Lower Airline Costs per Passenger Are Possible in the United States and Could Result in Lower Fares

Published by the Government Accountability Office on 1977-02-18.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                                                        u. .,            cJ~- 7 7 a$5
                                                                          a--/~ -‘4,7
                  ’ REPORT TO Tk7E’CONGRESS


.-                 BY THE             COMPTROLLER   GENERAL
 Y,                OF THE             UNITED STATES

                    Lower Airline Costs Per
                    Passenger Are Possible In
                    The United States And Could
                    Result In Lower Fares
                    Civil Aeronautics           Board

                    Airlines in the United States regulated by the
                    Civil Aeronautics   Board could have operated
                    at a lower total cost per passenger than they
                    did from 1969 to 1974, and passenger fares
                    could have been lower as a result. GAO esti-
                    mates that travelers could have saved over a
                    billion dollars annually   if regulated airlines
                    had ooerated with characteristics    of less reau-
                    lated airlines.

                    Accordingly,    GAO    recommends      that   two
                    things be done:

                         --The Civil Aeronautics        Board should
                           work toward improving airline efficien-
                           cy under existing legislation.

                         --The Congress should provide the Civil
                           Aeronautics     Board legislative guidance
                           defining current national objectives for
                           air transportation    and the extent to
                           which increased competition       should be
                           used to achieve those objectives.

                   COMPTROLLER     GENERAL       OF   THE      UNITED   STATES
                                 WASHINGTON,      D.C.      20548


To the President of the Senate and the
Speaker of the House of Representatives
      This is our analysis,       update, and extension    of
Dr.  Theodore  E.  Keeler's   study   entitled  "Airline   Regulation
and Market Performance."        It was requested by the Chairman,
Subcommittee on Administrative         Practice and Procedure,     Senate        Z "<
Committee on the Judiciary,        and is being sent to the Congress
because of its interest     in the effectiveness       of economic
regulation   of air transportation.
       We are sending copies of this report to the Director,
Office   of Management and Budget, and the Chairman, Civil
Aeronautics   Board.

                                               Comptroller  General
                                               of the United States
COMPTROLLERGENERAL'S                                           LOWERAIRLINE COSTS PER
REPORT TO THE CONGRESS                                         PASSENGERARE P3SSIBLE
                                                               IN THF UNITED STATES AND
                                                               COULD RESULT IW LOWZRFARES
                  The Chairman, Senate Subcommittee on Adminis-
                  trative  Practice   and Procedure,   asked GAO to
                  review a study comparing airline       fares regu-
                  lated by the Civil    Aeronautics   Board in 1969
                  with estimates    of fares the airline     industry
                  might have charged had it not been regulated.
                  The study was published      in 1972 by
                  Dr. Theodore E. Reeler, Professor       of Econom-
                  ics, University    of California   (Berkeley)  and
                  is referred    to in this digest as the Keeler
                  The Keeler study estimated     that unregulated
                  airline   fares were lower than the Roard-
                  regulated    fares, and Dr. Keeler attributed
                  this to lack of price competition     and too
                  much service competition--primarily      in the
                  form of frequent    flights.
                  GAO was asked to
                  --determine    the validity                  of the Keeler
                     study results,
                   --extend     the results             to an annual       national
                  --update  results                 to cover    the 6 years     from
                     1969 to 1974,
                   --identify            the critical        assnmptions      in the
                       study,          and
                   --determine   the influence                  each assumption        has
                      on the study results.
                   The Keeler study attempted    to determine the
                   effects   of domestic airline   regulation.     It
                   estimated   what fares would have been on a
                   number of busy routes if the airlines       were
                   less regulated   and compared those fares with

Jear Sheet         Upon    removal,    the report        i                               CED-77-34
cover   date   ‘should    be noted    hereon.
fares actually     cnarged by the regulated
trunk airlines.      GAi) found Dr. Keeler's        as-
sumptions and metnods generally           valid and
with some modifications,         used them to esti-
mate the annual effect         a less regulated     air-
line industry     might nave had in the long run
on air travelers,      airline     fares,  and airline
revenues from 1969 to 1974.
80th SACI and Dr. Keeler               assumed a less        reg-
ulated airline   industry              would allow

--new and existing      airlines               to more freely
   enter or leave particular                   markets (flights
   oetween two cities);
--airlines     to raise or lower fares                 largely
   at their    own discretion;  and,
--easier     creation      and certification            of new
During the 1969-74 period covered by G%0's
review the regulated  domestic trunk airlines,
representing  nearly 90 percent of all domestic
passenger services,  included:
American Airlines                      Delta Airlines
Eastern Airlines                       National   Airlines
'Irans World Airlines                  Northeast   Airlines I-/
United Airlines                        Northwest Airlines
Braniff Airways                        ?an American World
Continential      Airlines             Mestern Airlines
GAO's analysis          showed that       for     1969-74:
--Less regulated           airlines      would probably have
   charged lower         first-class       and coach fares.
--Nationally,     actual   fares exceeded esti-
   mated    farzs by an annual average of 22 to
   52 percent;    in dollar    terms, if airlines
   had been less regulated        during 1969-74,

-l/Northeast    merged with            Delta     on August    1,

                 passengers would have saved on the order
                 of $1.4 billion to $1.8 --
                                         billion,  annually.
               --To the extent lower fares would have in-
                  duced increased air travel,  the savings
                  would have been even higher.    (See p. 12.)
               --While    passengers on the average would have
                  paid lower fares,    they would have been re-
                  quired to give up certain       conveniences  the
                  regulated   airlines   no-w provide.    For ex-
                  ample, some flights     would have been more
                  crowded, and fewer flights       might have been
                  provided on some routes.        (See p. 14.)
               Because GAO's study was based on assumptions
               about the characteristics         that would occur
               in a less regulated      airline     industry    over the
               long run, the estimates        of savings should not
               be considered   exact, but rather          an indication
               of the magnitude of annual savings possible
               after  a period of adjustment.            The report
               also shows the individual         effects     of the most
               important   assumptions.        (See ch. 3.)
               The 5oard has recognized         that inefficiency
               exists    in domestic trunk airline       operations
               and has attempted       to reduce it by applying
               more rigorous      standards   in its rate-setting
               proceedings.       The magnitude of the difference
               between fares charged by the regulated             trunk
               airlines    and GAO's estimates       of what less
               regulated    airlines     would have charged appears
               to have gone down, but it remains large.
               The Board has concluded that the present form
               of economic regulation   of the airline    indus-
               try no longer serves the public interest        and
               that increased competition    would result    in a
               better  match between the kinds of airline
               service desired by travelers    and available
               from the airlines.
               The Board believes    that under its existing
               legislation  it could bring about some changes
               which would result    in increased comoetition,
               but that such a fundamental      change in the
               approach to regulating     air transportation
               should be legislated     by the Congress.

Tear   Sheet                              iii
(;A!2     study offers
        ’ 5               reliaole     evidence     that
airlines      could have operated        profitably      at a
lower cost per passenger           from 1369 to 1374,
resulting      in lower   fares    and therefore       sub-
stantial      savings   to travelers.         These results
could     have been achieved       mainly     by

--putting      more   seats       on each         aircraft;

--filling      more   of    the        seats    available          on each

--increasing      the      average         annual      use of        air-
    craft;   and,

--using     some of     the   more         efficient          aircraft

The airlines      could have charged    lower    fares
and, at the same time,      maintained     average
annual    rates   of return on investment       com-
parable     to those of the entire     corporate

Although     it finds      the arguments     for greater
reliance     on a more competitive         market    oer-
suasive,     GAO believes      its  study does not
answer a numoer of questions             about what
might    happen if the form of airline            regula-
tion    were chanqed or if regulation            were
aoandoned      completely.

GAO recommends that          the Board continue         to
work toward    improving         airline    efficiency     un-
der its existing        legislation        by emphasizing
the factors    identified          in this    report   and by
increasing   its    reliance         on competition     to
determine   service       and prices.

G40 also recommends            that as part of its cur-
rent reexamination           of the need for economic
regulation      of the airline          industry     the Con-
gress should        orovide      the Board legislative
guidance     defining       current    national      objec-
tives    for air transportation             and the extent
to which increased           competition       should    be used
to achieve      those objectives.

The board      and the Congress     should  allow     for
reasonable      transition  periods     to avoid    undue
disruption      of the air transportation        system.

               GAO solicited    comments on this report         from a
               number of interested       Federal agencies and ex-
               perts.     In general,   they agreed with the
               premise that increased competition            would re-
               sult in more efficient       airline    operations,    @ut
               suggested a number of ways in which the analysis
               could be improved.       GAG modified      its analysis   to
               incorporate    suggestions    where they would im-
               prove the study results.           (See ch. 4 and app.

Tear   Sheet
DIGEST                                                                  i
                   CAB-regulated      airlines
                   Intrastate     air markets
                   Airline    safety
            LARGE SAVINGS TO PASSENGERS                                 7
              Assumptions and methods                                   8
              Dr. Keeler's    study method                              8
              Our modifications      of Cr. Meeler's
                 study                                                  9
              Estimated savings to nassengers                          13
              More people would travel      if air fares
                 were lower                                            12
                    Possible effects    to passengers                  14
            STUDY RESULTS                                              15
               Load factor                                             15
              Aircraft    seating                                      16
              Aircraft    utilization                                  20
              Most efficient        aircraft                           21
              Rate of return on capital                                23
               Fare elasticity       of demand                         2G

   4      AGENCY AHD EXPERT CObWlENTS                                  28
              Load factors                                             29
                  Most efficient          aircraft                     30
                  Aircraft      utilization                            31
                  Seating densities                                    31
                  Rate of return on capital                            32
                  Fare elasticity          of demand                   33
                  'Jse of CAB's cost model                             33
                  Hawaii-Alaska         markets                        34
                  Published versus actual fares                        34
                  Net welfare         to travelers                     35
                  4irline     system as a network                      35
                  Combination of fares and trip             lengths    36
                  Stewardess expense                                   36
                  Passenger densities                                  36
                  CAB has attempted            to encourage airline
                     efficiency                                        37
CHAPTER                                                        Page
                                                               -',--   t
                Summary                                         38
                Conclusions                                     39
                Recommendations     to CAB                      40
                Recommendations     to the Congress             40

        I      Details of assumptions used in
                 Dr. Keeler's Airline  Regulation
                 and Market Performance Study                   41

   II          Details of the methods used in
                 Dr. Keeler's Airline  Regulation
                 and Market Performance Study                   42

 III           Details of our modifications     made to
                 Dr. Keeler's Airline    Regulation
                 and Narket Performance Study                   43
   IV          Estimated and actual airline          fares
                 for 110 city-pair  markets        in 1969      48

        v      Estimated and actual airline          fares
                 for 110 city-pair  markets        in 1970      56

   vi          Estimated and actual airline          fares
                 for 110 city-pair  markets        in 1971      64
 VII           Estimated and actual airline          fares
                 for 110 city-pair  markets        in 1972      72

VIII           Estimated and actual airline           fares
                 for 110 city-pair  markets          in 1973    80
   Iii         Estimated and actual airline            fares
                 for 110 city-pair  markets          in 1974    88
        x      Impact of various load      factors     on
                 passenger savings                              96
   XI          impact of aircraft     seating    on pas-
                 senger savings                                101
 XII           Impact of annual aircraft        utilization
                 on passenger savings                          106
XIII           Impact of "most-efficient-aircraft"
                 assumption on passenger savings               111
‘I   APPENDIX                                                                 Page
       XIV              Impact of various rates-of            return   on
                          passenger savings                                   116
           xv           Impact of fare-elasticity-of-demand-for-
                          air travel  on passenger savings                    121
       XVI          Statements of qualifications   of experts
                      who commented on this report                            126
      XVII          Comments by Federal            agencies     and experts   136

     CAB        Civil      Aeronautics     Board
     DPFI       Domestic       Passenger    Fare Investigation
     GAO        General       Accounting    Office
     IRS        Internal       Revenue Service
     RPM        revenue       passenger    miles
                                  ------  1
      The Chairman, Subcommittee on Administrative      Practice
and Procedure,   Senate Committee on the Judiciary,    asked
us to review a 1972 study entitled    "Airline  Regulation     and
Market Performance."    The study, authored by Dr. 'Theodore E.
     --compared fares regulated    by the Civil      Aeronautics
        Board (CAB) with estimated    fares computed on the
        basis of certain  assumptions    (discussed    in detail
        in ch. 2) about the way the airline       industry   would
        operate if it were not regulated,
      --found   the estimates      of the deregulated   airline   fares
         were significantly      lower, and
      --attributed     the higher fares in the CAB-regulated
         industry   to lack of price competition        and an
         excess of service-quality     competition--primarily
         in the form of more frequent     flights--fostered    by
         CAB regulation.
        We were asked to (1) determine the validity       of the study
results,     (2) extend the results   to annual national    estimates
for the years 1969-74, (3) identify       the critical    assumptions
in the study, and (4) determine       the influence    each assumption
has on the study results.        As part of our review, we con-
sidered comments and analyses of the study and, where appro-
priate,    altered   assumptions and methods to eliminate      problems.
      CAB is a five-member   board created by the 1938 Civil
Aeronautics   Act.   This act subjected    U.S. interstate    air-
lines to CYB's direct    economic regulation.       CAB's powers
and duties were reiterated    by the Federal Aviation       Act of
1958, which directs    CAB to consider the following,      among
other things,   as being in the public     interest   and in ac-
cordance with the public convenience and necessity.
      1.   Encouraging and developing     a U.S. air transporta-
           tion system properly    adapted to the present and
           future   needs of the foreign   and domestic commerce
           of the United States,    Postal Service,   and national
        2.    Regulating  air transportation       to recognize and
              preserve the inherent    advantages,       assure the
              highest degree of safety,       and foster    sound
              economic conditions   and to improve the relations
              between, and coordinate      transportation      by, air
        3.    Promoting adequate, economical,           and efficient
              service    by air carriers     at reasonable charges
              without    unjust discriminations,        undue preferences
              or advantages,    or unfair     destructive    competitive
        4.    Encouraging competition       to assure the sound develop-
              ment of an air-transportation       system properly    adapted
              to the needs of the foreign       and domestic commerce of
              the United States,    Postal Service,     and national   de-

        5.    Promoting       air    commerce safety.
        6.    Promoting,          encouraging,       and developing   civil
        CAB carries     out these duties by among other things,
controlling      competition      in the airline    industry.         To pro-
vide service       between two points,      an interstate        airline
must have a specific         grant of authority       from CAB to serve
each point anc!i not be prohibited          from providing        direct
service between the points.             This authority,       or certifi-
cate, requires       the airline     to provide a minimum level of
service     that can be discontinued        only with CA3 approval.
CAB determines       how many airlines      can operate between two
points    (often referred       to as a route or as a city-pair
market) and thereby determines            whether there will be com-
petition.       Similarly,    CAB must approve the rates an inter-
state airline       charges between each of its city-pair
markets.       CAB's regulation      of fare increases and decreases
has generally       resulted    in the interstate       airlines      charging
essentially      the same fares on the routes -where they compete.
--------------__--__-a airlines
        In 1974 there were 11 trunk airlines        and 8 local
service airlines.         From 1938 to 1974, CAB-regulated      air-
lines     (including   domestic and international     air service)
increased overall       operating   revenues from slightly     under
$76 million       to nearly $15 oillion.     During the same
period,      revenues from the domestic operations       of the
domestic trunk airlines       grew from $56 million        to nearly
$10 billion.       0ver the same period,    travel     on the CAB-
regulated    airlines    grew from 755 million       revenue passenger-
miles (RPM) in 1939 to nearly 230 times that much by 1974,
174 billion     FGM. That portion     of the travel       produced by
the domestic operations       of the  domestic     trunk    airlines  grew
from 430 million       RPM in 1939 to 121 billion        RI?3 in 1974.
     During the 1969-74                                                    period covered by this review,                 the
domestic trunk airlines                                                    as classified by CAB included:
        American Airlines                                                              Delta Airlines
        Eastern Airlines                                                               Xational     Airlines
        Trans G$orld Airlines                                                          Northeast     Airlinesg/
        'Jnited Airlines                                                               :;\iorthwest Airlines
        Braniff   4irways                                                              Pan American ;vdorld Airways
        Continental    Airlines                                                        Western 4irlines

     As shown below, scheduled domestic operations          of the
domestic trunk airlines    accounted for 58 to 90 percent
of domestic passenger operations       during 1969-74--both    in
terms of overall  capacity    (available    seat-miles)  and capac-
ity used (revenue passenger-miles).
                                                                                    Scheduled domestic operations
                Total domestic                                                      of domestic trunk airlines                  as
          passenger operations                                                      a percentage of total                 domestic
          of U.S. certificated                                                            passenger operations               by
          -----------m--m-------                                                              certificated
                                                                                    --------------------------w---- carriers
                                                                  Revenue                                             Revenue
            Available                                            passenger-             Available                    passenger-
----        seat-miles
            ---------                                              miles
                                                                   -----                seat-miles
                                                                                        ----------                     miles

1969                 212.7                                               108.5               89.4                         88.1
1970                 218.0                                               108.4               89.0                         38.4
1971                 225.8                                               109.8               89.7                         89.2
1972                 231.4                                               121.8               89.6                         88.9
1973                 250.0                                               130.4               88.9                         88.6
1974                 239.0                                               133.7               88.4                         88.1

l/Northeast                   merged with                                  Delta   on 5/l/72.

Intrastate          air
----------w----------w    markets

       A few airlines           operate    only within      a single        State    and,
therefore,       are not regulated           by CAB.     These include           Pacific
Southwest      Airways        and 4ir California        operating        in California
and Southwest         Airlines      operating      in Texas.        All three of
these airlines         compete with CAD- regulated              airlines       using the
same or s ilmilar         large   aircraft      over at least         part of their
route    system and have often             offered    lower fares          and more
frequent      service       than the CA%-regulated          airlines.

          In his study,           Er. Keeler        assumed that because            the
California         intrastate          airlines       were not regulated          by CA%
and because their              State       regulatory       agencies      were less re-
strictive        than CAB in some periods,                   some of their        experience
should       provide      a good approximation               of what airlines           would
do in a less regulated                   environment.        l/    He therefore         used
Pacific       Southwest        Airways’         operating-experience           during      the
early       1960s to adjust            the operating         and cost experience             of
the trunk        airlines        to estimate          what unregulated         fares     would
be in various           city-pair          markets.       For each of the 30 markets
he examined,           Dr. tteeler         estimated      the fare the airlines
would have to charge                 to cover all         costs    per flight,        including
a reasonable           pretax      rate of return          on capital.         A reasonable
rate of return            is defined          to be the longrun           average    rate
earned by all           corporations.              Dr. Keeler’s       estimated      fares
were lower         than the airlines’               published      fares,     and he attrib-
uted the difference                to CAB’s regulation             of competition.             (A
more detailed           description           of the study method can be found
in app. II.)

------------   safety
        Airline      safety     is the Federal           Aviation      Administration’s
responsibility.             The Aviation          Administration        maintains       the
airways       and certifies         aircraft,       pilots,       mechanics,     and air-
ports,      whether      or not they are subject                to CA% regulation.
Empirical        studies     indicate         economic      regulation     has little
effect      on airline       safety      and show

l/Dr e Keeler’s       report       was completed     in 1972 shortly  after
   Southwest    Airlines         began operations      in the Texas intra-
   state   market.       He felt      he was not able to consider       the
   ‘Iexas intrastate         airline    experience     with the same con-
   fidence    he had for Pacific           Southwest     Airways.
no correlation     between       fatality     rates     and profits.       330th
the Air Transport        Association      --a trade      organization      repre-
senting     the domestic     airline      industry--and        the Department      of
Transportation     have agreed that           the problem        of safety    argues
neither     for nor against        economic     regulation       of the airlines.

         Dr. Keeler          set out to determine                 the effects           of domestic
airline       regulation         by estimating             what fares          would be if the
airlines        were unregulated             and comparing              those fares         with
fares      actually        charged      by the regulated                airlines.          Our    pur-
pose was to determine                 whether        his assumptions                and methods
were valid          and, if so, extend               them to estimate                the annual
effect       a less regulated            airline         industry         might      have had on
air traveler            savings,      air travel,           airline         fares,      and air-
line      revenue       from 1969-74.            In Dr. Yeeler’s                study and this
adaptation          a less regulated             airline        industry          was represented
by modifying            annual    trunk      airline        cost and operating               charac-
teristics         to incorporate           some of the most cost-influential
airline       characteristics            as thev might              have evolved           over the
long run in a less regulated                       environment.               The portion         of
the airline           industry      addressed          was the scheduled                domestic
passenger         operations        of the domestic               trunk       airlines,       repre-
senting       nearly       90 percent        of all        domestic         passenger       opera-
tions.        he assumed a less regulated                       airline         industry
would have included                (1) unhindered             entry       into      or exit     from
the industry,             (2) freedom        for airlines             to choose the mar-
kets they wished              to serve and the manner in which they
provided        the service          (a market         is a route           between city-
pairs),       and (3) freedom            for airlines             to establish            what-
ever fares          they wished.

        As requested,       we validated,        modified,        extended,     and
updated      Dr. Reeler’s      study.       In particular,         we repeated
Dr. Keeler’s       study    to determine        the validity         of its    results.
Using Dr. Reeler’s          input     data and methods          for estimatiny
costs,      we found the same results             Dr. Keeler       reported.        How-
ever,     in the process       of reconstructing           his cost model,         we
encountered       some statistical          techniques      that     suggested
modification       or replacement         of a portion        of the model.          4
Inore detailed       explanation       of the study’s         assumptions        and
methods and our modification                is contained        in appendixes        I,
II,    and   III.

      We asked CA%, the Department     of Transportation,        the
Department   of Commerce,  and various   experts,     including
Dr. Keeler,    to comment on a draft   of this    study.      Where
the comments         were valid     we modified             assumptions        and methods.
The comments         and modifications       are          discussed     in     detail  in
chapter 4.

         As requested        by the Subcommittee,              we extended
Dr. Aeeler’s         study     to include       annual      estimates       of a number
of national-level            and market-level            characteristics           of the
domestic       scheduled       operations       of the domestic           trunk      air-
lines     for 1969-74        (see ch. 2).           These characteristics                in-
elude:       fares     in city-nair        markets       served by less regulated
airlines,        industry-wide         savings      to passengers,          average        by
which actual         fares     exceeded      fares    for less regulated              air-
lines     operating       at lower costs         per passenger,           increases          in
domestic       air travel        that might       result     from the generally
lower fares,         and airline        revenues      resulting        from the esti-
mated fares         and increased         air travel.          The most important
assumptions         were identified          and the individual             effect      of
each was calculated,                (See ch. 3.)

                                     CHAPTER 2
                 -----      AIRLIlilE EFFICIENCY COULD lvIEAN
                           LARGE SAVINGS TO PASSENGERS
        In his study "Airline          Regulation      and Market Performance,"
Dr. Keeler computed the difference                between published            1969
air fares and his estimated              unregulated      (cost-based)         fares--
defined as fares which might have existed                    if airlines        were
not regulated.        His basis for estimating            fares for unregulated
airlines    was the 1967-69 cost experience                 of CAB-regulated
trunk airlines       modified       to incorporate       (1) efficiencies
achieved in the 1960s by the less-regulated                      California       intra-
state airlines,         particularly      Pacific     Southwest Airways,            (2)
some efficiencies         demonstrated       by the trunk airlines,              and (3)
some of the longrun characteristics                 of a fully       competitive
industry.       The study estimated          nonstop unregulated            (cost-
based) fares for 30 high-density               routes served by C95-
regulated     airlines.         Dr. Keeler cornoared these fares with
published     fares and found that published                air fares for these
routes ranged from 20 to 95 percent higher than those which
might have existed           if the routes were flown by airlines                   with
certain    efficiencies         and other properties          Dr. Keeler believed
characterized       a fully      competitive     airline      industry      in the
long run.
       Our review of the study disclosed         instances where either
its methods or the assumptions could be improved.                For exam-
pie,   the  analysis   compared  estimated    fares   to  1969   published
fares for 30 high-density       routes whereas it could have been
extended to more routes of other densities            over a longer
time.     Xe modified   the study to include what we believe are
more reasonable approximations        of airline     experience.       These
modifications     made some changes in the differences           between
the two sets of fares;      however, they did not change the over-
all findings     of the study. Our results       showed that from 1969
to 1974:
       --The annual industry-wide     average by which
          actual fares exceeded estimated       fares was
          22 to 52 percent;  in dollar    terms, if pas-
          sengers in the 1969-74 period paid fares
          based on the assumed airline      characteristics,
          they would have saved from $1.4 billion            to
          $1.8 billion.
       --Lower fares would have increased                 air   travel
          and total traveler savings.

       --While   passenyers       would pay lower fares,    they
          might  be required       to give up certain    amenities
          CAB-regulated     airlines     now provide.

        Dr. Keeler        estimated      what he termed “unregulated”                or
“cost-based”         air fares       in 30 high-density          air markets        and
compared       them to 1969 published              fares   to establish       the
effects     of CAB regulation.             We modified        the estimating
process      to remove some technical               problems,      applied    it to
a more representative              cross-section         of the national        air-
line    system,      estimated      national-level          impacts    of esti-
mated fares        for city-pair         markets,      and updated       them to
include     the years 1969-74.             Our results        are estimates

       --Generally        lower   first-class                       and coach-class    fares
          in city-pair        markets      served                   by a less regulated
          airline      industry.

       --Industry-wide                savings          to     domestic         air travelers
           from airlines              operating              at lower         costs  per

       --Increases            in domestic              air      travel  that           might          re-
           sult  from         the generally                  lower fares.

       --Domestic           airline       revenues   resulting                     from       the
          estimated           fares      and increased      air                 travel.

--~___I-__-~           STUDY METHOD

      Dr. Keeler            assumed that  in the                         long run      the removal
of CAB regulatory             powers over airline                           industry      entry,
exit,  and fares            would result  in

       --a     seat     occupancy          rate    of         about         60 percent,

       --increased     seating     densities                           incorporating         all-
           coach seating     configurations                             on all     aircraft,

       --annual          aircraft        utilization                at     average          1968
          trunk        airline        rates,

       --use      of    some of        the      most         efficient          aircraft,

       --a     pretax       return-on-capital                     of      12 percent,          and

       --use      of    aircraft         only     over          their        design         ranges.

        Using cost data available       for the industry,   Dr. Keeler
computed the effect       of applying these assumptions to CAB-
regulated-airline       operating   costs over 30 heavily traveled
routes.      These costs became the basis for computing an un-
regulated     or cost-based     air fare for each route.     This
fare was then compared to the lowest regularly            available
daytime coach or economy fare published           in the "Official
Airline     Guide" and the percentage difference        was computed.
(See apps. I and II.)
       On the basis of our review of Dr. Keeler's        study, we
concluded that,      with some minor changes, its assumptions are
reasonable.      We found it necessary to modify Dr. Keeler's
method because of (1) what we believe are valid criticisms             of
the study,    (2) problems encountered    in applying statistical
techniques,    and (3) the need to expand the study to produce
results   representing    the national air transportation      system.
        For example, we substituted         parts of CAB's cost model l/
to estimate    both direct    and indirect       costs.     We believe the-
cost model Dr. Keeler developed and used provided reasonable
estimates    in total,  but certain      portions,      taken alone, were
subject to question.       Substituting       parts of the CAB model
avoided this and at the same time incorporated                major parts of
a cost model which is updated and validated                quarterly   and
which provides detailed       estimates      of the effects      of proposed
fare changes and changes in operating             and financial      charac-
teristics    of CAB-regulated     airlines.       CAB's model, however,
was not used to compute capital           cost because it did not ap-
pear to be an improvement over the method of computing return
on investment     Dr. Keeler used.
      We expanded Dr. Keeler's        study to simulate       the national
air travel    market by using samples of typical          city-qair
routes and their     characteristics.        we adjusted    the published
fares to show the effect        of (1) fare changes during the year
and (2) discount     and promotional      fares.    To project      our re-
sults to all scheduled domestic air travel             on the domestic
trunk airlines,     we made and compared two estimates           of total


L/A mathematical   or graphical   representation     of what
   actually  occurs in a given situation.        In Dr. Keeler's
   study and in our extension    of his work, the cost model
   is a set of equations   intended to provide estimates        of
   the lowest longrun cost of a unit of airline        output.

annual airline      revenue for air travel    in each year from
1969 to 1974.       One set of revenue estimates,    designated
"actual    revenues,"    was based on our estimated   actual    annual
fares;   the other revenue estimates      were determined from our
estimated     increased efficiency   fares and designated     "in-
creased efficiency       revenues."  Each type of revenue was
estimated    by:

         --Categorizing    by distance   and traveler  density 257 city-
            pair markets   that accounted for nearly two-thirds     of
            domestic air   travel   in 1969 on domestic trunk airlines.
         --Using air travel       characteristics       for the 257 city-
            pair markets modified by the distribution            of annual
            travel    with distance    to determine the annual air
            travelers     in each city-pair       market category for each
            year from 1969 to 1974.
         --Multiplying      for each category,    annual passengers by
            average annual fares (either       increased efficiency  or
            actual    annual fares)  to make annual revenue estimates.
         --Summing estimated   annual revenues over all market
            categories to obtain estimated   annual nationwide reve-
         We validated  this method by comparing our estimates  to
actual     revenue data CAB compiled and adjusting  where neces-
      Some of    the problems we encountered -were resolved by
using parts of CAB's cost model.           bVlealso found that the
CAB model introduced      a conservative      bias into our esti-
mates because it overestimates         fares for markets in at
least   the under-400-mile     range.     This conservatism   means
that our estimates     of annual savings to air travelers          in
1969-74 would have been as much as $325 to $925 million
higher if a cost model better         showing intrastate    opera-
tions had been used.        Removal of this conservatism      would
increase our estimated       savings to actual travelers      from
the cited $1.4 to $1.8 billion         a year to a range of $1.8
to $2.5 billion     a year.
ESTIMATED ----__-___~-

      As the table below shows,  the domestic trunk airlines
earned annual revenues of $5.4 to $8.5 billion     during
1969-74 from their domestic scheduled passenger operations.
Using (1) annual airline   costs and operating   characteristics
compiled by CAB and (2) the conservative     cost model we

constructed      by combining complementary parts of CAB’s                  Domestic
Passenger Fare Investigation             (DPFI) cost model and,
Dr. Keeler's      longrun cost model, we estimated              annual fares
based on the characteristics             we would expect in the long run
in a less regulated        airline     industry.     Using these fares
(for a set of 48 city-pair           markets that we used to represent
the national      air travel     system) and actual fares for the
same markets,       we estimated     the effect     a less regulated         air-
line industry       would have on the airlines'          costs,     fares,
revenues, and passenger traffic.               Our analysis       showed that
if the trunk airlines         had achieved the kinds of long-term
characteristics       we assumed for a less regulated              airline
industry,     travelers    on the CAB-regulated         airlines      might
have saved from $1.4 to $1.8 billion              a year during 1969-74.
These estimates        are the annual differences          between actual
trunk airline       revenue reported        by CAB and lower revenues
estimated     for less regulated         trunk airlines      operating      at
lower costs per passenger and proportionately                   lower fares.
                                                                                 ?ercent by
                                                                               which actual
                                                         Passenger savings     fares exceed
                                              Actual   from less regulated        estimated
---                                         revenue
                                            -----             airlines
                                                              ------         fares    (note a)

1969                                         $5.4                 $1.8             52
1970                                          5.5                  1.4             35
1971                                          6.0                  1.6             36
1972                                          6.7                  1.8             36
1973                                          7.4                  1.6             29
1974                                          8.5                  1.5             22
a/Percentage    computed by dividing     actual revenue by estimated
   revenue for actual     travelers   and subtracting      100 from the
   result.    The  computation    was done  in this   manner   to conform
   to and oe comparable with Dr. Xeeler's         results.     Another
   way to present these results       would be to express the esti-
   mated savings as a percentage       of actual fares.       These per-
   centages would be 33, 25, 27, 22, and 18 for the years
       The $1.4 to $1.8 billion     estimates  indicate  that
actual airline    fares were higher than they might have been
if the airlines     had been operating    in a less regulated

        The above savings estimates         do not include the effects
of additional       air travel     that might be caused by lower fares.
They represent       savings to actual travelers.        Airline   experi-
ence and studies        of how price changes affect      air travel   in-
dicate that when fares are lowered, more people travel               by
air.     An explicit     quantitative     measure of the amount of
change in air travel        to be expected from a change in air
fares is called the price elasticity             of demand for air
travel.     Airline,     CAB, and other estimates      of this measure
range from -0.5 to -4.0,           based on analyses of the effects       of
past airline      fare and travel       changes.
      A price-elasticity    value of -0.5 indicates    that if
average fares decreased by 20 percent,      air travel    would in-
crease by 0.5 times the same amount, that is by 10 percent
(0.5 x 20 percent).      The -2.0 value indicates   that if aver-
age air fares decreased by 20 percent,      air travel    would in-
crease by 2.0 times this amount, or 40 percent (2.0 x 20
      To estimate    the additional   travel   that might occur
because of lowered fares,       we used a price-elasticity    value
of -1.3 as shown in some CAB staff        studies and testimony.l/
The additional    air travel    caused by lower fares produces
savings because some travelers       are willing    to travel by

      "Traffic,   Fares, and Competition,   Los Angeles -
      San Francisco Air Travel Corridor,"     Staff Research
      Report No. 4, Research and Statistics     Division,  Eu-
      reau of Accounts and Statistics,    August, 1965.
      "Forecast  of Scheduled Domestic Air Travel For the
      50 States,   1972-1981,"   Office of Plans Study, \J.S.
      Civil  Aeronautics   Board, November 1972.
      Exhibit    EC-T-5, Docket 21866-9,       U.S.   Civil   Aeronautics
      Board,    November 20, 1970.

air at an intermediate         fare between the CAB-regulated          fare
and the lower fare.        The sum of the differences         between
each additional    traveler’s       intermediate   fare and the final
reduced   fare may be defined         as savings to travelers      at-
tracted  by lowered     fares.

        The next table    presents    estimates       of these savings
combined    with   the savings     to actual     travelers,     assuming
additional     passengers    are allocated       to aircraft     so as to
maintain    an average    60 percent     industry-wide       load factor.

                                                                                              Savings       to actual
                              Savings     to                                             and additional          passengers
----                   actual
                                             ---                                            for   -1.3      elasticity

1969                                             $1.8                                                    $2.5
1970                                              1.4                                                     1.8
1971                                              1.6                                                     2.0
1972                                              1.8                                                     2.2
1973                                              1.6                                                     2.0
1974                                              1.5                                                     1.8

        Shown below are the corresponding                                                   actual   annual
revenues     and estimated  annual     revenues                                             assuming    increased
efficiencies     and a fare elasticity       of                                            demand of -1.3.

                                        Actual                                               Estimated         revenue
----                                    revenue
                                        -------                                            for -1.3 elasticity

1969                                             $5.4                                                 $6.1
1970                                              5.5                                                  6.1
1971                                              6.0                                                  6.5
1972                                              6.7                                                  7.3
1973                                              7.4                                                  7.9
1974                                              8.5                                                  9.0

         Following     are the actual         levels       of annual    air travel
in revenue-passenger-miles              for 1969-74 for passengers               on
CAB-regulated        airlines      and the estimated             levels  of air
travel      assuming     increased    airline        efficiencies       and a fare
elasticity        of demand of -1.3.

                                 AC t ua I-            Zstimated        air travel
Year                       air      travel
                                    ------                    -1.3
                                                      lor--------     elasticity     I,      ,
                           ----------(jillions           of    FpMs)----------

1959                                 96                              165
1970                                 96                              14?
1971                                 98                              146
1972                                109                              151
1973                                11s                              163
1374                                112                              153
----      effects
to passengers
        Nhile    the passenger        tends to benefit           from savings      due
to lower      fares    under the assumptions             used in our analysis,
his flight       might    be suostantially          different        than under
CAB regulation.           For example,      increased         seating      means that
some    passengers      might    have less space.             3ecause our analysis
also envisions         higher    average    load factors           the probability      of
not outaining        seats on desired         flights       may increase       for some
passengers.         Passengers     might also experience                some incon-
venience      in that     hi,gher load factors           could mean flights         at
less    frequent     intervals      than are oresent1.y            available     on some

        On the other       hand, the Texas and California                 intrastate
airline      experience      on which these assumptions             are partially
aased,     included     both lower     fares     and frequent       flights       for
air traveler;        on medium and high density             routes.         The intra-
state    airline     experience     shows that        to obtain     lower air
faresp     and some-times      more frequent       flights,     travelers         are
willing      to accept     the somewhat        more crowded conditions
typical      of the intrastate       aircraft.

