of piggyback c3Vg(.I1Ls in the ILTL market.

The Intur-iCLI on with truck load factor" (!:C) acted to sLl lfc.atlv

2 I

-150-

decrease the amount of piggyback movements, while the interaction

with a change in the level of logistics cvnstraints (EF) acted

to increase the amount of pigg>lack movements in the TL/CL market.

Amount of Rail Movement,--The reduction in the level of system

capacity served to significantly reduce the number of rail boxcar

movements in the TL/CL market. The interaction with the reduction

in truck operating ratio (EA) significantly increased the amount

of rail movements. The interaction with a reduction in the level

of shippers' logistics constraints served to significantly reduce

the amount of rail movements the transportation company made with

respect to the average difference betueen the two systems or modes

of operation.

Explanations--The discussion of the causative effects of the

reduction in system capacity will begin with the TL/CL market.

The reduction in the amount of capacity significantly increased

the price paid for TL/CL movements, The reason for this is that

the restriction on capacity necessitated a greater dependence on

full cost pricing on the part of the single modal carriers. That

Is, in situation; ,-here the capacity was not great enough 'o

satisfy demand, as a result of poor market estimation and allocation,

the single r.odal. c-irriers could not nor did they vish to transfer

unMet tI'cmanc to the co.--cti tJv1c r-ode. The transpcrtation cc::.iony,

on t~h other hard, when it dd c:.ihaust the capacity of one rledo

wou Id ,;hJ f L th t L:;:' d :::'d to -o IAI.. r !:,(!thod of r-ow.ncn t..

ii

-151-

end result was that the transportation cotupany did not have as

much unsatisfied demand as the single modal carriers and that, as

such, the price paid for trunsportation was greater than that paid

to single modal carriers.

These findings are further supported by the fact that a change

In capacity resulted in a significant reduction In rail movem.ents

and a significant increase in piggyback movements. This resulted

from the fact that, in most situations, the rail mode was the

scarce resource. The transportation company would then shift the

unmet "boxcar de=and" to piggyback moves.

The significart TL/CL interaction effects with the change

in truck load factor (EC) on the various performance measures were

explained in an earlier section of this chapter. The interaction

with a low level of sh~ppers' logistics constraints (EF) sJgnJfi-

cantly increased the amount of pl;yback tiovcments and significantly

decreased the amount of rtl movements. This is the same reaction

as the main effect, the change in system c pacity created, except

to a lesser degree. This is as cie would expect. The change In

the level of logistics constraints ce.,trols the amount of traffic

which Is subject to JIntcr-;d, cc:- MCLjt1o:,. Thus th rcp;sulc cf

expandfi.g the co.petJij,;,. traffic :-.rkct aid reductng LW anu-unt

of capacity was to ca-,-c thc tranmpurLation ,:m~any to shift

competitive ,r,nd r.i:. scarce r',sourcc to a, availih]e next

most profi table z:uthc-? of LoLe!.L.

-152-

There were no significant interaction effects associated

with this te.;t factor in the LTL category of movement. 1he

reduction in system capacity significantly reduced the amount of

piggyback movement and correspondingly significantly increased

the amount of truck movements in the LTL matrket segment. This is

because the maximum contribution method, In general, was piggyback

movements in the LTL market. The scarce resource was also the

amount of flitcar capacity (See Table I, Chapter III). Thus reducing

the system capacity, in effect reduced the nurber of piggyback

movements the transportation company could make. The unmet "piggy-

back demand" was shifted to the trucking mode.

Effects of a Chnnre in the Level of Shipers' Ln,,-istics C:n.trai.ts

on the Performancu Measures

The effects which resulted from lowering the level of shippers'

lgZistics constraints are presented In Tables XVIIa and XVIIb.

The change in this test factor produced several significant results.

E.:pected Cu:r~ut~o:,,--In the 11./CL catcgory of moverenis,

the reducLion of the amount of traffic subject to shippers' logistics

constraints, sig:Jifcanrly increased thc CzpCcted contribution of

the transportatio:i compa:y over its SJng. e modal ,-ounterparts.

The interaction vith a .ha:ige In rail load factor (Fi) servud to

sign ifica;!'t1y 01CUCC /Cf. expecLod profiLt a.ure.

-Il

.4 VAg 4

IP

100

.93 w% V Vu. r . m

dig N

-of%pi P . I6 OvV4

1:2t

,.it'i 4

IM~ g9 g04 p4

.4 a4

M 40 r.* I

So 4

,,dai

%d in In 'M

00

as I.9" q 4 %a a

~ N 1 a1

054 *@

caS

uJk 0 'S n'

in r f % 94 Wm 1% %-1 .r VV4* 0% No m% 0 04

VA %n O0N

41 N 44 ''a 04 N '

, ,. , , , , , Lt. -, 4Zt l . I -. 4I I-I I I I

S-155- '

Actual Contribution--The change in the level of logistics

constraints had the same effects on this ;erformance measure

as it did on expected contribution.

Price Paid--The change in this test factor significantly reduced

the price paid for LTL trarsportation services. In the LTL Lategory

the interactions with truck load factor (FC) and rail load factor

(FD) served to sigaificantly increase and decrease the price paid

for LTL movements, respectively. In the TL/CL category of movement,

the interaction with truck load factor (FC) also served to signifi-

cantly increase the price paid for transportatipn.

Amount of Truck Move2nts--The change in the level of shippers'

logistics constraints had no appreciable cffect on the amount of

Cwt. novcd by truck for TL/CL uovements. In the LTL category of

movements, the reduction In the level of logistics constraints

significantly reduced the arzount of Cwt. the transportation company

moved by truck in relation to the single modal carriers. Also for

LTL movements, the interactions with truck load factor (FC) aid rail

load factor (FD) significantly increased and decreased the nutmber

of Cwt. moved by truck, respectively.

Anount if PJ' .'back..:"... ihe r. duced level (f loc'stic-,

constraints scrved to signilficantly increase the amount of LTL

6L. movcd by plggyback. The chan2c in this test factor hi; jurt

the OppofLic CffCLL GO Lhe 43 Qu;&L of 'i'/CI C',:t. ;o'.d by p,

r ,= :¢" "% %- ',," .-'-.. "c: := - -2. = -. 4 , .'- . .~- - s-..= - .*-.. - --...... .. ..

. ,

-156-

that is, It significantly reduced the piggyback movements of the

transportation company compared to the siL.gle modal carriers. In

the LTL market, the interaction with a change in truck load factor A

(FC) significantly de-reased the number of piggyback movements.

There were two significant interaction effects in the 'iL/CL market.

The Interaction effect with rail load factor (FD) acted to signifi- a

cantly decrease the nunber of piggyback movements. The interaction

effect of a reduction in logistics constratints with a reduction in

system capacity (FE) significantly increased the amount of ,wt.

moved by piggyback.

Amount of Rail!v-,nts--The reduction 1r the level of logistics

constraints served to significantly increase the amount of Cwt.

moved by rail boxcar in che TL/CL category of movement. The inter-

action with a low level of syntem capacity (FE) acted to significantly

decrease, the number of boxcar movements.

E.p]anations--,lI. of the significant interaction effects

caused by a change in the le1vel logistics constraints have been

discussed in previous sections of Ihis chapter. Therefore,

attention will now be placed o, the explaihaLiotts as to hDw the

change In this test LIcL r caused sjnJfJ(c:at nain effects in the

various performance i..ca.ures. she silificant a-in effects for

TL/CL movc...ints wil1 uc discussd fjr;t.

Wli the c.er i ,'c] of c gi :',cs co; *rzAAnts i. ',cd on LicO

S i.t|] .: ., t:, :;nda l-u ck r, ' . h :;It; 1.n0t tilc IC.: C -¢ S

~ / -157-

carrier, wculd try to compete for traffic whenever possible by

utilizing P'lan I movements or by pricing trucking movements at

the low cost method of movement. When the logistics constraint3

vere low, this meant cornsiderably rore traffic was moved by

piggyback In the single modal system than when the constraints

wete set at a high level (See Table 1, Chapter III). The

transportation compan7, when faced with low logistics constraints,

used the met*Aod of movct-,ent to satisfy unconstrained demand that

had the maxium, contribution. In a large number of situations

that maximum cnZ.1.t octhod of movcrnent turned cut to be the

rail mode. Fz-.r TL/Li. to-.': ents then, the change In logistics

conutralints ca--red the transportation compiny to use si-1ificantly

more rail m'ovements and less plZ ,ybiack movem.-c~ts than the singl~e

modal cairiers. This resulted in greater expected and actual

contributions for the transportatio-a con-janny than its sin~le. n odal1

counterparts because, I:a &eneral, the sinSIh rmodal cquipment

allocatiot, decisions di.. not result in the maxitgum systcm v;ide

contribution which the. trinportauion company did obtaln.

In thp LTL market sc,,-.cnt, the reducc~d level of logisticsAi

cons~traints found the tran'spo' .iLl CO-.;z'.iy shiftiig a significanti

amount of tunc'~nstrii:ie!d :.raffic, 0that tlc i 1 r.cz~ crrier.s

were moying, by truck, to p ba:oc:'t.Thetr~rain

company did tit,:s whe:, 6'; c3C;' nofp ;;hc1 :~oc~

greater than tha.t Of LTLC-. :OVc.- Altllcu .h thcu cp1-~. z.o~hc C!

-158-

in contribution as a result of a change in the level of logistics

constraints, it did decrease the price paid by users for transpor-

tation services. The reason for this is that, in most situatioas,

when the transportation company utilized piggyback movements this

method was the low cost (rate) method. Thus piggyback movements

significantly reduced the cost of transportation service for the

customers of the transportation company as compared to the cost

paid by users in the single modal system.

Sunmary

Because of the length of this Chapter, a brief summary of

the important results is appropriate. To begin with it was found

that most of the average differences of tle performance ieasures

between the transporttion company and the single modal carriers

were significant. More specifically, for TL/CL movements it .as

found that the expected and actual contribution of the transpor-

tation company were significantly greater than the sum of these

measures for the single nodal carriers. Furthermore, the transpor-

tation comppny made signiflcant~y loss pl;:gyback movements than

the singlc modal carriers. In the LTL catcory of t'over.ents, the

expected and actual coniriLution of the transportation company

was again s gnificantly larger than the c.,abined contributions o

the single rodal carriers. The transportaLiCn c ,mpany also had

the effect of significa!tly ruducing the price paid by zhlppers

for LTI. tra.--;,porratic:; !c:rvlcc-s. Ic tr;,n.portation co:,A; bl! *.ode -

• - --

-159-

significantly fewer truck movements and significantly more piggyback

movements thin did the single modal carriers.

The prime consideration of tblis chapter was the identification

and explanation of the manner in which the levels of the test

factors effe:t the average performance measures. The chapter

presented the analysis of the effects produced by the six test

factors, for each level of the size of ship7 ents parameter, on

each of the performance measures. The results of the simulation

are graphically displaycd in Tables XMla through XVIIb. The test

factors prod-aced several significant main and interaction effects

in the performance measures.

The following tables recapitulate the significant main and

first-order interaction effects. In the tables an arrow pointed

upward (f) Indicates the effect increased the difference between

the performance easures of the transportation company as comparedto the single modal carriers. An arrow pointed downward (+) indicates

the effect decreased the difference between the performance measures.

The absence of an arro-,, indicates the factor or factor combination

had no effect on the performince measures.

V.44

0 4404

ad

4 a

4.4

4 4

I4j

O'd 0

.Fri

'JR in

S ' 4I .I,

-162- '4 I I ;

9 Economies of eal,

One faqat of th possible economic advantages a transpor-5 I *

tation company might enjQy over its single upodal countirparts thdt

not directly considered in the preceding inalysis is the

possibility of achieving .economies of scile as a result of com-

bluing two different modes of transportation. As indicated in 5

Chapter I, there is wide disagreement on the part of transpor-

ration economists to whether or not a nulti-modal transportation.'lI 4 . l

company couldachieve signiticant economies of scale. The disagree-

iont centers on the question of how similar are the operating II

functions of the combining modes 'and is there enough sImilarity

to allow' the transportation company to centralize ftnctions and

elminate. duplicative functions, and so forth. Thete kinds ofI I

questions 4iiX be addresied in Chapter VI. As will bb pointed

out in Chapter VI, there really has not een enough research Jone

in this area of the transportation company concept tb answer the *

question of whether or not there are substantial economies of

icale to be bbtained from combining separate modes.,, ,*

'How does the fact that economies of scale were not directlyI II

eonsidered in the analysis affect the. results of the study? To"

understand what effect this has on the reaults of the study, oneI " I S

mt know exactly how tho simulation considered acooies of "

scale, The cost data, for a given sot of paramtors, which -the

Atramportatidn company utiltzod to make prfcfuig and equipment

allochtion decisions was *xa~ctly the saml'as the coat dataltho

I * II

I II I

-163-

single modal carriers utilized to make their decisions. As such

the fixed and/or common costs, which are the costs which economies

of scale would effect, for each method of movement were the same

across the two methods of operation, i.e., transportation company

vie-&-vis single modal carriers. In essence, this means that the

transportation company considered in this study was operating

with no realized economies of scale resulting from the combination

of two separate modes of transportation. This conservative approach

to the econonies of scale question means that the significant

effects described earlier in this chapter would still be valid

even if there were some economies of scale involved in forming

transportation companies.

Te reason why the results of this study would remain valid

even if there were substantial economies of scale involved in

operating transportation companies is because the nature of the

performance measures which deal directly with costs and rates

are based upon the total contribution to fixed and/or common costs

and profit margin. Thus, while economies of scale may significantly

effect the profitability of a transportation company, they have no

effect whatsoever on the contribution which was used as a measure.II

of economic impact in this study. One mv, however, think of the

total contribution as a measure of profitability. This is because

the same fixed and/or common costs would be subtracted from the

total contribution. Thus if It were finally determined that there

were no economics of scale to be obtained from the formation of, iA

-164-

transportation companies, the results of this study would be

directly applicable. If there are some economies of scale, the

sigrificant results would remain significant. If the economies

were substantial enough, they may have the effect of adding a

number of additional effects to the significant list. This is

because the economies of scale would reduce the fixed and/or

common costs as compared to the sum of the fixed costs of the

single modal carriers. Thus increasing the profitability gap

between the two organizational approaches.

As stated previously, further research must be directed at

determining the extent that functions may be combined or partially

eliminated In multi-modal transportation companies before reasonable

estimates of economies of scale may be made. Thus the performance

measures used in this study avoid this uneasy task and yet provide

valuable information as to the economic impact a transp'ortation

company would have.

The Significance of Interaction Effects

Because there was a number of significant interaction effects,

47 of 330, a discussion on how these interactions have been inter-

preted is warranted. It will be recalled that in the use of the

fract.onal •actoral design, the main effets and first order inter-

action effects are confounded with higher order interaction effects.

It has been assumed, as stated Chapter IV, throughout that the high

order Interaction affects are zero. That is, in the explanations

of hov a chitnge in the factor levels "caued" thc significant effects

-165-

it has been assumed that the aliases of the main and first order

interaction effects are zero. This autho believes this

assumption to be reasonable. The reader is left to his own con-

clusion about the validity of this assumption. At any rate, if

the reader is not inclined to adopt the assumption the investi-

gation has at least identified the groupings of possible inter-

actions which have caused the significant behavior.

