along with a small staff. Sooner, rather than later, human nature
asserts itself; he feels that _my man_ can be more useful if he is on
_my staff_. He builds up a larger staff with an inevitable retarding
bureau of correspondence. He perhaps has a $200 traveling engineer
finding fault with the division performance of the $300 superintendent.
Sometimes a general superintendent is located at a large city under the
theory that the importance of the metropolis demands an officer of
higher rank. There are various ways to skin a cat, and the method we
have seen is not necessarily the only solution. The Pennsylvania
handles successfully large cities like Cincinnati, Cleveland and
Chicago with a superintendent who has the authority of a general agent.
The unit system of organization, because based on sound fundamental
principles, solves several vexatious problems. Among these is this
matter of general superintendents' districts. Under the unit system
every assistant should have his office of record in the same building
with the head of the unit. For example, it is a violation of good
organization to give a district passenger agent the title of assistant
general passenger agent with an office of record at a city away from
the general offices. If such outlying office of record is necessary,
and it sometimes is, a complete unit should be segregated under a head
with some such distinct title as district or division passenger agent.
This does not, however, preclude having an assistant reside in the
outlying city and maintain his office of record at the general office
in the same file with the head of the unit.
If I were you I would appoint enough assistant general managers so that
you can have one reside at each point where you have dreamed district
headquarters are necessary. Give him a business car and a stenographer,
but let him understand that his office file is a part of yours. Let him
live on the road as a high class traveling inspector, superior in rank
to the people he is inspecting. He is your staff officer with line
authority available for action when in his judgment circumstances so
require. He can obtain all necessary information from the files at
division headquarters or by telegraphing your office. Your chief of
staff, the senior assistant general manager, will promulgate
instructions, while this traveling representative, like a trainmaster
on a division, will see that they are carried out. When he finds it
necessary to give instructions he should promptly notify your office,
that the record may be completed and confusion avoided. He can do all
this without becoming bureaucratic, without putting the company to the
expense of a great circumlocution office maintained under the feudal
notion of his royal importance. Railroad administration suffers from
too many offices and instructions, not from too few. The best
officials, and the best train dispatchers, give the fewest orders. It
is a qualitative rather than a quantitative proposition.
The moral effect of the presence of an official cannot be discounted.
We need more officials and fewer clerks. The railways are over-manned,
because they are under-officered. The great mistake of the past, due to
crude conceptions of organization, has been in creating offices rather
than officials.
The same line of reasoning applies to the handling of outlying
terminals on a division away from a dispatcher's office. The old idea
has been to locate a trainmaster with an office at such points. The
moral effect of his presence is unquestionably good. The objection is
that he must necessarily be on the road much of the time, and the train
crews are handled by a clerk. Duplication results because most of the
correspondence and records have to be referred to the superintendent's
office. The Union Pacific has found it better under the unit system to
have an assistant superintendent reside at such important terminals.
His office, however, is located with the superintendent, which
encourages travel back and forth, just what is desired, and discourages
sitting in an office and carrying on correspondence which can better be
looked after by the chief of staff in the superintendent's office. The
train crews are under the immediate direction of the yardmaster when in
the terminal, and of the train dispatcher when on the road.
The railroads of this country have suffered from rigidity in
administration. The unit system permits an elasticity of assignment to
take care of conditions as they come along. For example, your
non-resident assistant general manager can, if desirable, chaperon
three divisions when movement is heavy, and four or five, if you
please, during the dull season. You can on short notice throw all
assistants to the most exposed points. A non-resident assistant
superintendent can likewise be sent to an exposed district. A
permanently located trainmaster requires an official circular to have
his jurisdiction extended, and if suddenly ordered away can leave only
a clerk to represent the company. A railway has an ever-present firing
line. The more mobile the official force the more promptly can weak
portions be reinforced.
A striking violation of the unit principle in organization is to have
the master mechanic report to the division superintendent in
transportation matters and to the superintendent of motive power in
technical matters. This is a half-way attempt at divisional
organization which lacks the courage of conviction. Better have a
straight departmental organization with its divided authority and
expensive duplication than thus to straddle the question. If the
division is to be a real unit, it must be complete and self-contained.