                                 CHAPTER 3
        EFFECTS OF-1_--------
        This analysis    of th e effect  of CAB regulation     on
airline     fares, traveler    savings,  air travel,    and airline
revenue is based on certain         assumptions,   including    in-
creased efficiency,       a return on capital    equal to the long-
term average return for all corporations,            and an elastic-
ity of demand for air travel         of -1.3 as described     in
appendixes I and III.
       In this chapter we show how changes in assumptions
cause    the study outcome to vary.    This procedure,   called
sensitivity     analysis, allows a decisionmaker   to judge the
importance of each assumption.
        A passenger aircraft's         load factor      is defined as the
percentage of available          seats occupied by revenue passen-
gers.     An airline's    load factor        is determined as the ratio
of revenue-passenger-miles           to available-seat-miles.
Dr. Keeler assumed unregulated             airlines    could maintain
a 60-percent      load factor     in high-density         markets over the
long run.      We believe    this assumption is reasonable based
on average load factors          achieved by Pacific         Southwest Air-
ways in the California         intrastate       air market during the
196Qs, Southwest Airlines           in the Texas intrastate        air
market in the 1974-75 period,             and the trunk airlines        in
the decade before the airlines'             substantial      use of jet
       Our study included markets of various densities,
distances,     and load factors.         To approximate      load  factor
variations,      we assumed the airlines        would achieve an
industry-wide      load factor     of 60 percent,       but varied the load
factor    by trip    length in the same proportion           as actually
occurred during each year.            Therefore    flights     in medium-
distance    markets were assumed to achieve more than a 60-
percent load factor,         and flights    in short-      and long-distance
markets were assumed to achieve less than 60-percent                   load
factors.      We believe this adjustment         better     shows what
actually    would occur if the trunk airlines              were to achieve
an industry-wide        60-percent    load factor.
      To provide some understanding       of how our study results
are affected    by variations    in the load-factor   assumption,
we made computations       using 72 percent--Pacific   Southwest
Airways'   average load factor     in 196O-65--and the actual
annual coach class load factors        for the domestic operations

of the domestic trunk airlines--ranging         from 52 to 59
oercent during 1969-74.        Actual coach load factors   were used
instead of combined first       and coach class load factors    to
be consistent      with our assumption of equal coach-class     and
first-class      average load factors    in our conversion of all
coach seats into first-class        seats and coach seats.    The
results     of these computations    are summarized below and pre-
sented in more detail       in appendix X.
                                     Effect      of Load Factor
                               e-w------               on Study Results
                                           Actual                    Assumed industry-wide
                                       --------                            load factor

                                                                 -----         60%
                                                                               ---                                     72%
Annual savings to
actual passengers in                                               1.1                            1.4                      1.9
1969-74 (billions                                                   to                             to                       to
of dollars)                                   N/A                  1.5                            1.8                      2.4
Annual savings to
actual and induced
passengers in                                                      1.2                            1.8                  2.6
1969-74 (billions                                                   to                             to                   to
of dollars)                                                        1.9                            2.5                  3.4
Excess of actual over                                                     19                       23                       39
estimated fares in                                                        to                       to                       to
1969-74 (percent)                             N/A                         39                       52                       72
Annual air travel                in           96                   126                            141                  165
1969-74 (billions                of           to                    to                             to                   to
RPi\il)                                      118                   151                            165                  193
Annual airline                               5.4                   5.9                            6.1                  6.3
revenue (billions                             to                    to                             to                   to
of dollars)                                  8.5                   9.0                            9.0                  9.4
-------w-w--- SEATING
           Dr.   Keeler    assumed all-coach                  seating                      with    seating          den-
sities       similar      to     densities          on Pacific                   Southwest              Airways
aircraft         and seats        similar          to those      of the trunk                           airlines.

As shown, he assumed that only             one of the three     aircraft
types he used had a galley for             longer flights.
                                                Number of
Aircraft      Range: trip
              ---------        distance        coach
                                               __    seats
                                                         -        Galley
DC9-30       Short:    under   350 miles           110        No; two coffee
B727-200    Medium:     350 to 900                 158        NO; two coffee
              miles                                               bars
DC8-61       Long:    over   900 miles             251        Yes; small
       Full food service might be needed on some long-range                 and
medium-range flights.            To accommodate this,     we modified    the
seating assumption for the B727-200 to eliminate                 the coffee
bars, added a full         galley,    and reduced seats accordingly.         We
also modified       the DC8-61 all-coach       seating to that of Trans
International      Airline's       DC8-61 and X8-63 that have full
galleys    and all-coach        seating of 254 seats.      We also assumed
use of wide-bodied         aircraft    in the years they were actually
used by the trunk airlines.              The B747 was used from 1970 on,
and the DC10 provided substantial             scheduled passenger ser-
vice beginning        in 1971.      The B747 all-coach    seating figures
with full     galley service tabulated         below are taken from
World Airway's        5747 configuration;      the corresponding     DC-10
all-coach     seating with full-galley         service  is taken from
Trans International's           DC-10 experience.
       The aircraft     and seating estimates      we used in our base
case are shown below with the aircraft           class to which they
belong.     Although we assumed that both coach and first-
class service      would be provided,    the seating estimates       we
used are ta.bulated      in terms of all-coach      seats to facili-
tate comparison with Dr. Keeler's          estimate   shown above, and
the aircraft     manufacturers'    estimates    of maximum all-coach

                                      ~G‘ACJ          Manufacturer's
                                estimates:               estimates:
                Dr. Keeler       all-coach            maximum all-coach         Aircraft
-----em-         estimates
                 --------           seats
                                    ---                      seats
                                                             ----                class
DC9-30               110           110                       115          2-Engine,   turbo
                                                                              regular    bod-
B727-200             158           153                       189          3-Engine, turbo,
                                                                              regular    hod\
X8-61               251            254                       259          4-Engine,   turbo
                                                                              regular hod-
DC10                               376                       380          3-Engine,   turbo
                                                                              wide body
3747                               461                       500          $-Engine,   turbo
                                                                              wide body
       For our base case we assumed that                  in a fully    competitive
environment   the airlines    probably would                provide first   class
as well as coach travel.       Accordingly,               we converted    our esti-
mates of all-coach     seating to a mix of                first   class and coach
seating using the following       assumptions:
        1.     'The average first-class  seat uses 160 percent of
               the floor   space used by the average coach seat.
        7.a.   'The same load    factors        for   coach and first-class
        3.     The average proportions    of coach and first-class
               revenue passenger-miles    per aircraft flight     were
               the same as the proportions    for the entire     year.
      tie varied our seating assumptions to determine how
changes in seating densities      affected    our estimates.      In
one case, for low-seating,      we used the averaqe trunk air-
line seating per year for the five aircraft;            in another case,
for high-seating    density,   we used the aircraft       manufacturers'
estimate   of maximum seating for these       aircraft.     The high-,
low-, and base-case estimates      of aircraft      seating used in our
seating sensitivity    analysis   are compared on the following

‘.        7’

                  1969-74          Average
                                      ---- Annual           Seating     per-II_--
                       Low seating                                            High seating
                       (based on                  GAO base case                  (based on
                     actual trunk                (our assumed                 manufacturer's
                         airlines)               . seating)                       maximum)
                     Pirst       Coach            First  Coack                First     Coach
     -I_-            class
                     ---         class
                                 --               class
                                                  ----      class
                                                            --                class
                                                                              --        class
     DC9-30              16            73              14          89             14         93
     3727-200            22           103              19         123             23        152
     DC8-61              26           155              31         205             32        209
     DC10                40           188              42         309             42        313
     3747                50           285              53         376             58        408

          The effect of using these different                         seating figures       in
     the study is shown below.   More detailed                        annual estimates
     are shown in apppendix XI.
                              Effect   of Uncertainties                in
                                                      _-----_ Aircraft
                                                                    ----       seating    density
                                                                                 GAO      Manufac-
                                      Actual           Trunkline             assumed      turers'
                                   --__---              seating
                                                        --_I                 seating
                                                                             ----         maximum
     Annual savings to
     actual passengers                                      0.5                 1.4         1.7
     in 1969-74 (bil-                                       to                   to          to
     lions of dollars)                  N/A                 1.5                 1.8         2.1

     Annual savings to
     actual and induced
     passengers in                                          0.5                 1.8         2.2
     1969-74 (bil-                                          to                   to          to
     lions of dollars)                  N/A                 1.9                 2.5         2.9

     Excess of actual
     over estimated                                           6                   23         31
     fares in 1969-74                                       to                    to         to
     (percent)                          _hJJ/A              39                    52         61

     Annual air travel        in         96                 122                 141         152
     1969-74 (billions        of         to                 to                   to          to
     ZPM)                               118                 146                 165         178
     Annual airline    revenue          5.4                 5.9                  6.1        6.2
     (Pillions   of dol-                to                  to                    to         to
     lars                               8,5                 8.7                  9.0        9.2

        An aircraft’s    annual  utilization      rate                                     is defined    as the
average    number of block- hours it is used in                                              a year (block-
hours are measured       from the time the plane                                           first    moves under
its own Fewer for purposes          of flight    until                                     it comes to rest
at the next point      of landing).         Dr. Keeler                                     used 1968 trunk
airline    utilization     rates to estimate      flight                                       equipment  capital
cost in his study.

         We used the highest            annual     aircraft       utilization          rates
experienced        by the trunk         airlines      during      1569-74 to calculate
flight       equipment     capital      cost because         (1) our study covered
that period,         (2) we believed           maximum trunk          airline      aircraft
utilization        rates    are a better         measure for rates              that would
exist      in an interstate         airline      industry       operating         with orice
and service        competition        in a less regulated               market,      and
(3) we believe          the rates       are consistent          with the longrun
airline       cost approach        used in Dr. Keeler’s               study.

       To determine       how changes         in aircraft         utilization          rates
affect   the study      results,     we compared           results       obtained       by
using the minimum annual            utilization          rates      experienced         by
the trunk    airlines      during    the period          1969-74 with our basic
assumptions      which included         the higher         utilization          rates.       We
also compared       these to results            obtained       by modifying          our basic
assumptions      to include      actual      a.nnual utilization              rates.

           Some of    the     aircraft      utilization                            rates     used      were:

                        Annual       Aircraft
                        ----------------------------- Utilization                          Rates

                                            Highest       1969-74                               Lowest 1969-74
-------               Keeler
                      -----                    utilization
                                               ----------                                         utilization

DC9-30                3,294                             3,680                                            3,284
B727-200              3,426                             3,440                                            3,052
DC8-61                4,136                             4,007                                            3,282
8747                    (a)                             3,794                                            3,392
DC10                   (a)                              3,370                                            3,084
a/Dr.      Keeler’s     study       did   not    include                       these       aircraft.

      The effect          on our study     results                             of using different    air-
craft  utilization          rates   is summarized                                below and is presented
in more detail          in appendix    XII.

                           Effect       of Aircraft
                           ------------P-v------       Utilization    Rate
                                            on Results
                                           Annual aircraft      utilization
                         Actual         Actual----Lowest                  HigEesE
                      ---_I_            1969-74
                                        -_               1969-74
                                                        __I_--            1969-74
Annual savings
to actual pas-
sengers in                                 1.4                1.3          1.4
1969-74 (bil-                               to                 to           to
lions of dollars)                          1.7                1.7          1.8
Annual savings to
actual and induced
passengers in                              1.6                1.6         1.8
1969-74 (bil-                               to                 to          to
lions of dollars)                          2.3                2 .3        2.5
Excess of actual                            20                  20           23
over estimated                              to                  to           to
fares in 1969-74           N/A              47                  48           52
Annual air travel            96            139                137         141
in 1969-74 (bil-             to             to                 to          to
lions of RPM)               118            153                159         165
Annual airline              5.4            6.0                6.0         6.1
revenue (bil-                to             to                 to          to
lions of dollars)           8.5            9.0                9.0         9.0
        Dr. Keeler assumed that without              CAB regulation,     airlines
would assign some of the most efficient                 aircraft     to each
route in accordance with their              original    design distances.
On the other hand, trunk airlines               use various aircraft        over
a variety      of routes because of scheduling             considerations,
feeder routes,        and ownership of earlier          generations     of air-
craft    than the improved versions            assumed in Dr. Keeler's
study.      We believe       that in the long run, a less regulated
airline     industry      would probably not have the characteristics
of the current         industry    to the same extent because the ex-
pected intense price (fare)             competition     would cause more
emphasis to be placed on development and use of low cost
per seat-mile        aircraft.       We also believe city-pair         market
entry and exit freedom for airlines                would enable airline
management to better            match capabilities      of existing     and
new aircraft       to air travel      markets.


        Assuming use of some of the most efficient                aircraft    on
each route might result           in a higher estimate of savings to
passengers than could be achieved in a fully                   competitive
airline    industry.       On the other hand, we believe use of the
current    match of aircraft          to routes would probably result
in a lower estimate          of savings than would actually            occur.    To
obtain an estimate         of the longrun effect          that could reasonably        L
be expected in a less regulated              airline    industry,    we combined
two assumptions concerning              use of the least cost per seat-
mile aircraft.        tie based our analysis          on use of some of the
most efficient       aircraft     available      in the 1969-74 period be-
cause we believe that price competition                 in a fully     competi-
tive airline      industry     would probably have encouraged earlier
development of such aircraft              by aircraft    manufacturers.       Me
also used these aircraft            in our analysis      over ranges outside
their most efficient          ranges.      We assumed they would have been
used to the same extent and in the same way that the aircraft
classes they belonged to were used by the trunk airlines                      in
the 1969-74 period.
       In summary, comparatively        low cost second and third
generation   stretch-version      aircraft   were used in our analy-
sis, but the potential       savings from using these aircraft
were reduced by using them over the ranges actually           flown by
the classes of aircraft       to which they belong.      The aircraft
we used and the classes to which they belong are tabulated
Years       Aircraft     used              Class   to which aircraft
----        in   base
            --e---w---- case                         belongs
1969-74        DC9-30           2 Engine,      turbofan,    regular-bodied
1969-74       5727-200          3 Engine,      turbofan,    regular-bodied
1969-74       X8-61             4 Engine,      turbofan,    regular-bodied
1971-74       DC10              3 Engine,      turbofan,    wide-bodied
1970-74       5747              4 Engine,      turbofan,    wide-bodied
       de also determined    the effect  of using the aircraft
tabulated    here over those ranges at which we believe their
per seat-mile     costs to be close to their minimum values,      The
two results    are tabulated    below and presented in more detail
in appendix XIII.

                              Effect       of Uncertainties ---- in
                                     the Assumed Use of
                            the    M~-EfficTe~~?%?raft
                                               ------- Route
                                          Our base case
                                         Most efficient           Alternative
                                                                  -------------   case
                                       aircraft      used in
                                       same proportions               Most efficient
                                       as trunk airline               aircraft     used
                                           actual use of                only in their
                        Actual           corresponding                most efficient
                      me-----          aircraft
                                       ------------- classes          ranges (note a)
Annual savings
to actual pas-
sengers in
1969-74 (bil-
lions of dollars)           -              1.4 to 1.8                   1.6 to 1.9
Annual savings to
actual and in-
duced passengers
in 1969-74 (bil-
lions of dollars)           -              1.8 to 2.5                   2.0 to 2.7
Excess of actual
over estimated
fares in 1969-74
(percent)                                   23 to     52                 24 to    56
Annual air travel
in 1969-74 (bil-
lions of RPX)            96 to 118         141 to 165                   147 to 170
Annual airline
revenue (nil-
lions of dollars)       5.4 to 8.5         6.1 to 9.0                   6.1 to 9.1
a/We defined an aircraft's  most efficient     range to be that                   range
  over which we expected tne aircraft      to have close to its
  least average cost per seat-mile.
          Dr. Keeler assumed a 12 percent pretax rate of return on
airline      capital.  He expected this to yield a longrun aftertax
return      on capital of about 7.5 percent.   His assumption was

sased on his research and comparison of airline      and general
corporate    pretax and aftertax   return on capital and on
similar   work by Richard E. Caves which was re.ported in a
1962 airline     study. _1/ From this work Dr. Keeler concluded:
       --Average   annual      pretax rate of return on capital
          during 1960-66       for all corporations  was about
          11.7 percent
       --Average annual        aftertax     rate of return on capital
          during 1939-66       for all     corporations   was about
          7.5 percent
       --Average     annual    airline aftertax  rate of return             on
          capital    during    1939-66 was about 7.5 percent.
        From this and the airlines'        rate of expansion,          Dr. Reeler
concluded that (1) in the long run airlines                  had earned a
sufficiently       high rate of return        to attract      adequate capi-
tal for expansion and (2) the airlines'                 rate approximated
that of the entire        corporate   sector.        Because (1) the cost
model he constructed        represented     lonqrun airline        costs (and
capital      costs in longrun cost models are based on the longrun
rate of return on capital),          (2) historic        longrun capital
costs for the airlines         approximated       those of the corporate
sector,      and (3) he found no compelling           reason why this
should change, Dr. Keeler selected              the historic     longrun
pretax rate of return on capital           for all corporations          as the
appropriate       value for his study.
        We reviewed the work on which Dr. Keeler's                  conclusions
were based and found that he intended to calculate                      a return
estimate      composed of interest,         depreciation,       corporate     in-
come taxes, and corporate            aftertax    income.       By using this
return     estimate     to calculate      longrun annual capital          costs,
Dr. Keeler included costs representing                 airline    corporate
income taxes,       interest     payments, depreciation          costs, and
aftertax      earnings.      tie accepted and used !X. Reeler's
definition      of the long-term        pretax return on capital.             Iiow-
ever, we were unable to reproduce his estimated                     return
using Dr. Keeler's         data sources and stated procedures.
Instead,      using the same Internal          Revenue Service (IRS) data
and his definition,          -we obtained an estimated          annual pretax
return on capital

pives,-fmia?n:,               "Air Transport   and Its Regulators,
  An Industry       Study,"    Harvard University    Press, 1962,
  P* 394.

of 10.5 percent.         Dr. Keeler     notified    us that     the difference
in our 10.5 percent         and his 12 percent       was probably        due to a
set of approximate        adjustments      he made to consolidate           aggre-
gate corporate      financial     and income statements           for all    active
corporations.      Making these adjustments            required      disaggregat-
ing some of the published           IRS accounts     for all      active    corpo-
rations     used by Dr. Keeler        in his study.

        We examined       this      published        IRS data which consisted                of
various     disaggregations            for various          groups of corporations
but did not find          what we needed.                The IRS official           respon-
sible    for these statistics                of income data told              us there are
no other      disaggregated          data available            of the sort we needed.
Lacking     these data,         we felt        that adjustments            to consolidate
the IRS’ aggregate            corporate          income and financial              data were
too uncertain         to include         in our base-case            calculations.          In-
stead,     we included,         as a sensitivity             analysis,        the results
of using the 12-percent                rate of return            Dr. Keeler        assumed.
The results       are tabulated            below     for:    our base case using our
assumed     10.5 percent          rate of return,            our base case except
that    Dr. Keeler’s        12 percent           rate was used, and our base
using an 18 percent             rate of return            on capital.           The 18 per-
cent pretax        return     is included           as an approximation             of the
post-tax      12 percent        rate of return            that    CAB uses as a stand-
ard in ratemaking           decisions          for the trunk         airlines.         iNore
details     of the effects           of these rates-of-return                   are
presented       in appendix         XIV.

                           Effect      of Uncertainties
                           --------B---WI_           ------7 in
                      Pretax      Rate   of Return
                      _-------___--_____---------      on Capital
                                                  Pretax rate of return
                                                        on capital
                         Actual              10.5                 12       18
  -----    area       experience
                      ---------             oercent
                                            L----d-         percent
                                                            ----     percent
Annual savings
to actual pas-
sengers in                                    1.4              l-4       1.0
1969-74 (bil-                                  to                to       to
lions of dollars)          N/A                1.8              1.8       1.5
Annual savings
to actual and
induced pas-
sengers in                                    1.8              1.7       1.1
1969-74 (bil-                                  to               to        to
lions of dollars)          N/4                2.5              2.4       1.9
Excess of actual
over estimated                                 23               21         13
fares in 1969-74                               to               to         to
(percent)                                      52               50         40

Annual air travel          96                 141              139       128
in 1969-74 ('oil-          to                  to               to        to
lions of 2P%)             118                 165              161       148
Annual airline            5.4                 6.1              6.0        5.9
revenue (bil-              to                  to               to         to
lions of dollars)         8.5                 9.0              9.0        8.8
         ITo estimate    the effect    lower fares might have on the
number of passengers the airlines              served during the years
1969-74, we assumed a fare elasticity-of-demand                  factor   of
-1.3 based primarily          on (1) empirical       studies of elastic-
ities     demonstrated      by Pacific    Southwest Airways in 1961-64
and Southwest Airways in 1471-74 and (2) cross-sectional                     sta-
tistical       studies of airline      travel   in two sets of city-pair
markets.        The empirical     studies produced elasticity         estimates
Of    -1.3 to nearly -4; the statistical             studies produced elas-
ticity      estimates    of -13.96 to -1.37.       !qe also considered     other
studies      and testimony     that suggested elasticities          could range
from -0.5 to -2.0,          and found that CA5 uses an assumed value of
elasticity        of -0.7 in its rate setting         proceedings.

       In view of this.wide    range of elasticity    estimates
and assumptions and what we believe to be substantial
uncertainty   as to the correct     value, we made computations
using -0.7 and -2.0 to determine how varying the elasticity
factor   would affect   our study results.      These computations   are
compared in the following      table.    Annual estimates   are com-
pared in appendix XV.
                   Effect     of Uncertainties      in
           Fare Elasticity      of Demand for      Air Travel
                      Actual               Fare elasticity       of demand
                    experience      -0.7                 -1.3
                                                         --             -2.0

Annual savings
to actual pas-
sengers in
1969-74 (bil-                         1.4                 1.4             1.4
lions of dol-                          to                  to              to
lars)                  N/A            1.8                 1.8             1.8
Annual savings
to actual and
induced passen-
gers in 1969-74                       1.6                 1.8             2.0
(billions of                           to                  to              to
dollars)               N/A            2.1                 2.5             3.0
Excess of actual
over estimated                            23                23             23
fares in 1969-74                          to                to             to
(percent)              N/A                52                52             52
Annual air travel        96           118                 141             174
in 1969-74 (bil-         to            to                  to              to
lions of RPM)           118           136                 165             221
Annual airline
revenue (bil-           5.4           4.7                 6.1             7.5
lions of do1             to            to                  to              to
lars)                   8.5           8.0                 9.0            10.4

                                        --------- 4

                            A3EhiCY       AiJD EXPERT
                            _----_-_-I-_------------    COWGEk~1’S

        ijecause    any study of this           kind includes    both generaliza-
tions     and assumptions     that are          not certainties,     we solicited
comments on our proposed          report          from a numoer of interested
Federal      agencies   and experts.

       Federal        agencies
              Civil   Aeronautics    board
              aepartment    of Transportation
              Department    of Commerce
              Council    of Economic   Ildvisors

       _-L_---        -1/
              3r .   Theodore  5. Keeler
              Or .   George W. Douglas
              Or .   Nilliam  A. Jordon
              Dr.    George C. Eads

          In general,       these agencies         and ex?er,ts     agreed wit!? the
pre,mise     that     increased       competition       would result      in more
efficient       airline       operations,       but suggested       a nu,mber of ways
in which the analysis                could be improved.          Further,      there was
agreement       that the sensitivity              analyses    presented      in chapter
3 identify        the effects         of the most ismportant         assumotions     and
make availaole           alternative       results      in cases where a reader
may be interested.

        In several        cases,    the agencies          and experts       suggested
caution      in interpreting          the oroposed         report’s     results.       The
technique       used in the analysis              predicts     results      based on
assumed relationsnips              between independent             measurable     fac-
tors.      In actuality,         these relationships             are not the only 1
factors      that   influence       airline      operations.         Others we have
not included        may also be important.                 4s we have stated,          we
find    the assumptions          and    techniques        used   to  be   reasonable
but agree tnat          the results         should    be interpreted         and used

-               of experience        and training        are   included     as
   appendix       XVI.

      We did not solicit    comments from individual              airlines
or the Air Transport     Association.
       Some of the major comments and our responses are
outlined  below.   The complete comments are included as
appendix XVII.
        The 6G-percent       load factor    used in our study represents
the average industry-wide           annual load factor        we believe
would be achieved in the long run by domestic airlines                     operat-
ing with less regulation.            CAB doubted the ability         of the
trunk airlines         to achieve this level on all routes and over
the entire       airline    system while adhering to the (1) assump-
tion of only three types of aircraft                 which allows little
flexibility       in serving markets of various densities             and dis-
tances,     (2) assignment of aircraft            types to specific     mileage
blocks,     and (3) high aircraft        utilization      rates assumed.
Dr. Eads also pointed out that an average load factor                    does
not take into consideration           the many variables        in the market
place.      Dr. Keeler stated that the extension              of his assump-
tion to low- and medium-density              routes may overstate       poten-
tial    benefits      because 6G-percent       load factors    may not be
feasible     on such routes.
         We agreed that our assignment of aircraft           types to
specific      mileage blocks would have limited         an airline's
ability     to serve its markets,       so in12modified   the assump-
tion by assigning      aircraft      to each city-pair    market in
the same average proportions           as the proportions      of air
travel    that the aircraft      classes represented      by these
aircraft      provided  in city-pair     markets of comparable
distances      in each year.
        We did not modify our assumption of aircraft
utilization      rates because we believe chanqes we made in
two other assumptions would permit the airlines              to operate
at these rates.        One change was the use of each of our
assumed aircraft       over the same distances      and to the same
extent as the aircraft         classes they represented.       For ex-
ample, we assumed 3727-200 aircraft           would be flown over
the same distances        to the same extent    (same proportion      of
revenue passenger-miles          per year) as the aircraft     class
that it represents--        3-engine turbofan   regular-bodied     air-
craft.      This change would permit airlines        to fly smaller,
shorter     range aircraft,      such as the DC9-30, over longer

nonstop        and   ixultistop              --thereuy
                                        rout233         increasing      utilization
rates  over those                 achievable with an arbitrary     limitation.
Tne otner   cnange                was made in assumed load factors        for city-
pair markets.

         ‘IO incorporate        the variation           in load factors     that
actually       occurs    over routes         ( city-pair      markets)    of various
lengths       and densities,        we varied         the assumed load factors
per market        by trip     distance      in the same proportion            as
actually       occurred     in each year.             Thus our overall      load
factor      is 60 percent,         but average          load factors    over medium-
distance       routes    are higher       than 60 percent,           and average
load factors         over short-       and long-distance           routes   are less
than 60 percent.

         Dr. Douglas        stated     the report       is correct         to assume
significantly          higher      load factors       in an efficiently              con-
firjured      industry      and that his research,                while    indicating
tnat     load factors        would vary by market density                    and dis-
tance,      confirms      that an overall         efficient          load factor
for all       markets     would be in the neighborhood                   of GO percent.
Dr . Jordan believes             the GO-percent         load factor          is unduly
low in view of intrastate                 airline     results       and his assump-
tions      about the form an unregulated                  airline       system would
take in the long run.

        Cur bases for using the 60-percent                   load factor    include
(1) ?acific      Southwest         Airways    experience       in 1960-70,
(2) Southwest       Airlines’         experience     in 1974-75,      and (3) trunk
airline    experience         in the 195Os, when industry             load factors
ranged aDove 50 percent.                 Load factors      in particular     mar-
kets of course,         would vary depending            on, among other       things,
the value     travelers         place on their       time,     the market density,
and the distances          involved.

Host    efficient
-~--_---_-------_-__       aircraft

       Our report     assumes the airlines,            if unregulated,      would
use some of the most efficient              aircraft.       CA5 believes
this   assumotion     is unrealistic        in view of the existing
fleet   of aircraft,        scheduling    considerations,        and the man-
ner in whicn aircraft           are designed       and improved.       CAB
points    out  that   each efficient        aircraft      was preceeded     by
less efficient       earlier     models which were necessarily             pur-
chased and used before           the better      versions    were built.

          It   is true that   the existing       fleet     includes    aircraft
other       than those identified      as most efficient.             Cur assump-
tion      is not that    the present     airline       industry,    if regulation

were reduced in December of one year, would begin realizing
savings from using more efficient  aircraft in January of the
next year, but that over many years the industry      would tend
to emphas-ize efficiency and match the most efficient      aircraft
with the markets they serve best.
      Our purpose in selecting     the aircraft   we assumed for
our study was to represent     all the major classes of air-
craft  the trunk airlines   actually   used during each period.
Therefore,    we added two additional   aircraft,   the B-747
and DC-lo, in the years they became available.         The effects
of variations    in the assumption are detailed     in appendix XIII
and discussed and summarized in chapter 3.
--v---------m utilization
       be used the highest annual aircraft        utilization     rates
experienced   by the trunk  airlines      during 1949-74 to calcu-
late flight   equipment capital     cost.     CAB thinks the rates
are too high because of, among other things,             stage length
differences   between actual experience and our assumptions.
        Ne agreed with CAD's comment, but believe that by
changing the analysis      to allow assumed average load factors
in each market to conform to actual experience           and assumed
aircraft    use to approximate    trunk airline   actual   use (in-
cluding nonstop and multistop        flights),  we have relaxed the
constraints     to a point that could reasonably be achieved
in actual operation.
        Our sensitivity     analysis    (see ch. 3) shows the results   of
using    the lowest utilization       rates actually    experienced  by
trunk    airlines    and the actual utilization      rates for all the
trunk    airlines,     in contrast   with our assumed high utilization.
      CAB had several comments about the seating densities
we used in our proposed report.       It objected to our assump-
tion that there would be no first-class       travel    and seating
and to the inclusion   of savings generated by replacing         first
class with lower quality   service.     We agreed with its crit-
icism and changed the analysis      to include first-class      service
on all flights.
        CAB also objected to the assumption that full      galleys
would not be needed on 2-engine,      turbofan regular-bodied
aircraft   --it believed  food service might well be required

   on relatively   short   flights.           Ne left      our assumption     unchanged
   oecause we oelieve    airlines          operating       efficiently    would not
   serve food requiring      galleys        on short       flights.

  Rate     of return
  -------P-----e--%--^   on caoital

          The rate of return        on capital  Dr. Keeler    used repre-
  sented     the average oretax       rate of return   on all   active     cor-
  porations.          We considered   the rate a reasonasle     estimate      of
  earnings      airlines     would have to sustain   to attract      capital
  investinent       over the long run.

         CA3 stated       the rate we used is too low because the air-
 line    industry     is cyclical,     oligopolistic,        l/ and capital
 intensive.        CAL has concluded        airlines      musE earn 12 percent
 on invested       capital    after  taxes,       equivalent    to a 15 to 20
 percent     pretax     rate of return.

         Dr. Douglas     also stated    the 10.5 percent      of pretax
  rate of return      we used for the report        was too low.      He he-
  lieves    an 18-percent     rate would be more appropriate          consid-
  ering   airline   financial     risks   and current   interest    rates.

         C3e believe    an average    rate of return      in a competitive
 industry     would include      both successful      and unsuccessful         busi-
 nesses.       It would take into       account   businesses    consciously
 operating       at a low rate of return        or even a loss to gain a
 foothold      in the market,     and businesses      whose rate of return
 was so low the investors           would choose to cease operation.                de
 believe     the average    pretax    rate of return      for all   corpora-
 tions    is a reasonaole      rate to use for a competitive           airline
 industry     as a whole in the long run.

        Dr. Keeler       believes      we may have duplication                in our in-
come and rate base calculations                      because income and capital
may be reported          more than once in cases where corporate
assets     include     securities        in other        corporations        and earnings
from    securities       of Other      COrpOratiOnS.            We believe        Dr. Keeler
may be correct,          but   IRS   could       not   provide      us  data    that   would
permit     a calculation         of the incidence            and effect       of such a
duplication.          Dr. Reeler       believes        a 12-percent        rate of return
would be the average             pretax     rate of return           for all      corpora-
tions     corrected      for income consolidations,                  and we have in-
cluded     that    value as one of our sensitivity                     analyses.


L/An industry    in which             each of    a few companies        affects      but
   does not control    the            market.

          Cur sensitivity analyses (see ch. 3) show the study
results      at 10.5, 12, and 18 percent rates of return.
Fare elasticity
-------m----e------   of demand
       To estimate  the effect    lower fares would have on the
number of passengers airlines       served during 1969-74, we
used a fare elasticity     of demand factor       of -1.3.    Both
CAB and the Department of Commerce pointed out that we
applied the elasticity     factor    assuming that demand for air
travel   would vary in a regular,      consistent    pattern,   when
demand might actually    change in some other way.
       Our review of available       evidence including     CAB staff,
consultant,    and airline    studies and testimony      before CAB in
its DPPI, disclosed      estimates    of this elasticity     ranging
from -0.5 to -4.0.      Based on our review, we concluded that
-1.3 is the most reasonable estimate           of the average fare
elasticity    of demand for the entire       trunk airline    system.
         CAB and the Department of commerce are correct      in their
comments and the fare elasticity        of demand may vary in ac-
tuality.       Again, we believe the -1.3 factor    is a reasonable
approximation        of what might occur on average over the entire
airline     system and have presented alternatives     in our sensi-
tivity     analysis.
      CAB also commented that our method of applying elastic-
ity of demand resulted   in percentage ambiguity   and that the
use of logarithms   would result  in a better application.    We
agreed with its comment and changed our method of computa-

Use of CAB's cost
----------------_I_    model
       YJe substituted       parts of the CAB's DPFI cost model for
the complementary noncapital              cost elements of the cost
model Dr. Keeler used in his study.                  Dr. Keeler,   Dr. Douglas,
Dr. Jordan, and Dr. NacAvoy all expressed the opinion that
CAB's cost model is inaccurate;               or: Keeler because actual
block times are not used and because of arbitrary                   cost allo-
cations;    Dr. Douglas because he believes               the trunk airlines
do not minimize costs under CAB regulation;                  and Dr. Jordon
oecause he believes          airline    costs are increased by the cur-
rent regulatory        structure     and therefore      our costs are over-
estimated    for unregulated         airline     performance    and our sav-
ings underestimated.            Dr. Keeler pointed out that our

cost model estimated  fares for some intrastate  air             markets
that exceeded fares that Pacific  Southwest Airways              actually
charged and made a profit  on.
       We agree that C4B's cost model overstates             costs, partic-
ularly   for intrastate      air markets.     However, we believe it
provides computational        advantages and is frequently         updated
and validated.      Since the overstatement         of costs results     in a
corresponding     conservative     (underestimated)      estimate   of sav-
ings, we used the CAB cost model despite its flaws.                 We found
that our estimated      savings were from $325 to $925 million             less
than they would have been if we had used a cost model that
better  represented     intrastate    airline    operations.
-------------     markets
      CAB observed that in our original     computations we had
used fares applicable    to markets in the continental   TJnited
States for routes from the mainland to Alaska and Hawaii.
CAB said this overstated    savings because fares on mainland-
Hawaii routes are considerably     lower than fares for compa-
rable routes within   the continental   United States.
        We agreed with      C!IEI's comment and revised    our computa-
tions    accordingly.
Published versus actual
--------------------__I____     fares
        Dr. Keeler's    study compared estimated        fares to oublished
fares for CAB-regulated         airlines.      W;dE!
                                                   felt that published
fares generally      overstate     actual fares because of discount
programs which reduce the average fares actually               charged, and
therefore,    modified     published fares to include reductnons due
to the discount      programs in effect        during 1969-74.     Dr. Keeler
and Dr. Jordan both commented that published fares would have
resulted    in more realistic        estimates   of savings because these
are the fares available         to most passengers.        Dr. Keeler also
believes    published    coach fares are the fares for service most
closely    comparable to the service provided by less regulated
       These are reasonable arguments for using published
fares,   out we continued    to use the modified published       fares
because we believe they result       in actual fare estimates       that
better   represent   CAB-regulated  airline    actual annual fare
experience,    and because they are consistent       with our method
of transforming    fares per city-pair      market to national-level

Net welfare
-Y-P--------     to travelers
        As we have pointed out, the lower fares computed in
this study would be possible because of changes that might
not be to the passenger's          liking.    These changes (mainly
higher load factors,         denser seating,    and possibly    less timely
flights)    can all be expressed as costs (loss of benefits)              to
individual    travelers,      so that the net welfare to travelers
would actually       amount to something less than the amount of
savings from lower fares.            Both Dr. Douglas and CAB pointed
out that neither       Dr. Keeler's      study nor our revision     of it
attempts    to evaluate      the additional    costs of inconvenience
to individual      travelers    that might result    from more efficient
airline    operation.
       Such an evaluation        is beyond the scope of this report.
However, other studies          have attempted to evaluate these
costs.   I;/
Airline      system as a network
       Dr. Eads states that both Dr. Keeler's              study and our
revision    treat     the airline      system as an aggregation      of in-
dividual    city-pair     markets instead of as an interdependent
network.      He believes      airline     decisions   are made in terms of
maximizing system profits            rather than minimizing      costs on in-
dividual    segments, allowing          some segments to operate at low
load factors      and still     contribute      to network profits.
Dr. Eads believes        an analysis,       such as Fresented in this re-
port,    is best used to estimate           results   for a few carefully
selected markets rather thanjan entire               network.
      However, Dr. Keeler never suggested his study reore-
sented anything other than a few high-density                city-pair    mar-
kets, and he has some of the same reservations                 about extend-
ing the analysis       to the entire      network that Dr. Eads ex-
pressed.     iAlebelieve many of the changes we made in
Dr. Keeler's     assumptions and techniques           allow the analysis
to better    approximate     the airline      network.     These changes in-
clude (1) assuming that aircraft            would be used in the same
proportions     over various markets as the aircraft              classes they
represent   were actually       used, (2) assuming load factors           would
vary by distance       in the same pronortions         as actually     experi-
enced, (3) including        both first-class       and coach-class      service,


l/George      W. Douglas and James C. Niller,         III,  "Economic
   Regulation     and Domestic Air Transport,"          The Brookings
   Institution,     Washington, 1974.
(4) using the same oattern     of dilution    by distance     that
actually   occurred,   and (5) including   th e characteristics              of
both multistop     and nonstop service.
       Unfortunately,      existing  technigues   do not seem to pro-
vide a way to more realistically          approximate    what would ac-
tually   happen in a network as complex as the airline           sys-
tem, and we therefore         agree with Dr. gads' recommendation
that figures     resulting     from our analysis   be treated   as pos-
sible "order of magnitude" estimates,           not as exact oredic-
tions of what might occur in a differently            regulated  market.
Combination            of fares
------l-----------------_-------   and triD   lengths
       C4B commented that our use of passenger aggregates
based on segment RPM by mileage block overstates                  the number
of passengers in the short-          and intermediate-distance            blocks
and understates    the number in long-distance             routes.       In its
view, for example, a passenger originating               in New York des-
tined for Los Angeles but connecting             in Chicago would show up
twice in our analysis       as two intermediate        distance       travelers,
each paying an intermediate          distance    fare,   instead of a single
traveler   who pays one long-distance          fare.   Since fare-per-mile
rises as distances     shorten because of fixed costs,                the sum of
fares for two shorter       trips    would b e more than one long trip.
CAB believes    the overall      effect   of this situation         is to over-
state our total    revenue figure        by about 7.5 percent.
       We agreed with CAB's comment, although not with its                   esti-
mate of the size of the eEfect,       and obtained a conversion
factor    from it to eliminate  the problem by using origin-
destination    .RTN oy mileage block.
Stewardess exDense
     In its comments on the nroposed report,          CAB observed
that we misinterpreted      and misapplied   the ratio of steward-
esses to available     seats.    CAB was right and we corrected
this error  in our final      computations.
-------_--_-    densities
       CAS stated that our             estimate of the number of passengers
who actually    use various            classes of city-pair       markets (we
classed city-pair     markets            by distance    and passenger density)
is not representative     of           the domestic airline       system.   CfiB
believes    our method tends             to understate    the percentage of
passenger-miles    generated             in low-density     markets and over-
states   those generated in              higher density markets.        It con-
cluded that the effect      of           such a bias is unclear.
      CAB also observed that the number of passenger
enplanements estimated    by our model exceeded actual enplane-
ments in the same periods by about     5 percent,   thus building
an upward bias into our revenue estimates.        The adjustments
we made to correct   other problems in the analysis      eliminated
this effect.   The estimated   number of passengers per year
are now below the CA6 reported    passenger enplanements.
CAB    has attempted          to encourage
----------------------~~--~~~~---------       airline   efficiency
      In recent years, CAB has developed and applied standards
in its rate deliberations      intended to encourage more effi-
cient airline     operations.    CAB said it (1) predicates        pas-
senger fares on standard,      full-fare     load factors    to eliminate
the effect    of excess capacity      on fare levels,     (2) calculates
load factors    assuming optimum seating densities,          and (3) ad-
justs aircraft     utilization   rates upward when appropriate.
CAB believes    these adjustments      have resulted    in savings of
approximately     $750 million   to domestic air travelers        in 1975.
        The standards CA3 said it applies parallel                     the sources
of passenger savings estimated        in our report.                 We are pleased
that CAB recognized     the inefficiencies        present            in the trunk
airline    industry,  and our results      confirm that              CAB's actions
have reduced them.
      Shortly   before we issued our proposed report for commentp
CAB proposed legislation      to increase the role of competition
and decrease the role of regulation          in determining     the cor-
rect mixture of fares and services         for the domestic airline
system.     CAB specifically    recognized     the desirability    of a
gradual transition      to a domestic air transport        system
essentially    governed by competitive       market forces.