Another matter concerning the rejection of the null

hypothesis, that an interaction effect is zero, deserves some

attention. If an interaction effect has been determined to be

significant, is the rejection of the null hypothesis concerning

the main effect of a component of that interaction effect

meaningful? That ii once an interaction effect has been found

to be significant can the main effect be considered and if so

Is it meaningful? Guenther states that if the interaction is found

to be significant, the main effect can still be tested but the

results of such tests are usually of no interest. Guenther goes

on to say, "when interactions are present, the best treatment

combinations, rather than the best levels of treatments, are

usually the prime concern".7 This may be true in most experi-

mental settiigs, but this is not the case In this particular

project. The reason for this is that the object of this

dissertation is to obtdin a broad picture of the effects specific

factors have on the operating characteristicn of a transportatiot

7 uentLer, . p. 103.

'0110 Al ii O " '

I

-166- -

company. The intent is to show over the spectrum of possible

test factor levels the effect these facto:s have in general on 3the operations of a transportation company as compared to the

operations of single modal companies. This kind of broad or general

information would be useful in focusing attention upon a possible

change in public policy as it pertains to multl-modal transpor-

tation companies. That is, policy makers would be interested in•I

the main effects of this study in so far as if there were a changeA

in policy It would likely be general. That is, the Congress orI

th-t I.C.C. would be interested in figures as to how the transpor-

tation company concept fairs over tho average level of the factors.

These are the main effects. The carriers or. the other hand would be

interesLed in how the specific levelo of the factors affect the

performance measuren. In oth-r words the carriers would want to

know how they would benefit from the creation of a transporLat Ion

company with the specific levclst of the factors the intcr !nted

carrier have. ThIs study provides both of these Inforatioual 4

requira:wwto. 'hum both the rlgni l cart 1ntcrsLction and main

effectpl of' tL Ii tud ll, ta - f J:'terc;'t.

1.i

) -167-

CHAPTER VI

OTHER ASPECTS OF COMON ONERSHIP

In the last chapter it was found that the formation of

transportation companies can result in economic benefits for both

the providers and users of transportation services. Before one

can really interpret the meaning of these reoults and outline the

implications they have for the transportation industry, the signifi-

cance of the economics of common ownership must be related to the

other aspects of the issue. As mentioned in Chapter I, the

question of whether or not transportation companies should be

established in a multidimensional issue. That is, the formation

of transportation companies involves economic, legal, and social

issues which must be resolved before such companies can be

established.

The purpose of this chapter is to identify the other aspects

of the transportation company concept and discuss their relation-

ship to the research accomplished in this study. This chapter

consists of two parts. The first identifies the obstacles which

currently bar the establishment of transportation companies in

the United States. The second section of the chapter relates

these obstacles to the research conducted in th15 ntudy. The

implications end Interpretation of the results of this disser-

tation in the light of the total aspects of the concept will

be reserved for the final chapter.

WW

/ -168-

Obstacles to Integration

Sampson and Farris have noted three types of obstacles which

stand in the way of integration in domestic transportation. These

are regulatory obstacles, inherent obstacles in integration, and

1environmental obstacles.

Regulatory Obstacles

In this section the National Transportation Policy, the laws

applicable to common ownership, the case by case development of

the I.C.C. criteria for determining when a railroad may own a

trucking company, and the role of the Department of Transpor-

tation will be analyzed.

The National Transportation Policy (NTP)--The National

Transportation Policy, as stated in the Interstate Commerce Act

is as follows:

It is hereby declared to be the national transpor-tation policy of the Congress to provide for fair andimpartial regulation of all modes of transportationsubject to this Act, so administered as to recognizeand preserve the inherent advantages of each; to promotesafe, adequate, economical, and efficient service andfoster sound economic conditions in transportation andamong the several carriers; to encourage the establish-ment and maintenance of reasonable charges for transpor-tation services, without unjust discriminations, unduepreferences or advantages, or unfair or destructivecompetitive practices; to cooperate with the severalstates and the duly authorized offic~als thereof; andto encourage fair wages and equitable working conditions;all to the end of developing, coordinating, and pre-serving a national transportation system by water,

ISampson and Farrio, gopcic., p. 332.

-169-

highway, and rail, as well as other means, adequateto meet the needs of the commerce of the UnitedStates, of the Postal Service, and oZ the nationaldefense. All of the provisions of this Act shallbe. administered and enforced with a view to carryingout the above declaration of policy.2

After a careful reading of the NTP, one can conclude that

there is nothing explicitly stated which forbids the creation of

a transportation company if 'the creation of such a company could

be proven to be "beneficial." However, in the phrase fosters

sound economic conditions in transportation and among the several

carriers, the term several carriers has not been interpreted

by the I.C.C. as a reference to the carrier industries or modes,

but rather to the protection of the business stability of indi-

vidual carriers (firms) of a mode.3 Also, to preserve the

inherent advantages of each mode implies a sacredness of division

or separation of each mode from one another. Although the term

inherent advantages has never been adequately defined, it is

usually taken to mean that each mode has some unique character-

istics that enable it to offer services for transportation users

which other modes cannot.4 This statement should not, however,

prevent two or more modes from combining services to create

2The Interstate Commerce Act, (Washington, D. C.: Govern-

tent Printing Office, 1968), p. 1.

3National Transportation Policy, oocit., p. 37.4Locklin states that the low cost carrier, on a fully distri-

buted coat basis, has the "inherent avantago". (D. Philip Locklin,Economicsof Traj.potat io ., sixth edition, (ilomowood, Ill.:Richard D. irwin, inc., 1966), p. 865.

-170-

some new inherent advantage, but the implicit language of the

NTP does not seem amenable to such a cons deration. To sum

up, as one scholarly and oft quoted source puts it, "this state-

ment of policy, at least as it has been interpreted, has not

gone beyond the frame of reference of promoting the stability

of each mode as a basic requirement for a healthy and satisfactory

national transportation policy".

Applicable Laws--There are four mutually exclusive ways

in which a railroad and a trucking company may form a transportation

company. First, the tricker may acquire an existing railroad or

secondly, he may seek new rail operating rights and,.if granted,

build a new railroad. Since trucking companies, even the largest

ones, are much smaller asset-wise than railroads, it is unlikely

that either of these possibilities would occur. The second method

would be even less likely since the rail mode already connects

all manufacturing sites of any size. More likely methods of

forming the transport company would be thirdly, for the railroad

company to acquire an existing trucking firm or fourthly, the

railroad company could apply for new trucking operating :.'ghts

and if granted, purchase the relatively inexpensive motor

equipment.

If either carrier attempts to merge with the other, they

must receive specific approval from the I.C.C. as outlined in

5National Transportation Pollcy, op.cit., p. 37.

-171-

in Section 5 of the Interstate Commerce Act (I.C.A.). If the

railroad requests new.motor carrier operating rights, it must

obtain a certificate of public convenience and necessity as

described in Section 206 of the Act. If the trucker wishes to

construct new rail lines it must be granted permission by the

I.C.C. under Section 1 (18) of the Act.

Since a trucking firm has never attempted to gain control

of a railroad, there are no precedents which have been set by

the I.C.C. and it is therefore uncertain as to the posture that

the Commission would take regarding such a matter. If the

Commission had developed a generally appropriate decision

criterion for determining when transportation companies were

desirable or "in the public interest", it would not matter which

mode requested the merger or new operating rights as long as

the proposal met the criterion.

There has been over the years, on the other hand, numerous

attempts by the railroads to acquire control of motor carriers.

As a result, the I.C.C. had developed a very explicit standard

(perhaps unacceptable) for determining when a railroad may acquire

a trucking firm by means of the two methods for doing such as

mentioned above. Section 5 (2) (b) of the I.C.A. contains a

special proviso relating to rail acquisition of motor carriers.

This proviso stipulates that the Commission must find, if the

applicant is a railroad, or owned by a railroad, "that the

transaction proposed will be concistent with the public .,itrest

-172-

a.d will enable such carrier to use service by motor vehicle to

public advartage in Its operations and wi.1 not unduly restrain

competition".6

To ensure that the above proviso is met, the Commission has

evolved certain standards based on case precedents to apply to

such proposals. Much of the following sunmary of the specific

precedents was obtained from a study by Byrron Nupp. Thefirst

case was the Barker Case (Pennsylvania Truck Lines, Incorporated,

Acquisition of Control of Barker Motor Freight, Inc., 1 X.C.C.

101 (1936) also 5 M.C.C. 9 (1937) and 5 M.C.C. 49 (1937)) in

which the Ccmission decided to restrict or confine rail

operated =otor service to that which is "auxiliary and supple-

mentary" to that performed by railro.,ds disallciw .: t.uck

servkle which would be stricty conzetitive with the rail operations.

The hansasCitySouthern Case (KansasCity Southern Tran,"port Co.,

Inc. Co.on Carrier Application, 10 1(.CX. 221 (1938)) ext.endcd

the Barker Ddcrrize to original certificntes of public convenience

and necessity noL covered Jn.Sccrion 5. This case further re-

strctel rail. o'.- d truc. -.ri.cc to poit., on the r: il lIne,

aid to ihlIa.(rL; c(' ved ot d,.livtvrd by rnI, anid disallwo':ud

trucI nurvicc btt:{cIn "4v poi:ntu" ou Use raiJ. line., SiiCr- thu

K|,sIIln;[ C1ity Foutho'i Atit! 1.,I'rtur Cxrt,:,, v lvv,rL1 Sup r :.;., CuurtL

I,

ih iL i ' ' .. . ., AT ,|s ** '. I . (. .. ( ! =' =

9 J lI * . 9 "' ! , ' 9 , 9 ,, '" ( . ;

~III,. J, , , . I i ,

473.

cases (U.S. at-al., v. Rock Island Hotor Transit Co. et.al.,

340 U.S. 419'(1951); United States v. the Texas and Pacific Motor

Transit Company, 3.40 U.S. 450 (1951)) havc upheld the Kansas City

Southern Doctrine.

The Commission applies the above standards in merger cases

and In cases where new operating authority is sought. A very

important point can be drawn from the above anz.1-sis. As Nupp

puts it, intermodal combinations have beent treated more restrictively

IIthan intramodal combinations because in addition to the usual4

4."public Interest" test, addition~al criter--.a have been established. 8

These restrictions were spelled out at a time when railroads held

considerable monopolv power aiud Later wheii tle truclgin; industry iwas a stru;Zlin- flledclin-. This is clearly not t1w~ situation

now. The recent deterioratin& econoric sItuation of a number ofj

merica's largest raihroads can bear tusti.=mo)ny to this fact.

The I.C.C. has then strictl~y Int. rprt~ted the law as stated

in Section 5 of the t.-C.A. and "tno Jntenvt appears to be more

I ~ nearly to in~sure thrit cach jznodek rema.ins In compe~tlticn with

the others rad to itroserve all rncdcs,- ratctr thun to Frovide

the me.anrs fur 4dzyLi-.j Oki: to placts fla i nctord vith their relativ:e

eC(,ncjr-JC c.q i;' l HiLI a:." . Aio I I,,:: soirce conrlc I'cs ,

l.thnLthe Ow, r!' or H.- pre'nent rr ~ulataryPol r...y I~ :; t- j*.: c'JVr L".1. fi. ( i f

)'.. , p~. 31

* 174- , .9,r- r -=-..~~

4

I - .. a general program of preserving the s.tatus quo .--

which is in direct o positiun to the overa]lobjectLive of a d:'nanic transportatioL system

which can best serve the eccno-y and defenseof the country. 1

0

The Role of the Denartment of Transzcrtaticn--One of the

functions of the Departm.ent of Transportation as.-stated in the

Declaration of Purpose of the Departmant of Transportation Act...

is to provide general leadcrship in the identificat!Qn and solution

of transportation problems. As such, research into the concept

of the transportation co;'pany falls under the jurisdiction of

this Department. A search of gove;I:meut docuzents.or literaturereveals •that as of yet the Departmen;t has not published any

-materizl related to rese..Irch on the transportation company concept.

To determine if any work was currently belng-done on the subject

the author-wrote a letter to the SccreLar' of. "ransporta'tion

rioquesting information on research wzhich the Department had done

or is currently doingoris plnnin* to do on th subj.ect. The

reply was ncgative.

Inhcront Obstacies

As alluded to In the pre'V;ou5 secticn of Lh i 'chapter the

formation of transportaticn cvpa:: '.:.uld rcfuire volun'ary

action on Li c part of the particip .,.t. x Sa=pson and Tarri.

put it:

110: c,:. Tr - ti' :'.,ri : ol; ,', or.,it., p. 38.

' ...I . . 2_

I11

-175-

* Agreement .ar.onp the firms boe.Z mercd or bcinz InIto&rated is aot al".iyas vcs Since :.an are ccrporations,=4n dn, cerors uith Idivoere noals anc, Objectives mustbe satisffied. This Iia':clvos bcrh it.bt and ceuitv ou-nrs.Manap,,mrizt itsali may. di sarue. it is not easy 'to mer~coneself OLIL of a job, and a .r-orjed f rm can have but onepresldeint and set of, administratiNvC cificers. As wehave soz~n, acjust inrt in tnhco~t pariod a reailmost incvitabie ulzc- unificatioa or irt.-gr.tiontakes placo....The tendeucy to avoid ajustmort andchange; Is 'gruat in cll usiness..,s, and in some of. theolder transPortation fir=S 01his inertia itselfr i. amajor obStaCle. 1horefore, rhere are! obstacs. in, thevery idea.,.of Inte-ation.11

This conclusioa is supported by Carmane, Glaskowsky ard Ileskett

whcn they f')Und:

There is a stronrg tendency for railrc-ad exccutives tothink vitLIC,:V ia . ;4 a Ira'r; fo. trc-n

* ccnIpany ex<ecutives to thiiak only in tannis oi tuching,etc'. For, exxmn.;iC, j,, was n~o asy thinz for some rail-road e:xccutive-E to >.-arn to thinlk in. terms~ of opra~icin

* a coordinated service in parrnerc*;n uith trucers, agroup long de scrlbed by some raIroalftrs as "hereditaryI eeiies, a~tn- o the devil", or even -StLronger term:.sof opprobrium. Cai the other sida-, varicus trucking

* . .. executires have b-ecn de-eo1v suspiciou~s of railroadcrsand their motives ini offtrigp:b.c servlee. Thcseunfavorable rnilr-aadUr and .tru:k:cr attitud.:, have ser".ed

*1 ~to Slow theDevClont of coordinato-d servicc.

* Eveti when~ thure is . .. no inhercnt ar.imcsity betureen theranagerient groups ceniccerind, ... there sti3.1 re,:ains the

.. prolpl:s of co:mnunicantions. 12

Whil G~r~a:.Claskousky, and 1!cskvLtL were takn specifJi..dly

about c cord inat.Ld survices, their dis--ussion sce;:m., equally

app31ic:,;;)3r ,o tho cormon c.vlLrship .cd' .

111

-176-

I Environno.ntal Obstacles

The very nature of our poli ti cal and economnic system may

pose another obstacle! to the formatio oftasotto c~ es.

Again Sampsoii and F'arris have succinctly stated this proposition

as fal1oo s:

... The economic e-nvircon-7;ent of tht; nation is quite.oriented totuard indi-idual fir.-.s a:-d connectticn.Unification caid in'I czra .ion in volv2 -'rO*1z ac tio0.non the p..rt of firis and oftea a dccac' in com-ptitici. This may pr .escnz an obrtacle.