The lack of balance in this attempt at divisional organization comes
from the fact that units are mixed. The superintendent of motive power,
a general officer with jurisdiction over the entire road, is a member
of the general manager's staff. He has a rank and value superior to
that of a divisional officer, the superintendent. The poor master
mechanic is often puzzled which superior to please. His natural
inclination will be toward the man higher up, the superintendent of
motive power. Again, it is difficult for any three men to agree upon
what are technical matters. The chief of staff method is not applicable
to this phase of the problem, because units have been mixed. The master
mechanic and the superintendent of motive power are not components of
the same integral unit. The unit system of organization requires a
superintendent of motive power to transact all business of record with
the office of the superintendent of the division, a component unit of
the general jurisdiction. The senior assistant general manager and the
senior assistant superintendent, each, as a chief of staff for the head
of his unit, decides promptly in the absence of the head of the unit,
what matters are sufficiently technical to demand the attention of a
particular official. Clear-cut, definite and prompt action is possible,
with proper checks and balances, because units are not mixed. The
governor can introduce a balance without throwing the administrative
machine out of gear to avoid stripping its cogs. The splendid personal
equation of railroad officials often serves to carry an illogical
organization in spite of its fundamental defects. Similar violations of
scientific principles in material things would cause bridges to
collapse and locomotives to break down. The showing made by the
railroads is a tribute to the administrative ability of their officials
rather than to their knowledge of organization. The Pennsylvania a half
century ago, and the Harriman Lines in more recent years, are said to
be the only roads that have made comprehensive studies of the science
of organization. Both of these great railways are prepared to stand the
test of time. Both will grow stronger as the years roll by. So feudal
is the conception of organization on most railways that the essential
elements of self-perpetuation are sadly lacking. Fortunately their
traffic strength is so great and our country develops so fast that
errors due to preconceived misconceptions and personal caprice are
covered up by increased earnings. One encouraging sign is that railway
officials have ceased to be quite so cocksure of themselves and are
seeking the underlying reason for the faith that is in them. True
science ever finds its vindication in impartial inquiry and intelligent
investigation. The world advances by definite steps rather than by
leaps and bounds. Do not lament the fact that some roads are groping
ahead only to occupy the abandoned organization camps of the Harriman
Lines. Be thankful rather that they have moved forward at all, that
though lacking in faith they are coming to a position admitting of
enlarged perspective.
Affectionately, your own,
D. A. D.
LETTER XV.
THE SIZE OF AN OPERATING DIVISION.
Los Angeles, Cal., July 15, 1911.
My Dear Boy:--How many miles of road should one division superintendent
handle? Like the old lady's recipe for pie crust, it all depends. Some
superintendents in the east with two hundred miles handle as much
business as do their western brothers with a thousand. As a matter of
fact mileage has little to do with the question. On the ideal division
the superintendent is in the middle with territory extending one
freight district in each direction. If he happens to be at a hub he can
comfortably handle several freight district spokes, which will increase
his mileage accordingly. Under such a condition the advantages of a
seemingly large mileage are numerous. The superintendent can run his
power wherever most needed. He can hold back at the farther end of one
district cars that he knows the connecting district cannot possibly
load or unload for several days. He can preserve a balance which is
impossible when jurisdictions divide at the hub. In the latter case
each superintendent hurries freight to the end of the division to avoid
a paper record showing delay on his territory. The result is that the
next man has terminal indigestion because he has been fed too fast.
Therefore, divisional jurisdiction should, when possible, change at an
outlying district terminal away from a large city. This avoids the
added complication due to industrial switching, suburban trains,
restricted area, etc., etc. A congestion of cars is often caused by a
congestion of jurisdictions. You may avoid the one by diffusing the
other. Several roads in the country have saved heavy expenditures for
larger terminal facilities by more scientific organizations.
The amount of mileage a superintendent can economically handle depends,
then, for the most part upon the location of his headquarters. Such
location in turn admits of no hard and fast rule. Cities and towns
spring up and industries develop quite regardless of the limits of a
hundred-mile freight district and a speed of ten miles per hour on the
ruling grade. A railroad usually begins and ends at a large city which
is either a seaport or a gateway. It is normally better to locate a
division superintendent at such beginning and ending city. He can then
handle its terminals and the one or more diverging freight districts.
His division should include the terminal at the farther end of such
districts, to afford him opportunity both to hold back stuff whose
inopportune arrival might congest the more complicated terminals at
headquarters and to relieve such terminals promptly by movement
outward. In other words, owing to his important terminals this
superintendent should have less mileage than his country brother who
would be in the middle between the second and third districts.