                                     ---v---w 5

       Our study offers     reliable       evidence that airlines         could
have profitably     operated at a lower cost per passenger during
the 6 years from 1969 to 1974, resulting                 in lower fares and
therefore    savings to domestic air travelers              on the order of
$1.4 to $1.8 billion      a year.         These results     could have been
produced by higher load factors,             denser seating,      increased
average annual aircraft         utilization       rates,   and use of some of
the more efficient     aircraft        available,     while at the same time
yielding   average annual rates of return on investment which
were comparable to those of the entire                corporate   sector.      The
tradeoff   for lower fares would have been less space for some
passengers and fewer flights            on some routes.
        CA3 has recognized         that inefficiency         exists     in domestic
trunk airline      operations        and has attempted to reduce ineffi-
ciencies    by applying stricter           standards      in its rate deci-
sions.     Although the magnitude of the difference                    between
fares charged Uy regulated             trunk airlines        and more efficient
carriers    not CAB regulated          appears to have gone down, it
remains large.        CAB has concluded that economic regulation
of the kind that has been applied to the airline                        industry
is no longer in the public interest                 and that increased com-
petition    among new and existing            airlines     would result         in a
better    match between the kinds of airline                 service desired
by consumers and available             from the airlines.           CAB believes
it could bring about some changes that would result                         in in-
creased competition         by proceeding        under its existing           legis-
lation,    but that such a fundamental change in the approach
to regulating      air transportation          should be mandated by the
Congress.      Therefore,      CAB has recommended that the Congress
pass legislation       facilitating        creation     and certification
of new airlines,       permitting        new and    existing     airlines       to
more freely     enter or leave particular              marketsp and allowing
airlines    increasing      freedom to raise or lower fares.
        Under the existing    regulatory    structure,   CYB controls
competition      mainly by determining     (1) what airlines   can
operate,     (2) the markets they may serve, and (3) the fares
they are permitted       to charge.    Proponents of reduced regu-
lation    believe increased competition        among new and existing
airlines    would result    in a better   match between the kinds
of airline      service and fares desired by consumers and those
supplied by the airlines.         Increased price and service

competition  can be achieved by various combinations    and
degrees of reduction   in CAE's control over what airlines
provide service   in particular markets and the rates they
       The Government's         objectives     in regulating     the airline
industry    in the late 1930s were apparently              to provide sta-
bility    to an infant       industry     by protecting    existing     airlines
from the threat        of competition,        enhancing the national         secu-
rity,    expanding service,         and rationalizing      the airline
subsidy system, thus permitting               growth and public acceptance.
Today, however, the trunk airlines                are large,    well-established
corporations,      and intercity        air travel     has become an indisoen-
sible feature      of the transportation           system.    The only airlines
still    subsidized     are local service carriers           originally      created
as "feeder"      carriers      that would specialize        in providing      short-
haul, low-density         air service,      to expand air service        to the
smaller and more isolated             communities of the country.
       Even though the original    arguments for Government regu-
lation   no longer apply to the airline       industry,  opponents of
less regulation    have stated that to discontinue       CAB regula-
tion would be a needless disruption       of a good system and
would result    in loss of service    to many communities and se-
vere economic disorder     for existinq   airlines.     Xe believe   it
is questionable    whether many communities would lose air serv-
ice if CAB regulation     were reduced or discontinued.
       The fear that there will be a considerable                 change in the
industry    and some economic disorder         if regulation        were sig-
nificantly    changed is reasonable.          Efficient      airlines     would
put considerable      competitive     pressure on less efficient            air-
lines that are now protected          by regulation.         Investors     and
lenders who made commitments based on the stability                    provided
by regulation     would lose some of their          security      and new in-
vestors would likely        be cautious    at the outset.          The disrupt-
ive effects     of a change in regulation          however, can be reduced
by providing     a transition     period that would permit gradual
adjustment    to the new system.
      based on this report,     we believe    that markets now served
by the trunk airlines     could be profitably     served at a lower
total  cost per passenger,     and that passenger fares could be
lower as a result.     The arguments for greater       reliance   on a
more competitive    market to determine services       and prices are
persuasive,   but this report does not answer a number of ques-
tions about what might happen if the form of airline            regula-
tion were changed or if regulation        were abandoned completely.

------------------_-- ‘I3 CAB

        tie recommend that            CAB continue     to work toward           improving
airline       efficiency         under its existing        legislation         by empha-
sizing      higher      load factors,       higher    seating       densities,      and the
other     factors       identified       in this   report     and by increasing           its
reliance       on competition          to determine      service       and prices,

.---___---------------------- TO THE COIGRESS
        We also recommend that           the Congress,   as part of its cur-
rent reexamination         of the need for economic         regulation     of the
airline    industry,      provide     to GAS legislative      guidance   defin-
ing current      national     objectives      for air transportation,       and
the extent     to which increased           competition  should     be used to
achieve    those objectives.

         Both CAB and the        Congress    should  allow          for   reasonable
transition     periods     to    avoid   undue disruption            of   the air
transportation       system.

     APPENDIX I                                                                      APPEHDIX I
            -----l-l------------_______I_____       ______
            -.----------------------~-~----~-----~---~- STUDY

           Dr. Keeler's    study was based on the following    assumptions
     aoout characteristics      of the competitive airline  industry   he
     expected to develop in the long run after      CAB's powers to reg-
     ulate market entry and exit and fare changes were eliminated.
          1. Load
             -------- factors:          Load factors  (percent available                     seats
                                        occupied by revenue passengers)                      would
                                        average 60 percent.
          2. Most efficient             Three of the most efficient     aircraft
             -m----m-                   would be used in each city-pair       air
                                        market as follows:
                  --------                                           Blarket
                  DC9-30                Short-range:                           Under 350 miles
                  B727-200              Medium-range:                          350 to 900 miles
                  DC8-61                Long-range:                            Above 900 miles
          3. -----
             Seating:         All planes would have all-coach     configura-
                              tions with seating densities    similar    to Pa-
                              cific  Southwest Airways as follows:
                  Aircraft                     Number seats
                                               -----------                          Galley
                  DC9-30                                110                       2 Coffee bars
                  8727-200                              158                       2 Coffee bars
                  DC8-61                                251                       Small galley
          4. Return-on-capital:
             -m-------------w                  All capital     costs include the
                                               long-term    pretax average return
                                               on capital--12      percent--based  on
                                               IRS reports     for all active cor-
                                               porations    since 1938.
          5. Aircraft        utiliza-          411 aircraft  were utilized     (number
             ----                              of block hours oer year) at their
                                               1968 trunk airline  utilization

APPENDIX    II                                                                APPENDIX       II
             ---         REGULATION
                               --~--_ AND MARKET PERF9RMANCE STUDY
          Dr. Reeler’s study estimated unregulated    (cost-based)
air fares -from 1967-69 data in 30 high-density     air markets
and compared them to early 1969 published     fares using the
following    methods.
     1. ~-
        Identified          competitive        market ---efficiencies:
                 Identified     the long-term    competitive   airline     in-
                 dustry efficiencies       (see app. I) expected to de-
                 velop if CAB's regulatory        powers were eliminated
                 primarily    regarding    (1) fare changes, (2) entry
                 to or exit from any market by an airline,             and (3)
                 hindrance of certification        of new willing      and
                 able airlines.
     2. Constructed          airline
                             ----~-       cost     model:
                 Constructed      a long-run           airline   total-cost   model
                 from domestic trunkline                cost data modified to in-
                 corporate     the long-term            competitive     market effi-
                 ciencies    cited in step             1.
     3. -------------
        Selected        city-pair      air     market       sample:
                 Selected     a set of 30 nigh-density,                  city-pair     air
     4. --__------______
        Estimated unregulated                 fares:
                 Used the total-cost      model and the distance                     and
                 time characteristics      of each of the 30 air                     mar-
                 kets to estimate     unregulated  or total-cost                     based
                 air fares.
     5. ---i-----------------
        Comoared unregulated                 and regulated       fares:
                 Compared unregulated  fares to coach and economy
                 class fares published   in the "Official Airline

     APPENDIX       III
'.                                                                                   APPENDIX       III

              --_ll_---l--l-----_-----------w---avp-          MADE TO DR. KEELER'S


             In general,   the modifications                  we made to Dr.          Keeler's
     study    were as follows:
             1.     Efficiency
                    ---v--w-------    assumptions:
                    We used Dr. Keeler's               study assumptions regarding                   the
                    efficiencies           that could reasonably be expected in t1
                    long run in a less regulated                   scheduled domestic air.
                    line industry            with among others,           the following          six
                    exceptions.            We modified the seating assumption to
                    introduce        full     galleys      in two aircraft.             The effect
                    on seating was to reduce all-coach                       seating to 153
                    in the B727-200 aircraft                  and increase it to 254 in
                    the DC-8-61 aircraft.                  We also added representative
                    of the two groups of wide-bodied                     aircraft--the           5747
                    aircraft       (for the four-engine              turbofan,        wide-bodied
                    aircraft      group) with all coach seating of 461 seats
                    with full        galley      service and the DC10 aircraft                    (for
                    the three-engine             turbofan      wide-bodied         aircraft      grou!
                    with full-coach             seating of 376 all-coach                seats with
                    full     galley service.             In response to C4B suggestion:
                    with which we concurred,                  we converted these all-coat
                    seating configurations                 to mixtures       of coach and
                    first-class          seating with common costs,                 cargo costs,
                    and offset        cargo revenues allocated                on the basis of
                    floor     space to first           class and coach travel.                  After
                    reviewing        studies and testimony              regarding         aopropriati
                    elasticity        values,       and in view of the empirical                    evi-
                    dence from the Texas and California                        intrastate        mar-
                    kets,     we assumed the -most likely                 long run value of
                    fare elasticity             of demand for air travel                is -1.3.
                    We used Dr. Keeler's              method for calculation                of the
                    longrun pretax corporate                  rate of return on capital
                    except for modifying               one step in the procedure he
                    followed      thereby changing his estimate                     from 12 to
                    10.5 percent.             We assumed each aircraft                was utilized
                    at the highest average annual                    trunkline        utilization
                    level in the period 1969-74.                     We used aircraft            over
                    market distances             to the same extent as actually                     oc-
                    curred for the aircraft                  group they represented.                 We
                    included first-class               and coach-class           travel      to the
                    same     extent that it actually               ocurred.

                                                                              4PPEI'?DIX III
      2   l
                                   Long-Run Cost__-Xodel
              tie modified    Dr. Keeler's   airline     cost model--used                to
              estimate    increased efficiency       fares--to  increase               its-
              validity    oy:
                             a.   Replacing all out the capital         cost portions
                                  of the direct-cost       part of the model with the
                                  direct  cost portion      of a detailed   CAB cost
                                  model 1/ except for the CAB model's flight
                                  eguipment,    depreciation    and flight   equipment
                                  rental  cost.
                             b.   Replacing the indirect          cost portion     of
                                  Dr. Keeler's      cost model --except      for capital
                                  cost accounts --with       the indirect      cost portion
                                  of the DPFI cost model except for depreciation
                                  accounts.      Dr. Keeler's      inclusion    of a pretax
                                  return on capital       based on the annual average
                                  of returns     on capital     for all corporations
                                  for nearly 30 years made use of the DPFI de-
                                  preciation     accounts unnecessary.          Profits,
                                  interest,    depreciation,       and taxes are im-
                                  plicitly    included in Dr. Keeler's          and our
                                  computation     of the capital       cost accounts.

                             C.   Adding the DPFI cost       model's   methods of esti-
                                  mating block-time.
                             d.   Adding the DPFI cost model's method of esti-
                                  mating cargo revenues used to reduce expense
                                  and modifying    it to account for all reported
                                  offsetting   cargo revenue.
     3.        Increased
              -~---~-I__-           market   sample:
              We modified     Dr. Xeeler's sample of 30 high-density
              city-   pair air markets by adding medium- and low-
              density markets and separate Alaska-Hawaii      markets
              to obtain more representative     sets of 48 and 110
              markets.     'The 48 market set was used to provide a

&/Hereafter     referred    to as the "DPFI cost model," this model
  was developed as part of the CAB's Domestic Passenger Fare
  Investigation       (OPFI) in the late 1960s and is now main-
  tained,     updated, and used by the CAB's Bureau of Accounts
  and Statistics       to provide estimates  of the impacts of pos-
  sible fare and cost changes. See "Costing Methodology,
  Version 6," Domestic Fare Structure,        U.S. Civil  Aeronautics
  aoard Docket 21866-7.         Exhibit No. BC-3999.   August 1970.

     APPENDIX III                                                                   APPE.XDIX   I II
81     ,’
                 base for making national  level estimates.                              The 110
                 market set was used to provide markets for                              annual
                 fare comparisons similar  to Dr. Keeler's.
            4.   Estimated unregulated
                 -------a--a--------w--e--                fares:
                 We used each market's         distance,     the competitive    air-
                 line industry      operating     and efficiency     assumptions,
                 the modified      total   cost model, and model inputs ob-
                 tained from C\AB or estimated           by us or Dr.    Keeler to
                 estimate    for each of the years 1969-74, increased
                 efficiency     (cost-based     or unregulated)      air fares
                 (both first     class and coach class)          for each market in
                 the 48 market set and each market in the 113 market
            5.   Compared unregulated                   and published
                 ---------------------------------w------                  fares:
                 For each of the years 1969-74 we compared the esti-
                 mated increased efficiency            fares (both first        class
                 and coach class)      for each market to weighted averages
                 of published    fares for each market.              These weighted
                 averages of published         fares--designated         "actual    annual
                 fares" --were weighted to reflect:               (1) the influence
                 of CAB-approved fare changes by weighting                  fares by
                 revenue passenger-miles          by month by market distance
                 before and after      fare changes and (2) the influence
                 of discount   fares by weighting           first    class and coach
                 fares by dilution       factors     (that reflect       discount     fare
                 types, discount     levels,      and traffic       mix) by market
            6.   Constructed           fare-to-revenue
                 -A-------------------------------w--e------       extension   method
                 We constructed    a fare to revenue extension         method
                 that extended fare estimates        for 52 air market cells
                 to revenue estimates     for all domestic scheduled
                 passenger operations     of the domestic trunk airlines
                 in the 50 states.      The fare-to-revenue       extension
                 method consisted    of the following      steps:
                        a.     Represent total             market by a 42-cell
                               F-:-T------r'--------------T-------w----           distance-
                               aensity matrix for the 48 contiguous
                               ma3~essT---------------------------1---------       State
                                               and a set of 10 market --------
                               ~iasffaTHawaTi~~~~~~---------                 cells for

                               We represented    the entire    air market for the
                               airline  industry    sector in question by a 10
                               cell set representing      Alaska/Hawaii   air mar-
                               kets, plus a 5 x 7, 42-cell        matrix represent-
                               ing air markets in the 48-contiguous        states.

APPENDIX III                                                                  APPENDIX I;1    ,
                        In the 42-cell    matrix,  a six division   axis
                        represented    average passenger density     (aver-
                        age number of passengers per day)
                        and a 7-division     axis represented   market dis-
               b.       Estimate v--w-
                                  fares                   per
                                                          - -----cell:
                        tie estimated     first-   and coach-class      fares for
                        48 city-pair      markets that occupied 48 of the
                        52 cells.      Fares were either       increased effi-
                        ciency fares or oublished          fares,   that were
                        weighted to provide more accurate estimates
                        of actual fares and which were designated
                        llactual annual fares."         City-pair     markets with
                        sufficient     information    for our review were not
                        available    for 4 of the 52 cells.
               C.       --m--w--
                        Estimate             sengers
                                     eEz--~----~~--~~~~           per cell:
                        We allocated       the revenue-passenger-miles
                         (RPMs) for the airline           industry     sector in
                        question to the 42 cells using RPH data ob-
                        tained from the DPFI for 257 city-pair                   mar-
                        kets in 1969 in the contiguous               48 states
                         (these 1969 data represented              two-thirds      of
                        the 1969 air travel          in the 48 states).             In
                        1969 and in subsequent yearsp we updated and
                        expanded to 50-state           travel    using DPFI cost
                        model allocations         of all RP3i by market dis-
                        tances to modify the 1969 data and 10 addi-
                        tional     cells to represent         Alaska/Hawaii-main-
                        land travel       markets.       From these modified RPMs
                        data, we estimated         the number of passengers
                        per cell by dividing          RPM by distance         esti-
                        mates per cell.         These distance         estimates       were
                        calculated      to provide estimates          of passengers
                        by market distances          consistent      with CAB dis-
                        tributions      of RP?/Iwith market distances.
                        City-pair      markets from the sample of 257 mar-
                        kets, and therefore         market RPM as well, were
                        located in only 48 of the 52 cells,                   Conse-
                        quently 4 of the cells were considered                   to pro-
                        duce negligible       air travel       in the review.
               d.       Estimate
                        --------    revenue
                                          A--- per cell:
                        We estimated   the revenue per cell per travel
                        class by multiplying     passengers per cell
                        by each cell's    corresponding   fare estimate.


APPENDIX III                                                              APPENDIX III
                    e.     Estimate   total        revenue:
                         We estimated   total     revenue by summing revenue
                         per cell across the 52 cells        for the two
                         traffic  classes --first      class and coach.
        7.     Estimated     unregulated       and regulated        revenues:
               We estimated      unregulated     revenue from estimated      in-
               creased efficiency        fares (analagous to Dr. Keeler's
               unregulated      or cost-based      fares) and regulated   reve-
               nue from our estimated         actual annual fares.      We
               estimated     these revenues from these fares using the
               fare-to-revenue       extension method described       in six
               above for each of the years 1969-74.
        8.     Modify    revenues:
               Using results   from the previous    step we modified
               revenues to insure that estimated       actual airline
               revenues and estimated    offset  cargo revenues
               equaled CAB-reported    actual airline    revenues and
               offset  cargo revenues.
        9.     Estimated     industry-wide          excess    of actual    over
               estimated     fares:
               For each of the years 1969-74, we calculated           the
               industry-wide      excess of actual over estimated       fares
               due to the domestic trunk airlines'         lack of longrun
               competitive     market characteristics.       The excess
               was calculated      as the percentage   increase   in un-
               regulated     revenue reguired    to equate unregulated
               and regulated      revenue.
       10.     Estimated     impact   of unregulated           fares:
               For the years 1969-74, we used air travel         in RPM,
               air fares expressed as yields       (average fare expressed
               in cents-per-mile),     and a fare elasticity     of demand of
               -1.3 to estimate    (1) the impact on demand for air
               travel   of a fare reduction    from regulated    to unregulated
               levels,    (2) the associated   impact on actual and in-
               duced air traveler     costs and savings,     and (3) airline

APPENDIX       IV                                                           APPENDIX      IV

                     ESTIMATED AND
                     ---I__    --I-- ACTUAL AIRLINE         FARES __--
                                                                  FOR 110

                              --____---     MARKETS IN 1969

                                COACH CLASS FARES
                                             -___-                 FIRST CLASS
                                                                           -_I_- FARES
  MARKETS           (MILES)    FARES    FARES
                                        --I_       UP
                                                   --             FARES    FARES
                                                                           __-        -UP
CORPUS CHRIST1        351    $ 17.60      $ 25.35    44.0%      $ 25.40     $ 35.40     39.3%

WASH., D. C.          319      17.42        21.22    21.8         25.23       29.42     16.6

SPRINGFIELD,    MO 145         12.54        13.94    11.1         17.90       19.40       8.4

DETROIT                94      12.56        11.33   - 9.8         18.07       15.97    -11.7
PHILADELPHIA          215      14.70        18.69    27.1         21.16       26.06     23.2

BOSTON                370      18.83        23.62    25.4         27.35       32.97     20.6

NEW YORK              107      12.39        10.66   -14.0         17.77       15.04    -15.3
NEW YORK              179      13.98        15.65    12.0         20.11       21.87       8.8
PHOENIX              346       18.03        24.25    34.5         26.12       33.59     28.6

SYRACUSE              197      14.28        15.86    11 .,o       20.54       22.15       7.8

ROCHESTER             252      15.81        19.75    24.9         22.85       27.40     19.9
PHILADELPHIA         274       16.34        20.81    27.4         23.64       28.86     22.1

LOUISVILLE           271       16.27        20.16    23.9         23.52       27.96     18.8
LOS ANGELES          227       14.98        19.54    30.5         21.57       27.25     26.3

>.   APPENDIX      IV                                                    APPENDIX   IV

                                      COACH CLASS FARES        -- FIRST CLASS FARES
       MARKETS          (MILES)      FARES    FARES     UP
                                                        --        FARES   FARES     UP
     HOUSTON              222      $ 14.87   $ 19.54   31.4%   $ 21.42   $ 27.25    27.2%

     ST. LOUIS            256        15.93     19.09   19.9      23.03     26.50    15.1

     TAMPA                199        14.58     16.51   13.3      21.03     22.94     9.1

     NEW YORK             289        16.71     20.81   24.5      24.19     29.00    19.9

     CLEVELAND            312        17.25     21.22   23.0      24.97     29.42    17.8

     PITTSBURGH           329        17.89     22.73   27.0      25.96     31.50    21.3

     NEW YORK             191        14.20     15.00    5.6      20.42     20.96     2.6

     WASH., D. C .        215        14.70     16.72   13.7      21.16     23.34    10.3

     DETROIT              238        15.26     18.23   19.5      21.98     25.43    15.7
     MINNEAPOLIS          344        17.98     24.25   34.9      26.05     33.59    28.9
     SAN FRANCISCO        441        19.77     27.68   40.0      28.90     38.60    33.6
     DETROIT              540        22.22     31.84   43.3      32.58     45.71    40.3

     KANSAS CITY          643        25.64     38.72   51.1      37.79     55.78    47.6
     PITTSBURGH           406        19.13     27.28   42.6      27.97     38.04    36.0

APPENDIX       IV                                                    APPENDIX   IV

                                  COACH CLASS FARES           FIRST CLASS FARES
  MARKETS           (MILES)      FARES    FARES
                                          --       -UP       FARES    FARES    -UP
ST. LOUIS             448      $ 19.89   $ 32.44   63.1%   $ 29.07   $ 45.20    55.5%
PHILADELPHIA          749        28.67     40.98   42.9      42.38      58.79   38.7
MEMPHIS               485        20.80     31.13   49.7      30.44      43.96   44.4
PITTSBURGH            496        20.98     30.69   46.3      30.71      43.34   41.1
DETROIT               623        25.01     35.13   40.5      36.84      50.55   37.2
WASH., D. C.          540        22.22     33.86   52.4      32.58      47.79   46.7
CHICAGO               613        24.71     33.46   35.4      36.39      48.19   32.4
CHICAGO               777        29.35     41.18   40.3      43.40      59.07   36.1
WASH., D. C.          406        19.13     25.32   32.4      27.97      35.33   26.3
NEW YORK              410        19.20     26.17   36.3      28.08      36.51   30.0
WASH., D. C.          591        24.05     33.46   39.1      35.38      48.19   36.2
PHILADELPHIA          675        26.66     36.16   35.7      39.35      52.00   32.1
SEATTLE               671        26.10     37.64   44.2      38.46      54.08   40.6
NEW YORK              489        20.87     29.21   40.0      30.55      41.26   35.1

APPENDIX      IV                                                     APPENDIX   IV

                                 COACH CLASS FARES            FIRST CLASS FARES
  MARKETS          (MILES)      FARES    FARES     UP
                                                  --         FARES    FARES
                                                                      --        UP
CHICAGO              721      $ 27.98   $ 38.67    38.2%   $ 41.35    $ 55.54   34.3%

NEW YORK             755        28.81     42.89    48.9      42.60      61.49   44.3

MINNEAPOLIS          850        31.31     49.41    57.8      46.81      71.39   52.5
MEMPHIS              880        32.03     49.60    54.8      47.90      71.66   49.6

SAN ANTONIO         1041        36.16     55.78    54.2      54.15      80.47   48.6

MIAMI                946        33.64     51.45    52.9      50.32      74.30   47.7

TAMPA               1183        41.37     60.68    46.7      62.09      89.06   43.4

MIAMI               1194        41.70    62.54    50.0       62.59      91.76   46.6

NEW ORLEANS          831        30.20     46.43    53.7      45.05      66.48   47.6

BOSTON               946        33.64     50.36    49.7      50.32      72.75   44.6

MIAMI               1083        37.42     58.71    56.9      56.07      84.65    51.0

HOUSTON              932        33.30     52.91    58.9      49.80      76.39    53.4

TAMPA               1006        35.14     54.76    55.8      52.58      79.02    50.3

WASH., D. C.        1161        40.66     60.86    49.7      61.00      88.78    45.5

APPENDIX      IV                                                    APPENDIX   IV

                                 COACH CLASS FARES              FIRST CLASS FARES
                                                          ---___---___          ---
                                          ACTUAL MARK     ESTIMATED ACTUAL MARK
  MARKETS          (MILES)
                    --          FARES     FARES   -UP         FARES
                                                              ---       FARES
                                                                        -__       UP
CHICAGO               860     $ 31.55   $ 43.78   38.8%   $ 47.17    $ 63.21   34.0%
SEATTLE               959       33.95     51.83   52.6      50.79      74.04   47.3
MIAMI                1188       41.55     61.68   48.4      62.36      89.96   44.2
NEW YORK             1093       37.71     59.18   56.9      56.52      85.83   51.9
PdOENIX              1276       42.69     67.65   58.5      64.78      99.18   53.1
MCNNEAPOLIS          1300       43.45     70.52   62.3      65.96     103.34   56.7
SriN DIEGO           1532       47.32     73.87   56.1      71.64     110.07   53.6
SEATTLE              1398       45.22     72.20   59.7      68.57     106.43   55.2
SL'OKANE, WASH. 1508            46.92     75.54   61.0      71.06     112.54   58.4
LAS VEGAS            1229       41.21     65.18   58.2      62.48      95.02   52.1
SAN FRANCISCO       1587        48.28     13.87   53.0      73.04     110.07   50.7
MIAMI               1258        42.14    66.22    57.1      63.92      97.10   51.9
NEW YORK            1363        44.60    67.78    52.0      67.66      99.97   47.8
PHOENIX             1445        46.08     72.05   56.4      69.83     106.21   52.1

APPENDIX      IV                                                    APPENDIX    IV

                                 COACH CLASS
                                        ---   FARES          FIRST CLASS FARES
    MARKETS        (MILES)      FARES     FARES     -UP     FARES    FARES    -UP
  LOS ANGELES         1248    $ 41.84   $ 60.82   45.4%   $ 63.46     $ 88.72        39.8%

  SAN FRANCISCO       1493      46.62     72.89   56.3      70.61      108.62        53.8

  LAS VEGAS           1521      47.12     75.73   60.7      71.34      112.81        58.1

  SAN FRANCISCO       1915      56.87     93.56   64.5      86.45      140.10        62.1

  PORTLAND,    ORE.   1637      47.48     83.44   75.7      71.76      124.20        73.1

  SAN FRANCISCO       1647      47.77     81.94   71.5      72.20      121.99        69.0

  SEATTLE             1681      48.72     86.97   78.5      73.67      129.57        75.9

  NEW ORLEANS         1658      48.08     81.22   68.9      72.68      121.08        66.6

  ST. LOUIS           1710      49.62     80.53   62.3      75.08      120.05        59.9
  SAN DIEGO           1729      49.52     84.78   71.2      74.85      126.34        68.8
 .NEW YORK            1627      47.31     77.25   63.3      71.50      115.06        60.9
  PORTLAND,    ORE.   1748      50.62     84.78   67.5      76.62      126.34        64.9
  LOS ANGELES         1934      57.32     93.12   62.5      87.14      139.45        60.0
  SEATTLE             1730      50.14     84.78   69.1      75.88      126.34        66.5

APPENDIX      IV                                                    APPENDIX   IV

                                 COACH CLASS FARES           FIRST CLASS FARES
   MARKETS         (MILES)
                    I-          FARES    FARES    -UP       FARES    FARES     UP
 LOS ANGELES         1740     $ 53.84   $ 84.78   57.5%   $ 82.05    $126.34    54.0%

 SAN FRANCISCO       1853       55.13     84.26   52.8      83.75     127.21    51.9

 TAMPA               2153       60.84    102.35   68.2      89.83     156.02    73.7
 PITTSBURGH          2124       60.17    101.29   68.4      88.83     153.47    72.8

 LOS ANGELES         2046       58.36     97.92   67.8      86.15     147.54    71.3

 SAN FRANCISCO       2086       59.29     95.52   61.1      87.53     144.82    65.5

 SAN FRANCISCO       2141       60.56    101.73   68.0      89.42     154.12    72.4

 WASH., D. C.        2288       63.96    109.10   70.6      94.48     166.94    76.7

 SAN FRANCISCO       2589       72.29    123.53   70.9     110.50     189.67    71.6

 SEATTLE             2408       67.08    115.33   71.9     102.38     177.98    73.8

 SAN FRANCISCO       2526       70.37    112.30   59.6     107.49     177.48    65.1

 SAN FRANCISCO       2703       74.80    120.75   61.4     114.32     185.44    62.2
 PHILADELPHIA        2396       66.18    112.31   69.7      97.37     174.14    78.8
 LOS ANGELES         2600       72.45    120.75   66.7     110.73     187.09    69.0

                                                                        APPENDIX    IV

                                       COACH CLASS FARES         FIRST CLASS FARES
     MARKETS           (MILES)      FARES
                                    --         FARES     UP
                                                         -      FARES    FARES
                                                                         --        -UP
   SAN FRANCISCO         2574     $ 72.14   $115.33   59.9%   $110.30     $178.78        62.1%
   NEW YORK              2453       69.03    115.33   67.1     105.49      178.78        69.5
   SEATTLE               1448       44.64     86.10   92.8      67.11      124.00        84.8
   HAWAII                2408       63.40     99.66   57.2      94.86      150.00        58.1
   HAWAII                2573       67.07    99.66    48.6     100.38      150.00        49.4
   HAWAII                2603       65.65    104.07   58.5      97.91      140.00        43.0
   HAWAII                2677       67.28    104.07   54.7     100.35      140.00        39.5
   CHICAGO               2846       89.38    170.30   90.5     136.59      240.00        75.7
   NEW YORK              3278      119.91    226.59   89.0     181.19      284.00        56.7
   HAWAII                3785      101.81    158.08   55.3     153.88      234.00        52.1
   HAWAII                4248      107.67    188.61   75.2     161.96      261.00        61.2
   HAWAII                4974      164.99    235.58   42.8     254.95      305.00        19.6

 APPENDIX      V                                                                    4PPENDIX    V

                      ESTIMATED AND ACTUAL AIRLINE
                      ~---------~~--_~-_-~~~~-_~-~~__~--~_~         FARES FOR 110

                                 CITY-PAIR        MARKETS IN 1970

                               --I_ COACH CLASS
                                          ------m--e FARES                   FIRST CLASS FARES
  ---__            (MILES)
                    -__            FARES
                                   ----       FARES
                                              --I_        UP
                                                          --               FARES
                                                                           ----           FARES
                                                                                          ---   UP
CORPUS CHRIST1        351      $ 19.76      $ 25.80       30.6%         $ 29.33      $ 35.78   22.0%

WASH., D. C.          319         18.99        24.36      28.3            28.16        33.51   19.0
SPRINGFIELD,       MO 145         14.11        15.68      11.1            20.51        21.77     6.1

DETROIT                94         12.88        13.19          2.4         18.61        18.31   -1.6

PHILADELPHIA          215         16.07       19.00       18.2            23.58        26.26   11.4
BOSTON                370         20.50        26.64      30.0            30.52        36.92   21.0
NEW YORK              107         13.21                                   19.12        18.31   -4.2
NEW YORK              179         15.19       17.34      14.2             22.22        23.97    7.9
PHOENIX               346         19.64       26.01      32.4             29.16        35.78   22.7

SYRACUSE              197         15.62       18.17      16.3             22.89        25.11    9.7
ROCHESTER             252        17.29        20.65      19.5             25.51       28.41    11.4
PHILADELPHIA          274        17.85        22.70      27.2            26.39        31.23    18.3
LOUISVILLE            277        17.91        22.70      26.7            26.49        31.23    17.9

LOS ANGELES           227        16.39        19.83      20.9            24.08        27.40    13.8

     APPENDIX V                                                       APPENDIX V

                                     COACH CLASS FARES
                               ~------~--                         FIRST CLASS FARES
      --            (MILES)       FARES
                                  --         FARES
                                             ---       UP
                                                       -        FARES       FARES   UP
    HOUSTON           222      $ 16.24   $ 19.83    22.1%   $ 23.84   $ 27.40    14.9%

    ST. LOUIS         256        17.39     20.65    18.8      25.68     28.41    10.6

    TAMPA             199        15.85     18.17    14.6      23.28     25.11     7.9

    NEW YORK          289        18.23     22.70    24.5      26.98     31.23    15.7

    CLEVELAND         312        18.83     24.36    29.4      27.90     33.51    20.1

    PITTSBURGH        329        19.24     25.19    30.9      28.53     34.64    21.4

    NEW YORK          191        15.48     17.34    12.0      22.66     23.97     5.8

    WASH., D. C .     215        16.07     19.00    18.2      23.58     26.26    11.4

    DETROIT           238        16.67     19.83    18.9      24.50     27.40    11.8
    MINNEAPOLIS       344        19.59     26.01    32.8      29.08     35.78    23.0

    SAN FRANCISCO     441        22.03     29.92    35.8      32.95     41.48    25.9
    DETROIT           540        24.72     34.78    40.7      37.15     48.86    31.5

    KANSAS CITY       643        27.69     40.26    45.4      41.77     56.82    36.0
    PITTSBURGH        406        21.20     28.28    33.4      31.68     39.20    23.7

 APPENDIX      V                                                     4PPENDIX      V

                                         CLASS FARES      __--_IFIRST CLASS
                                                                       -__----- FARES
  MARKETS          (MILES)
                    --          FARES
                                -__          FARES
                                             ----   -UP      FARES       FARES
                                                                        --            UP
ST. LOUIS            448      $ 22.21   $ 30.74   38.5%   $ 33.22    $ 42.62      28.3%
PHILADELPHIA         74Y        30.45     44.56   46.3      46.05       63.05     36.9
MEMPHIS              485        23.30     32.35   38.9      34.95       45.44     30.0
PITTSBURGH           456        23.58     33.16   40.7      35.38       46.58     31.6
DETROIT              623        27.15     38.65   42.3      40.93       54.55     33.3
WASH., D. C.         540        24.72     35.59   44.0      37.15       49.99     34.6
CHICAGO              613        26.89     37.03   37.7      40.54       52.27     28.9
CHICAGO              777        31.20     45.25   45.0     .47.21       64.15     35.9
WASH., D. C.        406         21.20     28.28   33.4      31.68       39.20     23.7
NEW YORK            410         21.28     29.10   36.8      31.80       40.34     26.9
WASH., D. C.        591         26.34     37.03   40.6      39.67       52.27     31.8
PHILADELPHIA        675         28.50     40.15   40.9      43.03       56.81     32.0
SEATTLE             671         28.40     40.96   44.2      42.87       57.95     35.2
NEW YORK            489         23.42     32.35   38.1      35.15      45.44      29.3

     APPENDIX V                                                      APPENDIX V
-.     .'