From' another Doi-nt of view, the political. Climatemay be. a ve~ry real ob'netl .stace. Iranspor-tation sysrc::n.-s srve nr.. ton."'S,. COU*i..!i,, andstates. Fach is a political unit with it.s rcnre-scntartives cleczecd to promote the Iwefarc of eachparticular Sovernn1a±1 UniL... . epolicI C~I;A-Smuay .cr Dos a [inze~ration) not bec-use it iseconoically unso-und, bu"" beca'Use it is Politi4.1call,,expeiient Lo do sc. The political c31ate, thcn,may i~apose a very real obstacle tt6...inr:ration1.' 3

Relnticnshl;?, cf this Stucdv to the C'stz~cleS

From the above discussion it is zpparent that there are

formidablu obstacles which pr:esent barriers to the establish~.ent

of transpcxtation co-F.panic-s. This the!ris has .detionstrated for

tse f jrst tivne that thcre. are. eefinita cccnctnic advantaftes

rLSUILing; Ircn the forma'tioul of transj;ortati cn cot-panias !or

both1 the carricrs and Ltce wut:S. Is this kn~idcsuiicic"t to

r.ove the c o! :r..!jicnc 0SaCC;S LO c~~e~h

nndn a Yaris j.. 1. .. 1)- t

-177-

3efore this question can be addressed, the author must

acknowledge however that the research was undertaken in abstract,

that is, apart from the real world. Thus before the conclusions

Of the study can be translated to the rea. world, the critical

assumptions of the study must be analyzed for their impact. As

will be seen shortly the validity of the assumptions cannot really

be determined without further research in the appropriate areas.

The following discussion relates the results of this research

to the obstacles which bar the formation of transportation companies.

Attention will be focused on: (1) the need for further research

on the common ownership question and-how this needed research

is related to the-critical assumptions made in this study; and

(2) the impact the knowledge obtained in this study may be

expected to have on each of the obstacles to integration. The

discussion will begin with inherent obstacles after which regu-

latory and environmental obstacles will be addressed.

Relationship to Inherent Obstacles

The focus of this dissertationhas been on the determination

of the economic feasibility of establishing intermodal transpor-

tation companies. The hypothetical firms modeled in Chapter II

are assumed to be capable of effect:vcly .rganizing to obtain

their goals of maximizing their (profit) contributions. This

brings up a series of questior.s which are concerned with what

Q . may be called the mant.eri al feasibility of establishing

A1

transpcortat-au co;:ipacides. T!4-e managc.zIal fLsiL'ility of t ranspor-

tation compaiies raises r'Losticons of not only is it possible to]

organL-e And. oparate a multi-cuodal transportation cempany, but Iwhat is, the best method of ogniing a transport~ation ccon'*any

under a given set of .circumstances.- In tL~e follo-in ragapA

the author h-is posed a set of questions in the areas- or or~ari-

zation, operations, marketinZ, a,.d finarcc. tihat should be zst:erd

before atten~ts are mad a to CStablish mu1ti-moda1 transportation

companiez. .tt should be pointedi Out that the author does not

a oses t the lists of questicns, 'are complete .but shculd -ive the

read--. in ap?reciantion of the arzouat oi rc search re~maliing to

be accor.plislhed..

krt.nIzaionzil AI-terni :v-Tie qui.e;tlcns that must be

answered In Ois area are concerned with-the extent to u:hich h

company should bo prganized functionally as opposed to orgnizJ~n:

allong Qod1a1 lines. "lle Author Poses the .:ypothesi-, that, ifjthe transpor:ation com-saes are, or'~ned~ alonz-mod-1. ines and

each zode is operated as. a pro.fit center, .the tr~wrportatlon

or.-rpan-.- w;'J. not achieve the cecncmics. 7rcm posr.-!L: r~Uo'4.C-

of tr~affic f:L.m. high to 1-. co-;L ca"rcr 1t;.t are pessible

-Ithiourh afu:ict.ionll2 orga'e co CL.zn. jliiz seii'.; :zsj."e

because the ::Od-a1 profiL CC:1tt'rc would be. att ptin- to :2 e

profi J~ L:; 1:1 tir ovm areas wilthout -. Z rd Lo thce oblecrIve of

totlp(1 L:~ ri :!,%- I or I fi rr. Az; rc lt I. Lha

-179- :transporti~ii company or;;anized alon- m~oda~l lines would notj

result In snahfca:it econo:'ies, as was t.Le case in this study

where the objeccive wis totz4 firm-ulde profit maximization.

-Obvlously iztrsarc not this simple. Oae must determine to

vhat cxtent the organization may be functionally organized.

That is, in eOach N=tional -aroa, -It t.-uat be JcLLr-.ne3 the

extent 'to which the cipcratic.ns of the ::.des can be co.,bined with

the object Ii mitid 'of- mxiizin. the total profits or ccntri--

bution of the -firm as a whole.j

pcraEti6.S---Lhere -nre several quesciens Lhat come to rdiud

which. must be answered in this- aica. To What, CXLCet Ican or should

ruaiiLLCei~uct faiili ties be com.bined? V~ow will the ocperactnt, of

more than oc~e tode affect the schadul.nr, pro'ble.-s of transpor--

tation? Isliz~ will. be the effect on the ccmm"unicat-Jons system

of the firms? What are. thc izpaCtW of containerilzat'ion?

jrlein5--1-at Impa~ct will t -transportLatic-n company h,-:vc

on service 'offered, to cust6mers-T low: will thc fornatioa o1

transportati on coumpanies afctth(: fz-~ i y equan cy,

and speed of service? Sho:.3d' the ce:.-pa-gy prictice consu'1tative

markati-.n x~d If so ,o M:ii :zcn . at is Lho 1=portance 2c:~

on s2N;k ri. sarvicc 1:y s*;,Ip,,,Xr:;?

Tin~ci--Vi~r ni i:f~;t~ ~..l bt ., IlI such a co:-.x.,:y.

place on -,1(!zn f ~~c - itrol .uct -. ? iityc o'~e f i-fer-

;.-180-

Labor elatLions--ill the unions concerned violently oppose

the formati:n of these con..zanics? howw wi..l the top management

people be chosen in each futLctional area? 7

As indicated above, the list of questions is not a completer

list of all r;atterz that must be investigated, but should give

the reader an appreciation that a great dual of though nust be

given to the question of whether a transportation company is

managerially feasible. Assuming that these questions can be

resolved, hcw can the knou-ladcd galned from this study be

expected to affect the inherent obstacles?

The author would argue that the pote-ntial profitability of

transportation companies should act to reduce the inherent barriers.

Althouth the vested interests of single modal managerants may

not entirely disappear, the prospects of increased earnings for

stockholders of single nodal companies should act to encourage

the formaticn of transportation companies, given that the other' " "" : 14

obstacles can 1-e overcome.

ReXLt iv:,shi '. R egu.atory Obsacles

In the conduct of this investigation it has been assumed

that the I.C.C. would retain the role of protecting the public

interest a'.-r transpor'e,-lo o wre fon.i-d, This is

a4!iiis: prcpCuitiiur, S .'rteo, e some:.:tat, by C.r.:., .

" - " .. ?. 72).I _____ ] vIt:l:;p o .*: L:::e'~ £ t.

I . ",,___,

-181-

15

a completely reasonable assumptLion. Of more importance is hcw

will the Coz.:lission react to the information &tnerated by thiS

study?

The National Transportation Policy (.'.Tll) directs the

Co=Ission to prctote economical and efficient transporLt-cn

services while zaintai:.ing the In icrent 6dvantarCs of cach tmCk.

This study has shovn tr.at a transportarion company, with the

objective of "profit iaxi: zation", is economically superior

(at least for son:e test factor co:-binations) to its single .odal

counterparts. The tranrportati on company echieves econozies

by the allocation of the nodes to .-oVcemnts for which th y are

low cost and/oi the ma.'.u= contribution i-ethod of rnovc=.Cnt.

That is the tra.sportation company uses the modcs On the route

segmccnt in which they have the "inherent advantre".

It is the authrs ophion that if the I.C.C. were ccnf'onted.

with such qunnLitativo!y docu'ncrtcd infcrmation the Co-issio.1

would be hard prsse~d to deny thlar such a company would not be

"in the public intereSt".

lhe r.jor o;nvirc:iintal obs. tcles which n.ay pr.vc.n the

formtI:, o: tra'zpor:atioi c:p.:; J: .dii Co:nt eticn that

the ,;za.'A of L,¢:, cc .. ; :,ct to c'~duct: co::icrp::t'-'ticn

" , tl' Q v L--

.....................................

-182-..'

and correspoudintly result in the elimination of soze facilities .

and jobs in some locales, It is not clear that the formation of II

transportation companies woaid result in less conpetition in

the transpor*atlon industry. Whqat has been determined in this

study ils that a transportation company "ould use the modes in .

a d.ffercnt manner than they are being used currently by single

modal carrles. The dif..rent allocations may require some

relocation o' facilities but, until the extent of possible

econodes of scale can be datermined, -it Is not certain :hat the

formation of tran.portation ccrpanies .il result in the i

elliination of facilitiQs, duplicative jobs, and so forth.

It is unlikely tlat the creaticn.of a transportation company

would result in the eliminaLion of facilities, duplicative

functions, and so forth, .o the same 4ireo that occurs when

two carriers of the same mode are mrred. At any rate, the

I.C.C. is tha arbitrator of what is in the public interest

and as sucl nust hear the. testimony of those who. would oppose

the fori.ation of t ransport.ILon compai.'es. To the extent thaL

the creation of such companies would result in the .limniatcn

of facili.ie:, jcbs aud so forth, will dccer.inc th e .ppositicn

to . r.nspor ,i t ion c~ip. i*.. by- locale. ,,, poli ti ci aus., unions,

othcr cL-.")c-tn . F-I.n'lc modal carriers, and so forth. The end

result is ti!Lt, e tho!:O.h this stud:' ha., in:dicat:&c the

pOt.:,th: l econ .-. . ,vt...-.s' of , rt. on . ., -.- ,

fict I3i.l: 1 LtL. i::.p .• tOf 01C

]r :v.: . :'. '.::I.[L. .au:apaat .. S L.me lL'aS'+a€.]. Ol W.I

~-183-

I" environmentc. obstacles. Thus the Counssion will have to

balance the economic advantatgas for the t:nsportaton

companies against the economic sacrifices which some people and

locales may cuffer.

It must be pointcd out that further research Is required

to determine the extent to which economics of scale may be

obtained by forzi i' tratnportation ccmpanies. The greater the

economies of scale, the greater the uconomic benefit t. £ormiug

companies bvt correspondl- ly, the magnitude of the econc:ies

of scale i rc inversely proportional to tho economic sacrifice

some people and locales will be required to make. That is,

>1 I . the tore cccnonies of scale the compony achieves, the greater are

the reductions of facilities and jobs.

This chapter has placed emphasis on the fact that the

economic aspects of co±-,on ownerhip are but one facet of a

.ultidir:enslonal problcm. The cl.c pter:has not attempted tostate all areas of the concept w-hich need future Investiraticn

N' but has attcr.pted to list iicportaht, areas tha&* require further,

thought. The follorjuLg chapter uill interpret the result.;

of this study with respOcc to all Lhe issues of cor:.zor. c-':,er-

ship a:id Lt the Implications for tile* future Of the concept.

I.

I~.

-184-

ClIAP2ER VII

CONCLUSIOS ANL IMPLICATIO.S OF THE ST'UDY

The results of the simulation were discussed in Chapter V.

Attention was focused on the rf~ects upon the perfor;i.aace measures

produced by the test factors. Chapter V also described how the

average perfor.,.ance tca-urs differed. bet-cen the single modal

carriers and the tiansportation company. The analysis was .ainly

concerned with describing the i.atute of the effects and -dis-

cussing the model orerations wh"ich caused the.. -Chapter VI

identificd otlier aspects of the transportation conpany concept

and related them to the research accc:r.plithed in this study.

The obstacles %hich currently bar the establishment of transpor-

tation copanics in tha United States weore discussed. Chapter

VI also indicated that there was a need for further research

(other than economic analysis) on the concept and related the

need for this resezrch to sor.c of the critical assumptions

made in this project.

The purpose of this chapter is three fold., First 'the results

of the research will bt brou,1.t togccher in one convenient pl.ce.

Sccondly, the chapter will futur'. avenues tf researich

which sh1.ould bc Pursued. The "osted r prc%. od in

this chapter is c.rned w Lh c:W1 en0ionz of the simulaticn

modc-. This diffur. frc-a th,.. r.sezrch ,'cat, prcs ,;ted "n

.4

-1!85- ..

'1the last chapter wldchi were concerned with matters which were

external or not cotsidered In the simulatIon. Finally the

ImpiLications of thi-s research projecc for the real world con-

sideration of transportation ccmpanies will be presented.

Suma__% of t!-e Research

It was found that most of the averaie diffrence.s of the

perforimance reasurcs between the transportation ccrnpany and the

single modal carriers wer_ significant. More specifically,

for TL/CL mo:ements it was fQ!-end ti'at tha c-:e,:ted and actual

contribution of the transportation company was 11% greater

than the sum. of these mcasures for the sinle uma.il utricrs.

In the LTL c.itegory of movements, the expected an- actual

contribution of the rranspora:ion compan: was 17% larger thar.

the combined contributlons of the sir.le -..odal carriers. The

transportation company alzo had the effect of sini'Icantly

reducing the price paid by shippers for LTL transportation

services by 12".

The prim.e co-r d cration of the research was the identifi-

cation and c:;pJ~l .nai of zthe =nnr in w 1 c h the lvi-ls of' the

test f.ctors arf,-L a , ,.rf,.-.:.-xcc -- cnizurc-q. TLe

analysis of 11C cfrCeLU; producad by '.he ..'x t fct izLor., A""o

each l*v(:] 0! the I:,, of sh'':':.t p::r,,:c.tur, revoa:l.vd undcr

what operatin-.; co:dt'u (t,:t factor 1:ve1::) the t

company was "economicall y superior" to the single modal companies

and vice versa.

The major results of the analysis are presented in the

following statements:

1. The reduction of the truck operatin- ratio from 99%

-t 91% had no main' effect on the performance of the

transportation comi~any or the sin-le modal carriers.

2. The main cifHects of a reduction In rail operating ratio

from 85% to 65% was to increase the expected and actual

contrib, -i of the transportati.on company compared

to that of the single modal conanies for both TL./CL

and LTL traffic. For TL/CL traffic, this change also

had the elfct of reduciriF the total price paid by

transportation compny users thin when this factor

was at its high level.

3. The increase Jn truck load factor from 100 Ogt. to

300 Cwt. acted to increase thc price paid by users

of the trzasportaticn coq.4any for both TLI CL and LTL

movements. This thanc also decreased the ex~pected

contribution of toic tranfsportation company as ca-

pared to thc-. t irZlc modal companics in the LIL

market.

4. The Incre-i~e In rail load factor from 400 L%:t. to

800 Cvt. surv:d to tcc~:' he dtffurncc in

e~p 2tr~ioni ;~L c~1c~nrJ1%u, "oa ~e thez t,.o

-187-

, organizational approaches. This chan-e also had the

effect of reducing the price pa-d for transportation

company custo ers in the LTL market.