Some roads try to solve the problem by giving the superintendent the
first and second districts with headquarters in the middle. If in such
case the general offices happen to be at the initial point they soon
ignore the superintendent and do business direct with his terminal
subordinates. When this condition becomes intolerable, one of two
things usually happens. Perhaps the superintendent's office is moved to
the first terminal where it really belongs. Thereupon he loses full
touch with his freight crews on the second district, which is left out
in the air. The other attempted remedy is to appoint a superintendent
of terminals reporting direct to the general offices. The difference in
viewpoint thus legalized may cost the stockholders much money. To the
terminal superintendent the trains are always made up on time and the
power and road crews are seldom ready. To the division superintendent
the trains are seldom made up on time and the power and road crews are
always ready. Much energy of both officials and their offices as well
as that of the general superintendent and his office is then directed
to holding useless post mortems and negotiating unnecessary treaties of
peace. Remember, my boy, that typewriters exert no tractive power and
explanations move no cars. Self-preservation is the first law of
nature. We must so organize that this law will operate to keep the
company into clear, not to put some other fellow in the hole. All of
these questions are largely matters of opinion. After working with
every kind of terminal organization all over the country, your old dad
believes that the best is to have a division superintendent at the big
terminal with an assistant superintendent in direct charge of and
responsible for such terminal, the superintendent controlling every
diverging freight district to include the next terminal.
It should always be remembered that a large terminal demands preferred
consideration, because owing to restricted area its problems are
intensive and expensive. A dispatcher has a hundred miles or more over
which to keep his trains apart, while a yardmaster finds his engines
bunched within a mile or two. Again, if the cost of terminal switching
does occasionally happen to be reflected in a freight rate, the genial
gentlemen of the traffic department are prone to recommend its
absorption. I believe as a broad proposition that the management of
railroads is more scientific than that of most modern industries. I
would not like, however, to file much of their terminal operation as an
exhibit. A majority of the switch engines in the United States have one
superfluous man in the crew. This is partly because so few operating
officials have sufficient practical knowledge of switching to go out
and intelligently handle a crew all day. If you don't believe this,
make some time and motion studies of switching. Compare the relative
performance of your yard conductors. The tasks of road conductors are
relatively so well defined that comparison of individual performance is
not so difficult. The intense conditions of a terminal complicate such
differentiation as among yard conductors.
Another factor of prime importance in determining the size of an
operating division is the location of train dispatchers. The
dispatcher's table should always be considered an integral part of the
superintendent's headquarters offices. The train sheet is perhaps the
best record on a railroad. It is never fudged by being made up in
advance. It is a history usually unimpeachable because it is so close
to the actual transactions which it records. It deals with the essence
of railway operation, train movement. Few are the important records on
a railway that do not derive their primary data from the train sheet.
The sheet may be graphic, like a daily time card chart, or may be cut
up into card strips, as under the A B C system. In any form, it is a
fundamental of operating history.
The number of dispatchers to which a division is limited is, like the
number of miles, variable. With headquarters at the hub, one
superintendent and one chief dispatcher may comfortably handle three or
four sets of dispatchers. An outlying division with thin traffic may
require only one set of dispatchers. When it becomes necessary to
locate a set of dispatchers away from division headquarters, it is time
to appoint another superintendent and create a new division, perhaps
with only a light staff of all 'round officials. So important is the
train sheet and so much of vital, human interest centers around a
dispatcher's office, that the far away superintendent must refer much
correspondence to this detached portion of his office. The result is
expensive circumlocution and a lack of human touch. The superintendent
has in effect become a general superintendent too far away from real
things. A trainmaster or a chief dispatcher is really carrying the
responsibility of a superintendent without the title and authority
necessary for smooth administration. I know several railways that are
fooling themselves into the belief that they are saving money by having
one superintendent for two dispatching offices. One of them has five
superintendents and ten dispatching offices, really ten divisions in
fact, if not in name. By a logical arrangement of territory these ten
dispatching offices could be consolidated into seven division
headquarters and the road operated in seven divisions. In these days of
overtime and complex working schedules, a timekeeper should check the
time slips against the original train sheet, not against a copy, a
transcript or an excerpt. A division accounting bureau handling all
that it should handle has also much other use for the train sheet.