                                    COACH CLASS FARES         FIRST CLASS FARES
       MARKETS      (MILES)
                     --          FARES
                                 --         FARES
                                            --       -UP     FARES    FARES
                                                                      --        UP
     CHICAGO          721      $ 29.70   $ 42.95   44.6%   $ 44.89   $ 60.77   35.4%

     NEW YORK         755        30.60     44.44   45.2      46.29     63.01   36.1

     MINNEAPOLIS      850        33.32     49.26   47.8      50.58     69.84   38.1

     MEMPHIS          880        34.07     49.92   46.5      51.74     70.97   37.2

     SAN ANTONIO     1041        38.34     57.35   49.6      58.36     81.74   40.1

     MIAMI            946        35.81     52.69   47.1      54.44     74.90   37.6

     TAMPA           1183        43.01     63.35   47.3      65.64     91.39   39.2

     MIAMI           1194        43.29     63.35   46.3      66.07     91.39   38.3

     NEW ORLEANS      831        32.10     48.45   50.9      48.57     68.70   41.4

     BOSTON           946        35.81     53.49   49.4      54.44     76.04   39.7

     MIAMI           1083        39.44     58.59   48.5      60.08     84.02   39.9

     HOUSTON          932        35.46     53.49   50.8      53.89     76.04   41.1

     TAMPA           1006        37.37     55.75   49.2      56.85     79.46   39.8
     WASH., D. C.    1161        42.40     62.56   47.6      64.69     90.26   39.5

 APPENDIX     V                                                  APPENDIX     V

                               COACH CLASS FARES
                                             ---          FIRST CLASS FARES
                 --          FARES
                             ~-        FARES
                                       --        UP
                                                 --      FARES    FARES    UP
CHICAGO            860     $ 33.57   $ 49.12   46.3%   $ 50.96   $ 69.83    37.0%

SEATTLE            959       36.14     54.15   49.8      54.95     77.18    40.5

MIAMI             1188       43.13     63.35   46.9      65.83     91.39    38.8

NEW YORK          1093       39.70     59.38   49.6      60.47     85.16    40.8

PHOENIX           1276       43.95     66.09   50.4      67.04     95.95    43.1

MINNEAPOLIS       1300       44.59     67.66   51.7      68.04     98.22    44.3

SAN DIEGO         1532       50.69     75.39   48.7      77.64   111.30     43.3
SEATTLE           1398       47.30     77.13   63.1      72.26   111.28     54.0

SPOKANE, WASH. 1508          50.03     74.62   49.1      76.61   110.16     43.8

LAS VEGAS         1229       42.69     64.92   52.1      65.08     93.67    43.9
SAN FRANCISCO     1587       52.25     75.39   44.3      80.09   111.30     39.0
MIAMI             1258       43.45     65.31   50.3      66.27     94.81    43.1
NEW YORK          1363       46.37     70.36   51.7      70.82   102.77     45.1
PHOENIX           1445      48.58      73.07   50.4      74.26   106.73     43.7

   APPENDIX V                                                        APPENDIX V
II   .I

                                 COACH CLASS FARES           FIRST CLASS FARES
    MARKETS       (MILES)      FARES     FARES    -UP       FARES    FARES
                                                                     --        UP
  LOS ANGELES         1248   $ 43.21   $ 64.92    50.3%   $ 65.89    $ 93.67   42.2%

  SAN FRANCISCO       1493     49.54     73.85    49.1      75.84     109.03   43.8

  LAS VEGAS           1521     50.35     75.39    49.7      77.12     111.30   44.3

  SAN FRANCISCO       1915     60.00     91.43    52.4      92.87     135.15   45.5

  PORTLAND,    ORE.   1637     51.35     79.23    54.3      79.20     116.97   47.7

  SAN FRANCISCO       1647     51.59     78.42    52.0      79.58     115.77   45.5

  SEATTLE             1681     52.47     81.83    55.9      80.97     120.95   49.4

  NEW ORLEANS         1658     51.92     81.06    56.1      80.10     119.82   49.6

  ST. LOUIS           1710     53.24     83.36    56.6      82.17     123.22   50.0

  SAN DIEGO           1729     53.70     85.67    59.5      82.89     126.63   52.8

  NEW YORK            1627     51.11     79.23    55.0      78.82     116.97   48.4

  PORTLAND,    ORE.   1748     54.16     85.67    58.2      83.61     126.63   51.4

  LOS ANGELES         1934     60.61      92.20   52.1       93.85    136.29   45.2

  SEATTLE             1730     53.72      85.67   59.5       82.93    126.63   52.7

APPENDIX       V                                                    APPENDIX    V
                                                                                      ‘.   .

                                   COACH CLASS FARES         FIRST CLASS FARES--
  MARKETS          (MILES)
                    --          FARES
                                --         FARES
                                           --       -UP     FARES    FARES
                                                                     --       -UP
LOS ANGELES           1740    $ 53.97   $ 85.67   58.7%   $ 83.31    126.63    52.0%

SAN FRANCISCO         1853      58.33     85.67   46.9      90.26    126.63    40.3

TAMPA                 2153      66.29     99.87   50.7     101.70    150.51    48.0

PITTSBURGH            2124      65.33    101.01   54.6     100.16    151.27    51.0

LOS ANGELES           2046      62.93     96.80   53.8      96.37    144.07    49.5

SAN FRANCISCO         2086      64.07     94.49   47.5      98.16    141.50    44.2

SAN FRANCISCO         2141      65.83     99.87   51.7     100.95    149.56    48.2
WASH., D. C.          2288      71.01    107.94   52.0     109.21    163.38    49.6
SAN FRANCISCO         2589      78.48    116.78   48.8     121.53    179.11    47.4
SEATTLE               2408      72.81    113.70   56.2     112.50    172.86    53.7
SAN FRANCISCO         2526      78.29    111.40   42.3     121.38    170.10    40.1
SAN FRANCISCO         2703      al.41    118.31   45.3     126.11    181.47    43.9
PHILADELPHIA          2396      75.06    111.40   48.4     115.53    169.35    46.6
LOS ANGELES           2600      78.76   118.31    50.2     121.97    181.47    48.8

    APPENDIX     V                                                              APPENDIX   V

                                   COACH CLASS FARES                    FIRST CLASS FARES
      MARKETS        (MILES)      FARES    FARES
                                           _I-      -UP                FARES    FARES
                                                                                --        UP
    SAN FRANCISCO       2574    $     78.20   $113.70        45.4%   $121.11    $174.40    44.0%

    NEW YORK            2453          14.35    113.70        52.9     114.98     173.62    51.0

    HAWAII              2408          69.69     92.94        33.4     107.30     150.00    39.8

    HAWAII              2573          73.64     92.94        26.2     113.45     150.00    32.2

    HAWAII              2603          69.55     92.94        33.6     106.51     140.00    31.4

    HAWAII              2677          72.60     92.94        28.0     111.34     140.00    25.1

    HAWAII              3785        110.53     178.81        61.8     174.64     257.75    47.6

    HAWAII              4248        114.06     158.47        38.9     178.12     271.92    52.7

    HAWAII              4974        174.48     203.43        16.6     277.96     310.00    11.5

    ANCHORAGE           1448          48.17      95.71       98.7       74.25    124.37    67.5

    CHICAGO             2846          94.75    171.24        80.7     l-48.99     248.00   66.5

    NEW YORK            3278        114.35     185.44        62.2     177.09      264.00   49.1

                                                                               APPENDIX      VI
j,.   APPENDIX       VI

                          ESTIMATED AND
                          ----            ACTUAL AIRLINE
                                     __--__--__-_____---       FARES FOR 110

                                   CITY-PAIR MARKETS IN 1971

                                  -- COACH CLASS
                                               -----FARES                FIRST CLASS FARES
        -----         (MILES)
                       ----          FARES
                                    --       FARES       UP
                                                         -             FARES     FARES
                                                                                 ---       UP
      CORPUS CHRIST1      351     $ 19.62   $ 24.31     23.9%      $ 27.97     $ 33.72     20.6%
      WASH., D. C.        319       19.39     25.82     33.2         27.67       35.52     28.4

      KAN. CITY, MO.
        MO.          145            16.45     16.48        0.2       23.47       22.88     -2.5

      GETROIT              94       15.80     14.09    -10.8         22.60       19.57    -13.4

      PHILADELPHIA        215       16.89     19.67     16.5         24.00       27.19     13.3

      BOSTON              370       21.00     27.99     33.3         30.09       38.80     28.9

      NEW YORK            107       15.45     14.09    - 8.8         22.00       19.57    -11.1

      NEW YORK            179       16.40     18.08     10.3         23.31       24.99       7.2

      EL PASO
      PHOENIX             346       20.09     27.42     36.5         28.72       37.71     31.3

      NEW YORK
      SYRACUSE            197       16.53     19.45     17.7         23.48        26.89    14.5

      NEW YORK
      ROCHESTER           252       17.Y2     21.85     21.9         25.53       30.05     17.7

      PHILADELPHIA        274       18.37     24.23     31.9         26.18        33.33     17.3

      LOUISVILLE          277       18.43     24.21     31.3         26.27        33.30     26.8

      LOS ANGELES         257       17.96     21.04     17.2          25.59       29.09     13.7

    APPENDIX       VI                                                       APPEMDIX VI
c     L'

                                      COACH CLASS FARES
                                                   -~-                 FIRST CLASS FARES
      MARKETS           (MILES)      FARES    FARES
                                              ----      UP
                                                        --          FARES
                                                                    --         FARES
                                                                               ----      UP
    HOUSTON               222      $ 17.03    $ 21.62    27.0%   $ 24.20    $ 29.88    23.5%
    ST. LOUIS             256        18.01      21.85    21.3      25.66      30.05    17.1
    TAMPA                 199        16.83      18.88    12.1      23.94      26.09      9.0

    NEW YORK              289        18.68      24.23    29.8      26.63      33.33    25.2

    CLEVELAND             312        19.22      25.82    34.4      27.42      35.52    29.5

    PITTSBURGH            329        19.65      27.20    38.4      28.07      37.41    33.3
    NEW YORK              191        16.49      18.66    13.1      23.43      25.79    10.1

    WASH., D. C.          215        16.89      19.67    16.5      24.00      27.19    13.3

    DETROIT               238        17.36      21.04    21.2      24.68      29.09     17.9

    MINNEAPOLIS           344        20.04      27.42    36.8      28.64      37.71     31.7

    SAN FRANCISCO         441        22.70      31.15    37.3      32.70      43.18     32.1

    DETROIT               540        25.52      36.78    44.1      36.94       51.67    39.9

      YO.                 643        28.98      42.03    45.1      42.16     59.33     40.7

    PITTSBURGH            406         21.87      29.57   35.2       31.48      40.99    30.2

APPENDIX       VI                                                    APPENDIX    VI
                                                                                       '.   "

                                  COACH CLASS FARES
                                                  -~       ----- FIRST CLASS FARES
  MARKETS           (MILES)      FARES    FARES
                                          ---       UP
                                                    -          FARES
                                                               --        FARES    UP
ST. LOUIS              448     $ 22.86   $ 31.94   39.8%   $ 32.93   $ 44.28    34.4%

PHILADELPHIA           749       32.11     46.56   45.0      46.87     65.88    40.6
MEMPHIS                485       23.99     34.45   43.6      34.65     48.38    39.6
PITTSBURGH             496       24.23     35.22   45.4      34.99     49.48    41.4
DETROIT                623       28.35     40.48   42.8      41.22     57.14    38.6
WASH., D. C.           540       25.52     37.56   47.2      36.94     52.76    42.8
CHICAGO                613       28.03     38.93   38.9      40.73     54.95    34.9
CHICAGO                717       32.90     47.78   45.2      48.05     67.74    41.0
WASH., D. C.           406       21.87     29.57   35.2      31.48     40.99    30.2
NEW YORK               410       21.98     30.36   38.1      31.64     42.09    33.0
WASH., D. C.           591       27.36    38.93    42.3      39.72    54.95     38.3
PHILADELPHIA           675       29.96    41.92    39.9     43.64     59.31     35.9
SEATTLE                671       29.85    42.70    43.0     43.48     60.41     38.9
NEW YORK               489      24.07     34.45    43.1     34.77     48.38     39.2

    APPENDIX       VI                                                APPENDIX     VI
*    .'

                                  COACH CLASS FARES
                                               ----        - FIRST CLASS----
      MARKETS       (MILES)      FARES    FARES     -UP      FARES   FARES      UP
    CHICAGO              721   $ 31.32   $ 45.02   43.7%   $ 45.69   $ 63.69    39.4%

    NEW YORK             755     32.27     46.44   43.9      47.11     65.84    39.8

    MINNEAPOLIS          850     35.05     51.63   47.3      51.29     73.21    42.7

    MEMPHIS              880     35.88     52.14   45.3      52.54     74.11    41.1

    SAN ANTONIO         1041     40.36     59.67   47.8      59.26     85.05    43.5

    MIAMI                946     37.66     55.21   46.6      55.21     78.49    42.2

    TAMPA               1183     45.51     66.39   45.9      67.19     95.78    42.6

    MIAMI               1194     45.86     66.39   44.7      67.72     95.78    41.4

    NEW ORLEANS          831     33.78     50.86   50.6      49.29     72.12    46.3

    BOSTON               946     37.66     55.98   48.6      55.21     79.58    44.1

    MIAMI               1084     41.65     62.03   49.0      61.20     88.00    43.8

    HOUSTON              945     37.64     55.98   48.7      55.17     79.58    44.3

    TAMPA               1006     41.47     58.14   40.2      61.14     82.86    35.5
    WASH., D. C.        1161     44.87     65.63   46.3      66.22     94.69    43.0

APPENDIX      VI                                                       APPENDIX    VI

                                 COACH   CLASS FARES
                                     ---e-w-                    FIRST CLASS FARES
  MARKETS          (MILES)
                    --          FARES      FARES
                                           ---      UP
                                                    --         FARES    FARES    UP
CHICAGO               860     $ 35.31   $ 51.37   45.5%      $ 51.68   $ 73.02    41.3%

SEATTLE               959       38.01     56.60   48.9         55.72     80.68    44.8

MIAMI                1188       45.70     66.39   45.3         67.48     95.78    41.9

NEW YORK             1093       41.93     62.13   48.1         61.64     89.09    44.5

PHOENIX              1276       47.18     68.99   46.2         71.60    100.16    39.9

MINNEAPOLIS          1300       47.93     70.50   47.1         72.77    102.34    40.6

SAN DIEGO            1532       53.88     78.91   46.5         81.95    116.50    42.2

SEATTLE             1398        50.59     79.59   57.3         76.86    116.25    51.2

SPOKANE, WASH. 1508             53.24     78.17   46.8         80.98    115.41    42.5

LAS VEGAS           1229       45.71     67.90    48.4         69.41     97.97    41.1

SAN FRANCISCO       1587        55.40     78.91   42.4         84.30   116.50     38.2
MIAMI               1258       46.65     68.24    46.3         70.77     99.06    40.0
NEW YORK            1363       49.66     73.60    48.2         75.43   107.50     42.5

PHOENIX             1445       51.90     76.59    47.6         78.88   111.88     41.8

APPENDIX      VI                                                     APPENDIX   VI

                                   COACH CLASS
                                          ----- FARES          FIRST CLASS FARES
  --               (MILES)
                    --          FARES
                                --          FARES    -UP     FARES
                                                             ----      FARES
                                                                       __-       UP
LOS ANGELES           1248    $ 46.33   $ 67.90    46.6%   $ 70.27   $ 97.97    39.4%

SAN FRANCISCO         1493      52.84     77.43    46.5      80.37    114.32    42.2
LAS VEGAS             1521      53.59     78.91    47.3      81.50    116.50    42.9
SAN FRANCISCO         1915      64.89     95.71    47.5      98.37    141.47    43.8
PORTLAND, ORE.        1639      55.67     82.62    48.4      83.94    121.97    45.3
SAN FRANCISCO         1647      55.88     85.40    52.8      84.26    126.08    49.6
SEATTLE               1681      56.84     86.03    51.4      85.72    127.17    48.4
NEW ORLEANS           1658      56.24     85.29    51.7      84.80    126.08    48.7
ST. LOUIS             1710      57.66     87.51    51.8      86.98    129.36    48.7
SAN DIEGO             1729      58.16     89.73    54.3      87.74    132.64    51.2
NEW YORK              1627      55.36     82.62    49.2      83.46    121.97    46.1
PORTLAND,    ORE.     1748      58.65     91.41    55.8      88.49    135.12    52.7
LOS ANGELES           1934      65.56     96.45    47.1      99.41    142.56    43.4
SEATTLE               1730      58.19     89.73    54.2      87.78    132.64    51.1

 APPENDIX      VI                                                       APPENDIX      VI

                                COACH CLASS ---------
                           __-----------      FARES             FIRST CLASS FARES
  ------        (MILES)
                 ----         FARES
                              -.---      FARES
                                         ----         UP
                                                      --      FARES
                                                              -----            FARES
                                                                               -----   UP
LOS ANGELES         1740   $ 58.45    $ 89.73     53.5%    $ 88.17     $132.64      50.48

SAN FRANCISCO       la54     63.12      89.79     42.2       95.66      132.72      38.7

TAMPA               2153     70.34     104.24     48.2      106.16      157.08      48.0

PITTSBURGH          2124     69.40     105.72     52.3      104.71      158.32      51.2

LOS ANGELES         2046     66.86     101.45     51.7      100.77      150.98      49.8

SAN FRANCISCO       2086     68.15      99.06     45.4      102.77      148.35      44.3

SAN FRANCISCO       2141     69.94     104.24     49.0      105.53      156.10      47.9

WASH., D. C.        2288     74.25     112.92     52.1      112.15      170.92      52.4

SAN FRANCISCO       2589     82.01     121.85     48.6      125.13      186.89      49.4

SEATTLE             2408     76.77     118.89     54.9      117.02      180.74      54.5

SAN FRANCISCO       2526     81.37     116.67     43.4      124.57      178.15      43.0

SAN FRANCISCG       2703     85.73     123.94     44.6      130.95      190.10      45.2

PHILADELPHIA        2396     76.85     116.67      51.8     116.08      177.37      52.8

LOS ANGELES         2600     82.41     123.88      50.3     125.77      190.00      51.1

APPENDIX VI                                                       APPENDIX VI

                               COACH CLASS FARES           FIRST CLASS FARES
  MARKETS        (MILES)      FARES    FARES    -UP       FARES    FARES    -UP
SAN FRANCISCO        2574   $ 81.60   $118.89   45.7%   $124.52   $182.35   46.4%

NEW YORK             2453     77.96    118.89   52.5     118.86    181.54   52.7

ANCHORAGE            1448     52.17     35.56   83.2      78.62    130.00   65.4
HAWAII               2408     72.06    100.26   39.1     107.85    176.46   63.6

HAWAII               2573     76.01    100.26   31.9     113.79    176.46   55.1

HAWAII               2603     75.34    101.14   34.2     113.17    176.46   55.9
HAWAII               2677     76.52     99.73   30.3     114.54    176.46   54.1

CHICAGO              2846    104.04    171.24   64.6     159.74    248.00   55.3

NEW YORK             3278    113.32    185.44   63.6     171.54    264.00   53.9
HAWAII               3785    113.79    151.85   33.5     170.60    264.00   54.7

HAWAII               4248    121.87    158.47   30.0     182.81    299.00   63.6

HAWAII               4974    192.21    206.72    7.5     298.35    343.00   15.0

 APPENDIX      VII                                                           APPENDIX VII

                      ESTIMATED AND ACTUAL AIRLINE           FARES FOR 110

                              _--_I_----     MARKETS IN 1972

                                  COACH CLASS
                                         ---   FARES                 FIRST   CLASS FARES
 CITY-PAIR      DISTANCE     ESTIMATED     ACTUAL MARK           -mD           ACTUAL MA=
  --            (MILES)        FARES
                               --          FARES
                                           --       -UP            FARES       FARES
                                                                               --       -UP
BURLINGTON           182     $ 17.60       $ 20.13    14.4%      $ 25.76     $ 27.82     8.0%

WASH., D. C.         319       20.87         26.50    26.9         30.82       36.45    18.3

SPRINGFIELD          145       16.94         16.93    -0.1         24.17       23.52    -5.1

DETROIT               94       16.95         14.55   -14.2         24.92       20.20   -18.9

PHILADELPHIA         215       18.36         20.13      9.6        26.93       27.82      3.3

BOSTON               370       23.16         28.66    23.8         34.47       39.73    15.2

NEW YORK             107       17.03         14.55   -14.6         25.01       20.20   -19.2

NEW YORK             179       17.90         18.54      3.6        26.27       25.63   - 2.5

PHOENIX              346       21.59         28.09    30.1         31.93       38.64    21.0

SYRACUSE             197       18.03         19.89    10.4         26.42       27.49      4.0

ROCHESTER            252       19.43         22.28    14.7         28.60       30.64      7.1

PHILADELPHIA         274       19.87         24.91    25.3         29.28       34.26    17.0

LOUISVILLE           277       19.94         24.91    24.9         29.38       34.26    16.6

LOS ANGELES          227       18.64         21.48    15.3         27.36       29.69      8.5

APPENDIX       VII                                                APPENDIX    VII

                               COACH CLASS FARES           FIRST CLASS FARES
  MARKETS        (MILES)      FARES    FARES
                                       --        UP
                                                 -        FARES    FARES     UP
HOUSTON              222    $ 18.52   $ 22.04   19.0%   $ 27.17   $ 30.46    12.1%

ST. LOUIS            256      19.52     22.28   14.1      28.75     30.64     6.6
TAMPA                199      18.34     19.34    5.4      26.95     26.72    -0.8

NEW YORK             289      20.16     24.91   23.5      29.72     34.26    15.3

CLEVELAND            312      20.69     26.50   28.1      30.52     36.45    19.4
PITTSBURGH           329      21.15     27.85   31.7      31.24     38.30    22.6
NEW YORK             191      17.99     19.10    6.1      26.38     26.39     0.0
WASH., D. C.         215      18.36     20.13    9.6      26.93     27.82     3.3
DETROIT              238      18.85     21.48   14.0      27.68     29.69     7.3
MINNEAPOLIS          344      21.53     28.09   30.5      31.84     38.64    21.3
SAN FRANCISCO        441      24.20     31.82   31.5      36.28     44.11    21.6
DETROIT              540      26.78     37.80   41.1      40.30     52.87    31.2
KANSAS CITY          643      30.18     43.12   42.9      45.67     60.86    33.2
PITTSBURGH           406      23.34     30.24   29.6      34.93     41.92    20.0

                                                                     APPENDIX VII

                                  COACH CLASS FARES           FIRST CLASS FARES
      MARKETS       (MILES)
                     --          FARES    FARES    UP
                                                   -         FARES    FARES    -UP
    ST. LOUIS         448      $ 24.36   $ 32.61   33.9%   $ 36.52   $ 45.20   23.8%
    PHILADELPHIA      749        33.26     47.64   43.2      50.54     67.41   33.4

    MEMPHIS           485        25.25     35.45   40.4      37.89     49.58   30.8

    PITTSBURGH        496        25.50     36.23   42.1      38.30     50.68   32.3

    DETROIT           623        29.58     41.33   39.7      44.74     58.34   30.4
    WASH., D. C.      540        26.78     38.58   44.1      40.30     53.96   33.9

    CHICAGO           613        29.26     39.78   36.0      44.22     56.15   27.0

    CHICAGO           777        34.01     48.83   43.6      51.71     69.23   33.9

    WASH., D. C.      406        23.36     30.00   28.4      34.97     41.59   18.9

    NEW YORK          410        23.44     31.03   32.4      35.10     43.01   22.6

    WASH., D. C.      591        28.59     39.78   39.1      43.17     56.15   30.1

    PHILADELPHIA      675        31.19     43.00   37.9      47.28     60.84   28.7

    SEATTLE           671        31.05     43.78   41.0      47.05     61.94   31.6
    NEW YORK          489        25.35     35.45   39.8      38.06     49.58   30.3

     APPENDIX VII                                                          APPENDIX VII
.      *'

                                   COACH CLASS FARES
                              --_-_--------------                    FIRST CLASS FARES
                                                                ___--__---_----       -----
      --~          (MILES)
                    ----         FARES            FARES
                                                  ----  -UP        FARES
                                                                   -----        FARES
                                                                                -__-      UP
    CHICAGO           721     $ 32.51    $ 46.10        41.8%   $ 49.37    $   65.22   32.1%

    NEW YORK          755       33.42      47.52        42.2      50.79        67.37   32.7

    MINNEAPOLIS       850       36.24      52.92        46.0      54.81        75.03   36.9

    MElYPHIS          880       37.03      53.52        44.5      56.03        76.08   35.8

    SAN ANTONIO     1041        41.36      61.27        48.1      62.78        87.33   39.1

    MIAMI             946       38.75      56.60        46.1      58.70        80.46   37.1
    TAlvlPA         1183        46.14      67.91        47.2      70.45        97.98   39.1

    MIAMI           1194        46.43      67.91        46.3      70.90        97.98   38.2

    NEW ORLEANS       831       34.87      52.14        49.5      52.56        73.93   40.7

    BOSTON            946       38.75      57.37        48.0      58.70        81.55   38.9

    MIAMI           1083        42.60      62.97        47.8      64.73        90.30   39.5

    HOUSTON           932       38.38       57.37       49.5      58.14        81.55   40.3

    TAMPA           1006        40.31       59.52       47.6       61.14       84.83   38.8

    WASH., D. C.    1161        45.45       67.15       47.7      69.37        96.89   39.7

 APPENDIX     VII                                                APPENDIX    VII    r

                                COACH CLASS FARES         FIRST CLASS FARES
  MARKETS       (MILES)      FARES
                             --         FARES    UP
                                                 -       FARES    FARES    UP
CHICAGO              860   $ 36.51   $ 52.75   44.5%   $ 55.22   $ 74.99    35.8%

SEATTLE              959     39.09     57.98   48.3      59.23     82.64    39.5

MIAMI               1188     46.27     67.91   46.8      70.66     97.98    38.7

NEW YORK            1093     42.87     63.73   48.7      65.14     91.39    40.3

PHOENIX             1276     48.06     70.74   47.2      73.49    102.68    39.7

MINNEAPOLIS         1300     48.80     72.24   48.0      74.66    104.87    40.5

SAN DIEGO           1532     55.78     80.66   44.6      85.83    119.08    38.8

SEATTLE             1398     51.43     81.53   58.5      78.76    119.09    51.2

SPOKANE, WASH. 1508          55.02     79.92   45.3      84.62    117.99    39.4

LAS VEGAS           1229     46.72     69.66   49.1      71.40    100.50    40.8

SAN FRANCISCO       1587     57.30     80.66   40.8      88.20    119.08    35.0

MIAMI               1258     47.53     69.98   47.2      72.67    101.59    39.8

NEW YORK            1363     50.51     75.32   49.1      77.33    110.01    42.3

PHOENIX             1445     52.66     78.54   49.1      80.68    114.71    42.2

APPENDIX VII                                                       APPENDIX VII

                              COACH CLASS FARES            FIRST CLASS FARES
  MARKETS       (MILES)      FARES    FARES    UP
                                               -          FARES    FARES    -UP
LOS ANGELES         1248   $ 47.27   $ 69.66    47.4%   $ 72.26    $100.50   39.1%
SAN FRANCISCO       1493     54.76     79.18    44.6      84.23     116.90   38.8

LAS VEGAS           1521     55.49     80.66    45.4      85.37     119.08   39.5

SAN FRANCISCO       1915     66.25     97.80    47.6     103.34     144.57   39.9

PORTLAND,    ORE.   1637     57.79     84.59    46.4      89.60     124.88   39.4

SAN FRANCISCO       1647     58.05     87.33    50.4      90.01     128.93   43.2

SEATTLE             1681     59.01     87.96    49.1      91.51     130.02   42.1

NEW ORLEANS         1658     58.34     87.22    49.5      90.45     128.93   42.5

ST. LOUIS           1710     59.83     89.44    49.5      92.82     132.21   42.4

SAN DIEGO           1729     60.40     91.66    51.8      93.72     135.49   44.6

NEW YORK            1627     57.46     84.59    47.2      89.07     124.88   40.2

PORTLAND,    ORE. 1748       60.96     93.22    52.9      94.62     137.79   45.6

LOS ANGELES         1934     66.91     98.54    47.3     104.40     145.66   39.5

SEATTLE             1730     60.42      91.66   51.7       93.76    135.49   44.5

APPENDIX       VII                                                  APPENDIX    VII *I   I

                                 COACH CLASS FARES          FIRST   CLASS FARES
  MARKETS        (MILES)      FARES
                              --         FARES
                                         --       -UP      FARES      FARES    -UP
LOS ANGELES          1740   $ 60.76   $ 91.66    50.9%   $ 94.29    $135.49    43.7%

SAN FRANCISCO        1853     64.39     91.66    42.4     100.37     135.49    35.0

TAMPA                2153     71.01    106.66    50.2     111.25     160.74    44.5
PITTSBURGH           2124     70.08    108.14    54.3     109.77     161.94    47.5

LOS ANGELES          2046     67.49    103.72    53.7     105.61     154.36    46.2

SAN FRANCISCO        2086     68.85    101.27    47.1     107.79     151.65    40.7

SAN FRANCISCO        2141     70.61    106.66    51.1     110.61     159.73    44.4

WASH., D. C.         2288     74.59    115.54    54.9     116.91     174.88    49.6

SAN FRANCISCO        2589     83.64    124.63    49.0     126.20     191.15    51.5
SEATTLE              2408     78.67    121.67    54.6     118.55     184.97    56.0

SAN FRANCISCO        2526     81.87    119.23    45.6     123.35     182.06    47.6

SAN FRANCISCO        2703     87.19    126.67    45.3     131.71     194.28    47.5

PHILADELPHIA         2396     77.38    119.23    54.1     121.88     181.27    48.7

LOS ANGELES          2600     84.03    126.63    50.7     126.82     194.22    53.2

APPENDIX      VII                                                 APPENDIX    VII

                                 COACH CLASS FARES
                           ---__-----------                  FIRST CLASS FARES
                                                        -_------_---        --
  -----         (MILES)
                 -__-          FARES
                               ----         FARES
                                            -----  UP
                                                   --      FARES
                                                           -_--      FARES
                                                                     ----      UP
SAN FRANCISCO       2574   $ 83.14   $121.67    46.3%   $125.43   $186.61    48.8%

NEW YORK            2453     79.76    121.67    52.5     120.20    185.78    54.6

SEATTLE             1448     54.58      95.87   75.7      85.29    130.00    52.4

HAWAII              2408     74.02    106.56    44.0     113.29    190.00    67.7

HAWAII              2573     76.97    106.56    38.4     117.74    190.00    61.4

HAvJAII             2603     71.28    106.56    49.5     108.14    190.00    75.7

HAWAII              2677     73.32    106.56    45.3     110.62    190.00    71.8

CHICAGO             2846    106.10    171.24    61.4     164.39    248.00    50.9

NEW YORK            3278    120.45    185.44    54.0     185.72    264.00    42.2

HAWAII              3785    111.12    167.10    50.4     170.48    264.00    54.9

HAWAII              4248    113.62    162.55    43.1     172.78    299.00    73.1

HAWAII              4974    164.48    206.72    25.7     253.38    343.00    35.4

                                                                        APPENDIY VIII
                                                                                  '.                ..