5. The reduction in the amount of capacity available in

the system acted to increase the price paid for

transportation company users as compared to when this

factor was at its high level.

6. The reduction in the -. ount of raffic constrained to

the modes by shippers' logistics constraints acted to

itncrease the expected and actual contribution of the

transportation cor.pany over that of the single modal

carriers for TL/CL movaezans. Hs chance served to

reduce the price paid for transportation by transpor-

tation company users for LTL shipments.

7. There were 47 (of 330) sitnificant interiaction effects.

The manner in 'hich,these Interaction effects affected

the performance measures is graphically displayed in

Tables XVITI and XIX in Chapter V.

It must be stressed that the results just described pertain

to the analys!s of the simulation mod l pr so.. That is, those

results were ob:ained from the =aniplmaticn of a =athemat:cal

nodel which wz-; forul.ated to rescr-bl. the real. world ph!ca;c.,'non.

Before thc;c rc:-'ultr r:ay be statcd as truth;s research and allalysis

in the roa:l wcrld is required. "it ',1ould also bc poinLC-d ouL

7

-188-

. . . . . . . ..

:--.---ti.at the hypothetical nature of the study may be justified by

the lack of information on what the econom'c impact of establishing

* transportation companies would be.

These results may be restated as a set of working hypotheses

that may stand until verified or disproven by real world rescarch.

The analysis of the research points to the £ollc'rng hypotheses:

MAJOR IHYPOTHESIS--A multinodal tran.eortation cja v composed

of a railroad and trucking company will

Senerate. reater contributicns to fixdd

and/or co.-zon costs including a profit

allowance than. its single mo.-? counter-

parts at the snm- or lct:er costs to shippers.

SUB-IIYPOTHESIS --(I) The operating ratio of the trucker has

no affect-, on 'the econc-nic perforu ance :

of the transportaticn company.,

(2) The operatiun ratio of the composite

railroad directly affects the economic

purf, rmance of the. transportation

company. The lower the.operating

ratio the hilhcr .ill be the total

contribution.

(3) A tran, portlon cUmpann: with high truck

and rail load! factor.; .i l have a lar-er

contributicr. than a tr,,,rta.io;

cc.p);y Ui6h I,.: Lruc-. razil Ic.d.U

-189-1

* Iactors, if the rate Is held

Con'stdft.

-. . (4) The riore traffic is const-aii&ed along-

modal lines by shippers' logistics

I.... constraints, the. less Will be the

economic b&enefits of creacinZ a -

transportatlon cc.-.any.

(5) A. transportation cc:-.Pa-r.y is especially]

likely to increase the contr-bto and

reduce the tctLl price ;aid for transpcr-

tation Ii th-e LTL =arkat as compoared

withsli-o-le operticrs inthiFurter arkct s6--cnt.

Fu~erRsarch T'vu 1r d *

Chapter VI presented the -=or ci;tazles WhIC11 bar Lhe formation

of transportation companics in the United States. These were

identified as; regulatory, e., ct, aa *envonzxcn-al obstaclecs. .These obstacles verv' thun rclartc.e to so-.-. of the assump-itions, nade

in the project. As arc:lofteeosacs :dh su:pcn

made In the study, it dX.'tr:Lmiied that further research was

requir-d to .Iterine the Vali -LvY ----.:iics --d to

deter-iine if L. C . couli 1, crc.....Iir- *r sevcral

suggestioea. s d fo;r £u~~vin CIw~ arc. of .n:;ra

f easi b i11 Ly

*1

-190-

- :-.--. In addition to these research efforts, the nature of the

simulation nodel also sugges ts other avenu2s of research. It

must be urdCr1tood that the simulation model itself was the first

attempt made in this area to attempt to ansu7er the types of 4

questions addressed in this dissertatic. As such, the simulation

model was necessarily of a simple construction. The simulation

analysis could be enriched by expanding the number and categories

of test factors. As explained in Chapter ii, envirormental and acommodity factors were not considered in the thesis. Certainly, _q

enriching the model to consider these factors would advance the

model to a nore real-world-like abstractio:.. The model could

also be made more corj,3ex to involve the real world scheduling

problems carriers have to trying to place thei equipmcnt at

the right tiie in the rjZht place. Furthermore, the significant

interaction effects could be axa:ined to deter-ine which of the

aliases caused the significance.

iiAnother subtle aad implicit factor which should be brought

to the reader's attention is the fact that the simulation was

undertaken for a transp-rttrion s'stcm in which the route structures

of the for:n.:,a com p;.ie were parallel. U1,at: would be the effect

of different, route .tru-tures on t.he pcrfcr.;anae Of t0ar.sportatioi.

eompauies? Ihe aul hor I-'se thc Lpothcsis Lhat if the route istructures o;. the for.ir, carriers are not a;pprc:.,ately parallel

or, ir the route 5L*: c-f o.:.! carrier dcc.; ncl: lie W 114 in

Llh.t of G .. c..t, tcar.. t -" a :..anc:• ''n . j.\I

-791 7'

- would not result In significant economics. Tlae reasoning behind

this is that if the siznplc. modal route struictures we'cre not parallel,

but say end to end, there would be no opportunity for the tranLS~or-

tatoncoman toalocae rafie to the low cost mode. Thus in

effect a transportation~ company withi this type of route structure

would be nothin- uiorct7han a. holding Fcipn for two'-distinct -

modal oeprations. This question deserves further investig-ation.

Also thi's projct~ has been a shert run economic analysis.

The inVeStigation was conducted or. the basis' of current operating

plant 2nd using the current stata-of-the-art technology in

transportation. In the past technolo.gical developments have

Iplayed key rules not only in the developm.W&n of each mode, but

Ialso in the nanner in which the modes hiave competed with each

other. What relationsh p will future technologvical develop-ents

have on the M~lti-2oda1 transporLt1tin ccraoany? Will future

Itechnological developments reduce the aeconomic benef its of

transportation companies? A -ain these'are but a few of the

unanswecred quecstions concerniig cor~on c-w.aership wh-ich must be

addresscd bodorc serious attem~pts are m~ade to for-m %transpor-

cta~tlon.c~pii

-192-

- -of scale. This would allow one -to directly ascertain what the -

effects of tie factors are upon ptofitabil-ty of the firms for

varying economies of sccle.

*IMLICATTVo\SO~F E STDY

AlthoughL there remn a lot of unatiswered questio-s which

-stbe resenrched' concerning the ccmmon ovnershaip question, this

study has been a step in the direction of obtainirZ answers to

those questions. It hasb*een pointed out that the.<-resulmsof

this study must be verified In the r 6a world. -As such the study

has served to develop a- set of working hypotheses which should

be tested in the real wVOrld. In cthacr vords, it Is the author'sa

opinion tbat LIM economic benefi ts of tran!,portation companies,

over their single modal counterparts, discovered inthis thesis

*should pr6v1C e iMPOLus for a set of cxpcrii~cnts involving

transportation companies 'in this country.

If the, reader. is r.cnviuccd that-the results of this

disscrtation warraint such a dev;*tion 'from* current pollcyi at

tevr ni:mtis digs , r Ltatibn should serge to eliva-te

the concept of transportation ..om-)ani s-from the spl e re. off

emtZ21al sc:7.nt:s illogical :rZumnn thAt have surroun~ded -

Such this; d::Lr;:.-;j~n Aloh-1d. ;-- !.)LU ftus r vxlre ~a ~u

di:1 ~':c 'h!~ --.-. r 1.U.. LI-C hj-~.I

6193-

BIBLIOGRAPHY

Books

Baumol, William J., Economic Theor an prtin nlsis,Second Edition,' (Englewood Cliffs, N.J.: Prentice-Hall, Inc.)3965.

Bonini, Charles P., Simulation of Information and DecisionSystems in the Firm., "The. Ford Foundation Doctoral DissertationSeries",p (Englewood .Cliffs, N.J.: Prentice-Hall, Inc.) 1963.

Boversox, Donald J., Smykay, Edward Wand LaLonde, Bernard J.,Physical Distribution Manase et, (New York: The' MackllanCompany) 1968.

Churchman, 'C. -West, Ackof f Russel L.,9 and Arnoff, p . Leonard,.Introduction to Operations Research, (New York: John Wiley andSons, Inc.) 1957.

Cochran, William G., and Cox, -Gertrude M6, Experimental Designs,(New York: John Wiley and Sons, Inc.) 1950.

Dixon,, Wilfrid.J., and Massey, Frank J., jr., Introduction toStatistical -Analysis., (New York: McGramt-Hill, Inc.) 1969.

Pel:'.math, Robert C., Thbe Interstate Commerce Omission: ThePublic Iute rest and the I.C.C. (TheRalph Nader Study Group Reporton the Interstate Commerce Commission and Transportation), (NewYork: Grossman Publishers): 19*70.

Gas, Saul I., Linear ProsgtaiminZ Methods and- Applications, ThirdEdition, .(New York, N.Y.: McGraw-Hill Book Company, Inc.) 1969.

Germane,,Gayton E., Glaskowsky,, Nicholas A., Jr.$ Heskett, J. L.9RivaX Transportation Management, (New York, N.Y.: McGraw-HillBookeCompany, Inc.) 1963.,.

Guenther, William C., Analysis ,of Variance, (Englewood Cliffs, N.J.:Prentice-110l, Inc.) 1964.

Guenther, William C., Concepts of Statistical Inference, (New York:McGraw-Hill Book Company, Inc.) 1965.

-194-

H iadley, G.., Nonline-r -md Dynamic Pronzrzmrinn,, (r~eading, Mass.:Addison-lwcsley Publishin, Co.', .-,c.) 1964~.

Ileskett, J. L., Ivie, Robert M., and Glas'Aowsky1, Nicholas A., Jr.,Business Loi.,stic.; o f Physicz! Stip~l and _Djstrib-;,ion,

Hicks, Charles R.., rund&:ntntal Cnnc -ts in the Desitgn of F.xterf-v tents, (NCW York: BILt, RinchnrL :snd ;instoai, Inc.) 1964.

*Hillier, FredericK S. an Lieberm~an, Gerald J., Introduction tof peration,; 11i:search, (Sa n csoCai. Hlden-Day., .Inc.)

1969. Facso ai.

11oci, Paul G., Introduction to ~aic:tclStc.istics, Third Edi*tion,(New York: John EWiley and Sons, inc.) 1962.

Hunter, Holland, Soviet Tr= :.ort 7.prience: its Lessons for

The Broo' inZiz iuStitU i.a) 1963..

Locklin, Di. Philip., j.con;e clc ofTrarsportcatlcn, Sixth Edid'on,

Mc~illan, Claude, Jr., M t--J ?rv:.rinS: An Inrrc'uctlonto the Desif 0' Lci-sicn aci2s(XNwYork: J ohn n Wiley and zilt, ..nc. 1970~.-I

Meier,. Robert C. , Naewell, 'Villiam T., and Pazer, Harold L.,v ~Sirnulatic- In~:~s : 'Econonmcs, (En~lewood Cliffs, N.J.:

.Peticte-~Iall, inc.) 19 6 1 .

Mendenhall, William,- Th Dcsimand Sa~q Of Eprients,A (Belm~ont, Calif.: VWas.rti li~a C.p.n -nC.) 1. 8.

Meyer, John R.,, reckl, ifrtn q., ~:oC. John, an.d. A!ickCharles, The Lznc:: c of Cor:', -.' -i flhn in *-..:r1r 4o7ndustri.:.;, :as, 199

Naylor, Thomans Hi., llalnLfy, Jco-vphi L., Burd.,c%, Donald S-, and

Chu, n, iw .i S., -. (nc Yo, .s UN.Y.:

Ciizrlc';- L. %.vvti I. , .,1*l.~

l'vgru-im, 1)D-.(Oy a: - .~:~ .! *',1jCIo i'j t CIo, L4.IW n :c ) i1.:(

-195-

Sampson, Roy A. and Parris, Martin T., Do-.e.tic Trcnsortatlon:

Practice Theory and pcll-v, Secoad Editicn, (Bfiston, X.!s.:HIoughton X!i:i1in Company) 1971.

Taff, Charles A., .Co:i:crcial Motor Transtcrt.tIcn, Fourth Edition,(lomewood, Ill.: Richard D. Irwin Inc.) 1969.

Trans.Iportatlcn Economics, A Conference of the Universities -National iU .u COZ.=iLLu for ECRc.'ic Recarch, National BureauOf Economic Research, (New York: Cclu:.ba 'nlversity Press) 1965.

Wagner, Harvey %.,, r~i' Ole ra "s Research, (EnglewoodCliffs, N.J. : Prent. ca-H!all, I-nc.) 1969. :

I

Articles and Reports

"'A Better Cl1s of Traffic' Is Needed for New Grovth", RailwayA,_, (May 31, 1971), pp. 27-28.

"American "rucking Tretds 1970-71", DIpart'"ents oi Research andTransport Economics and Public A.-ttir d,-.-..an TruckingAssociations, Inc., 1971.

Econometrica, Vol. 20' "o.4 .Bickerton, Richard L., "Self-lHc.ip ore faorc " .tansti " o4

.and Distribution *nsnaterient, (Ma 1971), .p. 29-32.Bickerton, "Tire to Plan". Trap fttior a air-ibuton Manar(-o.and lstrbuc~n ". .na-.o~~ent (:.a. n1) and Dist.

ment, (Deccrber 1970) pp.. 23-24. - 1'Campanella, Frank, "On the R!h t Track", a:rrcs, XLVIII (October 2,

1968), p. 9.

Dodge, Vil!.iam H., ".he Dile.m=in of IRr.1 iatu Coc.peAtion",15..C.C. Pj:tyrs.".', X."VI .(ju. Au;Usit 1969), PP. ISOI-

Dodge, W 111 o ii., "Tie Inhe rn L ,v.n tat . of C;rri.r . ds Cndorthe ;Natjo:':l 'irn 't P1o.c,", ' XtIV (:f:vc;:;rr1968), pp. 492-507.

II;]

-196-

.Douias, Peter S., "he Economic Irrclevarce of 'Cc-.on Ownership",

I.C.C. PractlLionurs Juunail, XXVIp (July-August 1969), pp. 1794-1800.

Doyle, John P., "A Transportation System", Distribution M,.namcr,(December 1968), pp. 35-42.

Harbeson, Rcbert W., "The Supreme Court and Intermodal RateCompetition", I.C.C. PractitioncrsJon X.XXVI, (March-April1969), pp. 1487-1495.

Harbeson, Robert W., "Recent Trends in thc: Regulation. of Inte.odalRate Competition in Traisporatioan", dEccnoics, XLII, (August1969), pp. 315-326.

Houser, Fred, "Putting Profit Into Pig-ybr.ck", Railwar Age, (May 31,1971), pp. "q-31.

Levine, Harvey A., "The Railroad Industr's Exper -nce Under Section15a (3) of the Transportation AIkt of 195 Z, I.C.C. ?ractitc;:ersJournal, XX)XV, (January-February .963), pp. 252-266.

Naylor, Thoeas II. and Finzer, J. M.., "Verification of ComputerSimulation Models", Manatcriecnt. Scnca, X!V-, (October 1967), pp. 92-101.