Second only in importance to the train sheet as a record, and with
which it should be closely related, is the conductor's car and tonnage
report; what the men call the wheel report. This important report made
by a division man is sent to a remote general office in disregard of
the responsible head of such division, the superintendent. The result
is that a distant authority, the superintendent of transportation, is
telling the superintendent that certain cars are being delayed on the
latter's division. This profuse correspondence is often foolish,
because meantime the cars have actually gone. Some roads now have a
carbon copy of the wheel report made for the use of the accounting
department. Why not send this carbon to division headquarters and let
the division accounting bureau make up the ton miles and the car miles,
subject to proper check after the fact? Why not have the office of the
superintendent know so much about the cars on his division that he will
tell the general offices that certain cars are being delayed on his
division for lack of motive power, loading or disposition, conditions
which, perhaps, the general office, with its larger view, can remedy?
This would also permit, when desirable, the checking of the agents' car
reports against the conductors' reports. The more closely to actual
transactions we can do our checking the more intelligent should be the
process and the smaller its volume.
I wish that you would come out here and see the Southern Pacific run
its monthly supply, pay and inspection train. Before coming, re-read my
letter to you on the subject some seven years ago. I know of no place
where the idea has been better carried out. Ideas seldom originate with
any one man. They seem rather to float around in the air. They are
pulled down by those who happen to erect lightning rods or like
Benjamin Franklin to fly kites. To vary the metaphor, do not laugh at
people who ride hobbies. Sometimes they ride well enough and far enough
to demonstrate that the hobby is a real horse. Then it is the turn of
the horse to laugh.
Whenever I see an announcement that a division has adopted the
telephone for train dispatching, I always feel that there should be an
accompanying apology for being several years behind the times. For
years progressive young railway men advocated the telephone only to be
assured by old-time dispatcher officials of the unwisdom of such a
course. Time and practical tests have shown that not only is the
telephone practicable for dispatching, but it actually makes operation
safer because of the increased human touch. Whenever and wherever we
can replace a specialist with an all 'round man we are gaining.
The first train dispatching is said to have been done by Charles Minot
when a superintendent on the Erie in the early fifties. So seriously
was the matter taken that only the superintendent himself could issue a
train order, even though this involved calling him out of bed. Hence
the foolish feudal custom of signing the superintendent's initials to
all train orders. It soon developed that a regular dispatcher was
necessary. Accordingly, a conductor, a man who knew how trains were
practically handled, was taken off the road and brought to the
superintendent's office to dispatch trains. Stop off at Port Jervis,
N.Y., some time and in a local hotel see the portraits of some of
these old Erie dispatcher-conductors, their dignity being protected
by the tall beaver hats of the period. The dispatcher not being a
telegrapher, he wrote out his orders and handed them to a young
operator to send. This operator was a bright fellow, who, by and by,
graduated into a dispatcher, able to send his own orders and often to
do the work previously requiring both men. Too often it has happened
that the experience of the new dispatcher, a telegrapher specialist,
was limited to the office end, with no firsthand experience in train
service. The telephone, fulfilling the immutable laws of evolution,
will take us back to first principles. The dispatchers of the future
will graduate from the train, engine and yard service, through the
dispatcher's office to higher official positions. The man who gives the
order will be a man who has once carried out such an order himself. The
man below will obey the more cheerfully and the more intelligently
because of increased confidence in the man above.
When the record is made up by the future historian, with that
discriminating perspective which time alone can give, high will be the
place accorded the railroad officials and employes of America. The
military, the pioneers of civilization, the forerunners of stability,
have their periods of enervating peace. Transportation, the first
handmaiden of progress, is in active attendance every day of the year.
Those who worship at her shrine and follow her teachings must lead the
strenuous life and love the voice of duty. The splendid, virile
performance of the past, handicapped often by crude facilities and
forced expansion, must and will be eclipsed under the intense, trying
conditions of the present and the future. In no profession more than in
ours is there eternity of opportunity.
Affectionately, your own,
D. A. D.
LETTER XVI.
SUPPLIES AND PURCHASES.
Salt Lake City, Utah, July 22, 1911.
My Dear Boy:--Supplies and purchases are a feature of railroad
operation illustrating the tendency to overcentralization through
overspecialization. Please notice that I say supplies and purchases;
not as some roads do, purchases and supplies. Is not "supply" the
broader term, including "purchase" as a very important component? If we
happen to make some of our supplies from our own scrap, a question of
supply and accounts is involved, but not necessarily one of purchase.
The volume of work involved in purchasing for a large railway may be so
great as to warrant the segregation of the purchasing function.