                   ESTIMATED    AND ACTUAL AIRLINE
                           -----__--                   F4RES FOR 110
                            CITY-PAIR     MARKETS ----
                                                  IN 1973

                                COACH CLASS FARES
                           -~----------                          FIRST CLASS FARES
  ---          (MILES)
                __-           FARES
                              __--      FARES
                                        ---      -UP           FARES       FARES
                                                                           --     -UP
BURLINGTON       182       $ 19.31      $ 20.69     7.1%    $   28.36   $ 28.59               -8%

WASH., D. C.     319         22.71        27.18   19.7          33.40     37.39       11.9

SPRINGFIELD      145         19.01        17.50   -7.9          28.00     24.31   -13.2

DETROIT            94        18.63        15.12                 27.59     20.99   -23.9

PHILADELPHIA     215         20.08        20.75     3.3         29.36     28.68   -       2.3

BOSTON           370         24.19                21.3          35.52     40.67       14.5

NEW YORK         107         18.88        15.18   19.6          27.95     21.09   -24.5

NEW YORK         179         19.63                  2.7         28.76     26.39   -       8.2

PHOENIX          346         23.35        28.77   23.2          34.34     39.58       15.2

SYRACUSE         197         19.72        19.96     1.2         28.85     27.58   -       4.4

ROCHESTER        252         21.27        22.34     5.0         31.29     30.73   -     1.8

PHILADELPHIA     274         21.73        25.59   17.8          31.97     35.20       10.1

LOUISVILLE       277         21.80        25.59   17.4          32.07     35.20           9.8

LOS ANGELES      227         20.34        21.55     5.9         29.74     29.78         0.1

I,   APP.ENDIX VIII                                                          APPENDIX VIII

                                            CLASS FARES                 FIRST CLASS FARES
                                                                   -__----            ----
        MARKETS       (MILES)      FARES       FARES   -UP            FARES
                                                                      -__-      FARES      -UP
      HOUSTON           222      $   20.23   $   21.55      6.5%   $ 29.58   $ 29.78          0.7%

      ST. LOUIS         256          21.35       22.34      4.7      31.40     30.73    -     2.1

      TAMPA             199          20.16       19.89      1.3      29.67     27.49    -     7.3

      NEd YORK          289          22.04       25.59    16.1       32.42     35.20          8.6

      CLEVELAND         312          22.55       27.18    20.6       33.16     37.39        12.8

      NEh    YORK
      PI'I'TSBURGH      329          22.94       27.98    22.0       33.73     38.48        14.1

      NEW YORK          191          19.70       19.10    -3.1       28.83     26.39    -     8.5

      WASH., D. C.      215          20.08       20.69      3.0      29.36     28.59    -     2.6

      DETROIT           238          20.56       21.48      4.5      30.07     29.69    -     1.3

      MINNEAPOLIS       344          23.31       28.77    23.4       34.29     39.58        15.4

      SAN FRANCISCO     441          25.54       32.50    27.2       38.14     45.05        18.1

      DETROIT           540          29.11       38.47    32.2       43.77      53.81       22.9

      KANSAS CITY       643          32.35       44.39    37.2       48.71     62.66        28.6

      PI'TTSHURGH       406          24.52       30.92    26.1       36.56     42.86        17.2

                                 COACH CLASS FARES
                           ------___----------                 FIRST CLASS FARES
  ---__         (MILES)
                 -__           FARES
                               -----           FARES
                                               --__  UP
                                                     --      FARES
                                                             ---        FARES
                                                                        ----    -UP
ST . LOUIS        448      $ 25.74   $ 33.29      29.3%   $ 38.45   $ 46.14   20.0%

PHILADELPHIA      749        35.51      48.92     37.8      53.59     69.21   29.1

MEMPHIS           485        21.51      36.13     31.3      41.30     50.53   22.4

PITTSBURGH        496        27.83      36.91    32.6       41.78     51.63   23.6

DETROIT           623        31.77      42.07     32.4      47.82     59.38   24.2

NASH., D. C.      540        29.11      39.32     35.1      43.77     55.00   25.7

CHICAGO           613        31.48      40.52    28.7       47.38     57.19   20.7

CHICAGO           777        36.35      49.56    36.3       54.90     70.27   28.0

WASH., D. C.      406        24.52      30.20    23.1       36.56     41.85   14.5

NEW YORK          410        24.63      31.71    28.7       36.74     43.95   19.6

WASH., D. C.      591        30.82      40.52    31.5       46.35     57.19   23.4

PHILADELPHIA      675        33.30      44.28    33.0       50.18     62.65   24.8

SEATTLE           671        33.19      45.05    35.8       50.01     63.74   27.5

NEW YORK          489        21.63      36.13    30.8       41.48     50.53   21.8

APPENDIX      VIII                                                    APPENDIX   VIII

                                   COACH CLASS   FARES
                                            --~----             FIRST CLASS FARES
   -___              (MILES)
                      -__         FARES    FARES      UP
                                                      --      FARES
                                                              ----            FARES
                                                                              --__  -UP
 CHICAGO               721      $ 34.66   $ 47.37   36.7%   $ 52.29    $ 67.02    28.2%

 NEW YORK              755        35.68     48.79   36.7      53.86      69.17    28.4
 MINNEAPOLIS           850        38.74     54.19   39.9      58.38      76.83    31.6
 MEMPHIS               880        39.64     54.90   38.5      59.76      78.05    30.6
 SAN ANTONIO          1041        44.56     63.20   41.8      67.36      90.08    33.7

 MIAMI                 946        41.65     57.98   39.2      62.87      82.42    31.1
 TAMPA               1183         49.54     69.33   40.0      75.01     100.03    33.4
 MIAMI               1194         49.91     69.33   38.9      75.59     100.03    32.3
 NEW ORLEANS           831        37.58     53.42   42.2      56.47      75.73    34.1
 BOSTON                946        41.65     58.75   41.1      62.87      83.51    32.8
 MIAMI               1083         45.92     64.34   40.1      69.47      92.26    32.8

 HOUSTON               932        41.20     58.75   42.6      62.18      83.51    34.3

 TAMPA               1006         43.46     60.89   40.1      65.66      86.79    32.2
 WASH., D. C.        1161         48.84     68.50   40.2      73.93      98.83    33.7

APPENDIX       VIII                                                      APPENDIX       VIII

                                     COACH CLASS FARES               FIRST CLASS FARES
   --__               (MILES)
                       --          FARES
                                   ---       FARES
                                             -__       UP
                                                      --           FARES
                                                                   --__          FARES
                                                                                 ---   -UP
 CHICAGO                 860     $ 39.02   $   54.06   38.5%   $ 58.81   $     76.85   30.7%

 SEATTLE                 959       42.02       59.36   41.3      63.44         84.61   33.4

 MI4MI                  1188       49.68       69.33   39.5      75.23       100.03    33.0

 NE4 YORK               1092       46.23       65.10   40.8      69.96         93.36   33.4

 PHOENIX               1276        49.75       72.67   46.1      76.97       105.49    37.0

MINNEAPOLIS            1300        50.47       74.18   47.0      78.11       107.68    37.9

SAN DIEGO              1532        58.53       82.58   41.1      90.91       121.91    34.1

SEATTLE                1398        53.54       83.53   56.0      82.99       122.00    47.0

SPOKANE, WASH. 1508                57.72       81.78   41.7      89.61       120.73    34.7

 LAS VEGAS             1229        48.29       71.60   40.3      74.66       103.31    38.4

SAN FRANCISCO          1587        60.30       82.58   37.0      93.71       121.91    30.1

MIAMI                  1258        49.14       71.92   46.4      76.00       104.40    37.4

NEW YORK               1363        52.45       76.72   46.3      81.25       112.05    37.9

PHOENIX                1445        55.04       80.46   46.2      85.37       117.52    37.7

,+PPEqDIX     VIII                                                         APPENDIX    VIII

                                    COACH CLASS FARES                  FIRST CLASS FARES
   MARKETS           (MILES)
                      --          FARES
                                  ---       FARES
                                            --       -UP            FARES
                                                                    --          FAF@S
                                                                                --      -UP
 LOS ANGELES           1248     $ 48.87   $ 71.60        46.5%   $ 75.57   $103.31    36.7

 SAN FRANCISCO         1493       57.23     81.04        41.6      88.84    119.63    34.7
 LAS VEGAS             1521       58.15     82.58        42.0      90.31    121.91    35.0

 SAN FRANCISCO         1915       69.33    100.28        44.6     107.65    148.23    37.7

 PORTLAND,    ORE. 1637           60.59     86.25        42.4      93.85    127.34    35.7
 SAN FRANCISCO         1647       60.79     89.24        46.8      94.15    131.75    39.9

 SEATTLE               1681       61.73     89.87        45.6      95.62    132.84    38.9

 NEW ORLEANS           1658       61.09     89.13        45.9      94.63    131.75    39.2
 ST. LOUIS             1710       62.53     91.35        46.1      96.87    135.03    39.4
 SAN DIEGO             1729       62.97     93.57        48.6      97.56    138.31    41.8
 NEW YORK              1627       60.31     87.02        44.3      93.41    128.47    37.5
 PORTLAND,    ORE. 1748           63.50     93.57        47.4      98.38    138.31    40.6
 LOS ANGELES           1934       69.94    101.02        44.4     108.61    149.32    37.5
 SEATTLE               1730       63.00     93.57        48.5      97.60    138.31    41.7

APPE:JDIX      VIII                                                     APPENDIX    VIII

                                     COACH CLASS FARES
                                                   -----            FIRST CLASS ------
                                                              ---__--             FARES
   --                 (MILES)
                       --__        FARES
                                   ---       FARES       UP
                                                         --       FARES
                                                                  ----      FARES
                                                                            -___       UP
 LOS ANGELES            1740     $ 63.28   $ 93.57    47.9%   $ 98.03    $138.31    41.1%
 SAN FRANCISCO          1853       67.70     93.56    38.2     105.12    138.30     31.6
 'TAMPA                 2153       75.97    109.17    43.7     115.60    164.51     42.3
 PITTSBURGH             2124       75.10    110.65    41.3     114.24    165.70     45.0
 LOS ANGELES            2046       72.81    106.20    45.9     110.72    158.04     42.7
 SAN FRANCISCO          2085       73.98    103.25    39.6     112.52    154.62     37.4
 SAN FRANCISCO          2141       75.65    109.17    44.3     115.10    163.48     42.0
 dASH., D. C.           2288       80.02    118.11    47.6     121.86    178.77     46.7
 SAN FRANCISCO          2589       85.87    127.72    48.7     131.57    195.90     4ti.L

 SEATTLE                2408       81.42    124.71    53.2     124.71    189.60     =1 2 . fJ

 SAN  FRANCISCO         2526       84.66    121.75    43.8     129.79    185.51     43.2

 SAN FRANCISCO          2703       88.18    129.21    46.5     135.04    195.18     45.5
 PHILADELPHIA           2396       81.99    121.75    48.5     124.65    185.10     43.5
 L3S ANGELES            2600       86.17    129.20    49.9     132.03    195.17     50.1

' APP'kWDIX VIII                                                     APPENaIY VIII

                                  COACH CLASS FARES
                              ___----___----                   FIRST CLASS FARES
     ---           (MILES)      FARES        FARES  UP
                                                   _--       FARES
                                                             --__          FARES
                                                                           ---     -UP
   SAN FRANCISCO       2574   $ 85.58   $124.71    45.7%   $131.12   $191.28      45.9%

   NEW YORK            2453     82.74    124.71    50.7     126.77    190.43      50.2

   SEATTLE             1448     57.97     98.49    69.9      89.17    141.00      58.1

   HAWAII              2408     77.28    109.53    41.7     117.96    190.00      61.1

   HAWAII              2573     79.84    109.53    37.2     121.64    190.00      56.2

   HAWAII              2603     82.60    109.53    32.6     126.31    190.00      50.4

   HAWAII              2677     84.05    109.53    30.3     128.59    190.00      47.8

   CHICAGO             2846     78.40    171.24   118.4     118.19    248.00     109.8

   NEW YORK            3278    119.44    189.22    58.4     183.56    276.00      50.4

   HAtJAII             3785    120.36    170.14    41.4     185.13      285.00    53.9

   HAWAII              4248    112.99    188.75    67.1     171.12     319.00     86.4

   HAvJAII             4974    131.05    226.59    72.9     199.37    375.00      88.1

APPEhDIX        IX

                         ESTIMATED--- AND ACTUAL AIRLINE         FARES FOR 110

                                 CITY-PAIR     MARKETS IN 1974

                                     COACH CLASS FARES
                                -_--___-_----_~__-                       FIRST CLASS FARES
   ------            (MILES)
                      ----         FARES        FARES UP
                                                      --               FARES
                                                                       ---       FARES    -UP
BUHLIVG'I'ON           182      $ 22.26    $ 22.21    -0.2%         $ 31.90      $ 30.66    - 3.9%

dASH., D. C.           319        26.59      30.75     15.6            38.35       42.24     10.2

KRN.iAS CITY,        MO
SPHIUGFIELD            145        21.79      18.73   -14.1             31.28       25.97    -17.0

                        94        21.33      16.20   -24.1             30.17       22.46    -26.9

PHILADELPHIA           215        23.17      23.06   - 0.5             33.15       31.82    - 4.0
HOSTON                 370        28.33      33.03     16.6            40.84        45.72    11.9
NEd YORK               107        21.65      16.94    -21.7            31.17       23.50    -24.6
NEW YORK               179         22.59     20.43    - 9.5            32.34       28.20    -12.8
PHOENIX                346         27.33     32.44     18.7            39.43       44.56     13.0
SYRACUSE               197         22.71     22.21    - 2.2            32.47       30.66    - 5.6
ROCHESTER              252         24.73     24.75    - 0.1            35.63       34.00    - 4.6
PHILADELPHIA           274         25.38     28.97     14.1            36.58       39.79      8.8
LOUISVILLE             277        25.43      28.97     13.9            36.65       39.79      8.6
LOS ANGELES           227         23.54      23.90         1.5         33.70       32.99    - 2.1

LI APPYNDIX        IX                                                      APPENDIY          IX

                                    COACHCLASS FARES              FIRST CLASS FARES
     MARKETS            (MILES)    FARES FARES                  FARES FARES Up
   HOUSTON               222     $ 23.38   $ 23.31    -0.3%   $ 33.46    $ 32.18    - 3.8%

   ST. LOUIS             256       24.85    24.75    - 0.4      35.80      34.00    - 5.0

   TAMPA                 199       23.21    21.37    - 7.9      33.39      29.50    -11.7

   'JEW Y3HK             28')      25.76    28.97     12.5      37.13      39.79       7.2

                         312       26.39    30.75     16.5      38.06      42.24     11.0

                         329       26.86    31.59     17.6      38.75      43.40      12.0

                         141       22.66    20.43    - 9.9      32.43       28.20   -13.0

  NtW   YORK
  WASH.,       0. C.     215       23.17    22.21    - 4.1      33.15       30.66    - 7.5

  DETROIT                238       23.84     23.06   - 3.3       34.14      31.82    - 6.8

  MINNEAPOLIS            344       27.29     32.44    18.9       39.37      44.56     13.2
  SAN FRANCISCO          441       29.88     36.47     22.0      43.66      50.49      15.6
  DETROIT                540       34.01     42.82     25.9      49.97      59.79      19.7
  KANSAS CITY            643       37.85     50.04     32.2      55.74      70.49      26.5
  PITTSBURGH             406       28.75     34.70     20.7      41.95      48.03      14.5

APPENDIX            IX                                                      APPENDIX        IX
                                                                                             \   c

                                       COACH CLASS FARES           FIRST CLASS FARES
    MARKETS              (MILES)      FARES    FARES    -UP       FARES    --FARES  -UP
 ST. LOUIS                 448      $ 30.09   $ 37.31   24.0%   $ 43.97   $ 51.64    17.4%
 PHILADELPHIA              749        41.57     54.83   31.9      61.35     77.43    26.2

 MEMPHIS                  405         32.14     40.33   25.5      47.15     56.32    19.4

 PII‘TSBURGH              496         32.48     41.16   26.1      41.66     57.47    20.6

 DErROI'r                 623        37.13      47.37   27.6      54.65     66.74    22.1

 A rWN'l'A
 HIASH., D. C.            540        34.01      44.48   30.8      49.97     62.11    24.3

 CHICAGO                  613        36.78      45.73   24.3      54.13     64.42    19.0
CHICAGO                   777        42.11      55.51   30.0      63.10     78.55    24.5
kASH.,        D.   C.     406        28.75      34.70   20.7      41.95      48.03   14.5
NEti   YORK               410        28.86      35.64   23.5      42.11      49.33   17.1
kASH., D. C.              591         35.99     45.73   27.0      52.94      64.42    21.7
PHILADELPHIA              675         38.86     49.91   28%4      57.26      70.48   23.1
SEATTLE                   671         38.85     50.73   30.6      57.25      71.64   25.1
NEW YORK                  489         32.27     40.33   25.0      47.35      56.32   18.9

  APPENDIX      IX                                                               APPENDIX           Ix
-3    ,

                                   COACH CLASS FARES                   FIRST CLASS FARES
    MARKETS          (MILES)      FARES    FARES                      FARES    FARES    -UP
  CHICAGO              721      $   40.65   $ 53.19         30.8%   $ 59.98    $ 75.11      25.2%

  NEW YORK             755          41.91     54.69         30.5      61.88      77.39      25.1

  MINNEAPOLIS          850          45.70     60.73         32.9      66.46       85.94     29.3

  MEMPHIS              880          46.83     61.74         31.9      68.14       87.60     28.6

  SAN ANTONIO        1041           52.85     70.71         33.8      77.11     100.58      30.4

  MIAMI                946          49.24     65.10         32.2      71.73       92.36     28.8

  TAMPA              1183           59.14     78.12         32.1      86.52     112.45       30.0

  MIAMI              1194           59.60     78.12         31.1      87.20     112.45       29.0

  NEW ORLEANS          831          44.32     59.60         34.5      64.31        84.34     31.1

  BOSTON               946          49.24     65.91         33.9       71.73       93.51     30.4

  MIAMI              1083           54.54     71.91         31.8       79.65     102.89      29.2

  HOUSTON              932          48.71     65.91          35.3      70.94       93.51     31.8

  TAMPA              1006           51.47     68.28          32.7      75.05       97.12     29.4

  WASH., D. C.       1161           58.24      76.52         31.4      85.16       110.14    29.3

APPEADIX       IX                                                          APPENDIX        IX

                                    COACH CLASS FARES--           FIRST CLASS FARES
   MARKETS          (MILES)      FARES
                                 --         FARES    -UP         FARES    FARES    -UP
 CHICAGO               860     $ 46.04   $ 60.12       30.6%   $ 66.97    $ 85.29   27.4%

 SEATTLE               959       49.74     66.56       33.8      72.48      94.67   30.6

 MIAMI               1188        59.39     78.12       31.6      86.89     112.45   29.4

 NEW YORK            1092        54.92     72.72       32.4      80.21     104.05   29.7

 PHOENIX             1276        60.46     81.65       35.0      91.28     118.23   29.5

 MINNEAPOLIS         1300        61.42     83.25       35.5      92.76     120.54   30.0
SAN DIEGO            1532        71.02     93.24       31.3     107.55     137.25   27.6
SEATTLE              1398        64.90     94.06       44.9      98.10     137.02   39.7
SPOKANE, WASH. 1508              70.12     91.66       30.7     106.16     134.93   27.1
LAS VEGAS            1229        58.59     80.53       37.4      88.39     115.92   31.1
SAN FRANCISCO        1587       72.99      93.24   '   27.7     110.58     137.25   24.1
MIAMI                1258       59.71      80.85       35.4      90.11    117.07    29.9
NEW YORK             1363       63.63      86.02       35.2      96.16    125.31    30.3
PHOENIX              1445       66.58     90.09        35.3      100.68   131.24    30.3

    APPENDIX      IX
"      I'

                                     COACH CLASS FARES           FIRST CLASS FARES
       MARKETS         (MILES)      FARES    FARES    -UP       FARES    FARES    -UP
     LOS ANGELES         1248     $ 59.37   $ 80.53   35.6%   $ 89.60   $115.92   29.4%

     SAN FRANCISCO       1493       69.61     90.87   30.6     105.37    133.77   26.9

     LAS VEGAS           1521       70.56     93.24   32.1     106.84    137.25   28.5

     SAN FRANCISCO       1915       81.53    113.12   38.8     122.96    166.72   35.6

     PORTLAND, ORE. 1637            71.61     88.86   24.1     107.81    130.80   21.3
     SAN FRANCISCO       1647       71.94   100.42    39.6     108.31    147.82   36.5

     SEATTLE             1681       72.97    101.08   38.5     109.87    148.98   35.6

     NEW ORLEANS         1658       72.22    100.30   38.9     108.72    147.82   36.0
     ST. LOUIS          1710        73.83   102.66    39.0     111.17    151.30   36.1

     SAN DIEGO          1729        74.37   105.10    41.3     111.97    154.90   38.3

     NEW YORK           1627        71.28     97.97   97.4     107.31    144.21   34.4
     PORTLAND,    ORE. 1748         74.99   105.10    40.2     112.92    154.90   37.2

     LOS ANGELES        1934        82.15   113.91    38.7     123.91    167.88   35.5
     SEATTLE            1730        74.40   105.10    41.3     112.02    154.90   38.3

                               COACH CLASS FARES            FIRST CLASS FARES
                                                        _____-------       -----
 -----          (MILES)      FARES
                             -__-      FARES
                                       __-       UP       FARES
                                                          -__-       FARES
                                                                     ---         UP
LOS ANGELES       1740     $ 74.73   $105.10    40.7%   $112.52   $154.90     37.7%

SAN FRANCISCO     1853       79.50    105.16    32.3     119.88    154.98     29.3
TAI'IPA           2153       89.69    122.46    36.5     136.03    183.93     35.2

PITTSBURGH        2124       88.65    124.04    39.9     134.42    185.15     37.7

LOS ANGELES       2046       85.82    119.42    39.1     130.08    177.18     36.2

SAN FRANCISCO     2086       87.21    116.09    33.1     132.21    173.28     31.1

SAN FRANCISCO     2141       89.30    122.46    37.1     135.43    182.80     35.0

VJASH., D. C.     2288       94.50    132.77    40.5     143.42    200.27     39.6

SAN FRANCISCO     2589      101.08    142.34    40.8     154.82     217.51    40.5

SEATTLE           2408       94.71    139.93    47.7     144.94     211.99    46.3

SAN FRANCISCO     2526      101.16    136.71    35.1     155.24     207.99    34.0

SAN FRANCISCO     2703      108.00    145.59    34.8     164.00     222.48    35.7

PHILADELPHIA      2396       94.33    136.71    44.9     144.35     207.10    43.5

LOS ANGELES       2600      101.56    145.52    43.3     155.57     222.37    42.9

                                                                        APPENDIX IX

                           -------COACH CLASS FARES
                                                 ----      ----- FIRST CLASS
                                                                           ----- FARES
  ~-            (MILES)
                 --            FARES
                               --         FARES       UP
                                                     --        FARES
                                                               --        FARES
                                                                         -__-         UP
SAN FRANCISCO       2574   $100.60    $139.93     39.1%    $154.08    $213.83      38.8%
NEW YORK            2453     96.26     139.93     45.4      147.34      212.90     44.5
SEATTLE             1448     76.29     107.82     41.3      114.71      148.00     29.0
HAWAII              2408     92.12     121.53     31.9      137.60      195.50     42.1
HAvJAII             2573     88.23     121.53     37.7      131.79      195.50     48.3
HAWAII              2603     93.11     121.53     30.5      138.94      195.50     40.7
HAWAII              2677     95.50     121.53     27.2      143.21      195.50     36.5

CHICAGO             2846    113.87     181.63      59.5     172.88      257.50      48.9

NEW YORK            3278    178.97     200.95     12.3      279.17      280.00       0.3

HAWAII              3785    151.22     187.66      24.1     232.78      293.50      26.1

HAWAII              4248    123.91     208.87      68.6     182.27      328.50      80.2

HAWAII              4974    174.94     250.46      43.2     267.39      386.50      44.5

 APPENLJIX X                                                                                                                       APPENDIX        X

                                  IMPACT OF VARIOUS                         LOAD FACTORS
       LOAD                            ON PASSENGER                         SAVINGS
YEAR   FACTOR                 ESTIMATED             SAVINGS         TO ACTUAL            PASSENGERS
       (Percent)                                        (Billions    of Dollars)


1969      60                                                                                1.8

          72                                                                                                 2.2


1970      60

          72                                                                                      1.9


1971     60                                                                        1.6


1972     60


1973     60



       Base case except actual    industry-wide
                                                                                    Base case:      60 percent     industry-wide     load factor
       trunkline load factor

       Base case except   72 percent    industry-wide
       load factor

      APPENDIX               X                                                                                         APPENDIX   X
*:          ,’

                 IPercent)                              (Billions of Dollars)

     1969         60

                   72                                                                                                 3.4









     1974          60

                  72                                                                                   3.0

                        I                      I                    I                             I                           I
                        0                      1                    2                             3                           4

                                                                           Base case:     60 percent         industry-wide
                                                                          load factor

 APPENDIX             X                                                                                                      APPEND'J,X X ~

YEAR     FACTOR                         EXCESS           OF ACTUAL       OVER ESTIMATED           FARES
         (Percent)                                                     (Percent)


 1969      60



1970       60


1971       60

           72                                                                         53

1972       60

          72                                                                          53


1973      60

          72                                                                43

          59                                         1

1974      60                                         23

                I          I                 I               I         I         I         1           I          I           I     I
                0          10               20              30         40        50        60         70         80          90    100

        Base case except       actual    trunkline       load factor                   Base case: 60 percent   load factor

q       Base case except       72 percent        load factor

APPENDIX               X                                                                                                                 APPENDIX            X

         FACTOR                                           ANNUAL       AIR TRAVEL
YEAR     (Percent)                            (Billions     of   Revenue-Passenger-Miles)


          60                                                                                                                      165


          52                                                                                               129
          60                                                                                                        146

          60 1                                                                                                                1161



          72                                                                                                                                     181

        Actual       experience                                                                  Base case except   actual   trunkline       load factor

        Base case:        60 percent   load factor                                               Base case except   actual   annual      utilization
cl                                                                                0

APPENDIX             X                                                                                                        APPENDIX                X

         FACTOR                                 ANNUAL       AIRLINE   REVENUE
YEAR     (Percent)                                    (Billions of Dollars)

         60                                                                          16.1

         53                                                                           .9

         52                                                                                 6.4
         60                                                                                 16.5
         72                                                                                        6.8

         56                                                                                                .2



         72                                                                                                                             9.4

              0           1          2             3        4         5          6                 7              8              9            10

                                                                           Base case except              actual   trunkline     coach   load factor

       Base case:    60 percent   load factor                               Base case except             72 percent       load factor

    APPENDIX XI                                                                                                    APPENDIX XI
Y     r’

                    IMPACT         OF AIRCRAFT             SEATING           ON PASSENGER                SAVINGS
     YEAR                             ESTIMATED      SAVINGS          TO ACTUAL          PASSENGERS
                                                         (Billions     of Dollars)






            Base case except   trunkline   seating                                    Base case:   GAO   assumed   seating

APPENDIX                                                                                                APPENDIX        XI

   YEAR               ESTIMATED            SAVINGS            TO ACTUAL AND INDUCED         PASSENGERS
                                                               (Billions of Dollars)











                 I                                                I                         I                                I
                 0                                                1                         2                                3

           Base case except   actual    trunkline       seating                Base case:   GAO   assumed     seating

   liII.Il Base case except   maximum         seating

.. APPENDIX       XI                                                                                          APPENDIX               XI

     YEAR                                EXCESS           OF ACTUAL       OVER ESTIMATED       FARES








                    I            I               I             I         I      I          I        I             I              I    I
                   0             10             20            30        40      50      60         70         80                90   100

              Base case except       actual   trunkline     seating                  Base case:   GAO   assumed       seating

APPENDIX        XI                                                                                                    APPENDIX *XI              '

                                                        ANNUAL      AIR TRAVEL
 YEAR                                       (Billions      of Revenue-Passenger-Miles)








                                                                                         Base case except   actual   trunkline        seating

 cl      Base case:   GAO   assumed   seating                                            Base case except   maxlmum         seating
.I   APcPENDIX XI                                                                             APPENDIX           XI

                                            ANNUAL      AIRLINE   REVENUE
       YEAR                                      (Billions of Dollars)







                    I          I       I        I       I        I          I           I         I       I           I
                    0         1       2         3       4       5           6           7         8       9       10

               Actual experience                                      Base case except actual trunk airline seating

       cl      Base case: GAO assumed seating
                                                              q        Base case except maximum seating

APPENDIX            XII                                                                                                APPENDIX              Xi1           '
                            IMPACT           OF ANNUAL  AIRCRAFT  UTILIZATION
                                               ON PASSENGER SAVINGS
 YEAR                               ESTIMATED             SAVINGS TO ACTUAL             PASSENGERS
                                                             (Billions of Dollars)














               I                                    I                              I                              I                             I
               0                                   0.5                            1.0                            1.5                          2.0

 /’     Base case except

        Base case: aircraft
                                actual   annual


                                                   is highest      actual
                                                                                        Base case except
                                                                                        annual trunkline
                                                                                                                        utilization     is lowest actual
                                                                                                                             per aircraft    in 1969-74

        annual trunkline    aircraft      utilization      in     1969-74


11APPiNDIX        XII                                                                                        APPENDIX                  XII

   YEAR              ESTIMATED              SAVINGS           TO ACTUAL      AND INDUCED       PASSENGERS
                                                               (Billions of Dollars)


   1969                                                                                                     2.3








                                                                                   Base case except   aircraft    utilization     is lowest actual
          Base case except       actual   annual    utilization                    annual trunkline   utilization      per aircraft    in 1969-74

          Base case: aircraft     utilization        is highest actual
          annual trunkline    aircraft      utilization      in 1969-74

APPENDIX                XII                                                                                                  APPENDIX                   XII          .-

      YEAR                                EXCESS         OF        ACTUAL           OVER   ESTIMATED        FARES

     1969               -48






                    I                                                                                  36



                    0                        IO                        20                  30                40                    50                     60
      II ,                                                                                        Base case except   aircraft    utilization      is lowest actual
 ‘/                                                                                               annual trunkline   utilization       per aircraft    in 1969-74
             Base case except       actual    annual    aircraft      utilization
 0                                                                                                period

             Base case: aircraft     utilization        is highest actual
             annual trunkline    aircraft      utilization      in 1969-74


     A;PENDIX        XII                                                                                       APPENDIX                XII

                                                         ANNUAL  AIR TRAVEL
       YEAR                                     (Billions of Revenue-Passenger-Miles)






                                                                                   Base case except   actual    annual   utilization

       cl       Low:
                            Base case except actual annual trunkline
                                 per aircraft In 1969-74 period
                                                                                   High:    Highest actual annual trunkline
                                                                                   per aircraft   in 1969-74
                                                                                                             period (base case)
APPENDIX          XII                                                                                                 APPENDIX                Xf-I   o

                                               ANNUAL      AlRLlNE   REVENUE
  YEAR                                              (Billions of Dollars)



                      I                                                           (6.0







                  0                    1   2          3             4     5      6                       7            8             9           IO

         Actual           experience                                          Base case except               actual    annual   utilization

          Low:     Base case except lowest actual         annual trunk-       High:     Highest actual annual trunkline   utilization
         line utilization   per aircraft in 1969-74        period             per aircraft    in 1969-74 period (base case)

     APPENDIX          XIII                                                                                  APPENDIX           XIII
                               IMPACT OF “MOST-EFFICIENT-AIRCRAFT”
                                 ASSUMPTION  ON PASSENGER SAVINGS
      YEAR                       ESTIMATED           SAVINGS        TO ACTUAL       PASSENGERS






                   I                             I                             I                         I                          I
                   0                           0.5                            1.0                      1.5                         2.0

             Base case except most efficient     aircraft   used only               Base case: most efficient  aircraft   used in the same
             over their least cost ranges                                           proportions as the awcraft    classes they represent

APPENDIX       XIII                                                                                    APPENDIX              XII.1         .

  YEAR              ESTIMATED         SAVINGS         TO ACTUAL       AND      INDUCED       PASSENGERS
                                                       (Billions of Dollars)









               I                  I                    I                   I                 I                   I                 I
               0                0.5                   1 .o               1.5               2.0                  2.5              3.0

         Base case except most efficient   aircraft   used only                  Base case: most efficient   alrcraft   used in the same
         over their least cost ranges                                            proportions as the aircraft    classes they represent

     APPENDIX       XIII                                                                                      APPENDIX              XIII

      YEAR                        EXCESS        OF ACTUAL            OVER ESTIMATED           FARES









                   1                   I                       I              I                    I                   I                  1
                   0                 10                      20             30                   40                   50                60

              Base case except most efficient     aircraft    used only                Base case: most efficient    aircraft   used in the same
              over their least cost ranges                                             proportions  as the aircraft    classes they represent

APPENDIX              XIII                                                                                   APPENDIX           XIII   v

  YEAR                                                    ANNUAL    AIR TRAVEL




                                      , , “;;;‘:y:
                                 I , /,/,’                          /‘y/AA         108
                               ‘:.....:/.,....,../~I.                        “1


 1974                                                                                                            156

                  I                                   I                        I                        I                          I
                  0                                  50                      100                       150                       200

                                                                                   Base case except most efficient   aircraft
         Actual       trunk   airline   experience
                                                                                   used only over their least cost ranges

         Base case: most efficient      aircraft   used only in same
         proportions     over distances    that trunk airlines
         actually    used the aircraft  classes they represent

    APPENDIX XIII                                                                                                 APPENDIX XIII
*     f

       YEAR                                               ANNUAL           AIRLINE       REVENUE
                                                                   (Billions     of Dollars)

       1969                                                                                            6.1





       1974                                                                                                                                 9.1

                       I              I             I        I             I          I            I         I           I           I                 I
                       0              1            2         3            4           5            6         7           8           9                10

                                                                                               Base case except most efficient   aircraft         used only
              Actual       trunk   airline   experience
                                                                                               over their least cost ranges

              Base case: mast efficient      aircraft   used only in same
              proportions     over distances    that trunk airlines
              actually    used the aircraft  classes they represent

APPENDIX           XIV                                                                                               APPENDIX               XIV
                                                                                                                                                  -.       e

                                  IMPACT           OF VARIOUS RATES-OF-RETURN
                                                   ON PASSENGER SAVINGS
 YEAR                              ESTIMATED                SAVINGS           TO ACTUAL    PASSENGERS


 1969                                                                                                                                     1.8







              I                                         I                            I                           I                                     I
               0                                   0.5                              1 .o                       1.5                                2.0

        Base case:       10.5   percent   pretax   rate-of-return                          Base case except   12 percent     pretax   rate-of-return

        Base case except        18 percent     pretax       rate-of-return

*   APPENDIX XIV                                                                                                        APPENDIX XIV

    YEAR                 ESTIMATED           SAVINGS              TO ACTUAL       AND      INDUCED          PASSENGERS
                                                                   (Billions of Dollars)














                 0                                            1                                        2                                            3

           Base case:   10.5 percent    pretax   rate-of-return                            Base case except      12 percent   pretax   rate-of-return

           Base case except    18 percent    pretax    rate-of-return


APPENDIX       XIV                                                                                                    APPENDIX                    XIV

 YEAR                         EXCESS               OF ACTUAL                   OVER ESTIMATED             FARES











              I                      I                             I                          I              I                     I                          1
              0                     10                            20                         30             40                    50                        60

        Base case.   10 5 percent        pretax        rate-of-return                             Base case except   12 percent        pretax   rate-of-return

        Base case except   18 percent         pretax        rate-of-return


     APPENDIX          XIV                                                                                                  APPENDIX                       XIV

                                                                 ANNUAL      AIR TRAVEL
      YEAR                                           (Billions      of Revenue-Passenger-Miles)




                       I                                                                      ------I131




                       0                                  50                               100                              150                                     200

              Actual       experience                                                                  Base case:   10.5 percent   pretax          rate-of-return

              Base case except          12 percent   pretax      rate-of-return                        Base case except    18 percent         pretax     rate-of-return

APPENDIX           XIV                                                                                     APPENDIX                      XIV           -      '-

                                                 ANNUAL      AIRLINE    REVENUE
  YEAR                                                (Billions of Dollars)
                              ““““‘,“““”                   /“‘/“/‘/‘//’     .I, /I -







                   I                                 I                             I                          I                                           I
                   0                               2.5                           5.0                        7.5                                       10

          Actual       experience                                                      Base case.   10.5   percent      pretax       rate-of-return

          Base case except          12 percent   pretax   rate-of-return               Base case except    18 percent       pretax       rate-of-return

     APPENDIX xv                                                                                                                 AP?EN!3IX      XV
I/     .*
                         IMPACT          OF FARE-ELASTICITY-OF-DEMAND-FOR-AIR-
                                          TRAVEL ON PASSENGER SAVINGS
       YEAR                      ESTIMATED            SAVINGS         TO ACTUAL            PASSENGERS
                                                          (Billions    of Dollars)


       1969                                                                                                                            1.8




       1971                                                                                                                      1.6










                    I                                I                                I                                  I                      I
                   0                               0.5                               1.5                               1.0                    2.0

              Base case except   elasticity   = -.7                                          Base case:   elasticity         = -1.3

              Basecaseexcept elasticity = -2.0
APPENDIX     XV                                                                                                         APPENDIX       ZV   ,

                         ESTIMATED                    SAVINGS   TO ACTUAL       AND      INDUCED          PASSENGERS
                                                                 (Billions of Dollars)






                 I                                                                                                               2.5


   1974                                                                                            1 .a

           Base case except              elasticity    = -.7                               Base case:      elasticity   = -1.3

           Base case except              elastlclty    = -2.0

APPENDIX     XV                                                                                                    APPENDIX       XV

                                    EXCESS            OF ACTUAL     OVER ESTIMATED   FARES
     YEAR                                           FOR ACTUAL     PASSENGERS   ONLY


     1969                                                                                                                    52


     1970                                                                                35

                  t : .. ”                                                               35


     1971                                                                                 36



     1972                                                                                  36


     1973                                                                  27



     1974                                                             23

                  I                         I               I                   I                 I                      I             I
                  0                    10                  20                   30                40                    50             60

            Base case except   elasticity       =   -.7                              Base case:       elasticity    = -1 3

APPEhDIX                  XV                                                                                                                    APPENDIX.,XV    ,,

                                                                          ANNUAL       AIR TRAVEL*
      YEAR                                                    (Billions       of Revenue-Passenger-Miles)




                          I                               I                               I                         I                        I              I
                          0                              50                              100                      150                      200            250

                 Actual       experience                                                                        Base case except   elasticity    = -.7


                 Base case:

                          of actual

                                                    = -1.3

                                                      and additional         travelers     induced
                                                                                                     q      ” Base case except

                                                                                                     by lower     fares
                                                                                                                                   elasticity    = -2.0

APEZNDIX       XV                                                                                    APPENDIX       XV

                                               ANNUAL      AIRLINE   REVENUE
   YEAR                                             (Billions of Dollars)









           Actual   experience                                             Base case except   elasticity   = -.7

           Base case:    elasticity   = -1.3                               Base case except   elasticity   = -2.0

APPENZIX     XVI                                                                          APPENDIX        YVI

                       !NdO    COMMENTED Q!d THIS
                       --------------------_-------                           REPORT

                                    --        E. KEELER
-----      position:       Assistant                                Professor   of Economics,
                           University                                of California   , Berkeley.
-----___                  Ph.D.# Economics, Massachusetts                                     Institute
                          of Technology,   1971.
                          S.M.,  Economics, Massachusetts                                   Institute
                          of Technology,   1969.
                          B.A* p Economics, Reed College,                                   1967.
-------                    "Airport  Costs and Congestion,    "American
                           ---_---_-  14  (Spring, 1970),  pp.   47-53.
                          "The Economics of Passenger Trains,"
                          Journal of Eusiness 44 (4pri1,
                          ---------_----_                 1971),
                          pp.               148-174.

                           "Airline     Regulation       and Market Perfor-
                          mance," -------------------------
                                       Sell Journal of Economics and Man-
                          agement Science 3 (Autumn, 1972), pp.

                          "Railroad  Costs, Returns to Scale, and Ex-
                          cess Capacity,"   Review of Economics and
                          ------_--   56  (Yay,   1974)~-~~-%jiZYK
                          Published          Working
                          ---------------------,II-----,-                    Papers    Yonograohs,
                          "Regulation    and Modal Market Shares in
                          Long-haul Freight     Transport:        An Interna-
                          tional   Comparison,"   Department of Econom-
                          ics Working Paper #47, University            of Cali-
                          fornia,   Berkeley,   December 1973; revised
                          July, 1974.     Submitted Journal
                                                      --------e---e of Law and
                          The Full Costs of Urban Transport.
                          ------ III:      Automobile
                                                   -----7------ Costs and Flnal   In-
                          termodal        Cost
                          ----y-------v----L-------Comparisons,       Monograph   821,
                          Institute        of Urban and Regional Develoo-
                          ment, University              of California,     4ugust;

APPENDIX XVI                                        APPENDIX XVI
               "On the Economic Impact of Railroad
               Freight  Regulation,"  presented at the In-
               dustrial  Organization  Workshop, Univer-
               sity of Chicago, May, 1975.
               "On the Environmental         Costs of the Various
               Transportation     Modes," processed,       Insti-
               tute of Urban and Regional Development,
               University     of California,      February 1974
               (with Kenneth A. Small).
               Resource Allocation
               -----                 in Intercity    Passen-
               ger Transportation  Ph.-EDissertatioc
               M.I.T.,  August, 1971.
               On the WelfareI---r-------
               ----                Impact of --------
                                                 Urban Residential
               Property  Taxation,
                           --             S.M. thesis,  M.I.T.
               February,  1959.
               Railroad Cost Functions:   %n Empirical
               gtudy, E.A.tKesrs,   ----
                                    Reed College,  May

APPENDIX XVI                                                     APPENUIY XVI
                         STATEME?JT       OF QUALIFICATIONS
                               NILLIAM A.
                               --I--          J3RDAN _

---l___l     position:      Professor,   Faculty  of Administrative
                            Studies,   York University,  Oownsview,
                            Ontario,   Canada.
------                      Ph.D., Business Economics, University      of
                            California,    Los Angeles, 1968.
                            M.S., General Administration,     Columbia
                            University,    1955.
                            B.S. * Business,    Engineering, Antioch
                            College,    1950.
---l-l_                     Books:

                            -----       Regulation in America:
                            and Imperfm<G(Bai?!?%ore:
                                    --------                        The Johns
                            Hopkins Press, 1970), pp. xvi, 352.
                            Patterns   of Performance,    Vol.     3 of "The
                            ?Ke~cZZZZZK~-Lessons                 for Manage-
                            ment and the Nation,"     junior     coauthor un-
                            der Eli Ginzberg (New York:          Columbia
                            University   Press, 1959), pp.       xix, 340.
                            "Producer Protection,    Prior Market Struc-
                            ture, and the Effects    of Government Requ-
                            lation,"  Journal of Law and Economics/
                            Vol. xv, No. 1 (April    1972), pp. 151-76.
                            "Airline  Capacity Agreements:     Correcting
                            a Regulatory   Imperfection,"  Journal of
                            Air Law and Commerce, vol. 39, ET?--
                             (Spring 1973), pp. 179-213.
                            "Air Transportation     Markets:    Definitional
                            Confusion,"   Journal
                                          ---       of  Air  Law and Com-
                            ---     Vol.  41,   no.
                                                 1(WZter        1975),-
                            PP. 33-56.
                            P----Y-      Conference ---t---
                            "Competition--   A Two-Edged Sword in Improv-
                            ing 9ir Transportation    Performance?"    in
                            John De S. Coutinho,    Chm., Transoortation:
                            A Service (Hew York:     New York Academy of
                            Sciences,    1967), pp. 153-68.