Nupp, Byron, "Rweulatory Standards -in Cor.-..n .0:nership in Transpor-tation", I.C.C. Practitionors Journal, "CXIV, (.oveober-Dece.ber1966), pp. 21-37.

Nightingale, Edcund A., "A Critique of e.ri.can TransportationPolicy Develop, ents", Third onnual Amr-rcan Transpaorta:icn ResearchForum, Pit.sburgh, .Pennsylvania, Decenber 29, .9.62..

"Panel Discussicn: The Effec't of Diverrlficatici int; Other Mfodcsof Transportation", .C.C. Prnct tlunor -curn, XXXVI, (Sepre.bar-October 190), pp. 1969-2011..

"Piggib-ckers All Fall Dc-.n But 1ailrcaCZ Look, for Pickup",BusIc-s '.c-:, (.cov:r.er 22, 1969), p. 15b.

Rays, J. M. 'N., "A Sclection P1cblc.. of S .are Fixed Costs and Nt...orkFlo;vs", ..... v.c.nce, Vc-.. 17, No; 3, (:cv ..ur 1970), pp. 200-207.

Suelflo~, Jame.s E., It11113, sL.:Inky J., "The TIransportation Company:An Economic Af-U;M1L' fur 11Atr dutI 0.-ncrship", Laid Econmcs,(August 1970), pp. 275-2S6.

"The Astro..Plan for Progress", reatc and Distribution

'Transpdrtntlon racts and Trends", Trailspcrtation Association ofAmerica, Qishiaeton, D.C., Quarterly S upplceant, Janua±ry 19721.

Tucker, Willian It,, "Vantado. Tha Air - Truck - Rail - Water -Bu

-Comnpany ', Co1'iibid journal of Lovid Busincss, 11, (NoVeMbcr-Deccnb ar1967),, pp. 47!!54:0

I'Vore Asks Intermodal Carrier Experimecnt, Creedy Favc~rs Replacemc:at-Cost. Rate Base", Traf fic 1.1orld', (h1"ay 17, 19 71),p. 35-56.

'9i11iams, Ernest W.,. Jr., "'Traffic Compoition and Railroa. EarningPower", Tr~nisport-ti en Journil, Vol. 9, '1c. 1,(Fall 1.969), pp. 17-

*"Ycarboek of Railrond r.acts", 1970 tditic.., 'Ecnonic _--d FinarzcDepartLnei, ikl;OC~a~lbn of American 1Hailrcadls, April 1970.

Government Dcuments

"Explanation of the Deve).opnnt of 114tor Carrieor Costs with State-mant as to TerMi and SiEgnificancc", T.C.C. Statement '.o.4-59, Washing~on, D. C., 1959.

"Explanation of Rail Cost Fiiud- n Procedurcs and PrInciples Relit-ingto thc bse of Costs", I.C.C. Si tee.t No. 7-63, If.ashingLon,, D. C. ,1963.:

"Eighty-Thi rd Annual Recport on: Traasport Stat istics In the UnitedStates for the Yozr 'Ended December 31, 19 9", I.C.C., CPO,Washin&LC:1, D.C., 1970.

"flessajec o:i 1'rins-,oriaLlon", (i'residcul: K ci.y' d-j),~rConn*:, 2 .*.5 . Ei.c.us& '0! JZ OCUI.*nt No.

WashiiilZcn, D.C. , April 5,1'2

The int~ V:Itr- V> rxAt . S. Cove rwa-ettt Prin Cffice,

IS.

-...

-198-i

'U. S. Congress House of Representatives, Comdttee en ArmedServices, Ad cw : of r'-portatlon S vtrs it S-u-,ort of theNational D; i. se Eff-ort i n Event of " l zat; n"" H.,r:!s beforea SubcommitLee under authority oi il. Rcs. 19, 86th Coagress, 1st ...Session, 1959.

U. S. Congress Senate Covmittee on Interstate and Foreign Coimerce, 2Cby. .Rqrulamd Carrier- hqar.-s before the SurfaceTransportation SVabcc.ttc un S. 432, S. 1333, 1354, 1355, 2189,86th Congress,. 1st Session, 1961. .4

U. S. Congress Senate Committee on Comerce' v ato.al TrzmportationPolicy, S. Rept. 445, 87th Congress, ist Session, 1961.

Unpublished I-ateria!s .

Cherveny, Norran. Lee, A Si.ulatio: A..-vsis of.Cash Flcw P.atterns;Within a :.anufacturinr Or-i U.. tion, unpubiisiaed D.B.A. dissertaticn,Graduate School oi Business, Indiana .Un.vrsiry, 1968.

Glaskm,,sky, Nicholas A., Jr., AnAn s ._ and Evaluatlon cl theDevelep,.,ant of- CoordinartoA. ,iat.. ,r .. L... , ... Sneci alReferencq to ,orz!:ast: . c., uulS' P11.1).dissertation, Graduate S chool, , ."ani'd Uriversity, :1960.

Hay, G. A., .orlok, E. K., and Charnes, A., "Tcward Optimal Plan-ning of a T'o-.:od Urban Transportatizn System: A Linear Prcgra.min;Formulation," The Transpor:ation Ce::ctr at Ncrth%estern University,Evanston, Illinois, 1965 (rin aograp'a).

Morlok, E. K., Nihan, N. L., and Sulijvan, R. F.,"A Multiple-ModeTransportation :ctwork Desin ModelThc Transportation Center c.'eNorthweStern Lniv erity, E'anston, Illi:ols., 4 VoIu.s, 1969(mineogr:.phed).

I ,

-199-

APPENDIX 1

A. THE t TEST INVOLVING PAIRED OBSERVATIONSOF TWO POPULATION MANS

To determine if the mean differences between the performance

measures of the transportation company minus those of the single

modal carriers were statistically significant, a t test involving

paired observations was utilized. This test was used because the

observations (from each simulation run) of. the performance measures

for each organizational alternative are related to one another.

That is, the random components and parametric values were held

constant acrcss the organizational alternatives for one run and

varied between runs. Thus, the obser-atioas occurred in pairs.

The exact hypothesis which.was tested for each performance

measure was:

R0 d 0

H1 . Yd 0

whereA 4 isthe mean difference of the particulary

transportation company performance measure mi.as the

performance measures of the single modal companies.

The null hypothesis was rejected if

sd

-200-

--where t 1 is the t statistic with n -I degrees of

fredom associatced with a particilar level of significance a,

n

d~~ i - x+ Ri)

XM X.,X =one observation of a particular performancic

measure of the multi-modal transportation

company, single tmceal trucker, and r .1aroad,

respec tively,

2-

d n-1

u ~a numbor. of paired observatios which was 32 for b~th

the TL and LTL categor- Of. rovc-:42et.

3. : Thf ALYSIS OF '.n2 ,: 01 T:;E KAIIA,.1

nhe f~x-2d effects anaiLysi:; of variance (ANOVA) rodel was

USCd- in t'-,L %.issercta-1on tc. eatrmno if the: L-ain an~d 4f.rsL-cordet

interac.k-.1 LffeCLs 16er-C! btat ;stt'zIIy sit:ificanit. Th±e A:;O;A

t~ch:Ii'--- 1:.I3C 0:1 thL -.! C'. 0'Lh~ t( ta L:z 0f a r C.'

d i* iia tr :r'-7 tihe r;nc vao can b,: part i L -2d I :LO thC

a (r ore r: . ' d i o ti cT s t..

-201-

~smOf squars assoclaitcd with thc, main effects, Interaction

effects, and error.2

MatlaomaticallY the &MOVA model used in this project is

SST SSA + SSiB + SSC + SSI + $S,1 +I SSP

SS + SSAC +SSC J+ S Sa.+

SISAF + SSBC + SSBD + SSB*-: +

SSF' +t SSCD + SSCL + SSCF +

SSDL + SSDr 4. SSE- + SSE-

where SST =total sun~ of squarcs,

SSAS-..SB, SSC, SSD, SSE, SSF sum of squares associate~d

with eachi mai of f-Ct,

SSA~ SSC DSr.SE: SSBC, sun of square iac

SSBL', Sabc, ssiw, S$00, SSCE, sscF$' with the respective

~SSM, SSDF-, "ISIA interaction oeflects

SSE sum of squares as.-ocdatedi

with the error tan..

The test statistic utilized In tkie analysis of varicincc to

determine if the rull 1hypotheses concercr1v the main and first

order Intaractic~n effeects should be rejected Ais tha F ratio.

The F ri.tio is silily zie men stluarce of ;ho tratrment of

squaires arL2 .:i:.--,,y tlla t rf the s~uarcs ol cich tormn ld-ULifjt~d

{ ~-202- :.*1 .4

the F r.-Io is greater than the Fr(1 c&) statistic the n~ll

hypothesis that tu f.ator had no effect 13 rejectd. r and-r 21 2

are the degrees of f reedoz in the numerator and denctlinator of the

F ratio.

The AMOVA data obtained in this study for the 'L/CL expected

profit performance measure is presentcd in Table.XX for the Jrtrested ..

reader. The de;racs of froodc. are thle same for each of the other

.3.4perforr-ance 'ncasure.3

ionua fn .: r|.iofSur.;myb-

f r-' -6

. -

..4

A I

, . 1I '

3i

~1& c.:: L~:...1 zf,u!::. i.r t:...: .s:: of s~urt:'c: nay b.2iI} ~fou'n3 3 u (';~I4LhV, :[i c:,.,, u: :;,.:xjL:;:I~ul!.

r!1II

.n x a fl

-203-

TABLE XX

ANOVA F R L/CL LXECTZD PROTIT

Source of degrecs of Sum of F ...Variaticn f r e6cc Squares ratio

A 1 86,064 .00024

B 1 2,481,612,162 7.06678

C 1 77,484,194 .22012

S2,928,671,473 8.31974

E 1 7781271,971 2.21091

F 1 8,977,039,91.5 25.50187

AB 1 77,054,877 .21890

AC 1 491 .00000

AD 1 7,202$549 .02046

AE 1 631,307,171 1.79341

AF 1 l9,119,695 .02591

BC 1 790,168,790 2.24470

BD 1 2,62-,74S,2'2 7.45920

BE 1 214,529,719 .60943

BY 1 1,032, 1lI,683 2.93201

CD 1 17,847,9$2 .05070

CE 1 1,613,393,219 4.58331

CF 1 6,116,di5 .01738

DE 1 80,086,535 .22751

Dy 1 1,9A6,3.89,605 5.52871

EF 1 44,227,059 .1256-4

EPJ~Oi: 10 3,520,149,339

TOTAL 31

* . . ' - -..-..2

" I II I I Il I I II I 1

-204-A

APPENDIX 2

A. FLOWCiIART OF THE SMU!ULTION MODEL

A verbal describption of thL simulation model has been .

presented in the text of this study. Thlus, discussion of the

simulator . will not be included in the appodix. In the ...

discussion uf the si.ulation model, the flo.:chart. of. the model

was not included tc increase tho raadabiiit:: Of the project.

For the interested reader, the flowchart of the simulator is

presented in Figure 3.

B. THE S1",ULMU 'ON' C,,_±.'ROGIPO'l I

The co-=uter pr graz of the simulation model iS presented

in *'i6ura 4. The coding is ,rittcu in the (V:ITAB computer

language.

II1

I

'; . I .' " . . "

i~ur..,: ¢ ..~i ...= ";. ...: .,.: i /.[

-205-

Figure 3

9 Flowchart For Simulation Model

1 detrines the fullcosts of each modeon each segment

multiply thefull cost

-mode (t de contribution(2) low omp ~e, " O 0 by the amount

cost? (3) to (4i>, 'TZw h must- goby this =ode

YES YES and store -i" Y--qf ..... needed capacty

multiply .6 of multiply fullthis segment cost contributio-traffic subject b]to internodal (5) by that: which

must go by thiscompetition mode and storePlus t.at needed capacityreserved $or (A')this mode dueto logisticsconstraints bythe contr.butiona mult.py €ompet.-Iand store needed tive contributioncapacity . by utimates of

com-30titive(6) trafiic he will

obtain (.6) plus'

multiply the what must go bycontribution by this mode and -

amount of- traffic store needed(8) which must go by capacity

this mode and (B')store needed __________

capacity(10)

CE) (9YA>BG

YES S

.0 To

C -206-

re18) stop~ 3 throu9 for .t (method

H

(15)(16)

(19) ES 3 D YES r 0 0 E ( oB IB > D Store E G

NTO

Store' DToand R Q)

2(

YE tr Go(20) BI E D To(20) I' • Dand B1\ ,

(22) and A

(12)" Store. A\ N.+ O ".

and D iathod vwhi is

no low ost

YESrepeat routineuntil all segments (13) et steps 3 throughhave been exhausted (13) mepeet ohe 3thod

(3rank the contri-butions of theI methods

subtraci the capacities requiredfor movements fron total capacity,

(. ( beginning with mode and route' _kToI segmnt with laraest contribution,.ntil capIity IU :xhausted

, H ,|,

-207-

( Repoa.t routine -'Or truckcr

< j k **T R jk3i

repeatc rouitinefor truckerset =0

(27)

do i- .tis print the.ocatcd YZS 1. carriers capacity allocations

(26) -ariers.2. odes ected and actual

(26)f p by sipper<hi~p I 13. price paid by shippersdman ?

NO

lp. s -E- hos of¢,. .

. .he. tp:

c arriage, orc go1 riyte ...

-208-

OMNITAS ALGQRI?,* FOR RATE DEt:'deINA11ON ANfl tOUIPu&4 y ALLOCATbO4 NotE 0?a ~flguto 4

U)1 LIST Cf CONMNDSe OATA ANO DIIONOSICS

LIST 2 TeIaaa~nCse rgao CiwEchblri 10 SooFrIEO aREAD 302o394

*Ste? 0e14? 0.0 .02110. 0. 0 74 0t 0 04192

411AUL T ,LY I S12 S

~jvUTiP IO .71591

* 'LY 39S 2110 too 19SA.'E1,2 4113 10~9

')11;E~ !,?21?13 1.333 1,2

l4P 90bN'243$4 t

l6n.tw12.2.3,4 20

V/4jAJVOIt~ ,2's O 10.1 ,4221DIVP .5 2410 2.2' 2

V 'Af.'t~vAle I.? a i c X

24/1 4alJr~ 4.? 3; : 1. 12 Z.

30,~~E .15 llO . 06371.-10 3.10- Z31 *S

ACT 12 lae I i .1

2.15 110 4,4

* -209- .

000"ItAl £LGOmll.'M tOA PATC DEWVMINATtON ANII E0UIPMNCt ALLOCATION P44C s

LIST Of COM"ANDS. DATA AND0 DIAGNOSTIC$

WIROVE lot 1110 504

4o/Awovc 3e 1110 ssA)/4"OVE 34s t'1 s.5s

461DIN07E S ?S 3 5tS 361 ? 37IS34 S1 T 0I 419 'o5 71 41 ? ) S b

17 5~735 if It ? T1 tdq ~ 7714 P Il 741 70 1103?44*710EPUTE? S 7065 78 66 766? 78S 6 69 78 ? 74 71 H0 ?Z TO ?3'u 14

SoVErko 1913.

mdove 1,35 6944 totalaDE5UC Itz3.&OPIS 3013 P1)3 *5) 31

DOfon 11 @

£A0VE 123413,9

£4uU ;eel) hI03 , 1,22£00 345 010 842S1.2

solvo 1' 113 116 1.2351%I

AIVIOE 1:111 ?al to3,

PgEMO is 4 .