     APPENDIX XVI                                              APPENDIX XVI
.I      .=
                    "Survival,      Profits,      and Resource
                    Utilization,"       in Joseph F. Vittek,            Ed.,
                    Proceedings of the NASA/MIT Workshop       ------     on
                    AirIine ---__-
                                               7----             II    (Cam-
                    bridge,     Mass.:      Flight   Transportation          Lab-
                    oratory,      Massachusetts      Institute       of Tech-
                    nology, FTL Report            R72-7, 1972) D. 79.
                    "New Departures    for Regulation,"   in Fed-
                    eral Bar Association,    Air Transportation
                    --7---        (Wash., n.C,~-?~~e~~iPubii-
                    cations,   Inc.,  1972), pp. W-l to H-16.
                     "If We're Going to Regulate the Airlines,
                    Let's do it Right,"        in James C. Miller
                    III,     Ed., --
                                  Perspectives      in Federal Trans-
                    i---y---      PolE-@ZKT-KC,T---           4meriZn
                    Enterprise       Institute   for Public policy
                    Research, 1975) pp. 57-70.
                    "CAB Regulation                and Airline  Efficiency,"
                    in Transportation                Research Forum, Pro-
                    ceedings --Sixteenth
                    -------------------------------    annual fleeting-hi-
                    cage, Ill.:              The Transportation       4esearch
                    Forum, 1975), pp. 187-93.
                    "Of 'Academic'   Interest,"      Letter  to the
                    Editor,  The Wall Street---e-7Journal (Hay 27,
                    1971), p.-TnPa&Tfic         Coast "odltion)
                    "Testimony of William A. Jordan,"            Exhibit
                    No . DJ-RT-1, -w---w
                                           7------------w- Flerger Case,
                    CAB Docket 22916 (WashIngton,          D.C.:    U.S.
                    Dept. of Justice,.June      1971) p. 29.
                    "Some Predatory     Practices  Under Government
                    Regulation?"      University  of Toronto-York
                    University    Joint Program in Transporta-
                    tion,Research     Report No. 26 (January 1975).
                    "Results of Civil          Aeronautics    Board Regu-
                    lation,"        Prepared Statement on Febru-
                    ary 14, 1975-in Hearings Before The Sub-
                    ---------------- on AdrninistratT~e-?~c~ice-and
                    Procedure of the Cornrnit~e-~~t~e-;~~-
                    Congress, First         Session, on Oversight     of


      APPENDIX XVI                                   APPEMDIX XVI
                                                                *a             /a
                     Civil Aeronautics  Board Practices  and
                     Procedures,  Volume 1 (1975), z)p. 464-87.
                     "Airline    Deregulation:           Chaos in the Eyes
                     of the Beholder,"            Supplementary Testimony
                     on April 21, 1975 in Hearings Before the
                     ------              on AdrninrsGtivePractice‘-
                     and Procedure of     --- the Committee on the
                                     TJnited States Senate,Kgtv-
                     Fourth Conqress, First             Session,'on  Over-
                     sight of Civil         Aeronautics      Board Practices
                     and Procedures,          Volume 1 (1975), pp.
                     Review of Economic Reuulation              of Domestic
                     Air   TransDo?~=--~~~~r~-~n
                     --w------L---             --me-------- PsE<-&?
                     G.'&. Douglas and J.C. Miller            III,  in
                     Journal      of    Political
                     ~-o.----‘~----------             Economy Vol.  84,
                           1 (Feoruary 1976).

    APPENDIX XVI                                                   APPENDIX XVI
-      .
                           -----    --- QUALIFICATIONS
                                 ------l--_l- W. DOUGLAS
    Present ---position:      President
                              Southwest Econometrics,  Inc.
    Education:                Ph.D., Economics, Yale University,    1967
                              N.A.I Economics, Yale University,    1963
                              E.A., Physics, Yale University,    1960
               ---            Book
                              Economic Regulation
                              --------                  of Domestic
                                                                 ------ Air Trans-
                              port:--  Theory-andP~r~~-~Tth
                                                     ---               James CT--
                              GTler    III)  Brookings,    1974.
                              "Regulation     of the U.S. Airline    Industry:
                              An InterDretation."       in PersDectives     in
                              7------    Transportation   Poiicyr-S%ies-c
                              Niller   III ed~,AmerTcan-~~~~rprise        In-
                              stitute,    1975.
                              "Quality     Competition,  Equilibrium      and Ef-
                              ficiency     in the Price Constrained       Airline
                              Market" (with James C. Miller,         III)    Ameri-
                              can     Economic
                              -----------       Review, September    1974.
                              "The CAB's Domestic Passenger Fare Investi-
                              gation"         (with James C. Miller, III) The
                              Bell Journal of Economics and Management
                              Science, Spring 1974.
                              "Excess Capacity and Fares in North Atlan-
                              tic Air Transport,"     Proceedings of the In-
                              ------------   Conference on Transportation
                              ----         Transpor~a~Ton-Res~~~~~Forurn,
                              1973 (copyright     1974).
                              "Price Regulation         and Optimal Service
                              Standards:         the Taxicab Market,"    Journal
                              of    Transport
                              ------II_---------   Economics and Policy:-gay-
                              "Quality        Competition   and the Structure   of
                              Airline       Fares," Proceedings of the Trans-

                              portation        Research Forum,~~i~x??,-~o,l
                              (1971).         Reprinted   in MIT/NASA Workshop,
                              Airline       Systems Analysis,
                              _------_----_-____                 July 1972.

                     "Risk in the Equity Xarkets:          An Fmpirical
                     Appraisal  of !dlarket Efficiency,"      Yale
                     --------- Essays,   Vol.  9   (Spring   1969).
Papers Read:
------I--            "Equilibrium    in a Deregulated Air Trans-
                     port Market,"     Seminar on Problems of Reg-
                     ulation    and Public Utilities,   Dartmouth
                     College,    August 21, 1972.
                     "A Review of Selected Aspects of the Su-
                     personic Transport    Program," Southern
                     Economic Association,    November 14, 1969.
Book Reviews:        Ernest W. Williams,             ed. The Future of
                     American         Transportation
                     ------7------i-----.-               %?-%??%iEx?ust
                     -------            1974.
                     Lewis PI. Schneider,             The Future of the I_-U.S.
                     ~----------7-----Air    Freight-~ndus~ry,~n-~~~
                     -m-m-            Economic Journal,
                              ---------------__                41, 1 (1974).
Testimony (written   Refore the U.S. Senate administrative
and oral unless      Practices    Subcommittee,    (Investigation          of
noted):              Airline   regulation),     February 6, 1975.
                     Before the Civil Aeronautics  Boardl CAB
                     Docket 22908, Exhibit  DOT-T-5 (Capacity
                     Reduction Agreements Case) qpri.1 25, 1974.
                     iBefore the Civil Aeronautics     Board, CAB
                     Docket 24353, Exhibit    D3T-T-1 (‘rflainland
                     U.S. - Puerto Rico/Virgin     Island Fares)
                     July 21, 19720
                     i3efore the Civil Aeronautics   Board, C.9B
                     Docket 21866-9, Exhibit    DOT-T-3 (Fare
                     Structure   Investigation)  May 17, 1971.
                      (written  only)
                     Before the Civil  Aeronautics    13oard, CAB
                     Cocket 21866-7, (Load Factor Standards
                     Investigation),  Exhibit    DOT-RT-2, Sep-
                     tember 22, 1970.   (written   only)
                     Eefore the Aorth Carolina   Public Utili-
                     ties Commission, (Extended %rea Telephone
                     Service Investigation)   March 15, 1971.
                     (oral only)

 APPENDIX XVI                                                        APPENDIX XVI
,.   e
                         STATEMENTOF QUALIFICATIONS
                                --------- C. EARS
----         position:
              ----         Executive Director          of the National
                           Commission on Supplies and Shortages,               a
                           temporary study commission composed of
                           four Members of Congress, four Senior Ad-
                           ministration      officials,      and five individ-
                           uals from the private          sector.    This com-
                           mission is scheduled to report to the
                           President      and to the Congress on Decem-
                           ber 31, 1976, on a variety            of materials-
                           related     issues.
Education:                 Ph.D., Economics, Yale University,     1968
                           M.A. ‘ Economics, Yale University,    1965
                           B.A., Economics, University    of Colorado,
----------                 "A Long-Run Cost Function for the Local
                           Service Airline            Industry:     An Experiment
                           in Nonlinear           Estimation,"     (with Marc
                           Nerlove and William Raduchel),                The
                                                                         ------ Review
                           of   Economics
                           ------e--------------w and   Statistics    (August    1969),
                           pp. 258-270.
                           Review of Mahlon Straszheim,        The   Interna-
                           tional    Airline  Industry    in The Journal of
                           "Statistical   Biases in Aggregate Time
                           Series of the Demand for Air Travel,s'
                           (with Philip Verleger),    Proceedings,
                                                      7----           Busi-
                           ness and Economic Statistics      Section,
                           American Statistical    Association    (Decem-
                           ber 1970)r pp. 120-126.
                           "Governmental    Support of Advanced Civilian
                           Technology-- Power Reactors and the Super-
                           sonic Transport,"     (with Richard R. Nelson),
                           --      Policy    (Summer,  1971), pp. 405-427.
                           Review of Alfred R. Kahn, ----------m-m-    The Economics of
                           T----------           in The  Bell      Journal
                                                     p--7---,---------       of Econom-
                           its     and    Management
                           ----------i----------           Science      (Autumn,  1971),
                           PP. 678-682.

APPENDIX XVI                                                  APPENDIX XVI
               "The Effect     of Regulation  on the Cost     *               vf
               Performance and Growth Strategies        of the
               Local Service Airlines,"      Journal of Air
               Law     and
               ------------ Commerce (Winter 1972),FG.i?4.
               The Local Service qirline                   Experiment,
               Lzs3iziq~nT----- the Brooms                  Institution
               Review of 2. Griliches           and A. Ringstad,
               Economics   ?---- of Scale   and the Form of the
               --v--e------------ Funct?%in           Jourxof-
               __-------         Economy  (Januaryflebruary,
               1973), pp. 243-244.
               Review of Herman Nertins,       Jr.,   National
               ------------?- Polic_y_in--------
               The Journal of Business (January 1974),
               PP* 112-114.
               "The Search      for an Efficient       Short-Haul
               Transport:      A Case Study in Conflicting
               Government     Incentives,"    The
                                              ------- Engineering
               Economist,     (Winter 1974), pp. 87-104.
               “U. s. Support      for Civilian Technology:
               Economic Theory Versus Political        Prac-
               tice,u Research Policy (Summer 1974), pp.
               "Subsidies       and the Local Service Airlines:
               The U.S. Experience,"               in Karl M. Ruppenthal
               ed., -----------------------------
                       Some Ramifications             of Transport    Sub-
               sidies,    Vancouver:             Center for Transporta-
               tion Studies,         the University        of British
               Columbia (1974), pp. 81-95.
               "Airline         Capacity Limitation    Controls:
               Public Vice or Public Virtue"?             The Ameri-
               can     Economic
               -----------------    Review  (May   1974),~?Y%Z
               "Railroad          Diversification:      Where Lies the
               Public Interest"?               The Bell Journal of

               Economics          and Management Science (Autumn,
               1974), pp- 595-613.
               "Economists Versus Regulators,"                  in James C.
               Miller,         III,   ed., Perspectives
                                             s-------y-       on  Federal
               ---.II---------------L    Pollcv,     Washington:    The
               American Enterprise               Institute   for Public
               Policy Research (1974), pp. 101-109.
    APPENDIX XVI                                                            .APPENDIX XVI
,      ,a
                                "Telecommunications             Regulation:      Nhat Can
                                We   Learn From Other .Regulated Industries?"
                                in Charles F. Phillips,               Jr.,  ed., Tele-
                                communications           Regulation,
                                ---~------7--L-----~~7~~---I--             and  Public
                                ---           Lexington,       Virginia:     Nashington
                                and Lee ?Jniversity            (1975), pp. 57-77.
                                "Competition           in the Domestic        Airline    In-
                                dustry:         Too Much or Too Little?"              in
                                Almarin Phillips,            ed., Promoting
                                                                    ----7------     Compe-
                                tition       in   Regulated
                                -------------~-------~---~-     Industries,        Washing-
                                ton:        The Brooklngs Institution              (1975),
                                pp. 13-54.
    Unpwlished         Papers   "Price Discrimination     and the Elasticity
    ------- Reports:            of Demand: The Case of Air Travel,"
                                Harvard Institute     of Economic Research,
                                Discussion   Paper Number 72 (April     1969).
                                "Economic theory As An Aid To Regulatory
                                Agency Decisionmaking:   The Capacity Re-
                                duction Agreements Case," presented at
                                The XXII International  Meeting of The In-
                                stitute  of Management Sciences, Xyoto,
                                Japan (July 26, 1975).

APPENDIX XVII                                                                           APPENDIX XVII              _

                                CIVIL     AERONAUTICS            BOARD
                                        WASHINGTON,     D.C.   20428

                                                June 30, 1976
                                                                                       INREPLYREFERTO.   B-60-64

  Mr. Henry Eschwege
  Resources and Economic
     Development   Division
  General Accounting      Office
  Washington,    D. C. 20548

  Dear Mr. Eschwege:

          The Chairman has asked me to thank you for the opportunity                              to
  comment on the GAO draft              report    entitled       "Impact of Increased
  Efficiencies     on Air Fares and Travel"                  and has further       requested      that
  I furnish    you with the staff's              observations         on the study.        Essentially,
  the draft    report     concludes        that the airlines           could obtain substantial
  efficiencies     by increasing           load factors,         aircraft    seating     densities,
  and aircraft     utilization,           and employing         optimum equipment        types.      These
  efficiencies     would, according            to the draft         report,   result     in savings
  between $1.5 and $2 billion                annually      to the traveling        public.      On the
  basis of this analysis,             GAO concludes          that the Board should work towards
  the achievement        of the kinds of efficiencies                  assumed in its study, either
  under its existing          legislative        authority       or by seeking appropriate
  legislative     changes.

         We would be in general            agreement with the draft            report  to the
  extent     that it recommends that the Board work towards the achievement
  of a more efficient         air transportation           system.       However, the report      is,
  we believe,       deficient    in its failure         to recognize       the actions     which
  the Board h&., in fact,            taken along these lines,             both in the exercise
  of the Board's current           legislative       authority,      as well as in the Board's
  regulatory      reform legislative          program.       Moreover,     we believe    that the
  level of savings predicted             in the draft        report    from operations      under
  the assumed conditions           is speculative         and inflated.

          With respect    to the Board's actions       under its present   legislative
  authority,     we believe    that the draft   report     should take cognizance      of
  the very substantial        adjustments  which the Board makes to the carriers'
  operating    results    for ratemaking   purposes.       Specifically, in employing

     APPENDIX XVII                                                                                 APPENDIX XVII
.I    I.

      Mr. Henry      Eschwege        (2)

      standards      developed      in the Domestic Passenger Fare Investigation,                               the
      Board predicates          passenger       fares on standard,                full-fare      load factors,
      in order to eliminate             the effect        of excess capacity                upon the passenger
      fare level.         Moreover,       in calculating           passenger          load factors,       the Board
      assumes optimum seating               densities,       including          the substitution          of 9 and
       lo-abreast      coach seating         for wide-bodied            aircraft          in lieu of the 8 and
      g-abreast      seating which is currently                 employed by the airlines.                     The
      Board also adjusts          the carriers'           aircraft        utilization          in cases where
      actual utilization          appears inadequate.                 The combined result              of these
      adjustments       resulted      in a saving of approximately                        $750 million      to the
      domestic     air traveler         in 1975.        Although        these standards            are not
      mentioned      in the report,          it is interesting              to note that the draft
      report     does indirectly          provide      documentation            for the impact of the
      implementation        of these standards,              since the differences                 between the
      actual fare level and the "increased                       efficiency           fare level"      significantly
      diminish     from 1969 to 1974.

              The report    also fails     to take account of the Board's proposed
      regulatory      reform program.        This is a detailed      proposal,        the substance
      of which was submitted          to the Congress in April of this year and was
      supplemented       by detailed    proposed legislative        amendments this month.
      Copies of these documents are enclosed              herewith.        As indicated      in
      these documents,        the Board has taken the position             that a substantial
      relaxation     of regulatory      controls    would, inter     alia,     result    in a more
      efficient     air transportation        system.    We believe      that the report        should
      take note of the Board's program.

              Finally,       we are concerned with the draft              report's        conclusion         that
      air fares could be reduced from present                     levels by several            billion       dollars
      a year through operating              efficiencies,               improvement           of load factors,
      seating     densities      and utilization          rates,    and the use of the most
      efficient        aircraft    type for each market.             The draft       report      assumes that
      optimum value% for each of these factors                      can be achieved in every
      market throughout          the domestic         system -- an assumption             which we believe
      is not realistic.            The system we have today is not a composite                           of
      mutually      independent        markets,      but a complex and inter-connected                    network
      designed       to maximize traffic           flows and, accordingly,             profits        throughout.
      An example is that the vast majority                    of short-haul        traffic,        perhaps as
      high as 75 percent,            is connecting        to some more distant            destination.            To
      rigorously         apply a combination          of GAO's optimum assumptions                  to this
      segment of the system could severely                    constrain     the ability          of this
      traffic      to move.


APPENDIX          XVII                                                                      APPENDIX          XVII
                                                                                                              .I     I.

    Mr. Henry      Eschwege       (3)

            The level of savings            to the consumer that is developed                 in the
    study from superimposing             optimum conditions           on the entire         domestic
    trunk system is also a source of concern                     to us D For example,            during
    calendar     1974 it is stated            that the consumer could have saved $1.9
    billion    if the optimum fares constructed                  in this report         were in effect.
    This level of savings is, in our opinion,                      overstated.        For example,
    the report     assumes the elimination             of all first-class           service,       and
    counts the difference            between first-class           and coach fares as a savings
    to the traveling         public.        Moreover,     the rate of return          found reasonable
    for the airline        industry      under free market conditions               appears too low
    to attract     capital     in a competitive          marketplace.          It is actually         lower
    than current      interest       rates.       In addition,      the Board’s       staff     has
    found several       mechanical       deficiencies        that also overstate            savings.
    The net result       of the above is to reduce 1974 savings                     from $1.9 billion
    as computed in the study to about $400 million,                         or by roughly        80
    percent.      We are certain          that there are other factors              that weigh
    both in favor and against               the level of computed savings,               but time does
    not allow further         analysis.

           Enclosed you will  find the detailed     comments of the Board’s   staff.
    If you have any questions     concerning   these comments, please contact      the

           We hope that in finalizing               the report  you will     find our comments
    constructive   and include    them,           to the extent    possible,     in any revision
    you may wish to make.

                                                                  Bureau of Economics

.)        APPENDIX XVII                                                                                                                                APPENDIX XVII

                                                                          Staff       Comments
                                                      “Impact  of                Increased     Efficiency
                                                          on APr               Fares and Travel"


               The staff               is      in     agreement                with        the         general            conclusion                 of       this        study

     that air             fares should                    reflect          an efficient                         industry            governed                to    the       ex-

     tent      possible              by competitive                       market           conditions.                       This        is        evidenced              by

     the     Board’8            comment8              on regulatory                      reform                and    the        Federal            Aviation              Act

     of      1975       presented              to     the        approprfate                    committees                 of     both         the     U.S House

     of     Representative8                     and U.S. Senate                          where            it     stated:

                “Economic              regulation                   should         be redirected                       so domestic                    air        transport

                is,       in       time,       essentially                   governed                 by competitive                     market             conditions.

                In      the     longirun              we believe                  this          can       result            in    a more efficient,

                lower-cost                  system          which         will          successfully                   respond                to     public           need8

                for       air       travel.”

                The Staff              is      also         in      agreement                  with       the        recommendation8                        of       GAO that

     the Board work toward the kinds of efficiencies                                                                        assumed            in     this           study.

     In      fact,        since        the          inceptfon             of      the Domestic                       Passenger                Fares         Investigation

     in      1970,        the       Board           has     moved         in      this          direction              and         now includes                      in    its

     rate-making                   procedure              adjustments                    for       load         factor,            seating            density,                 air-

     craft           utilization               and        discount               fare          traffic.               This        has         not     been           recognized

     by GAO, but                certainly                 should          be      In any finalization                               of this               study.

                However, the predl.cate                                for        GAO’s           conclusions                    and recommendation8                                  --

     an update                and     sanitization                   of      an earlier                   study        prepared by Dr. Theodore

     E. Keeler entitled                             "Airline           Regulation                     and Market Performance”                                    --       contain8

 APPENDIX XVII                                                                                                                                       APPENCIX XVII
numerous            untested               and,           in         some instances,                          unrealistic                  assumptions                 as well             I

as other            technical                  difficulties                            that        must       be made abundantly                             clear          to     the

reader.             While          at         first        blush,                 it        appears           that          the     consumer                could      have

saved         between             $1.5         and $2.0                    billion                annually            in     air        fares        for       the     period

1969-1974            had          the      airline                   industry                 been       a replication                     of      that        depicted

in     this     study,             it      must           be noted                     that        these           savings          are      extremely                sensi-

tive      to    the      particular                       assumptions                          imposed,              as well            as to        the       definitions

and procedures                     followed.-                  l/     For         example,                  savings          of     $l.?          billion            computed

by GAO for             1974             are         reduced                by almost                  80 percent                  to about           $400          million

if     only     corrected                     for      quantifiable                            deficiencies.

          There        follows                 a detailed                     comparison                     of      the     Board’s              rate-making

adjustments                 to     those              espoused                by GAO as well                          as     the        staff’s             comments              on

assumptions              and procedures                               used             in      the      study.

                       Standards                      Recommended                       By GAO and Those                           of      the      Board

          Included                in      the         Board’s                computation                     of      the     rate-making                    rate      of         return,

as stated             above,              are         adjustments                        for         load         factor,          seating           density,                aircraft

utilization                 and discount                            fare       traffic.                     A discussion                   of each            and     their

impact         is     set         forth             below.

Load      Fat tor

          In Phase                6B of             the        Domestic                 Passenger                  Fares          Investigation                     the Board
set     a long-term                     standard                    load         factor            of       55 percent               for         determining                 over-

all     revenue             need          together                   with         a standard                  variable               by distance                    (averaging

55 percent)                 for         determining                        the         fare        structure.                     The purpose                 of     this

I/    The sensitivity                            analysis  performed                                    by GAO demonstrates                                the impact
of the assumptions                              and should  be given                                    far greater  weight                           in      the final
product .

     APPENDIX XVII                                                                                                                          APPENCIX XVII

     lrtandard          i$     to ensure                 that      the consumer             doe8         not        bear       the         burden       of       excess

     capacity.                A comparison                  of     the     Board’8         etandard             with          that         assumed         in         this

     8tudy       is     Shown below.

                                                                                          Standard             Load        Fat tor
              Midpoint              Mileage                                                                                CAO/Keeler
                       100                                                                                                     60.0
                       200                                                        57.2
                       300                                                        57.6
                       400                                                        57.6
                       500                                                        57.6
                       700                                                        57.2
                      1000                                                        56.9
                      1300                                                        56.2
                      1600                                                        54.7
                      1900                                                        53.5
                      2200                                                        52.2
                      2500                                                        50.4
                      2800                                                        48.8
               Average                                                            55.0

               The Board’s                standard                by mileage             block     and in             total           is     applicable

     to aggregate                operations                 of aircraft              serving          all       city-pairs                   within          a

     given       mileage            block.               GAO’8 standard,                  however,             is     applicable                 to each              city-

     pair      in      each      mileage             block with the constraint8                                 that          DC-g-30            aircraft,

     will      operate           only         in     the         O-340     mile      interval,              B-727-200                 aircraft            will

     operate           only      in     the         341-900         mile      interval,            and DC-8-61                    aircraft            will

     operate           Only      in     market8             over         900 miles.

               Since          the      DO8ESti.C            Passenger             Fares      Investigation,                          the      structure               of

     the     airline           industry              hae changed              markedly.               Fuel prices                    have        more        than

     doubled           and      the     need         to consewe               fuel        ha8    become             paramount.                   Accordingly,

     the Board           ha8        docketed              a re-examination                   of domestic                   load            factor8        to

     determine               whether          the        current          standard          is   still          realistic                   or should             be
     changed           to reflect                  the     changing          structure           of      the        industry.                  Copies            of

     the     Board’8            order8         are         attached.

APPENCIX                    XVI I                                                                                                       APPENDIX                  -XVII.

Standard              Seats

           In Phase                6A of       the        above       mentioned                investigation                     the Board             determined

that       standard                seats       for        the     purpose             of      rate-making                  would        be based          upon

6-abreast                  densities           in     narrow-bodied                     jet       aircraft              and 9- and              lo-abreast

seating          in         wide-bodied               jet        aircraft.                 A comparison                    of    the Board’s               stand-

ard      for     wide-bodied                   aircraft             with        the        range        of      seats           actually          provided

by the          carriers               is     shown         below:

                                                                                               Number           of     Seats
           Aircraft                Type                                                    Standard                   High             Low
           -B-747                                                                              38%                      371            339
                 DC-10                                                                         276                      250            209
                 L-1011                                                                        276                      250            227

While          the         Board’s           standards             allow        for        galley            space         and mixed            configuration

of     first-class                   and      coach         seats,           GAO’s standards                     assume            complete             elimination

of     first-class                   service          and coach                densities               only          slightly           less      than          the

manufacturer                   ’ s maximums.                    A small         galley,               equal          to the        space        of      5 seats,

is     allowed               by CA0 in              the     B-727-200             and DC-g-61                   aircraft.                  No galley              space

is     provided               in     the      DC-g-30            aircraft,              a fact           that         is     discussed            in      some

detail          later,

Aircraft              Utilization

           To ensure                 that          the public            is     not        burdened             by the           under         utilization

of aircraft,                    an adjustment                    is made in                the        Board’s           rate-making               formula.

Each       carrier              has         been     held        to the          level         of      utilization                 it      achieved             in

each       of        its      aircraft              types        during          calendar              year          1972.         The effect                of

this       adjustment                  is     not     unlike           the      adjustment                   implicit             in    GAO’s          study.

A comparison                    of     the     utilization                    factors            is    shown          below:

*1    8PPENDIX         XVII                                                                   APPENDIX          XVII

                                                           Utilization       (Blk.     Hrs./Dsy)
              Aircraft   Type                              High          Low                   GAO
                 DC-g-30                                   10.43         9.48                  s.03
                 B-727-2&                                  10.82         7.42                  9.39
                 DC-8-61                                   10.90         8.56                 11.34

     Moat Efficient         Aircraft

              The Board does not directly               impose a specific           atandard for the moat

     efficient       aircmft.          However, the pressures            on coata created by the above

     adjustments       for load factor,          aeats and utilization              provide   incentives

     to the carriers            to make the mat         efficient      uae of their       aircraft    as poa-

     aible.       The incentives          here created have been further               strengthened        by
     the current       and recent-past          fuel    situation.

              The moat efficient           aircraft     type criterion        used by CA0 ia a standard

     of perfection          and one that ia practically                impossible      to meet at any given

     point       in time.       Further    comment on this           can be found in the section

     concerning        the assumptions         of the GAO atudy.

     Impact of Board Imposed Standards

              The savings to the public                generated by the Board's imposed standards

     can best be seen by comparing rate of return                         on an actual        basis   and as

                                                                Domestlc Trunks
                                                              48-State Operations
               Rate of Return                                    Calendar 1975
               Actual                                                 2.41%
                      Adjusted For:
                        Utilization                                       2.72
                        Load Factor and Seats                             5.38
                        Discount Fares                                    9.20

      Thus, the rate of return               used for rate-making           purposes reflects         an
APPEZilCIX                  XVII                                                                                                               APPENOIX                    XVII
                                                                                                                                                                             .I   1.

increase           of        .31        percentage               points           for         adjustments                of utilization                         and

2.66      points             for        adjustments                of     load          factor             and seats.                  An additional

3.82      points             is        added         by a discount                 fare            d sallowance.                       This        adjustment,

not     included               among           the      recommendations                        of        GAO, is designed                         to      ensure

that      the      normal               fare         payer       does       not         cross-subsidize                         the      carriage               of

discount           fare            traffic.               The cumulative                       impact            of     these          adjustments                    is

a $750 million                         saving          to the          consumer.

          It      is        interesting                 to note           that          the        spread           between            reported               revenue

and     the       "increased                   efficiency"                revenue              estimated                in      this        study            narrows

considerably                   over           the      period           1969-1974.                   From an excess                      of 46 percent

in     1969,       reported                   revenue         is       shown       to         be only            28 percent                 in     excess             by

1974.           A major                explanation               of      this       18 percentage                       point          reduction                is         the

standards               now imposed                    by the          Board        in        determining                 air         fares.

                                               Assumptions                of The GAO/Keeler                             Study

          The basic                    assumptions               of      this       study            --      high        load         factors,               dense

seating,           high            aircraft             utilization                 and use                of    the      most         efficient                aircraft

type       --     appear               reasonable             when considered                             individually                 or when               applied

in     isolation                  to one            particular            market.                  However,              the     ability                to     obtain

a combination                      of     these          over      every          domestic                 route        is      remote.

           The domestic                       route       system           is     a complex                  and       interconnected                        network

of many diverse                         markets.              As such,              to obtain                   a combination                     of      the        above

assumptions                  in        every         market           situation               is     not        a simple              task.             For

example,               in     thinly            traveled           markets               it    may be possible                         to        obtain         high

load       factors                in    densely           configured                aircraft,                   but     not       at     the           utilization

rates           assumed            in     this         report.             Further,                 in     these        same markets,                         the     use of

           APPENDIX       XVII                                                                      APPENDIX               XVII
.1    10

     the most efficient            aircraft       as dictated          by GAOmsy constrain              the ability

     to achieve high load factors.                   That is,          smaller     aircraft      that more closely

     match the demand in lighzly                  traveled      markets would not be available                        since

     their     unit    costs would be higher               than those of the most efficient                    air-

     craft     as assumed by GAO. These are but a few examples, and many more

     could be cited.             However, they are sufficient                    to shed considerable              doubt

     on the ability         of the industry          to generate optimum operations                       on all

     segments of an integrated                  system.

              More detsiled        comments on each of the assumptions                        of this       study

     follow     below.

     Load Fector

              A basic premise of the Keeler study and GAO's update is that                                      a

     sixty     percent passenger load factor                   could have been obtained                   in 811

     domestic         trunk markets        if   free market conditions                had prevailed         during

     the 1969-1974 period.               This assumption              is based upon the experience                    of

     PSA for 1959-1970 and Southwest for 1974 and 1975.

              Apparently     Keeler      had no difficulty              in making 8 direct           loed factor

     correlation         between an isolated              short-haul,         point    to point,     densely

     traveled,         sometimes monopolistic              airline      and a multi-company,               varied

     haul,     integrated        system.        Though GAO raised             some questions        regsrding

     this     assumption by stating              that while          "the California          experience       ...

     may not be representetive                  of long-run          price    competitive       airline      in-
     dustry      experience",        they felt      that      the results         of operations           in the

     intrastate         Texas markets during              1974 8nd 1975 could justify                retention

     of the 60 percent            load factor       assumption.              GAO apparently        gives     little

     weight           to the fact     that Southwest operates                  a monopoly on 8n airport-
       APPENDIX                   XVII                                                                                                       APPENDIX                XVII   '

to-airport                basis.            The real           problem,            however,             does        not     pertain            to     the

basis         for      establishing                  the sfxty            percent           load         factor           standard,            but      to

the      ability            of     the      carriers           to    obtain          this        level       over          their         system        fiile

adhering             to     the      constraints               imposed          by       the     study         on    aircraft                selection,

seating             density,             and    aircraft            utilization.

          The        fact         that      only       three        aircraft             types are employed leaves manage-
ment      no        flexibility                to    serve      markets            with other aircraft                             of    more        optimum

size      and        cost         characteristics.                     That        is,      markets            that could be mre
economically                   served by smaller                     or    larger           aircraft              would       be penalized                      by

the      selection                of only the three types.                                Less dense markets would                                   suffer

through frequency                         reduction            and     possible                loss of traffic                     to other modes

of transportation.                             More densely traveled                            markets would pay through                                   a
surplus             of departures                   and unneeded congestion.                                The problems that                         the

constraints                 on aircraft                selection           generate are even more evident                                            when

considering                 the introduction                    of wide-bodied                     aircraft               during         the ftr8t

year of this                   study.           Whereas         GAO would                only allow               operations                 of air-

craft         with 110,              153,       and     239     seats,          the trunk carriers                           during           1974      had

available              a variety               of aircraft             with seats ranging                           from 67 to 370 per


          A    second factor                    which constrains                     the achievement of a straight

sixty         percent             load factor              is the assignment                       of     each       aircraft                type to

specific             mileage blocks.                       Keeler's            plan would effectively                               elilninate              all

through plane service                               from secondary cities                          to distant                points.                Under

his assumption,                      the short segment from the secondary point                                                         to    the     hub

would be performed by a X-9-30                                         and the segment from hub to distant                                                  hub

would utilize                     a 727-200 or DC-8-61, thereby requiring                                                  a change             of    plane.
.I       ASPPENDIX XVII                                                                                   APPENDIX            XVII

     Neither     Keeler nor GAGmade any attempt                       to quantify         the increased
     reservation        and sales and traffic               servicing       expense (passenger,                 baggage,

     and cargo) that would result                  from additional           deplaning       and enplaning
     caused by reductions                in through plane service.

           The following               chart displays       a representative          sample of the

     common carrier         flight        scheduling       practice      of building        load factors

     for long-haul         operations         by providing       through plane service                   with     large

     equipment from secondary cities                    to hubs.

     Carrier           Flight      #                                   Itinerary

     United                59               Kichmond9LbWashington                     2288>Los                  Angeles

     American              73                           244 mi. Cleveland----,2053 mi. Los Angeles

     Eastern       *       11               Boston %!!'>Providence                 emi;       New York log7 mio~ Miami

     TWA                                                       134 mi.                       2288 mi.
                                                                                             11113Los                   Angeles
                           99               Philadelphia       ?Washington

     This produces lower load factors                      in the short-haul           but compensates by

     increasing        long-haul         load factors.         Secondary cities            are thereby            receivers

     of mgre frequent             and less time consuming block-to-block                      service.             It     also

     provides      local     passengers on the first                  segment additional            service        fre-

     quency.       This process would not be possible                       under GAO's assumptions.

              The third         condition     which makes the sixty                percent load factor                  un-
     realistic         is the extremely          high aircraft          utilization        rates     assumed.

     GAO diverted          from Keeler's         methodology here and selected                     the
     highest     utilization            rate for each aircraft             type experienced              during         the

     APPENCIX                 XVII                                                                                                                 APPENDIX                       XVII   8,

years       1969        to     1974.               This         assumes            that          there            is     no       inter-relationship

among aircraft                      types.               Changing           routes,               aircraft                 mix,       purchases,                    retire-

ments,         length          ofhaul              ) traffic              growth,                maintenance,                     charters,                   and

positioning               flights               are       all     contributing                        factors              to aircraft                   utilization

for      a given         period               of      time.           Therefore,                     the     dynamic              nature           of         these

factors          does         not         allow          individual                aircraft                 utilization                  rates            to be se-

lected         from       differing                  years.            Additionally,                          CA0 selected                     utilization

rates       from        a data             base          that     reflect                actual             operations                  over       a greater

variety          of     stage             lengths             (especially                  longer             stage           lengths)             than            have

been       considered                optimum              under        GAO’s own criteria.                                       The result                   is    that

an aircraft               is        placed            into       average                daily         service              for       longer           than          is

perhaps          physically                   possible.                For         instance,                  based           upon       (a)       GAO’s

assumptions               of        a DC-g-30                being        utilized                   in     all         markets          of       0 to         340 miles,

(b)       average         passenger                   trip       distance                of      266.2            miles,          (c)        average               air-

craft       stop        and         cruise            times,          and         (d)      a turnaround                       time       of       forty            minutes

between          flights,                 the      average            aircraft                  would         be employed                    in    revenue                 ser-

vice       an astounding                      average            of      18.4           hours             (ground          and air             time)           per         day.

Beside         physical              capabilities,                     a utilization                          rate         this         high       becomes                 more

untenable             when          considering                  the      curfews                which            exist          at many metropolitan

airports.               A review                of       PSA’s        current                 time         tables          shows         that           all        regular

flights          are      scheduled                   between            7 a.m.            and 9 p.m.                      This         is     only           a 14 hour

period         for      all         airline              operations.                     PSA’s             individual                aircraft                 utilization

is      even     less         because              all       planes          are         not         used         every          hour        or every               working

day       (especially                weekends).

           men        utilization                     rates        are       as high                 as indicated                    here,         excess                 capacity

is      provided          at        off       peak        hours.             It         is hard             to         imagine          that       short-haul

III    APPENDIX                XVII                                                                          APPENDIX         XVII

      passengers would be willing                            to travel    at four in the morning or at mid-
      night      as would be necessary to achieve a sixty                                percent load factor             at

      the utilization                 rates        assumed.        It would also be difficult                to acconnnodate

      all     passengers at peak periods because of the inability                                      to substitute

      aircraft          types during               prime time.

               We again wish to emphasize that a 60 percent,                                      system-wide     load

      factor      is probably                  both reasonable and attainable                if     management has

      the flexibility                 to match capacity              and its     use to the demand of every

      market.           It     is only when management's flexibility                         is limited         by condi-

      tions      like         those imposed in the GAO study that                         a 60 percent         load factor

      becomesdigficult                    if    not impossible         to achieve while,            at the same time,

      providing              sufficient           service      to meet all      passenger demand.

      Most Efficient               Aircraft             Type

               Under the Keeler                    and GAOmethodology,            all     operations         are carried

      out by what are labeled                           "the most efficient"            plane types;         the DC-9-30,

      for short-haul,                 the B-727-200 for medium-haul and the DC-8-61 for                                  long-

      haul.       How each trunk                   carrier       could be operating         the most efficient

      plane types at one point                           in time is left        unanswered.          In the real         world,      '

      the possibility                 of this           happening seems remote.             Just as in other             compe-

      titive      industries                   today,    and for various        reasons,      all    firms     do not have

      the same or the very latest                              in equipment.

               The three aircraft                       types selected       for the Keeler/GAO model are
      second and third                    generation           regular-bodied     turbofans.           The DC-g-30 fol-
       lowed the DC-9-10,                       the B-727-200 followed           the B-727-100 and the K-8-61

 APPENDIX XVII                                                                                                                             APPENDIX XYXI                  /I

followed           the         DC-8-10,               20,      30,      and 50.             The three           aircraft               types         selected

built       upon         and         profited               from      the     operating             experience             of         predecessor

aircraft           0     If      carriers               had not            purchased           first         generation                turbofans,                   the

%ore        efficient”                  second           and third              ger.eration             equipment              would          not      have

been       developed.

           Purchase              of     aircraft               is     usually         a long-term               commitment.                      The only

way all           carriers              in      the         industry          could         have       the     latest          equipment               is      if

they       sold        “old”          and purchased                     new equipment                  every       couple             of    years.             This

would       put        such          a glut           on the          market         that      the      excess          two and             three          year

old       planes         would          probably               be sold          at    a loss,           a loss          that          would         have       to

be absorbed                   by the         paying            passenger             or the         stockholders.

           Use of             only      the        three           “most      efficient”               plane       types         in        the      Keeler/

GAO model              seems,           therefore,                   an unrealistic                 assumption             and seriously

understates                   the     cost         of       producing           airline            service         under         a deregulated


Rate       of Return                 on Capital

           To derive                 an applicable                    rate      of    return           on capital               for        the      airlines

under       competitive                  market              conditions,              Keeler           concluded           that            the      pretax

airline           rate         should           equal          that         of an all-corporation,                         long-run,                 historic

average.               In      reviewing                Keeler’s             work,        GAO accepted              these             definitions

and conclusions                       but       found          errors         which         lowered          the    pretax             estimate              of

return        on capital                    from        12 to          10.5     percent.               The major           question                 here       is

       APF;‘ENDIX               XVII                                                                                                                 APPENDIX                    XVI I

     not       whether          either              percentage               is        the      correct           all-corporation                         long-run

     average,            but      if      this            average         would              be applicable                    to the           airlines            in      a

     free       mafket          situation.                       The all-corporation                             average            is      just       that;         a

     center        point          of      a spectrum                    which          reflects            the      average               rate       of numerous

     industries                (comprised                   of companies                     providing            services,                 producing              goods,

     and in        some cases                   both),            varying              degrees           of      regulation,                   contrasting

     debt       to equity               relationships,                       and differing                       levels          of risk             and        compe-

     tition.             Further,                  this         average           is        constructed                over      the        27 year             period

     1939       through           1966 where                     inflation                  was generally                 under           control            and        interest

     rates        were         low.           It      is,        therefore,                  not      representative                      of       today’s         world
     where        uflation                is         a major            problem              and interest                 rates           themselves                 are

     higher         than        the       long-run                 average              return           used      in     this           report.             Thus          the

     conclusion                that       the         airline             rate          in     a free         market           situation               would            approxi-

     mate        or equal              this          average            is not              realistic.                 The very             nature           of      the

     airline’industry                          (cyclical,                 oligopolistic,                        capital          intensive,                  etc.)          leads

     one to believe                      that             the    rate        of return                required            to attract                  capital            would

     be considerably                      higher                than      this          average.                This      theory            was supported                        by

     various         rate          of     return                studies           in         Phase       8 of      the        Board’s              Domestic
     Passenger             Fares          Investigation.                               In     those       studies             the        after-tax              rate        of

     return         ranged             from          10.5         to     13.5          percent           based          upon        “optimum”              debt         to equity

     ratios,         interest                  on debt             approximating                      6 percent,               and a cost                  of      equity             of

     approximately                      16 percent.

APPENDIX          XVII                                                                            APPENCIX          XVII   I1

       The 12 percent          after-tax         return       found most aIe)ropriate               for
regulated       airlines      in the Phase 8 decision                   of the Domestic Passenger

Fare Investigation            is equivalent             to a 15 to 20 percent pre-tax                     return
depending on the assumption made ,regarding                             the effective            tax rate.         The

argument could be made that,                    for a deregulated              airline      industry,        ii 15 to

20 percent pretax            return     might have to be increased                       to attract       capital

because of the greater                risk     involved       to the investor.              If    an 18 to 24

percent pre-tax            return     is required,         the "savings"           to the passenger under

deregulation,        which GAO has developed,                      would be cut by between $300 and

$900 million.            (See page 31 of GAOstudy.)

       Further,      GA6 has understated                 total      aircraft      investment          due to

unrealistically            high assumptions of load factors,                       seating        densities        and

utilization        rates.       Considering         that      it    usually     takes approximately                 two

to five       years from the date when an individual                           aircraft          is ordered to

the day it        is placed in service,                 Reeler and GAO are saying that                     they

have the ability            to predict         future     passenger traffic               with such precision

that   they can order capacity                  consistent          with   a 60 percent            load factor

at the optimum utiliztion                    rates assumed.

       All     of these factors              produce an understatement                   (in terms of

investment        and related         return)     of the aircraft              which are needed by the

trunk carriers           to carry      the current          level     of revenue passenger traffic.

      APPENCIX XVII                                                                                  APPENOIX XVII

                                                 Consumer Savings

             GAG concludes         that      the consumer would have saved between $1.5 and

    $2.0 billion          annually         had average fares been equal to those determined

    in this      study.        The savings are markedly                overstated,      however, even if

    one accepts many of the hypothetical                        assumptions     of the report.                 Among

    the more important             reasons are:            (1) "forced     savings'*      imputed for first-

    class travel;            (2) an incorrect         combination        of total      trip     fares     and

    flight      basis passengers;             (3) the application          of 48-State          fares     to

    50-State       passengers;         (4) the omission          of added costs through inferior

    quality      of service;         and, (5) miscalculation              of stewardess expense.
    "Forced Savings"             From First-Class           Travel

             The savings         in this      study reflect          the difference       between:          (1) the

    sum of acfual            first-class       and coach revenues received                by the trunk            air-

    lines     for each year 1969-1974; and (2) the revenue that would have been
    received       had all       passengers (first-class               and coach) used the "increased

    efficiency"        fares      developed in this            report.      Thus, included             in total

    savings are "forced              savings"       from first-class         travelers         assuming they
    useithe        "increased       efficiency'       fares.

              This seems to us an erroneous use of statistics.                                To impute "forced
    SaVingS" for first-class                   travelers      from a cheaper and lower quality

    service        is like     arguing        that a person who purchases a Cadillac                      could

    have saved thousands of dollars                        by buying a Volkswagon.              It     is not

    a meaningful          calculation.            There is a unique and distinct                     demand for

     first-class        service,       just     as there is for Cadillacs,               and this        demand

    will      be satisfied         whether the system is as presently                    conceived

APPENDIX           XVII                                                                        APPENDIX             XVII

or in any other               form.      Therefore,       it     would appear that the "forced

savings"          attributed          to first-class           travel     must be removed from GAO's

statistics          before meaningful             comparfsons can be made.
           The error       resulting        from including              the "forced     savings"        of ffrst-

class travel             is shown below:

                                      Savings ($ Bil.)                           Savings (X)
         Year                      GAO         As Revised                     GAO         As Revised
         iis                      $1.6            $1.4                        46             41
             70                    1.5              1.2                        38            30
             71                    1.7              1.4                        40            33
             72                    1.8              1.5                        38            31
             73                    2.0              1.6                        36            30
             74                    1.9              1.6                        28            24

Dollar       savings       are,    therefore,          reduced by $200 to $400 million                    or 4 to 8

percentage          points.        The revision          is based upon the assumption                    that all
passengers (both first-class   and coach) moving in each of the years used
actual coach fares.    Though more logical than the assumptions of GAO,

the correct           procedure         would    have    been     to     develop     optimum    costs     and fares

for first-class               and coach separately               and derive         savings    for each class.

Total       Trip,Fares         and Flight        Basis Passengers

           Passenger savings             are computed by GAO for each year in the study as


           (1)     the ratio       of actual       revenues reported                by the carriers to
                   "estimated"          actual    revenue        developed         by GAO is computed;

           (2)     "increased       efficiency"          revenues developed             by GAO are
                   adjusted       by this ratio;          and,

           (3)     savings are then defined as the difference between overall
                   revenues reported by the carriers and "increased efficiency"
                   revenues as adjusted.

APPENDIX        XVII                                                                 APPENDIX           XVII

An example of this           procedure for calendar          1972 is shown below:

              Revenue ($ Ml.)                           Calendar 1972
                   Reported                                 $6.7
                   "Estimated"                               6.8
                      Ratio                                  0.98529
                 "Increased Efficiency"
                    Unadjusted                              $4.9
                      Above Ratio                            0.98529
                    Adjusted                                $4.8

                 Savings                                    $1.9

Any error       in GAO's "estimated"         actual     revenue,      therefore,        is translated
by this       procedure     to both the adjusted          "increased      efficiency"       revenue

and savings.

           Average fares used to compute "estimated"                  actual     revenue are

 developed from:           (a) published     coach fares:          (b) first-class        fares     esti-

mated a; a ratio           of 1.3 times the appropriate              coach fare;        (c) first-class

 and coach revenue passenger-miles               as a percent         of total;       (d) first-class

 and coach discount          fare dilution      factors     from Phase 9 of the DPFI; and,

 (e) across-the-board          fare changes that          occurred      during the period.              They
 can be defined,therefore,as            total    trip     fares.      That is'       they are based

 upon weighted          average costs of non-stop,          multi-stop         and connecting

 0 & D trips.
           Passenger aggregates by mileage block               used with        these fares        to

 estimate      actual     revenue, however, are on a flight               basis.        That is,
 passengers moving beyond the destination                   of a given flight            are not

 included      in the mileage block of the origin                  and destination        of the

 flight.       The impact here is to overstate              the number of passengers                in the

 short-      and intermediate-haul         and understate          those in the long-haul.

APPENDIX            XVII                                                                              APPENDIX           XVII

For     example,       a passenger originating                      in New York ,destined for Los

Angeles but connecting                       in Chicago wculd show up twice in a flight                           basis

matrix      of passengers-- in the New York-Chicago                              mileage block and in the

Chicago-Los Angeles mileage block.                                But in a total           trip     matrix     of passen-

gers he would appear once in the mileage block of his true 0 & D--

New York-Los Angeles.

         Total      trip          fares combined with              flight     basis pessengers,              overstate

actual      revenue          since each connecting                  passenger is charged the costs of at

least     an additional                   departure     and a line-haul          rate that is higher               than
that of his true 0 & D.                         For example, during             calendar          1974, the origin
and destination                   trip     length     of all     connecting      passengers was 1050 miles

but the flight              stage trip              length was only 532 miles.                    GAO's methodology

would charge the connecting                           passenger on average for z                      532 mile     trips

(a conservative               assumption             that each connecting           trip      has two legs) whereas

the correct          procedure would have been to charge him for one 1050 mile                                             trip.

As calculated              by GAO the fare for a 1050 mile                       trip      is about $72 while              that

of a 532 tile              trip          is about $45.          Accordingly,       revenue received              from con-

necting      passengers is overstated                          by 25 percent       (2 x $45 f $72) and, since

connecting          passengers represent                      30 percent      of total      travel,      total     revenue

is overstated              by roughly           7.5 percent.             This would amount to a $600 million

reduction          in 1974 revenues as computed by GAO before validation                                         and a

$500 million           reduction             in associated          savings     after      validation.

48-State         and 50-State               Passengers

        A definitional                    inconsistency         exists      between the fares and passengers

used to derive              "estimated"              actual     revenue.       Fares are 48-State.                That is,

the city-pairs              used and the changes in normal fares adjusted                                    for over the
 APPENDIX XVII                                                                         APPENDIX XVII
#period are         for the 48-State,         rate-making     entity,.        Passengers,    on the

other hand, are SO-State for all                 years except 1969.             They contain     opera-

tions      of the 48-States,        Mainland-Hawaii,         Mainland-Alaska         and Alaska-Hawaii.
        Mainland-Hawaii          passengers,     the most substantial            part of the added

traffic,       move in the "2400 mile           and over" mileage block at substantially

lower fares         than their     48-State     counterparts."            Accordingly,      by applying

48-State       fares     to these passengers,          revenue is overstated.
        During calendar          year 1974, for example, 3.8 million                 passengers

moved between the Mainland and Hawaii generating                           $401.4 million      of revenue.

This produces an average fare of about $106.                         GA.], on the other hand,

assumed that           these passengers would have paid a fare equivalent                      to that

computed for . the New York-Los Angeles market of $143.83 generating

approximately $549 million    of revenue. Thus, GAO's estimate of revenue

before validation           and savings after          validation        is overstated      by about

$148 million.            This bias exists       to a similar         extent    in each of the other

years except 1969.

Omissiod of Added Costs Through Less Service

           The dollars      of savings assume that          all     passengers carried         during

the years 1969-1974 would have moved at the lower fares constructed                                     in

this       study.      But for this    to occur under the constraints                 imposed on load

factor,       seating     densities,    utilizations        and aircraft         usage, frequency            and

convenience of service             would have been substantially                 inferior    to that

actually        provided.

L/      The breakdown of 50-State traffic                 by percentage         for calendar     1975
                    Entity                                Percent of Total
                    48-State                                     92
                    Mainland-Hawaii                               7
                    Other                                         1
 APPENDIX         XVII                                                                        APPENDIX              XVII

         Each traveler        using the airways             places some value on his               time.        ,          ,(
The businessman places a premium value on his                              time, while       the pleasure

traveler     may value hfs time less.                    No matter       which, hov:ever, the less

frequent      schedules and delays associated                     with     the assumptions        of this

study have a definite              cost to the consumer.                 Such costs should have been

estimated      by GAO and offset              against     the derived         savings.

         Dr. George W. Douglas and Dr. James C. Miller                           III,      in their      recent
book entitled,           Economic Regulation             of Domestic Air Transport:                    Theory

and Policy,       performed an exercise                 similar     to that OF this           study but with
one major exception.               Savings generated by an optimum fare structure                               are

offset     by the costs associated               with     reductions        of service.         They conclude

that     the consumer, during              1969 paid excess fares ranging                  from approximately

$366 to $538bmillion.               For this       additional        price,     however, the consumer

purchased reductions              of delay time through superior                     service--valued

approximately           at $118 million         at an assumed value of time of $10 per

hour and $182 million              at a value of time of $5 per hour.                         Thus, the cos,ts

of additional           service    could range from 22 to 50 percent                       of the excess

fares paid to support              this      service.

         Had GAO's computed savings been offset                          by the costs of less frequent

service,      similar      to the percentages             derived        above, the results            would have
been far different.               The $1.5 to $2.0 billion                 range would have been markedly


Miscalculation           of Stewardess Expense

         The costing       technique         used by GAO in this            report       duplicates,       for the

most part,       that of the Board.               Many of the cost inputs,                 in fact,      were pro-

vided by the Board's              staff.       One of these-- the ratio                 of stewardesses             to

available      seats--has         been misinterpreted             and misapplied           by GAO.
     APPENDIX       XVII                                                                 APPENDIX          XVII
I*     3.

             The number of stewardesses assigned to a particular                       aircraft      type

     in the Board's        costing     methodology is based upon the ratio                of steward-

     esses to available         seats.      For coach service,        the ratio        is one stewardess

     for every 45 seats.             GAOmisinterpreted        this   as one stewardess            for every

     45 passengers.         Accordingly,        GAOhas understated          both the number of
     stewardesses     and their        associated    costs by 40 percent.

                                                                      Percent Understatement
             Aircraft         Number of Stewardesses                         of Number
             Type              GAO                GAB                        and Costs
             DC-g-30          ix7                2x                            40%
             B-727-200         2.04              3.40                           11
             DC-8-61           3.19              5.31                           11

             The impact on total         expense is an understatement             of about two
     percent,    since stewardess expenses are approximately                    five     percent     of
     total    exnenbes.       Thus, in 1974 for example, GAO's "increased                    efficiency"

     revenues are understated            and savings     overstated     by about $150 million.


             The savings      to the consumer computed by GAOwould be considerably

     lower if    adiusted      for the quantifiable          deficiencies      discussed      herein       and

     adjusted    to reflect      a more reasonable        rate of return.          For example, as
     shown in the following            table,    the savings     for 1974 would have totaled

     about $402 million         or only 21 percent of the $1.9 billion                   estimated

     by GAG.

     APPENCIX      XVII                                                               APPENDIX          XVII

                                                                        Calendar 1974
         GAO Savings ($ Mil.1                                                $1900
           Reduced for:
           (1) "Forced Savings" - First Class                                   300
           (2) Mixing Total Trip Fares and
                  Flight Basis Passengers                                       500
           (3) Mixing 48-State Fares and
                  !50-State Traffic                                             148
           (4) Miscalculation     of Stewardess
                  Expense                                                       150
           (5) An 18 Percent Pre Tax Rate
                  of Return                                                     400

           Savings as Adjusted          ($ Mil.)L/                            $ 402

         These adjustments        do not include       many items that cannot be quantified

within     the time constraints          of our comments.         For example, the added

costs to the consumer from less frequent                     service   as postulated         by the GAO

assumptions      are not included.           Additional       passenger handling       costs

(including  food)
               . resulting from increased passenger enplanements and

deplanemeqta have not been estimated.   On the other hand, no dollar value

has been placed upon the deficiencies                  in the passenger-mile          densities

used by GAO and the overstatement                 of passenger enplanements both of which

would probably         increase    estimated      savings.      There are undoubtedly           other

factors      th&t would be affected          by the assumptions of this             study.      However,

just     the differences      cited     above are sufficient           to warrant    major revisions.

         Differences      in savings     as computed by GAO and as corrected                  by the

Board, should be similar              in m&tude       for each of the other years included

in the study.

lJ     The adjusted savings are somewhat understated since the effects of
       items (2) and (3) are not directly    additive.  They would tend to
       compensate each other in the mileage block affected by Mainland-
       Hawaii traffic.   The understatement,   however, should be slight.

     APPENCIX        XVII                                                                APPENDIX        XVII
                                  Other Technical       Co-mments

         Several     problem areas in addition          to those discussed            in connection
with revenue and savings            computations       deserve comment.           Included     are:

(1) densities         assumed in the construction              of estimated      passengers;

(2) comparisnJn of estimated           passengers to actual;                (3) the application

of elasticity;         and (4) the need for food service                  in short-haul      markets.
Assumed Densities

         Actual     passengers for each year 1969-1974 are estimated                       as follows:

(1) the domestic            trunk system is disaggregated            into     42 distance/density

cells ;.!L/ (2) the percent of non-stop                revenue passenger-miles              per day in

each cell         is computed from a sample of 257 city-pairs                   provided     in Phase 6

of the DPFI; (3) the percentage for each density groupfng within each
mileage bl9ck is adjusted by the actual mileage distribution  of flight

basis passenger-miles;            (4) the adjusted           percentages      are applied     to each

year's      total    revenue passenger-miles          to obtain      the number in each of the

42 cells;         (5) the passenger-miles          in each cell      are divided         by a weighted

average firbt-class            and coach trip       length     to yield      estimated     passengers.

         Densities      used in this    methodology are not representative                   of the

domestic      sys tern.      The percent of revenue passenger-miles                   per day generated

in each of the six density             intervals      for a given mileage are derived

from a sample of 257 city-pairs              provided         in Phase 6 of the DPFI.            Under

the ground rules            set down in Phase 6, this           sample reflects          a combination

of the top 20 markets served by each carrier.                        The tendency here,           there-

fores would be to understate             the percent          of passenger-miles          generated      in

A/     A 7 x 6 matrix was developed with                7 distance        intervals      and 6 density
       intervals within each distance.

         APPENDIX      XVII                                                                         APPENDIX       XVII

    thin markets and overstate                  that of the more dense markets.                     For examples

    this     is demonstrated       by a comparison of the percent                        of non-stop     passenger-

    miles by density          and distance         interval        actually       experienced       during

    calendar     1974 to those used by GAO.
                               Percent Non-Stop Passenper-Miles Per Day By Density L/
           Mileage             O-300 Psgrs. Per Day          301 Plus Psgrs. Per Day
           Block               GAO           Actual         GAO
                                                            --                Actual
               O-400          0Y-E             8.00        13.04              12.20
            401-800           0.63             8.50        22.46              15.30
            801-1200          1.44             7.50        18.46              10.00
           1201-1600          0.79             6.20        10.55              10.30
           1601-2000          0.79             2.50        13.36               5.50
           2001-2400          0.51             3.40         4.23               3.31
           2401 Plus          0.19             3.00        12.97               4.40
               Total          4.87            39.10        95.07              61.61

    The percent     of revenue passengers generated in the less dense segments

    by the GAOmqthodology               is therefore          grossly     underestimated.            The net

    impact on "estimated"              actual     revenue, "increased                 efficiency"    revenue and

    savings     is unclear.        Initially        it would appear that GAO's revenue would

    be understated        since the bias implicit                  here tends to understate              short-

    haul traffic       and overstate            that of the long-haul.                  However, some correction

    of this     was made by adjusting              the percent          of non-stop         passenger-miles        by
    the mileage distribution              of actual       flight        basis     passenger-miles.           But

    whatever and in what direction                  the remaining             bias,     the deficiency       should

    be corrected.
    Actual     and Estimated       Passengers

             A comparison of actual              passenger enplanements reported                    by the carriers

    to those estimated          by GAO is shown below for calendar                          years 1969 and 1974.
                                                                   Passengers Enplaned (Mil.)
             Calendar Year                                         GAO                 Actual
                 1969                                              132                  126
                 1974                                              155                  148
    lJ     The percents       differ     from 100 due to rounding.

APPENDIX                 XVII                                                                                                                         APPENDIX                     XVII

In both         years,                  GAO’s estimates                          are        about         5 percent                  overstated.                      This

builds         an upward                  bias          into         GAO’s revenue                        estimates.                       The degree                 of         bias,

however,             would              depend          upon          the        location                of     the      passenger                    overstatement

within         GAO’s            42 cell              matrix.                 Some correction                           procedure                  should             also         be

devised          for          this          deficiency.

Application                   of        Elasticity

            GAO has             added             a new feature                        to    the         study         originally                     prepared              by

Dr.       Keeler.               That         is,         the         numbers            of       new passengers                           that        would          have          been

.induced         into           the         air      travel            market               by     the         lower          fares          are        determined

using         selected                  price-elasticities.                                  A technical                      problem              exists            with          the

application                   of        these        elasticities.

            Price            elasticity                  is     defined                as the            ratio          of     a 1 percent                      change             of
price         to-the               corresponding                      change            OF quantity                     (in         this         instance,                 traffic).

It     is     the        science              of     small            changes.                    The application                          of a point                  estimate

of elasticity                        to     successively                      larger              arcs         of      the     demand              curve         or,         stated

differently,                       larger          percentage                    changes             of        price          create             at     least          two

problems            :        (1)        an assumption                       of     linearity;                    and,         (2)         the      ambiguity                 of

percentages                    changes.                  If     we except                    the     presumption                      that         elasticity                     is

 linear         or constant                       over         all      increments                   of        the      demand              curve,          a percentage

 problem            still            exists.                  That      is,        depending                   upon          the     base          used         to     determine

 the        change           of price,               where            the        increments                    are      large,              the         changes             of

 quantity               at     a given             elasticity                    will         vary            substantially.                          Many       conventions

 have        been        developed                 to         correct            this            percentage                  problem              but     perhaps                 the

 most        logical               is     the      use of             logarithms.                         This         not         only          corrects             for         the

     APPENDIX      XVII                                                                            APPENCIX            XVII
percentage ambiguity             but ensures that               the condition              of unitary     elasticity

(e.g. unchanging revenue regardless                        the change of price)                  is met.

          The percentage        problems exist           in the change in traffic                    determined

by GAO that would have resulted                     from its            lower fares.           For example, GAO

shows that a 37 billion                 or 41.6 percent               increase        in passenger-miles         would

have resulted         from the 31.5 percent reduction                            in average fares         for 1969

assuming an elasticity                 of -1.3.      Done correctly,                 however, the increase              of

traffic      would equal about 64 percent                  not 41.5.1/                 The proof is simple.
If unitary       elasticity        is assumed, the GAOmethodology would have pro-

duced a 31.5 percent              increase        in traffic,           and revenue would be only

90 percent       of the original           amount (131.5 x 68.5 = 90.1).                          This is

inconsistent       ,with the definition             of unitary               elasticity.       Using the loga-

rithmic      method, howevar? the change of traffic                               would be about 46 percent

and revenue would be unchanged (146 x 68.5 = 100.0).

Food Service         In Short-Haul         Markets

          GAO assumes that         there is no need for galley                         space on the DC-9730

aircraft-assigned             to short-haul         markets (those within                    the "O-340 Mile"

interval).        This assumption            is predicated              upon the proposition              that

traffic      moving in short-haul             markets does not need meal service                           since

time on-board is minimal.                  Indeed, if           all     passengers moving in the short-

haul were turnaround,              0 & D passengers,                  this     might be true.           However, the
facts     are that     the vast majority              of these passengers,                   perhaps as many as

75 percent,       are inter-lining            or using the short                    stage to connect to their

final      destination.         Further,      under GAO's assumptions                       of aircraft     use,
connecting       services       will     become even more dominant.                         Thus, meal service

L/      The 64 percent is derived from the following  equation:
        LogT = 4.6 -1.3 LogF, where T and F are expressed as indices                                        at
        base 100.
     APPENDIX XVII                                                        APPENDIX XVII

      might well be required      for this    segment of the market.

            It   should be noted that a contradiction         exists   in GAO's assumption

      concerning    the use of the DC-g-30 aircraft.          On the one hand, no galley

      space is allowed,    but,   on the other,    food expense is included.          The

      expense input    is a weighted average of all      food service      provided    the

      short-haul    passenger including      that of meals.

 APPENDIX             XVII                                                APPENDIX   XVII
                             OFFICE OF THE SECRETARY OF TRAW=ORTATION
                                        WASHINGTOW.   D.C.   20590


                                                                     July 20, 1976

      Mr, Henry Eschwege
      Director, Community and Economic
           Development Division
      United States General Accounting Office
      Washington, D. C, 20548

      Dear Mr. Eschwege:

      Thank you for the opportunity. of commenting on the draft study
      entitled “Impact of Increased Efficiencies on Air Fares and Travel. ”
      I believe that your draft modeling effort for the domestic airline
      industry is an extremely thorough and well done analysis of the
      inefficiencies presently existing in our scheduled air transporta-
      tion system.

      Your staff has done a careful job of reviewing the methodologies
      used in Dr. Theodore Keeler’s earlier model along with various
      comments, criticisms and alternative suggestions. In refining
      Dr. Keeler’s earlier work and using more recent data, your study
      makes a valuable contribution to the present debate over the efficacy
      of the economic regulation of air transportation.

      The use of extensive sensitivity analysis greatly strengthens the
      conclusion of your study that passengers pay between $1.5 and
      $2.0 billion in excessive air fares each year. As is pointed out,
      the benefits from increased efficiency would be even greater if
      potential passengers who are discouraged by high fares from
      actually traveling had been considered. Also important is the
      fact that these efficiencies and resulting savings to consumers
      could be achieved in conjunction with increased profitability for
      our nation’s airlines.

      Perhaps the most important new ground broken by this study is
      your treatment of substantial improvements in efficiency and lower
      air fares possible in low- and medium-density, as well as in high-
      density markets.     Many people in the aviation industry have conceded
      that substantial efficiency gains are possible in high-density markets,

APPENDIX XVII                                               APPENDIX XVII

  but in general they have not yet recognized that such efficiency
  gains are also possible in less dense markets as well.

  I hope that you will release this excellent modeling effort’as soon
  as possible so that its forecasts can cc&rib&e to the current
  deliberations on airline regulatory reform.


                             iTz=sb            5./w
                                       William S. Meffelfinger

APPENDIX        XVII                                                                                              APPENDIX               XVII

                                   f’*pt                         UNITED      STATES      DEPARTMENT              OF   COMMERCE
                                   .                         .
                                           “4,rr Of r*
                                                             T   The Assistant
                                                                 Washmgton.   DC
                                                                                                  for   Policy

  JUL    21 1976

  Mr. Henry Eschwege
  Director,     Community     and Economic
     Development     Division
  United    States   General    Accounting                          Office
  Washington,      D. C.     20548

  Dear    Mr.   Eschwege:

          Thank you for sending              Secretary     Richardson    a copy of the
  draft    report     by the General          Accounting      Office  entitled       "Impact
  of Increased        Efficiencies         on Air Fares and Travel."             That
  report,     which     is an analysis          of a study by Dr. Theodore             E.
  Keeler     on the topic        "Airline       Regulation     and Market      Performance,"
  makes an important            contribution        to our knowledge        of the benefits
  that    consumers       could obtain        from more efficient        operation        of the
  airline     industry.

            The GAO report,           as well as much of the other                 recent
   literature         on the economics           of air transportation,              emphasizes
  that      up to the point           of full      occupancy     there     is a tendency         for
  the pro rata           cost of air travel             to decline     as the number of
  persons        traveling       on a plane        increases.       By increasing         seating
  densities         and load factors,            it is possible,         therefore,       to reduce
  the costs of air travel                per person.           In addition,        by more care-
  fully       matching       the size and types of aircraft                  used to serve a
  given       market     with    the size of that          market,     it is possible          to
  obtain       further       reductions       in costs      per person.

          Because changes        in seating        densities      and load factors         have
  an impact       on the quality        of the service         provided    by an airline,
  it is possible        that   such changes would cause a shift                    in the
  demand for air transportation                 service.      At a given       level    of
  fares,     the quantity      of air transportation              service    demanded may
  differ    depending      upon the quality           of the service       available.         Such
  a difference       is especially         likely     to exist      in short-haul       markets
  in which air transportation                is in close       competition       with   auto-
  mobile,      bus, and rail       transportation.

          If changes in the operation          of the air transportation
  industry    which make possible         lower costs   and fares     per person
  also cause a shift       in the demand for air transportation,              it
  becomes more difficult        to estimate      the benefits      to consumers
  from increasing    efficiency      in the airline      industry.     Under
                                                                                                                          "~) ,'C

APPENDIX XVII                                                                          APPENDIX XVII

   such a situation,           it may not be sufficient             to assume,      as the
   GAO report      appears       to do, that      the benefits        to consumers
   can be calculated          as if the demand curve had not shifted.
   In particular,        if there        is a reduction      in the quality         of service
   which causes the demand for air transportation                        to become ldwer
   and more price        elastic       at the previous       equilibrium       price,    the
   techniques      used by the GAO, and by Dr. Keeler,                   to estimate       the
   benefits     to existing         consumers     from lower      fares    may overestimate
   those benefits        while      the benefits      to the additional         consumers
   attracted      by lower       fares     may be underestimated.           There     is no
   reason    necessarily         to expect     these two amounts         to cancel      each

           In view of the difficulties              involved,        we would not fault
  the GAO report          for its treatment        of the issue of estimating                 the
  benefits      that   would result      from an introduction               of greater
  efficiency        in the air transportation              industry.        We would      suggest,
  however,      that    the "Recommendations"            section      ought     to take some
  implicit      account      of the issues      raised       above by indicating           that
  it is desirable          to have a market        test      in each separate          market     of
  what the public          believes   to be the optimum              combination       of fares
  and service.         Such a market       test    would help to ensure              that
  quality     of service        would be traded        off     for lower      fares    only up
  to the point        at which there       is a maximization             of the benefits
  to consumers.

          Rather      than recommending           either    an increase      in competition
  in the industry            or changes      in the rulemaking         and enforcement
  procedures        of the Civil         Aeronautics      Board,   we believe      the report
  should     support       an increase       in competition       with    the accompanying
  greater     flexibility         with    respect      to fares   and fewer     restrictions
  on entry      into      and exit     from the industry.           This in turn       would
  permit     a market        test   in each of the separate            air transport
  markets     of the appropriate             combination       of fares     and service.

                                                   Sincerely,              ,

                                                   Robert   S. Mil     '
                                                   Deputy Assista +2&L             etary     for
                                                     Policy     Development        and
APPENDIX            XVII                                                                               APPENDIX     XVII

                                          COUNCIL     OF   ECONOMIC       ADVISERS



                                                                                     June       14,        1976

          Dear       Mr.    Eschwege:

                   This    is to comment                on the draft      report      on the "Impact
          of Increased          Efficiencies               on Air Fares      and Travel."
          Although        I have a few              questions       the report      is,     overall,
          interesting          and helpful.                In particular,        I have had
          limited       confidence         in       the DPFI cost model           and the inability
          of your projections                to       predict    the intrastate           fares
          supports       the view that                we should       use it with       caution.

                   As a further          test      for your model,           you should     do your
         calculations          excluding          the Hawaii         markets    and then       see how
         close      the predicted            fares       are to the actual          fares    in these
         markets.         Intense        competition          from charter        operators        in
         earlier       years     encouraged            the CAB to substantially              relax
         restrictions          in this         market       so that     it may be the most
         unregulated         markets         in the country           at present.         You might
         try     the same test          using        the New York-Puerto            Rico market
         as a similar          situation          exists      there.

                      I look    forward        to    seeing         the    final        draft         of     this


                                                                                     Paul W. MacAvoy

         Mr. Henry Eschwege,         Director
         Resources    and Economic
          Development     Division
         U.S. General     Accounting        Office
         Washington,     D. C. 20548

I AP2ENCIX XVII                                                                                                         APPENCIX XVII

UNIVERSITY            OF CALIFORNIA,                 BERKELEY                                            _,_.....


DEPARTMENT      OF ECONOMICS                                                               BERKELEY,   CALIFORNIA       94720

                                                                                           June 30, 1976

           Mr. Henry Eschwege, Director
           Hesources and Economic Development Division
           U. S, General Accounting Office
           Washington, D. C. 20548
           Dear Mr. Eschwege:
                  Enclosed are my comments on the GAOanalysis of my earlier article,
           "Airline flegulation and Market Performance." I am also sending a copy
           of these comments to Mr. Ihomas Dooley. Overall, I believe that the
           GAOanalysis is an excellent piece of work.
                 If you, kii. Dooley, or anyone else on your staff have any questions
           regarding these matters, please do not hesitate to contact me.

                                                                Theodore .& Leeler
                                                                Assistant Professor of Sconomics

APPENCLX XVII                                                                                                              APPENDIX XVII

                                  "AIRLINE              REGULATION AND MARKET PERFORMANCE"

                                                            Theodore E. Keeler
                                            Assistant          Professor      of Economics
                                           University          of California,      Berkeley

            This         GAO study,          which          involves          an analysis             and extension            of my earlier

paper        on airline             regulation,                is to my mind a most impressive                             piece       of work,

and I agree                with      its     basic          conclusions.                 However,        I do have a few comments.

            1.      The extension                  of my assumptions                     to low and medium-density

routes            may overstate               the potential                  benefits        of deregulation,              because        the

60 per            cent     load     factor              assumed,          along     with     relatively           high-capacity

aircraft,            may simply              be infeasible                   or inconvenient              on low-density            routes.