£800 1.35 Otto lob5 ISM0&&Do I.&$ Otto loss 1231£aac lobs Si10 1%b5 1,321*100v 1#.9 81)01.6 lea 1A10 IE 3021il 1,29 3£500CI 1,10 ?Xl§ A,~ 1,51

1800C %*) 91 24 1.261 31AsCO 10Se 8110 3.2 101 33&A00 lo?) $110 1.21, 131

*400 1*271 $110 1,201 13SI4

1011A&*z We Alt 1*1 1 030

0iag AL00mM PC rATWE 0tERMINAIoIN AN4D IQgJ5pm('y ALL.OCATIONq ofG

LI1ST 0f COP'4&NOSO DATA AND 0 DIANOITICS&AOIa441011 01

401ATP&A41IOSE 16101 1116 lotSO/ I COEvTNI 49 0.20 010 014

$1,AIRANSOOSE 4,101 1310 1,6SSi#I'.C'I Sw? 3io1ooo .

RISC? haAX 10INjSERT 1 2 2 49INaSERT 49 73 3 49IE~ 49 1 4 4 491.41*? 4 it 2 a sot SEAT SO 11 3 3 50?ISERT So 3 4 450oINSERT 4 13 2 2 S1TmNSEDT St133 3 3 StI'NSE2. 51 1 4 4 StINER 2 1s 2 2 12PIc' 5SR 2 IS 3 3 St##,SEAT SZ 3 4 4 S2'SEAT 19 1? 2 1S3INSERT 20 1? 1 2 4lo;S!RT 19 IS 2 2.05INSERT 123 11 t 2 5?INSERT &0 S 1 2S

1NS4T 23 22 2 2 59

INSCAt 23 1? 2 2 60loScfIt 24 IT 2,2 021t4$EAt 14 11 2 2 04loiscat 23 i1 2 a 63

low$, at 20 21 2 2 06MotE~ 39 it 2 2 0?M'eat to 22 z 2 6S

Al"NEI 3? 3 21S3

'f Fo~m A1 02 4

W'~A. 01 6? 4*21I'NSENT S? 3? 3 St544II.C-eEENT *3 1 0 0 1

PCVro*m 03 44 it: FSTOOE 63 5T 37 425fRO~ 63 04 It

Amoy( ,9 3012 if102AMOVf 1,25 1014 3.100

OUIIL LS0P17"m FOR *ft OEIERMINAVION &NO EoUlpMIte? &L60CAIO'. PAGE( 140

6.35? OF Co"NO.SI Omt A-40 DIAGNOSTIC$

AMOVE t%4% 14 IOAFARASE 14?l *Ozpg4u'VE 1*9 toll 1.24&MOVE 1.101 10') 1.44

1,10) IO.I hlO 1,49

I Z 3 4 1 1 3 4 1 a3 341 a. 3 4 1 2 34 1 23 4123~I1e123SET 293st40 at

ski11 34

11#10 Woo 37810 900 7200 $0400 3700 9000 ?200 S0410 37RO0 9046 Ioftbiust36 000 4 9009 I'moO 25200 000 6040 30800 31400 90110 1200 10400

W1 0 0 M O 9600 0 0 14400 MOD0 0 0 P480 40 O0 00 43100 24600 0 0 '3200114400 0 0 '3200 20400 0 0 *3z00 mWoo 0 0 43?00 moh0 0 0l4*40 ?SOOI 0. a 432003 00. 0 10600 30444 06400 0 6000 400 32400 0 13200 42000 41000 0 16900 5"0040- 517 4%*00 0 14200 190*1000 0O 14200 S9404 '400 0 IvSot 43000 '6500. 0 1200 964boSit 33.as

31106 0 12000 4000 1000O 40 S~O 9400400 3600,100 23 11100 410000, 23 114A0 A04940. 13 1114ar, 1*40006, 23 11sAPO 140000. 23 117All) 300000. 13 114Ape 300000. 23 %isAmovE III)? 40164 1071

7SOiINlMUN 3? 61I.#AwOVE 1.20 04 1*32

U.'skayact 3p 3 *11gS 3'0e 840.

*1,~".11Lv1.44 101 2.'2 1,a,

01M ItAI ALGO~hbtt ON M WE OETERI"l'AT10N AND coutpp#ct? AILOCATION #P4t1h

V LIS? Of COMNNANOS. IT A140 AhDIACNOSTZCS

R PISUPIACItoi 11 1045 1044/ "MaOE JOGS II 4a.4

ft4i1CC~1'. 09l &0 010 94001W010041 3 36

2341 39

Sp.6ficMCEN7 92.5 JO.0 o 091/I4IlZPL 41 31 1

46/11,CREW!h7 9S 0.0 l.0 2.0prrp ? 94 10

A&IOYE ti6s 4011 1030ArSTOEE 76 lei 414' sl

1.20E fXM 72 77, PEpr ON 72 66

0QIvv( 333 's 1779.*31ICOEMEKY 99 Co* 010 6,0

PI'NFCAA 9? 99.3 10

Sa0 400 3 0 0 4 0 0 S 0 0 6, 9,0 7 0 6 0 0 9 0 a to0 0

03020 3 0 4 0 S 0 60 7 0 1 0 9 0 10

1 0.v4*&St :~g

1007/AOVIE 40 400 3*6? tot1 OI.5 ..jalACY 12 3 37

I 00o?/AMOCE )ill 301 1#23

300,9/sul7qACW S4 57 5?09MRC14 100 106.9

1 01A/ilI'EtDM S? 66 50 at 51 66 52 69 S3 7 0

10.11t 07 2.014nj/17/&Co 37 64 5? 99

ll'4po'cn 5? IV 0.0 ilog,5us "AcyTo7 72 733@q'Crusc 76hI uAmTC* 99 66 21 S6I)l3iNATC. 99 60 52 S3

93 610:1.S

0O4MITAS ALGORITHMf FOR R ATE OETERDAIN.ATION AND EQUIPMENT ALLOCATION AE o

4,4, LIST Of COM4MANOS, DAIAA.'4O OIAGNOSTICS

0 11?'2/"ACH S7 *66 S7 99112s3/14AICH S7 99 0.0 S711?.'I1FLE 71 0,011MA,'PLE 720.0112#6tSUBTRACI 70169 16lll0.SURTRACI 76 71 TI11208,PERFOR4 106 112113/IPHE 6? 40114/MAte" S7 66 ST 991lI/MA!C" 57 99 0.05?T116/lrtE 730It0116.1SURTR1CtTo7 69 7G116,?,54IBTRACT 76 73 731t?/PERFOP" 107 11211*/PERFORM 101 101.4116*11PERFRM~ 312,1 Ila's1191PEAtORK 113 117119*1/1LtE St 0.0MVFORN 110 19.1 30ERmASE 46G6 697075 76RMET 211l10.4AOI 1.65 2011 IM2

ftRFOM 119.4 119.5,1 ItRMMIUM 6S 66 56 67 59 68 60 69 61 70 98 96

1?024PNtE 67 lot120.3/0MATCi 96 9 6 65 9Ito ',MATCm 95 960. ItoG121/IfL-E 72 0.01221IPLE 76 0.0123/SUBTRACT 70"69 76125/SUOTmAV 76 72 72126/SUBTRACT 76 74 76

127/NAI it 99 212S/MAICk " 0 '1126.1/MATCq 59 68 0.0 65128.3/RESET 3012S.'/4"OVE 1,96 10o1 21.94l?6.7tmATtbI 97 96 21 S4

IP41/myc"97 96 52 53129!*41IME 67 1.01?9.9/MATCH 05 66 6S 9*129 4,6148TCH 65 990. co 5I130/lrtt 7404%0131l/SUBTRACT 70 69 T61331SWJaACT 16 74 74134/PERFORM 127 126.1

-214-

ONNITAS ALGOmlykAs 0 FDAt WECIEA.9z'dajOp 441 £rUPNtt LOCilp AO

.1) ~ ~ LST OF COaaas4OSta TA &N IAGN~OSTICS

0) 125/PEarpO 120 tlq.z1*PAON129.4 13% ..

'a &sun 02 123,"CAD p/. PAIL. V0OE5Of AD M/I a?UIE

*fA "I EST OP:?FI1'rAD 12

5.cuU EST oftNrAU SS.~uoiADJ RATE

MESAD 601 EST MARKET

WEAU W2 EST PROFITWWAD 3Z3/CV.rn EST PF?

14940 7PI'?5A9 r0 RsODE*4EA O/ 6, OUEHEAD ??/ EST MARKET84FAo I0/Ani E57 uM?

"EAO IP4.'CUP' EST Pr?ME& ND S hUih RATEp 0 PONT So S9 0 61 62 123 3

*R E ~5T 30P&RSuW S4 122PRINT 50 S1 S2 53 So 122 s5Of SC? 40iiERASE 66 6? to 69 -.0@7636iI O NL P N .17 1126 : 0.0

PAFORsa 131 13# 10IYO0VE lea? 401i 1071

13.,~v .e4 4011 1.24

zV1oSU473A~ ~l~1)9'"AIjnIm b4 6b ?? 67 76 68 79069 So 70

a9 13ffA C" 84 66 44 so

I 1AhiLE 72 0.0I.~I*l?OTACT 10 69 16

I"'/W4eTmAcT 76 73 7aI4/,.,e1oc? 10 72 72

14/NIC ?a6 004

tmAT~ 84 6215-9

LI 4O COIdOAT A i 0 99~lOTC

oTH 99 040 2 $699 0~900

/)69.3u3"dt ?70 96

390/PLEAC 73 0.0 7

10/pr~f 70 139 7*9

1067*/I7* A3 7M

KDs(/ItC Te 06.1 3POINT 771AC 7 T9 S 16$It 7*mSv 71. 9*16061REOPE 1 ke7 1,4'9l2 4I- /ESUO@ 7 13 ,8 ab 0

pPSj~?0%f 20 1* 49 1*4. 30/ 4'. fact 93 12' 2 x1I

99117?NV 70OS 79 OXI 1282143.5j'IdC12LC 16 144 1 .~1

petpar,0" ? q . t .o

I6 1.28 /U(4 3.32I ~ I2.329 118 19 150

I *VEPSC 95 1.3' 60# 4 1 S,13410SOT 2 5143124

1,2/0gar03. 1 1 9

t~gn " 3 01" 2I?9.l/vA20flN 300 10

I oa livy So 47 0 $1 70 123 74I~el/tl 20

'

t$W/A 'J 04 123

OMMMWWLZ.. 400

-216-

OUNNIT*S ALSORITIHN FOR RATE OCELoINATb0N ANO EOVIPNTs £LLFIC*11ON PAGE 181

40 LIST OF COOMANDO 0814 ArO CIAGRgOS11CS

146*3PRINjT SO S1 S2 53 S4 122 55

104faCAA$ ftb 40 8 6o0 I 16109faE0P(Iftm 13? 1314 to

1.b'AUD 17 '01 * 11

1~lI/P~f0E413".3 134.41 921PEWPO0CM 140 162.1 30

19~omI' T~?69,80 It 124 6241404( 111 4021 1.26&MOVE Is$ 4011 1.2?

A;0 E 1.100 IOXI 1,44VSI.8ACT 1.100 loll I.I,10 49

£S~dIRACT 19101 091llf 0ie49&MOVE 1.112 *01* 143lSET3036000 1200 4000 $2260 39400 10800 10800 4000 39190 10400 720u S0400 3960 10800Sri Is 3010000 boO3000 301000 IQOv 54000 3900 10800 l~beft SkOnO 3*06 9040 1200SET 28 3C50400 MOO3200 3000 00114CC 26400 7200 6000 34400 12600 3U09 2400 16BO0SFT 3)P0400 0 0 43100 2SA00 0 0 43200 Z1*00 0 0 43200 24614 'a n 432o0 RR06 0 a 43t00SC? 21 317400 0 0 43P00 26A00 0 0 *3200 900 0 ; 14400 1920C, 0 0 2*WOO 95.00 0 C 14408

40090 0 18200 S9400 $0*00 0 21008 6.800 *4640O 0 14000 61200 f-O00 0 ?.1613 46*00

S40 671416 64800 'J400 2160 o 0 4 0a 16200 59600 WOO0 0 600 21000

Sf7 33 as32400 0 13200 42000 1500v 0 S40i0 198007fwfLIup 372 193

AU4IOi~ ACt 1.10? l011 1.100 1#68.AtWaOaf 1. 01 tl 1'0 0614&MOVE tea9 6011 1.2?

AMOVE 1.1110 0X 10' ISE.3043200 36000 30000 660

0 33600 28000 26000 31200 1?004 900c 9100 Mon0 33600

Sf11Is3016300 Z600 36000 10400 14400 12000 1S600 12400 9000 900 1wo0 ?980 2efOOSri 27.3o0.. ,. .

19200 2000 I6009 1*600 12000 15000 20000 ?4000 2S04~ 366CC 360004800 36oSC? so 3048800Sri1311200 0 a 10880 4800 0 0 1204 2*00 00 3000 4900 0 0.7200 2400 *0 upi0240000aSET 2431

OM2A AL60*11eOMNFOR f LA E OCkEMNA11h AN* CQUIP'toofALLAIM #Ac 166

LIST AFE9b 4NSPDIAA4 OIAOOOSICS

o400 00 o20 i2oo c 0 @3000 40%00 q o vao20 0 100074200 0 7?000 82000 57600 0 $200 go o~ 21400 0 M~oo 22kq: 62600 0 S101) £49Q200 00 00 10Sir?I? 4S2OT13 05 1640300eboa14Pa 20 ?,0.. 4'00 000 21580. 0 2b'0 3008o .5700 0 S?Bo0 00000o ?2000 0 'o400 7500

OfFR4172 193

40k)1 CT19C 1.01L .0

aOvE 1614 40311 1 2;12.oo ?VE0 1.1 * 132c

S'30000330 9600 1 2~Q 60404300 12000 3900l4o1' 00130

SO7 31 1.44 00 0 0 7290 1200 0 0 l0000 2 soo 0 0 3100 2400 o a 3&00 *J'00 0 0 ?too, :~1?00 24,00 0 0 3000 2.00 0 0 300 2600 0 0 300 @. 0 300

SPO 500S20: 7920: 0 T2000 NZ000 21600: M 1 22 6 ::::1:: :,io P21104

SUMO 0 lb2doo 600$o*a*UCS202600140zA0241 70Sri 3385 O

2P.400 0 ?4000 70 60 Z40760PC~90om 27Z 293

&MOVE 1)52 409 124

&M£OVE 220 011 1.2?

306 oo !2(10 900 52200 12000 3000 2400 16-400 12000 2400 3000 14000,394 0 08ie)PR') oo00 2$200 0000 7200o 300cc1200 3000 2690 16400 ?4400 TZ0 MOO 34640.49,-19 301?000 2400 2600 1bd00 2524,0 000 *m00 3)60o 13200 3000 3400 16000'Sr? 31.,

* 28000 0 0 43200 0b00 0 0 l 4o00 0600 0 0 14400 Z010 0 c.13200 1?001 0?8P6SY21 3)00 00 a 4400 1102n 0 0 20goo 0600 o.0 14404) 11240 t * paO 060 ** 0 4

61;060 a 20200 594600 11000 0 5400 19600 15400 0 0043 P04PR $0400 0 ?1410j 041f005ff 1? as

-218-

0"I3?a&O hJ.6ORITNWP0 FR WE DCt~I'~N&TO Awn1 COULP4AP? ALLI2CAtION*t, 10?