Thus,        it     is quite          possible             that       fares        would     go down by considerably                     less

than     estimated                on low-density                   routes.          It     was to avoid           such problems           and

ambiguities                that     I restricted                   my study         to higher-density                routes,       where        the

feasibility                of a 60 per cent                    load        factor        is documented            by California


         2.         For all         but      direct          capital          costs,        the GAO team has substituted                         a CAB

cost     model           derided           from     the Domestic               Passenger             Fare Investigation             of 1971

in place            of my own.               It     is true           that        the CAB-GAO model            has some advantages

over     my own, and that                         the     latter          was not        available        to me when I did              the work

for    my paper,             which          was completed                  in 1971.          However,       similar       models        were

available            to me, and I rejected                            them,       mainly      because       the      CAB models         have

some serious                shortcomings                 when applied               to the      specific       routes          I was

concerned            with.          My broad             objections            to the CAB-GAO model may be divided

between           two groups,               each pertaining                   to a different              category       of costs:

                     a.      Direct          Operating              Costs.          Both     the CAB-GAO model                 and my own

         base their               cost       estimates              for      a given       route       on the assumption               that     the

APPENDIX XVII                                                                                                                           APPENOIX XVII

   cost     per block-hour                     for     each plane                  type     is constant               --      it        is        applied

   to an estimated                    flight          time        for      that      route.           However,              flight                times

   are     estimated             differently                 in     the      two approaches.                        I use actual                    average

   scheduled             time     on a given                 route,          plus         an allowance               for      lateness.

   The CAB-GAO model,                        on the          other         hand,          estimates           a relationship

   between         distance            and flight                  time,       and assumes                that       this          relationship

   is     the     same for            all      city        pairs.            The CAB-GAO model                       has the             weakness

   that     it     totally            neglects             higher          flying          times      occurring               in congested

   areas.          Thus,         in     1968,         the      fastest             planes         between           Los Angeles                    and

   San Francisco                 made the             350-mile             run      in     a scheduled               block          time           of

   45 minutes.                  On the         other         hand,         average          flying           time     on the                 189-mile

   run between              New York            and Boston                 was 55 minutes.                       Thus,        the            CAB-GAO

   model         would       grossly           underestimate                   direct            operating           costs          on the New York-

   Boston         run,       and probably                  overestimate                   costs      on the          Los Angeles-

   San Francisco                 run        (and the           GAO-CAB models                     have serious                problems

   predicting              unregulated                fares         in      California;              more about               that            below).

   I fail         to see why the                     substitution                  of estimated               aggregate                 times

   for     a route(will                 yield         better             results          than     use of actual                     times.

            b.       Indirect               Operating              Costs.           The GAO report                   dismisses                    my

   model         of indirect                costs      with         the      assertion             that       the     variables                    used      in

   it     are multicollinear.                          Multicollinearity                          of variables                does not                  in

   and of         itself         render         an econometric                      model         invalid;           this          is        especially

   true     when the model                     is used             for      predictive             purposes.                On the                other      hand,

   the CAB-GAO model                        contains              some     arbitrary              allocations               of       cost

   categories              to variables                which            my study           does not           (the         latter             aggregates

   indirect          costs,           and then             allows           them to vary              with          whatever                 of    the

   independent              variables                the     data         might       dictate).               There         are many possible


APPENDIX XVII                                                                                                                  APPENDIX XVII

 specifications                  of indirect                 airline          cost      models,      and I would               be the

 last     to claim             that      my own model                  is the best           of all       possible             such

 models.          But I believe                     that      in the task               of predicting            non-CAB-regulated

 fares,      it     fares            considerably                  better       than     the CAB-GAO model.

          More specifically,                          the CAB-GAO model                   has a much higher                    cost     per

 passenger          (at        a given           load        factor)          which      is fixed        with      a trip        and

 independent             of distance.                      That       this      is the case may be seen by

 comparing         costs             in the GAO study                   with      my costs        for     a llO-mile

 trip,      Los Angeles-San                     Diego         in my model,               and New York-Hartford                        in the

 GAO model          (remember                that         differences            in congestion,            aircraft             type,

 and load         factor             do not         show up in the GAO model,                           so that         the GAO

 estimate         for     New York-Hartford                          should           be the same as its                estimate          for

 Los     Angeles-San                 Diego).              My model           predicts       a fare       of $6.50             on the

 Los Angeles-San                     Diego      route,             compared       with      an actual           fare      (of all

 carriers         in 1969)             of $6.35.                   On the other           hand,      the GAO model               predicts

 a fare      of $12.18                on that             route,       despite          a much lower            assumed cost

 of capital             than         my own (see below).                         If     the GAO-CAB model were correct,

 PSA and the other                     carriers             on that           route      would    not only             lose     money

 at a 60 per c'ent load                         factor             (and a lot           of money);        they         would     also      lose

 money at a full                     load      factor.

          The Los Angeles-San                             Diego      route       in 1969 is not            a fluke.              Consider

 the Los Angeles-San                         Francisco              route,       just     under      350 miles           in length.

 The 1968-69             intrastate                 fare      was $13.50,               and my model            came within             two

 per cent         of predicting                     it.       On the other               hand,    for     an equivalent                 route

 (Minneapolis-Chicago,                          345 miles),                  the CAB-GAO model             predicts             a 1969 fare

 of $24.66,             82.7         per cent             above      the actual           Los Angeles-San                Francisco

 fare.      Again,             all     the carriers                  on the Los Angeles-San                      Francisco             route
    AP?ENDIX XVII                                                                                                                          APPENDIX XVII
            would        lose         significant                    amounts        of money if             they       had costs               predicted            by

             the      GAO-CAP model,                        even with          a full        load       factor.              In fact,           PSA earned

             a considerable                      profit          in 1968,           and it       did      so with            an average               load

             factor       of about                60 per cent               on the Los Angeles-San                             Francisco              route.          It

            might        be objected                    that         the CAB model              still       predicts             costs         well      for

             the      trunk          carriers,                and that         PSA is just              more efficient                   on short-haul

             routes.            If        that      were         so,     and if         the CAB model                 were correct,                   so that

             the      trunk          carriers               really       did have such high                         costs      on these           routes,

             then      why would                 the        trunk       managements             want      to compete                in these           markets             as

             vigorously               as they               have,       when every           extra        flight            would       lose      money,

             even at a full                      load         factor?          It    would       seem more likely                       that      there        is

             something           wrong            with         the GAO-CAB cost                  model       which            tends      to bias

             upwards           estimates                of potential                unregulated             fares           on short-haul                routes.

             3.       The question                  arises            as to the appropriate                          "actual"           fares         to compare

    with     "unregulated"                   fares.              There      is something                here         to be said            for

    selecting           a regulated                    fare      which      corresponds                 to the        same product                quality

    as is offered               in the             intrastate              markets         (with         perhaps            meal costs            added        on

    with     longer       distances).                         The intrastate              fares          used involve                  coach      service

    with     no time           period            or round-trip                 restrictions.                   It     is only           reasonable             that

    the     interstate               regulated                fares      selected         should           correspond               to the        same

    product         quality.               That             is one reason             why I selected                   standard          unrestricted

    coach daytime               fares            on CAB-regulated                     routes        as the basis                 for     comparison

    with     unregulated                  fares.              To the extent              that       the     GAO analysis                 mixes          in

    other       types     of fares,                    it     is no longer              standardizing                  for     product           quality.

    (Incidentally,                   it    has been objected                        to my work and to the GAO analysis

    that     they       do not            take         account          of the        superior            service            quality       which         CAB

    regulation           has induced,                       most       specifically             in the area                  of reduced           schedule

  APPENDIX                 XVII                                                                                                  APPENCIX                 XVII

delay,          or greater          ease of getting                 a seat         at the time             when a passenger                     wants

to leave.               But for       coach passengers,                   there         is little          evidence             that      CAB regulation

has reduced              schedule          delay      relative          to intrastate               routes;         research              currently

in progress              at Berkeley              would      tend      to indicate           hhat        to the extent                  that

CAB regulation                  has improved              service       quality,          the improvements                      have accrued

to first          class         passengers          alone).

           4.     The GAO study              finds         a pre-tax         opportunity             cost      of capital                 in the

corporate           sector         of only         10.4      per cent,           compared         with      12 per cent                 in my

own study.               The difference,                  I believe,         stems from differences                             in assumptions

as to the           capital         which         should      be included               in the      rate      base,        and the

specific          income         which      should         be included             in the return.                  The GAO analysis

includes          all      investments             of all       corporations              in the rate              base,         and all

sources          of income          in the return.                  But this            may make for           double            counting:

consider          the case where                  Corporation           A owns some stock                   in Corporation                     B.

The stock          held         by Corporation               A in B represents                   a share       in some of the

physical          assets         of B.       But the assets                of B are already                   included             in total

corporate          assets.           If     A's     holdings           in B are          included          in A's      assets,             there

will       be double            counting       when total              capital          is calculated.

           In order            to avoid      double-counting,                    and to assure              that      the        profit

figures          calculated          actually          represent           income         earned         from direct               operations

of firms          in the         corporate          sector,         I attempted             to exclude             outside             holdings

of debt         and equity           from      the rate          base,      and to exclude                 earnings              from      securities

of outside              corporations           from my income                figures.            The only           exception              to this

was in the              area     of current          assets,         where         it    was assumed that                  it      is necessary

to hold         a certain           quantity         of liquid           assets          (such      as government                 bonds)

as a close          money-substitute                   for     daily       operations            of the company.                        Thus,       the

interest          on short-term              loans         and government                bonds was treated                      as income

APPENDIX                   XVI I                                                                                                             APPENDIX                       XVII

(i.e.,          a by-product                    of holding                money-substitutes).                           Making            these

corrections                     to the         Statistics                of    Income          figures          involves            some arbitrary

decisions;                 for      example,                to what           extent       do rent            and royalty                 income        represent

operating              bncome?                 Nevertheless,                   I believe            that        an attempt                to make corrections

for       the    problem                will         yield         more accurate                estimates              of    the      opportunity

cost       of capital                   than         ignoring            them.

           Because               I have not                 kept      the      detailed          worksheets                 for     my dissertation

calculations                     (the      work            was done over                five      years         ago,        and this             part      of my dis-

sertation              was not             published),                   I do not          recall          exactly           whether             certain

accounts           were or were not                              included         in     the      income         and rate             base,         but        on

the      basis         of some spot                        recalculations,                 I have managed                    to     come quite                 close

to my original                      figures,                and the           12.per       cent        opportunity                 cost      of capital

figure          which            I used             in my airline                cost      calculations                 appears            to hold             up

nicely          with            several             alternative               assumptions              as to which                 accounts             to include.

I therefore                     stand      by my 12 per                   cent     assumption,                  and do not                believe          that        it

is     a mistake.

           5.         For        the      high-density                   routes         I considered,                  the        extra      costs         incurred

by stops              along         the’way                are     irrelevant,             as GAO figures                     bear        out,      because

the       routes           are      of sufficiently                       high     density             that      practically                 everyone               rides

nonstop.               However,                it      is appropriate                   that      GAO should                take      account             of    the

extra          costs            of stops             on lower-density                    routes,           as it        has done.

           6.         It        appears             that      the     GAO study            has double-counted                         aircraft             capital

costs,          for        it      has calculated                     full       capital          costs         at a going                interest             rate

and utilization                         rate         for      aircraft           as they          stand,         and then             added         on aircraft

rental          costs            as paid             by the         carriers.              This        procedure              would        only         be appropriate

if       the    GAO had used                        actual         airline        accounting               figures            for     aircraft             ownership

costs.           My procedure                         involved           calculation              of     full      capital            costs          for       a given

aircraft              trip,         regardless                   of who owns the aircraft.

APPENDIX XVII                                                                            APPENDIX XVII

         "Impact    of Increased          Efficiencies         on Air   Fares   and Travel"

                                    Dr.      George W. Douglas

Introduction       and Summary:

       This study, a recreation           of a previous          analysis    by Keeler,       is on the
whole a competent       piece of work which confirms                 conclusions      concerning
industry   efficiency       drawn by Keeler and other recent                  investdgators.       Any
study of this sort,         of coursep reflects           the assumptions         taken by the re-
searchers,    any or all of which are subject                  to challenge.         Based on my own
experience     in this field      it fs my judgment           however that the assumptions
taken are generally         reasonable      (and typically         supported      by sensitivity
analysis)    and that the studyss           results      are qual&tatively          correct. In
the discussion      which follows        I will     point out some thought%I              hava on the
reasonable     range of assumed input$and              on the study's        methodology.        While
there exfsts      of course considerable            uncertainty       as to the quantitative
measure of inefficiency,          I am satisfied          that it lies withfn           the bounds
of the study's      sensitivity       analyses.


A. Load Factors:
         A critical       source of inefficfency          in the regulated         industry    is the
excessive        level    of slack capacity,        or the low load factors            which have
typically        prevailed.      The study is correct            in assuming significantly
higher     load'factors        in an efficiently         configured     industry,      perhaps averaging
60%. While the study assumes that all markets will                           tend to have ALF's on
the order of 60%, my research               fndicxs         that the efficient         level of slack
capacity      would vary by market distance               and density.         (See Douglas and Plfller,
Economic Regulation            of Domestic Air Transport,            Brookings       1974.)    It may be
noted,     however,       that the overall       average efficient         load factor      for all
markets by my calculations              are also in the range of 60%. Hence, I should
not think        that this assumption         constitutes       an important      bias in the estimate
of total       cost savings,       but rather     that the pattern         of efficient      fares would
be somewhat different            than those generated           by this assumption.           (Speci-
fically,       long haul and/or high density              markets may have load factors               greater
than 60X, and therefore             lower efficient         fares than those calculated            here,
while other markets may have the converse.)

       Moreover,  it should be noted that while   the efficient    market would be
characterized    by lower fares,  the convenience   of the service   to the traveller
would be somewhat less9 so that the net welfare       burden is less than calculated.

APPENDIX          XVII                                                                APPENDIX        XVII

B.   Most Efficient        Aircraft:

         The study assigns    the "most efficient"     aircraft     to each market.        The
aircraft     so chosen (esp. DC8-61) are not representatjlve             of the existing
fleet     composition.    f have also observed      the superior      cost efficiency      of
the DC8-61, but wonder whether         there may be some statistical           or accounting
anomaly involved       -- else why wouldn't      the carriers    use them more?

C.   Return     on Investment          and Capital   Recovery:

      The "base case" assumption   of the competitive   ROI is definitely       too
low; I would think   that the value of 18% used in the sensitivity        analysis
would be more appropriate    as the "base case" assumption.    This is based on
the following:

      1.      The use of historic       averages is clearly            inappropriate in determining
              current  competitive      costs of capital.            (e.g. Could one conclude
              that since average bond yields        1939-66          were 4.5% (a guess) that
              current  yields     should also be 4.5%?)

      2.      The industry     does face greater    than average financial       risks,            which
              requires   higher    than average ROI.     For example,  the average               beta
              of airline    common stocks     is on the order of 1.55,reflecting                 their
              above average risk (see table).

              One can show that the before   tax return               on investment,      or cost of
              capital,  to a firm with a cost of equity                r, and interest      cost i
              can be given as

                              r = re/(l+d)(l-t)        + id/(l+d)        ,

              where d fs the debt/equity            ratio    and t is the tax rate.           Current
              Interest    yields    on airline      bonds are roughly         10%. The cost of
              equity,   re, cannot be directly            observed,     but it clearly      exceeds i.
              A lower bound estimate           of currently      required     ROI could be calculated,
              for example,       by setting     remi, and setting         t at some typical       level,
              say l/l.      This would yield        an ROI 2 15%. A similar           calculation
              could be performed         for the years studied          could generate      a more
              realistic     lower bound estimate.

      3.      The annual capital    recovery    depends as well on the economic life
              of the capital.     While twelve years may be typical,          it should be
              considered   that the economic life,      due to technological        obsolescence,
              may be considerably     less than the physical     life.      A shorter     life,
              of course,   would increase    the annual capital      recovery    factor.

D.   Use of      the CAB's DPFI Cost Model:

      The study employs the CAB's cost model in                     preference    to Keeler's
      "since   it has been validated."          It is not           clear what in the model is
      changed (only seating       desnities?)     but its           use could seriously       overstate
      the estimate   of efficient       carrier    costs.           One considerable     source of

APPENDIX                XVII                                                                           APPENDIX   XVII

           Table          1.      Risk      Characteristics           of Airline   Common Stocks

                 Carrier                                      Beta*                --Price   Stabil*

 American              Airlines                               1.40                           30
 Braniff              International                           1.60                           ,30
 Continental                Air     Lines                     1.55                           30
 Delta          Air     Lines                                 1.40                           50
 Eastern          Air       Lines                             1.40                           30
 National              Airlines                               1.65                           25
 Northwest              Airlines                              1.60                           35
 Pan American                                                 1.50                           15

 Trans         World       Airlines                           1.85                           15
 UAL, Inc.                                                    1.60                           30
 Western          Air      Lines                              1.55                           30

          The beta is simply the regression            coefficient     of a time series
 regression      ofi returns   to an individual      stock against       the average of
 all stocks.        The beta coefficient       measures,      then, the "non-diversifiable"
 risk;    a value of one would indicate          a stock of average risk while values
 higher     than one indicate     greater    than average risk.         Source - The Value
 Line Investment        Survey, January 16, 1976.

             The price    stability   index measures the gross risk of the stock
  (including        that which can be theoretically       reduced in holding  a dfver-
 sified      portfolio).      This particular    index,   devised by Value Line Investment
 Survey,      is sealed from 5 (highest       instability     or risk) to 100 (lowest

APPENDIX XVII                                                                        APPENDIX XVII

inefficiency      in the industry         is what has become known as “X” inefficiency,
or in mundane terms,            sloppy management and waste.          Hence, since the DPFI
cost model is calibrated              on existing   cost experience      of regulated      carriers
on the average,        it     tends to include     whatever     “X” inefficiency      Is imbedded
in the system.         (What is interesting,         perhaps,    is that the study shows slg-
niflcant     cost savings         even using this cost model.)         Carriers     in the intra-
state markets       seem to show significantly            lower costs/higher      productivities,
thus confirming        this effect.         Other factors     may also change observed           costs
in a competitive          Industry.      For example:

            --higher        advertising      costs would be incurred
            --higher        legal     fees from litigation
            --lower       regulatory       compliance      costs
            --differences           in passenger       service    costs ranging   from
                different       marketing     of onboard services         to innovation
                in baggage handling,            ticketing,       etc.

       While these factors  are difficult              to estimate,    they would most likely
cause efficient    costs to be lower than              those generated     by the CAB cost

E.   Without   regulation     the route structure       would change somewhat.       Some
     flights   currently    required   to have an intermediate         stop would proceed
     non8 top ; some markets not currently          served by direct      flights may be
     served,   others may be abandoned.         Airlines     could rationalize    their
     route structures     to integrate    more effectively        the stray bits    and
     pieces of routes.

F.   The definition      of “most efficient”         aircraft    type      and fleet    composition
     is more complex than represented            in this study.            That is, the efficient
     aircraft     type for a market is not defined             as that      which provides      the
     lowest cost/ASM or cost/RPM.            Rather,      it is that       type which given the
     density    and &stance,      can,provide      the least     cost      service   when conven-
     ience of the service       is included      as a cost.

G.   Some input    costs may be higher            in the regulated   industry.     Labor cost
     in the airline     industry      reflect       the power of each union to successfdly
     shut down the firm.         The sheltered         markets of the industry     has prompted
     management    to accede to wage demands which can be passed along via CAB
     sanctioned    fare increases.           Presumably     in a competitive   market each firm
     would have greater       resistance        to wage inflation.

H.   Has the cost model been validated       for this application?        That is, If
      the CAB base case inputs    are used, does the model generate          costs or
      fares consistent  with those denoted as “actual”          In the study?      A useful
     means of reporting    sources of inefficiency      might then be to begin from
      the CAB “base case,”   and change the crftical        assumptions   serially    and
     report  changes in costs (X) generated        by each.

I.   It should be noted explicitly          in the study that the part of the                 industry
     studied  is only trunk service.          While the local   service carriers                doubt-
     lessly  have much greater     relative      inefficiency than the trunks,                their
     share of the total   airline     market is relatively      small.

    APPENCIX          XVII                                                                         APPENCIX         XVII

                                                                         P-0. Box 455
                                                                         Nobleton, Ontario
                                                                         LOG 1NO
                                                                         June 24,        1976

Mr. Thomas E. Dooley
U.S. General Accounting             Office
Room 6122
441 G. Street,     N.W.
Washington,    D.C. 20548

Dear Mr. Dooley:

        As we discussed    by phone on May 27, 1976, I have reviewed          the restricted
draft    of "Impact   of Increased     Efficiencies     on Air Fares and Travel"    (May 6,
1976).     Since you appear to have been too busy to call back to discuss                certain
matters    further,   I will   outline    my thoughts    on those matters   in this letter       and,
unless you advise to the contrary,             will consider   my work on this project      to be

        Overall,   let me express my general    agreement   with the findings      of your
study.      By and large I think your findings      provide  good estimates     of the minimum
fare decreases       that would result   in the long run following     the abolition      of CAB
regulation.       Having said this,    however, let me explain    why I believe      your
estimates      are minimums:

   1. The assumed        60 percent          load   factor         is unduly     low.

   2. The cost data are based on the performance                               of CAB-regulated    airlines
      and, therefore,   are inflated by the extent                             to which regulation       per   se
      serves to incrvse    operating costs.

   3. The use of average fare yields                    (rather than lowest widely-available
      published   coach fares)    results              in your percentage   reductions     being
      based on "average    fares"    that             are lower than those available       to the
      majority  of passengers.

These   conclusions          are based   on the       factors         given     below.

         Load Factor.      60 percent       is conservative.          Evidence:       the consistent       achieve-
ment of load factors         in excess of 70 percent              from 1955 to 1964 by the California
intrastate      carriers    (p. 202 of my book).             Air California's         70.2 and 70.1 percent
load factors       in 1974 and 1975 (Air California,                 Annual Report,       1975).      Southwest
Airline's     58.4 and 62.5 percent            load factors       in 1974 and 1975, increasing             to
66.5 percent       for the first       five months of 1976 (Southwest              Airlines,       Annual Report,
1975 and "Media Release,"            n.d.).       It seems likely        that Southwest's        trend will
continue     until    it too achieves         an average 70 percent          load factor.        70 percent
load factors       are possible      for airlines       providing      a specialized,        homogeneous
service.      Without    regulation,        this is the kind of airline            that would prosper           in
the U.S.

                                                                                                                    e.*.   2

     APPENDIX XVII                                                                                   APPENDIX XVII

            costs.      See my Testimony       before   the Kennedy Subcommittee              at pp. 477-82,
    Vol. 1 of the Hearings.            Lack of specialization,            relatively      shorter     aircraft
    lives,     higher wages and payments to other inputs,                  are all examples of how costs
    are increased        by the current      CAB regulatory       structure.         Therefore,      basing your
    cost estimates        on CAB data results         in overestimating          the costs of unregulated
    airline      performance.      This means that the percentage                decreases      in fares follow-
    ing deregulation          are underestimated.

           Average Fare Yields.           The use of fare yields             results      in percentage          reductions
    being based on fares that don't                actually      exist.     They are a composite              of relatively
    low promotional      fares and the higher              coach fares (which are used by the majority
    of passengers).        Those passengers           now using promotional            fares with the largest
    percentage    discounts      (around 35%) will            experience     relatively        little       decrease     in
    fares due to deregulation            since these extreme promotional                  fares come close to
    approximately      marginal     costs.       Chapter      8 of my book shows that there are few
    promotional     fares without        regulation.          On the other hand, the majority                   of
    passengers    who are paying regular              coach fares will         enjoy larger           percentage
    reductions    than your figures           indicate.         See pp. 60-62 of my book for reasons why
    I used "lowest      widely-available           fares"     rather    than yields       for my fare comparisons.

            The effects      of these factors     are partially       handled      in your report     by the
    alternative      computations     that you made (the 72 percent              load factor,      for example),
    but you did not run a computation             showing the combined results             of a series       of
    these more extreme assumptions.              Doing so would have provided             estimates     of much
    larger      fare reductions.      Some of these percentage           reductions     would have strained
    the credulity       of your readers,      but I think      they would have been closer             estimates
    of fare reductions         than those you actually         achieved,     especially      for high-density
    city pairs.

         I made some marginal  notes                on the draft.         These pages           are enclosed     should
    you wish to check them over.

           I will be traveling    on the West Coast from June 25 to July 16.      Should you
    be interested   in discussing    any of the above you can telephone    my secretary   for
    information   on where I can be contacted.      Of course, I will   be happy to discuss
    any aspect of this with you at your convenience.

             Thank you for the opportunity              of reviewing        your       interesting      study.    I trust
    it   will   be well received.

                                                                       Very    truly       yours,

    dc                                                                 William         A. Jordan

    Enclosures       (2)

APPENDIX          XVII                                                                      APPENDIX          XVII   ,

                                                                            5412 32nd Street,  N.W.
                                                                            Washington,  D. C. 20015
                                                                            May 27,.1976

Mr. Henry Eschwege
Resources     and Economic    Development     Div.
United    States   General   Accounting   Office
Washington,      D. C. 20548

Dear    Mr.   Eschwege:

      Thank you for requesting             my comments on the Draft      GAO Study,  "Air-
line   Regulation     and Market Performance."           As you indicate     in your letter,
my comments on this         report    reflect   my own views and not those of my
employer,      the National     Cotnnission    on Supplies    and Shortages.

        Ted Keeler's        1972 study represented                 an interesting          attempt      to quan-
tify    the effects       of airline          deregulation.            Its basic approach was to
estimate      for a limited          number of markets             the fares      required        to operate
a profitable,       efficiently-run,               high load-factor         service        and compare
these fares       on a market-by-market                basis with the fares              then offered          by
the airlines.         Since these actual                fares were (and still              are) premised
upon lower load factors               than Keeler's           "constructed        fares,"       and since
they are set to allow             profitable          operation       by the average           (as opposed
to the most efficient)               carrier,        Keeler     naturally      found that         deregulation
would lead to large            "savings"         to consumers.

        As I understand       it,    the purpose       of the current        GAO exercise       is to
update    and expand Keeler's           results     and to correct       what were felt         by some
to be certain       errors    in the original.            One can evaluate        this   study     in
various     ways.     On one hand, one can ask whether                it indeed      accomplishes
the task I have just          stated.        It is my judgment        that,     by and large,        it
does.     The   modifications        that   GAO    has  made  seem    reasonable,       and   I   have
little    doubt that Dr. Keeler            himself     would have made such modifications
if he had had the time and resources                   and if the DPFI cost models had
been available        to him at the time he originally                performed      the study.
I am encouraged         to see that all the modifications,                  when taken together,
do not produce        results     that are substantially          different       from the results
he originally       reported.

        On a different level,  I have certain        problems    with both the Keeler
study and with the GAO extension      of it.       I would stress that these prob-
lems are not so great     that I would urge the suppression              of the GAO piece.
Instead, I would urge that the figures          resulting     from the analysis     be
treated    as possible "order  of magnitude"       estimates,     not as exact predic-
tions    of what might occur in a differently-regulated             environment.

APPENDIX XVII                                                                            APPENDIX XVII

         My first        problem      derives      from the fact that         the Keeler,        GAO, and
CAB cost models all               treat     the airline         system as an aggregation            of indi-
vidual       city-pair       markets.         But it is more than this.               It is a network.
Aircraft         routing      decisions       are made on a network           basis--not       on a dis-
connected,          segment-by-segment             basis.       The aim of a scheduler           is not--
and should          not be--to minimize              segment operating        costs.        Instead, he
should       aim to maximize            network      profits.        Thus, we see certain          segments
being operated             at very low load factors                because such operations            are
profitable          when viewed in the context                  of the entire      network.        While in
a deregulated            system,      load factors          would,    on the whole,       almost     certainly
rise,      even in such a system,                 certain     low load-factor        segments      would con-
tinue      to be operated.              The existence         of such segments        would not be
evidence         of "inefficiency,"             though      GAO seems to infer        this.

       The violence      to reality      done by treating              the airline       system merely
as a collection       of city-pairs        is minimized           if a few carefully-selected
segments    are analyzed.          By looking       at airlines'           aircraft    routing     diagrams,
for example,     one could uncover           certain       routes      that are operated         on almost
a pure "turnaround"         basis.      For such segments,               the analysis       Keeler     and
the GAO perform      may not be too wide of the mark.                          But for others,       the
analysis    may provide      a misleading         indication         of how that       specific     market
would perform      under "deregulation."               Expanding         the sample as GAO does in
its study likely        compounds     the problem.

        Second,     I hiEve problems with the model of optimal                       load factor       deter-
mination      that    is implicit          in both the Keeler           and the GAO pieces.          Those
of us who believe          that      fares would come down and load factors                      go up do
so because we believe              that,       on balance,       the current     and potential       flying
public     would accept        higher        load factors        if they could,      in turn,      be com-
pensated      for this     by lower fares.               Keeler--and       the GAO--apparently         assume
that    this    balance   would be struck              at a load factor        level     of about 60 per-
cent.       They justify       this      by appealing        to the California         intrastate      exper-
ience where load factors                 of this     level     have been operated          in an essentially
deregulated        system    (though         the citations         in the GAO report         are a bit
inconsistent        on this      point).

        However,     as Douglas        and Miller      have shown, the optimal             load factor
is determined        by balancing         the value of empty seats (in terms of in-
creased    scheduling        convenience)        against     the cost of operating            those
empty seats.         Factors      which influence         the value people          place on their
time and factors         which raise         the cost of operating            empty seats will
combine to shift         the optimal         load factor.          For example,       the price      of
aviation     fuel    recently      doubled.       All else held equal,            this    should     serve
to increase       the optimal        load factor       since     it raises     the cost of flying
empty seats.         And, just       as would be expected,             load factors       have risen.
 I know that they have risen in the interstate                         system.      I imagine      that
they also have risen            in the intrastate           system--though        not by as much.

 APPENCIX             XVII                                                                                        APPENCIX             XVII

        Indeed,       on page 14 of the draft,              the excess of actual          over estimated
fares     is shown to be steadily              shrinking.        One might    infer    ,from this    that
the CAB is already             moving in the direction            that GAO suggests.          And to a
degree,       this    is true.      Certainly       two events      that account      for part of the
decline       in this      "excess"   are the adoption           by the Board of load factor
standards         and the increase        in the fare taper           that was instituted       as a
result      of Phase 9 of the DPFI.               However,     the key point        to note is that
the target         has moved also.          If 60 percent        was indeed an appropriate           load
factor      target      for a deregulated         system in 1969, it most certainly               would
be higher        today.       Thus, the "savings          to passengers"     may be substantially

          This question               of what is indeed                     the "optimal"               load factor            in a
 deregulated            system raises               an additional                 concern         in my mind about the
 GAO study.            Throughout             the study there                   is an implicit                equating         of higher
 load factors             with "efficiency."                       However,           there      is no reason why an
 "efficient"           airline          system might               not be one that operated                           at a 40 percent
 load factor--provided                      costs       and the value travelers                           placed       on their         time
 combined         to make that the "optimal"                              load factor.                In such a case9 adoption
 of regulatory              policies         that would produce                       60 percent            average        load factors
would be highly                 inefficient.                People would be getting                           a product          that     as
much failed            to reflect            their        tastes         as would consumers                     who would prefer
a systemwide              load factor             of 60 percent                 and who were given an actual                            load
 factor       of 40 percent.               . Indeed,          in a deregulated                   system          (and even ignoring
the "network             effects"           I have discussed                    at length           above)         I would not be
 surprised          to see ;ertain                routes --routes                 catering          primarily          to business
travelers--to               have load factors                   in the 40's.                 Other routes--routes                     cater-
 ing primarily              to vacation             travelers          --might          exhibit         load factors             in the
70's.         Some routes            with large             numbers of both types of travelers                                     might
have two types of service --one offering                                          load factors              in the 40's and
fares       to match;          the other            offering          load factors               in the 70's and fares
consistent            with this          level        of service.                 All would be equally                     "efficient"
 in that they would be equally                              reflective              of the tastes               of the particular
consumers          they catered              to.        The key problem                  with regulation--aside                       from
 its sluggishness--                is that          it substitutes                  decisions          of bureaucrats               for
decisions          of consumers.                  Now there           is no reason why, if the bureaucrats
are conscientious                  and hard working,                     they cannot             arrive         at the same
decisions          as consumers              would have and create                        a fare structure                   that
produces         a structure             of services              that        is identical             to that which would
evolve        naturally           under competition.                        But why engage in this                       elaborate
exercise         to simulate             the workings               of the market                unless         there     are reasons
to believe           that     reliance            upon competition                   will      fail       to produce           the
desired        result?           The thing            that the research                   of the last               ten years
shows--and           the thing           that the CAB now admits--is                                that      the rationale
that      have been advanced                    to support            regulation             in the airline                industry
don't       stand up to the facts.                          "Destructive               competition"               is not a feature
t    APPENDIX XVII                                                                                    APPENDIX XVII

    of an unregulated        air transport        system.    There is not now widespread
    cross-subsidy.         Therefore,    there     would be little       or no loss of air
    service     by small communities.            Carriers   would not enter        and abandon
    routes    on a totally       random basis.         To be sure,    transition      to a system
    of open competition          should  be accomplished        slowly,      but there   should   be
    no fear about the ultimate           result.

             This last       point      gets at another             quibble       I have with both the Keeler
    and GAO pieces.             By attempting            to put a dollar              value on the savings              to
    consumers,        both have substituted                   their      judgment      for that         of consumers.
    If Keeler and GAO wanted to be more honest, they should call their
    results      not "savings           to passengers            with      increased        efficiencies,"         but
    "savings      to passengers             assuming       we are guessing             correctly          about what
    passengers'         preferences          truly      are."        For that       is what they are really
    estimating.          To give GAO credit,                  the calculation             of a sensitivity
    analysis      employing         different         possible         load factors           (page 80) is one
    way of getting           around       the problem,           though       I would question             the validity
    of the estimated             "savings"         derived        in the base case using the actual
    trunkline       load factors.               In view of the failure                  to consider          the "network
    effect"      I have referred              to above,         I would judge that much of this                      is
    illusory,       arising        as it does largely                from the assumption                that the most
    efficient       aircraft        is used on every route                    segment and that              higher  seating
    densities       are used.

            I would cloqe by returning                      to what I said at the start                   of this       letter.
    Ted Keeler's          original        piece was an interesting                  and useful       speculation            about
    how a deregulated               system might          differ      from the system that existed                    in 1969.
    To arrive      at his results,                Keeler      was forced        to make certain           simplifying
    assumptions         that      diverse       from reality          in certain       aspects.         The GAO draft
    study relaxes           several       of these assumptions                in a reasonable           way and performs
    useful    sensitivity             analyses.         Yet the GAO study retains                 many of the crucial
    assumptions         that      Keeler     made. Further,              the GAO's expansion              of Keeler's
    sample may compound the problem caused by Keeler's                                    failure       to treat        the
    airline     system as a network                  rather      than a collection            of individual           city
    pairs.      I don't         know how, given the state                   of the art of modeling,                 a
    better    estimate          could be generated.                 Nevertheless,         in view of this
    problem,      and in view of the problem                       caused by the need to guess what
    airline     consumers           on average         would consider           the optimal       load factor           to
    be, I would treat               the estimates           with an appropriate             level     of caution,
    realizing       that there           are some factors             which,      on balance,       should       serve
    to inflate        them (the network                problem)       and others       which should          serve      to
    depress     them (the problem                 caused by the fact that                optimal      load factors
    probably      have increased             with      increasing        costs),      and that      there      is no
    reason to claim             that these two effects                  should     cancel     each other         out.

APPENDIX    XVII                                        APPENDIX   XVII   *

      I am returning the draft report to you with mdrgjnal notes. I
would be glad to elaborate further on anything I have said, should you
desire me to do so.


Copies         of GAO           reports          are available            to the general
public       at a cost of S1.00                    a copy.        There      is no charge
for     reports        furnished            to Members               of Congress               and
congressional             committee               staff      members.          Officials           of
 Federal,         State,       and       local       governments             may         receive
up to          10 copies            free       of charge.            Members             of the
 press;      college        libraries,          faculty          members,           and        stu-
dents;and             non-profit             organizations             may      receive           up
 to 2 copies           free of charge.                  Requests       for larger          quan-
tities     should        be accompanred                    by payment.

Requesters       entrtled    to reports                    wrthout              charge      should
address    their    requests    to:

                    U.S. General             Accountrng         Office
                    Distribution            Sectron,       Room      4522
                    441 G Street,              NW.
                    Washington,             D.C.     20548

Requesters              who         are     required             to     pay        for      reports
should       send         therr        requests          with         checks          or    money
orders     to:

                    U.S. General       Accounting                      Office
                    Distribution     Section
                    P.O. Box     1020
                    Washington,       D.C.    20013

 Checks       or money       orders     should       be made                         payable          to
 the     U.S.     General       Accountrng            Office.                       Stamps            or
 Superintendent           of Documents             coupons                         will    not        be
 accepted.       Please do not send            cash.

 To expedite             filling       your   order,             use the           report         num-
 ber   in the           lower         left  corner              and    the         date      In      the
 lower   right         corner        of the front               cover.

 GAO    reports        are         now      available        on microfiche.    If such
 copies   will     meet            your       needs,        be sure to specify     that
 you want       microfiche                copies.
,”           UNITEDSTATES                         POSTAGE   AND   FEES   PAID
          WASHINGTON, D.C. 20548

               OFFICIAL BUSINESS