IIsa CoOM*0sv CATA P.0 0tAGNU.T1CS

3p;0oo 0 13Z00 42000 1500O0 C 500 198qo 3?OoU 0 W3.OO 4O~OO 146flo 0 4400 19?00

36001oo~oo aoo 100 MO 100 oo 21600

-aM*)V9 1.10 1.0vI 1A.40 14&SUS'AAC 1 10611 1.110 1.46

aMdILV 1@1 011 1.2!..

lefrIVE 1.116 '4014 1.71

26400O 12.10 7204 36100 37dO 000 9000 ~ S2200 36600 ??q0 10800 S060 30000 1200JOAO 10000 040 720 4:0 33600 26000 4900 46oo 334-JO 12600 .000 2400:6O

SF1 21 31

30340004 000 366 0500 W 1200 46330 3 6000 001 20 S400 10$S 31 6

lQO 0 0800 '320o 0 0 4O 28800 43024A006?912004200

Sri~V 21C1 1.1

£V?PC 3.0e02 .05.

13 o40 mo ,'0 )2k0 16 000 20400 3" 0 100 20.iOD lO 32.00 13200 4OOset? 3' @6

6908" 16102 10400 o4o 30 80 6802801430C ~C 40 2059*ItTOC 27 1300 01 1.1O200 380 000.66ia~~ 4001 2

12000F Ise .000 .2

Sri@ 30

280 00eO 360 40 MO 1 8 600 1600 17200 1000 20A0 720 ne 72000 300. 24100Sri 24 333APO 2000 23800 $o 0##3200 3600 480 16O 14C '0 0 10 8 0 69 ro

SF0 21 30.1200C0 70280 9000 2800C' 10d00 300. 26040 40 1*tOr 18060 11000 C14CR0 1440005ff 33 30

sr., /

..............................-- 2..=0O a 0?Zoe 200 00 I.C.W. 200 .

-219-1

ONN 110 ALOWIII4 FOR NATE O)tL,"2IAIVIN AND [0011PENT AW4411(Wfl PACE 10

LIST of COOMANUSs DATA AMC DIAGNOSTI1CS

-HE?4 1,1300 313 1 .'

IsOV 1? 401 1.27

)0"0 3600 WO 2 000 ? 4000 10200 24000 36000 3baO0 14600 1240h IS400 S@0@Sri 14 34ROO *37,0 S%009 SO*01o 432ov ifilloo 3960 1440o 12000 14'00 )boot 30,04 ?RaceSri V 3020000 J9610 *J2qQ 36000 3fto10 4600 12000 4600 9600 13?oo 50400 43700 36000SE, 60 30,

7200 a a 10400 00 0 0 74ov 240 0 0 3600 7200 0 0 10000 7204 0 0 1060,0 2400 0Sri 23 31 .1 3.00 U00 0 0 1l0400 7200 0 0 10600 2400 6 0 3660 71go 0 0 10400

Sri .4 b .7?3d0 72t-00 746200 W400 0 000 $SZOQ p*000 0 *,o04 3900.0 93200 0 72600 47200

W1 3 65211f,00 '. tol00 22k000 4*4104 a T9.200 94000prUUnp- 17? IV]

AsVE 19102 3I113 1.4..Ide?2 lubi 1#10, 104

A J~A1,103 1611 1,111 1*9Av0VE lot %OKI1.2A..ovF l.12 b0~l 1.27

Sri 30MObO VOOCOg 00M03 3000 U40 '410 b2289 3'i~go 14400 9000 6220 1?400 2400

.Vf IA 3v3M0O lU00 M3CI) tP00 390 1 2 0 ?00 3000 3600 10040 39400 1000 1200 40400SET 29 30.12000 200 ?too 16"qi 3700 "0.0 WcOO S440C 20.00 7200 700 34000.sri 31TASOO a Q 43200 2104, q 0 43Zt0 ZtAno 0 0 43ZCC 9tCO 0 0 10400 1'400 0 0 2aFO0

90 0 0 14400 P6800 0 6 63200 9000 0 0 16660 t6040 0 0*3C00 19ZC0 0 0 2000A4f00 0 10.000 63200 65000 0 16400 %9400 484(-0 0 jV0 0 6360 Isw0 0 wo0 "0400

%T17 P%.0th 00" kip 13." 0ft3 19We 63000 J3600 0 14*00 4320

_____P0 1219

- 4

-220-

7V M

9whIlAs ALt.ORU1? FOR uAt DOCSIN~AI1N AND~ E(W'IPMEN4 LLKnATh1ON PAOC 109

LIST af CC1rn4APIDSo CAIL AND D1A'iNO3STIC

2 AwUfot .1.030 lul.44 4 0

S1.101 1491 1.106 1049

£MflVf 1411 *0XI .1

1P600 340 3000 1!40) 26400 7240 72o@ 36000 Atpoll 10400 10000 rloco 120.4f30 loSr? IS 3024eO lbftoo IMO 300) 3000 1?440 36050 7203 72001 5,4140 30..00 1oAcq 1"00 StOfto

206404 720 7200 3600) 3604n 7?v@ 1o00 54000 30000 7200 7700 60440"b SF? 31

0400 0 0 14400 39200 0 0 MO1000 0o 0 43200 '0640 ft 0 16400 401)0 8 0 14406

2440943200 2bdQ) 0 0 1#3200 19200 0 0 2kCD 2boeo 0 43 *3200 2'.0W 6 a 4370'3

15000 0 6000 20400 33600 0 1640 4320C 5040, 0 Z1000 6'441n IS400 a %60 1940tiSri 17 scIs000 0 t'0oo 20400 4JZ00 0 14%00 s7bao 50400 0 gl'~cp 0'k0o S0600 0 Z10 64600

01600 JOOO 120 0320 014403 97600

AMOV' 10 4014 .71

5'? 303'.e0O ?"'00 ?86*00 390(o '14400 12003 12000 Isbco 10'e00 140 4hoo 13200 24400

S!? 27 Jt'19POO eamcso ?4000 194noo 1929. 26.*01 '3200 )6000 ?bt-OA 3W)OQ 10000h 14400 144041

14004Sri 3140' 0030 . 30 ... .- ..- 01~0.i7200 0 0 I0m32 -~o -)a34o20 604 2070tp01*0niSri14311000J0 4.00 a0 7200 .600 Oco 0 iQ 2oc s 0 liCov 41906 0 0 ma0

6409 0 67%00 60403 46o) '0 2460 27600 2m6.00 0 ?4(14,1 M'V8 0' 0468000 C'6200

70201 0 72400 d20 d'Goo00 9 e 9000 4&I000, a #JUfla %04#)o 43200 0 3"ln 4S00SrT 33 Ili7,004 0 '-&0@ 7Moo $2600 0 S7600 64400

PrwfOS)4 M?2 193

#I~rsooolllI&

4.1%T111 ol11514

. ......

-221-

Os'NjtAS AL640111- 14 h AUE Offts-11I4A?1CJ(N ANDO (IUI11 £LI.flC4?Tfl 110~~)

6151 Of CO&WiANDSo DATA A'.U D0Ue'STICS

...NP~t 1410 tettil0103 1hc 1 tI~OPM203033)0.00IalO ~ J~ B~ 00 4

A .c 3OojO. 3b01§ lilt bou~03364 80 EM t02A0 9C

.Sr? 23 1*' 300 Me10 0 4 200 Uc. ) J 100P 464000 flHO 148'41 14t ?8004240)10

wt~oo mo 36')00 111( 000 j0u A6.40- 33o0 mo 2d ?m,& 00 Z~00'oo 3I9PC 24

Sri 31 A7204Q * u 24a04 4O a 00 0 ?2e04 30 Aofo jloo 0 U'ffo 7Zo.P a2f 019 ~44-1 TIO

Sri 2S 31-10' 36. 67600 700 7200 -6 - 16040 ROC0 Z 6)0t170SET 4cl. . 1.93

Sr 17 % z 32

?064 0 ?tot* 02oo A'J tPlO 9"60 3?00Ij. 4604 "'*00 IfiGo0 7200 0 400 Mtoo

Pruen" 10"@ 721 5.

A . o i Ji tel41649

-222-

0'ihlTl ALUIR.tflhlO R ATE Dft-huINATION Aafn I.;jvmloifN &LIACAIo' VAAL ItII

0 - LIST OF C('AMkUS. CAU A11 MA' 'tGNOSTICS,

am &MOuVE 101A. 40'1 1.27o3 &-'Ive lt1 4021b lilt

srT 1 3

400O 4j*'g'n?*00 !MC ?200 360&0o 1260n 30an 2'Usi 1..Af0 IZ490J 3001 2600 1.000"S'1 29 3-.IP0a~ 3000 3604 1000v W0?0 3040o Z01) 104n 210oo 000O 4150 13500srT 1I on' 0 ( 28000 lftOg 0 0 2040q 46ne0 a 1440. 19200 0 0 fab00 11200 0 0 24400i?21 31

45.0'f 0 16400 94041 (1.Q 14601; '6(0 0 0 14440 9640 1 0 140 1"e04 0 01 1446$

3itiv, 0 144ov '320 31200 0 MtOD 40400 lSAljft 0 0041 io4sAO 32.00o 0 132:0 4200

33A(. a 14400 43200 15000 0 b-0 19R00 150001) 44 1EO "4 0 0 9 6460 11000SPY 33 b%.

094f.10"' IT? 114)Alku010hL1 k*1G2 1011 1.109 1.4"At~t.Ijj.CT 1.1c.3 loll 1,110 1,.0

A.O(vt: loll? *014 1,71

vbo i-o vbt 13., %Oft 4%5so '3a62 40 4 14440 o 192*,n 2400s 360G S440

4Ui04 2..moo 3964C 36boo 140. 14400 11000D 366400 2d'~0ft 43!10 $6@0 164,00 14400

1,000o 15480 36C00 2Seep 2telCt. 39104t 24001 19200 12ona ?t,40 %(-"I 43POO ?SAOO

Sf1 4a S

%F7100

SIP7 31 0P40 0 63ft00 4500 0 0200 0 -.00 0 20 ?#o70 eZ01 6Cw

SOV 1.0 0' I.1 ILt.b?04L 2400 0 360 UA .0 ~0 10 AO0 70)70 0sr .1 4~333

oA'0'0w 12 193 .7

4-t -0 0114

-223-

(jd4 4 £LOOPP,.NM FORSAT 041 CL?"MATIO% £54') tUYIPMNT &LlfC&Tjn 0ae; lip

LM .List or coWNANdus, c&1b A%o oIwr.oIPcs

01 SET :oo l#4600" 14&00 4'04C 132..0 14400 Ro0ut 0 I" 170. i44co 144cf* 901 W3oo jAAfin 12401

Sri is 3e*0 Do0 !3oo 28*4fk 192oo tAm(oe 2#%son T4Doo 24ooo 31?oo 333oe 2e&66 lo~po

Set P,4 3co* ?*Ofl 43P00u 3bo@c adoton 3*ou 10*110 ll4too 1*0C *360ftn u'no 2&qo* 19p00 PA408

9r1 31V430 0 0, 31J00 2444 0 0 3&oQ Pub1 0 0 3600 2400 o 0 3(.00 4600 0 0 12qo wlo0 00Sit 24 317 100 4von 0 4 7200 ?tot 0 a 11hoO P400 0 0 3600 0400 0 0 1200L

PA40.1 .00 06444 _2400 . 0~ *o Zioo ic%200 ;6400 0 14300 17JI1 21100*0 ? 1O'0 30090it 1? 6%suc. 0 bteOC alovoo %240O0 f~O tioc O0 s 200 0 46011 ssp0 Um00 a %,btoq 754,0sri) 33 e

31?0I at -Ant& 32400 44000 0 03too swo0

*."iV ~.10 100?3Ob,14

Sri 30i2s100 241.0 3000o le l2boo 304,o 3-Ao lo0 l2*0.j 3000t 3'oo 1740C12c 266O~a0SFT is 3.)10.0 t*P.eo 37'i,A 90o.' ??a0 I04, 12040 24n 0 * bl 1sooo ?'00 tocoo *Noo6 33AfteSET 29 3v

ID-, 6OO 72C0 03J 20 24',9 )MI 17400 11*00 340 3^0O 1160

Sri 21 !9600 0 0 141) MOO0 0 0*240 IV00 o a 14400 90%Q'ga0 14*09 t9600 0 c jiw0

SFT ? 2) 79*0 0 C D t9*04, 15600 0 60u40 10401) 3000 *eo )f 34,. A koc o lu**n 9* 0 40Srt3 as

186005 0 S400 MOO0 15660 0 6004 11.0

AL~T 13s3 4.It) 1.4m

An s 3ls 0X a

"090 36800 36610 !20:* 10.04 "t~oo 39AD7 1000') 94J00 SPZ') )0goftf ilo .0 no Snag

-224-

u'ii11 A61,0 11 "Pe 040~ PAIL OClIAA1IMP4 A't. 101I00-ifty £LOCAIII% IIIJ

C- LIST sir COMUA)'.). fUAIA AIM 0Ih0O4CILS

oJ Cu I,C t0 U-21~ 1".!01 U~ 10000 Sf0 2e4UC M41 021 14000set 312Aoo c o opo .iio t~o e pum awn a o t epio to a n ibec.o ra.omA e 43poo

if121 31*. u i k0L 006 'Ifc 4 19io~ eo0i0 -'0 0o zoo-tw it4D 000 0oof I .fato o00

632Cr. 0 1404 P7000 JOapo 0 IC0o 3"A00 31200 9 .1 n 4to.q .oo 31~ 0 I?010 AAROO

t Uo 0 10OC- 61200 32e,00 0 13200 41000

*vrrci-1 172 193

PRO, 4 10.0 2910. 303voo I20i. 9600 12040 350 tl0oo 2014r 2400 36fo~ 324406St? t I t,91 )17 3s

9&00 Je~j 3'OC.o 19,d@ 26boo JbfCo Ao 1A OOo )40 Vp 1320n 11,800 144111 4610'.FT At )

IRV0Sri il41k0o 0 0 7200 a01U 0 0 3000 %Ava 6 0 ?for) 4600 a 6 7200 ?&a* e a 3604 %410 48 0Sri A' 3172M Z4,00 a 0 3600 4,140 0 0 U410 kt'00 0 4 -4010 k%00 a 0 3,500Sri -AsSh"*1 0 '%21O1 '10400 Z4,.iO 0 iljpo 25? 7 .oo S41 J) WC.1 6C00n S2'~f*' 0 %1M. SSP40

?Uh0A A 0 30coo 4.2A00 0 khvoo 55700 234.00 0 Ivg04 22AI10 %?"no8 0 6111111f ci%?afl0t1 43 10-P46006 0 P.flCI' MOO0 Z6400 6 r'too 27400Of)-8tP61 4 172 103A"6Cf 1-100) 1011 1*44

TALT Woo0 lol1 1.106 1466A1%%7AF1,101 1051 1.106 1.49

A-00, let .'."1 1.20%

42. 30two.i 25Ms 39to~ 1614o1 1'4tso. 124100 Ist00 33'Ai 2#0100 19740 26461UO 1f4lo0

1*40-1 1?iUCo ISP60,' S0oo 4320,, 34;tIo 460~4 43d0 30441 2f.'oo 3VA.gi 2401 19140

260,0,. .112,10 31ju0g 20048o 3600.1 00uQQ 12000 tt 0010 ig 1320*1 Mot3 P14,01 43"A

-225-

OWIN I A &LGOOII FCA PATE fl-IN~,1AVION Vat?) f.(OUIOML*.? ALCATIOnW P8 114

LIST Of COMPANUS, UAA &.to OIAokfISTJCS

?Poo 0 0 lose# t400 0 0 3010 -04O 0 1 7210 1401 0 0 3600 Me0o 'I IOAO 7240 0 ftSF A* 31 . ..-

)4to0) '40Oo 0 * 1200 ?100 a 0 POS'O 2.40 .) 0 .40(1 ?2501 0 1 50100

?Pool 4 b4.10 4 00 lp"00 0 'ob%@0 $0400 S26(00 Q 4'3j0 %%8040 It.0'b * P2I 3.00

a 1420J0 900 ?20 a .400 ?S$,04 44000 0 OiJM 1a~ft 12000 0 A440! IIACOaSri %3 ftm

)1e40 0 ~o PP 2200 70200 0 72.0Q 0200 4

AwnV.: 16142 lost 101

&£1;'otkAC?1 10,IJ 1X I.10 J.,09

as-.e. 1.11 *uI Jon

26'01 tn.4 6tfo 3%4104 13200 J690 2641 1644 oo 200 41 3A40 16004 3?0 O.00Sr.? *& 34sAM101 Uoio 30453 ?2J 7,!00 1%0-3 13100 3(43 p-j 4 0 1 0~ 4001 400 WO) 14410%FT Zv Jo-0P~o )'4,60 Me0 300:-i P6404) '200 l00') 361) E400 720 7?AO 30000 9-Sri 3114203 0 0 161100 "41 0 0 11"10 900 0 a 14.0Q 20,444 q 0 -1104 ?490 1 4320453? ' .111400 0 ii 4400 1900 0 0 944)li IV00 0 4 2010o 19SOo 0 0 ?nou 1.260 0. 0 2440osty Is

Ip. I3iou 4?.)00 11606 0 "040 18400 1143* 0 .031 P041% 4 d"Oe 4 14fl',0 fw33004? t 7 qv II

4320's 0 1140.) PlOO SO 4D 81 00i 2400~i 3o a 104-m 316 a 31k33 0 12)4o1 &A*00

3"40-- 0 13104) *'064 J£YOOO 0 14.00 43100 .

*3I43stq 17d 193IU~V~1.541I0 ~ 1.411

Ai 41s*A:t 10101 101 )11.1 1.6q

*)(V. lot1 401 121

3461 1.)00 10000 54 )Of. 1260) 3000 3500 11011' ?2213 4000 A061 1400.19 e )0'3400113? lb 3)1",0 16100 31401' vo.'0 *O0 $240 12,0 0200 540atal4 3F IOC 3630 r404) 1&0t30..

"T 9 303401 50.00o 10000 ujoa 12003 0400 J'.00 1100o 224, '00 6003 344005!? 4120800 0 0 4324000 oo 0 05*ij3 1,0260 a 0 Z-400 ,#6jQ d) a 14ava P4400 1 0 ',PO4

04Nia &I.C41i'-'4 F-a W~t Ew~f41~ *4 C'J-jIP4CN? ALLOCAI WE ts

Lis UOF C0044AMOS, CATA AND OIA'1'OSICS

101q400 a 0 16*90 4160 i, q 1440) Z8100 0 0 43po.) 900.a m 0 14404 11?00 o 0 P440'~

S54o0 0 1100 6440 1#,200 0 $620 2)000 31o t 12)0 #A~wf 1"0~0 ' 10 TWOSset 17 46444oo 0 IAOQQ WzOO 14#400 4 '04 *a9oo ISS130 0 0031 27,41i j 430 110 swoo

IA6E03 0 sooo 20102 31231 0 i1.000 40$00ero,'I. 11t 14) 14sc.h7 -AV I. z 1191 If'44 .4

£AIACV' 1,112 4014 lollSri 3341?at J3'30) 70,100 39ooA 3io 20oio 19e0) z6'oo 33004 MA(10 V-800 36*00 I2000

9t.00 2mo Isto0 14*30 12000 9^00 13ag03168#00 ?044~. 4320: sfo0- Paefit 24AOOSV? 21 Jo1Q200 ?"66lo 14400 U*joi 000 I32M 3%ooo 21000 36voo 4.in 36Coo 2(433o 24000Sr1 40 J030600set 31??to 4 0 1000 4300 30 7200 -'CO 0 '3 7200 2%,10 0 j 3600 Z443 0 0 MtU0720 0 0sty?24 31I"t40.3j0 0 0 ?70 Z60O0 a 0) 12)0 0 0 10400 124', 0 3 10400SI., 4s?PC0 0 0 ) IS60 Z0 *%130j S%!00 6400o 0 57,Oq i44',#1# 2%o00 0 ?169" vs200

?4ir 0 M140) M1000 Pozoc 0 ?zvoo oriloo 44000o 4o ) Swi 4n '001 a 1460 P200Sri 338 ab?P01 0 64900 t?54o00 o&0@ 0 b7605 60400

WOVE.~ 172 4113lt

&WAVE~:o~ 12000 15400 '3OE3,tf :::z

10It 20 slajo vpo11o ~o 3oi3,,a-I4jS OS11SFT 14-3

oec- 4304 o;* ZlaofjV201 C4QC3600 140. 1401 14toI'lo'.33670 4POSty It StPn~o 3-,0o t%4o Ioto 1404 1ZO 3300 M O t?(v 2ton 1460jj -Q ,%0

set so --

-227-

U

OMWITAA £LO0Iloo FOR RATE 0ETE0Al%ATtON AND CQUIPNLIJT ALLOCATION PAGE Ii.,

:0 LIST OF C0t4MAIJVSo CA14 A4O DIAGNSTiCS,

p Sr.T24 31340C0 00 0 0'7200 4800 0 0 1200 4419 00§ 7200 J40,o 0 0 390

P"00 aC 26410 30000 14000 0 41600 21200 06400 0 19291 96140 93ooo o 064 07200

520CO 0 48000 11200 MOO0 0 t26.00 30000 62400 0 $1600 64440 46000 0 43Mo 40400SVT 33 SISP4GQ 0 400101W200 d4000 0 21600 21200P7PFf)A" 1721113

£51 jbya~Cy 10102 101-1 14109 i~

AUOVF 104 40XI 1*26&£CIVE 1919 4051 le21

2152to boot 4400 33bo; Pf12o0 6000 7280 36goo 132(ot 3600, 3oo 17408 13700o 3690Srt I5 303646 1600 3200 2403 P400 10840 13200 360@ 3000 17400 12EI00 3000 3000 1140SF1 29 302A400 t2ib 6000 3440 12b00 3000 3.ies 16000 26400 7200 6000 MOO@0S" 3130200 100 Rose* 1920S 00 18800 960o0 00 14400 V600 a a 1440C 0600 a00 14406SF1 21 319#1, 09 14400 9"08 0 14400 10200 006 2000 9e00 4 4 14400 19200 0 0 239000SFT IS3AO8CO 0 1064 O 30 00 0 13100 42000 10200 0.61 26105 16600ac 0 1200 21600

144(0 0 4600 19200 16200 0 6006 21000 15600 0. 6000 20400 32431t 0 13200 429005'0 33 Is-14200 0 4000 21090 32400 0 132vo 42000

pe.s.1?? 193AWAQVI Igloo 1121 104*SI~b7Oa#CT: I,1ti 1011 1.104 21#11£.UIRtCT 1,193 j1.a3 1,111 1,69ANOVE Mt 421111.264manb 3934 ox1 o2APOVE 1,112 6014 1.71S9*T It13001-t 10900 7200 5041.1 36000 ?ROO 9000 S??oc 12000 2400 P2600 .3660 24000 4000WT I 130600o vfiOo 1264C 3VCj 2400 logito Met0 6000 7P0o SosoG 12609 eOOp 3400 19000WT " 30MO920 00 7200 36o13200 31. 00 3 74000 24'0 ?400 IWOf

SET 31.M669 0 0 43200 26#J03 0 0 *3200 96b04 .60 14400 1020$ 0 645600 9500 0 0 14400VT .21,3110200t 0'0 2801 900 $0 14400 19kO00 00 2A800 Y60O0 0 34401-9600 0 0 1449C

44000 0 13000 t1200 45090 0- 14,200 $9400 14400 0 -6800 19POO 34to 0 1WO0 3%600

-228-

C5 ,I.S1 fIF COMAhOSo OATA £'.O 0ACpuSTICS

02..r I1 ? bsO 15~~Ote C S*oo 30AbO )C*oo 6 aIM %?OO 16260 0 OW~i Ploto 3?'*.o 1 13?40 42600

It..'f.0 C 160C PN(.I: 1,406 C ,eov 3'ic",c-rurof-.. 172 19j

£A,9rulf 10102 IAI 1113

400VE 10? 411 1.26A-0. 1.4 *,hal 1.2?

MCI)q PtIGOO 1920(;- 26f*tI

5 0',C.o '3200 420(_ S4000 33601, 21 boO 24001 31poo 3440C 4

SPY 1 3024btou 437.10 5Su,r 120001 969&i 1440C 1000£ 1440p &14j~f 14&frl 114rCI 366.( 2;%406Sri 2? 3

4U3ro. b ~tO 11000 V6 100a t ' %c 140 Inca v2o , 0 1Gn~cP no@' 14o 94C'O 1!40f0

30'0' MO0 0 0 IWO0 Z400 0 C 300C 7400 0 C 3000 24Oo0 3P0

$1A t411fQQ b~pp0 0 170 3t00o 0%4C bt fOo 57t.0e o.5"'S o o"o ~o io

?'..0 740(o j!MrCo PIMC0 Q jb'(Q ]C000 74,?Cp 0 7efloo G~,0 ?'01 0 10t nI70,%r; t.3 rS

;'040 tMOP i 000 264CO 0 44O 276.00

OP.4

Sri40340

Z16)240 0C 30 p lo9CC 1 0 7200 ?24c.0 10o 0 432Q0 0 '2'4 tIC 2 *00 4 3p..,~f

Si 231200 0 c 360 00 1'.0 0 7200 0Z 1.1'0 .109 0 ,POC 30' ?at.( woA031074 p

471l- IZ7 * 0 0 fde7O e .; o 3611 *'.'o '0 0,74P 4 201 00

33200 0 V-%Or 37.00 1920.0 0 ?>Co .'2400 1464c 0 7'Pno ooCp ?igfi 6 Akof 754.9

-229-

fLf C0014.Ub, LAIA &.() 11461 CJSIICS *ZUe",' f. i2bUC 'C0 L 5!700

&*4ME 1.114 40#4 1.P7SC t Ic21444-AQO 10I% 41 33600 Pt'400 ?2I.t 6CC0 Men~ 3600c1PI 72A0 *0Q(73 3dc4qSri Is 30??esl b0440 12acC 30e.a 2%oo lb'"" 1i~ ?'00 24010 IbARO "40 400 ?2*1. 341)1%

Pi4(t- *IA3 1200~ 3 0C0) WOO0 IZC* ?Pta 5040t% .13206 16.0 3Ei-(' 1.t'a01 p4' (0 0 21a11L 11024a' 0 0 2.ro(t Zt..hfo 6 1. 4s200 ?a-gvo 0 f 432C.. IEsO.Or C f 144POSir1 11 31

94%', '. C fc( 0 t 0P1O 1'Wt.' C 0 *?"OhO ef-SP1 0 ( &32.1h.4&l0 'f 0 14441'Sq T '.S3121 t. 0 17000 Ofit*00 .J2440 0 POO1 62Mf 46ro.A 0 IWA ~ S0.ci IW4 te -98,0

SfT 13 $5

)&# MOOV 57b0 IAO 0 72ve 116000*-i J# C-' 1??1s -4

O

P41&cr '2CC fct 3.bcv 3J00 7240 Sacn 2 2i Mo iii io 0III."e ,1i? e' A.Sri is 3c4344c, 33#40o 37lo YCOj 0000 4~2o 36fu1 MfC 906A W2?10 gi'40 00 ; 3(-we@

7.~r 40r. 3C 19(tOJ 37000 Ver0 7?2,..fCo C1 S ' I VO3o .j 3 )3Sf1 31~i@* @c k*2tfo- it a*':.zo ac ott v r P100 *11A;. It c 0 ~~~.S"7 21 31pAPF&i f'C'73?.I 0 F@ 96C.0 0 0 V.400C e -Dlft Q ( 'j2A* I'3 ' .f

llp.oe0 20 *2*00 .5400 0 lobe0o3 $*'C3 '5000 0 .Ieei Ss4" 1off 2- 0 Cttt 1.4ti

Sfrc t7 I~PVPO *1200 -fPOP0 It-CG's.5t 0 3i)?40 l2to-itA t$~ep cc#) ?.wt

4if.4' it-POO S94cC IwoF 0 mo~ 21309vfW(.1.- p 19

'10OI joft2 ell j,14

-230-

We 011 1#27A.U1.4 10112 4614 loll

SfT 1& 30VIA4 r ),?Cc ?4,40 14000 1209.0 021'Cv tibe(' 4i200 30floA 3fPaR 0A Lmkf&.f'

14, ZO,)%Q(, 280o0 t'CtV lqeoo 20%ft 3.1000 jej"eo ?414% ia' 3 'eoo)tfjj ~0(.Ofr 31 35437ID1 $4000

Mu to 0 14401) 7200 j 0 1I 26000 0 a 300f ,o 0e 0 72c.C . o ' 3F'4 ?rc 0 0sF? pa 311i1447 C P , 6 10 W00 VAOC 0 E IhO &POO0 C 7200 7200 0 IVIIJA'

91w. 0 46400 S7eoc 1440c 4 50100 Af4o0 ?16&10 0 M~e, p;PF~c irt! f, As,44f. %S2440Art 17 64

ph fip , 000 fl1?00 0 ?2100 A?r00 ?"POO 0 ?been~ 3ce'.r4 41tA C £370' 9!00Sri 7) 14% *

S~j 4?nPCO OCLC.0 7%21.O0 7 2M0 MOOCarba.sh' 372 If,

1*100 161W1 t ela CI.511

SFI .It1*7M!200 11 C eti 1 9bC~ 0 06 @ . 700 0IO20.0 114~bo 63o *eFC ?.1 ?'

Poto 144"m 16cr 0 100 ?osto 2,0 C JO 00 C1fl C 6wJ 7700I t?.2C ;I~ 260C013

770CC V-404.0 12000 o0C 9500 0-o P4OC2000 wiioC 1020 ;?t: '~ 164W. C*ti 12c.' 49'

ear? ?I?778C a c )*og 21000 0 3C c k0 1oc*0 '0 F(406 0 424 '.'C 10f 0 0 M&T1 70C 0Sf 1 '4310

96,601k' 172 Is, ST011