Factory Expense
Now if our product is simple and all of one kind, the
determination will be easy enough. It is when product is
diversified that accurate cost accounting becomes difficult
and at the same time becomes more important. Suppose,
for example, we are running a cotton-seed oil mill and mak-
ing a single grade of oil. The cost per pound is very
simply found by dividing total expenditures by the total
number of pounds made. But suppose, further, we decide
to branch out and work up our own product. We install
a refinery and begin to put out a fancy grade of oil for
table use; we get up a " lard substitute "; we install a soap
works and make several grades of toilet and laundry soap ;
we follow with a glycerine plant; and finally we manage
to do something with several kinds of by-products. Now
we have a number of different products, selling at very
different prices, in different markets, and under different
conditions of competition. There may be big money in
lard compound, while the soap market is so hard pressed
by competition or so captured by large manufacturers who
lavish money on advertising that we can not sell soap at
a profit. But unless we know accurately what lard com-
pound costs us per pound, or what soap costs us per box,
how can we tell that there is a profit in one and a loss in
the other? How can we know that we should put all our
raw material into lard compound and cultivate that market,
and that we should shut down the soap factory? Knowl-
edge of costs is the guide to success and, indeed, a necessity
to existence in modern commercial manufacturing.
In this exact determination of costs the most troublesome
factor as already stated is the element of expense. Material
and labor are fairly concrete, definite and tangible things.
We can see them, weigh them, measure them, and connect
them directly with the product they assist to form. If we
take any single article in the whole output of our plant,
whether it is a pound of cottolene, a cake of soap, a hat,
a globe valve or a dynamo, we should be able by com-
paratively simple records and accounts to know exactly the
value of the material that went into it, and exactly the out-
lay for the direct labor that has been expended upon it.
But in the total expenditures of any manufacturing busi-
ness there is a very large outlay (usually a very large frac-
tion of all the outlay) that is not for material, and is not
for labor, and yet we must get it back from our customers.
A proper proportion must be repaid to us in the price we
get for each bit of product we sell. If each article sold
does not repay us for its just proportion of these general ex-
penditures, as well as for its just proportion of material
and labor, our business will be headed toward failure and
not toward success.
It is these miscellaneous expenditures, not of themselves
productive of anything and yet necessary to the production
of things, that make up the expense account.
Among them are rent or interest on the cost of land
and buildings, insurance, repairs, salaries of general officers
or officials, of clerical staff and all unproductive labor,
power, light, heat, legal expenses, advertising and selling,
etc. The total is a load bearing upon the extra business,
and each item of product must carry its share hence
the figure of speech, " burden."
The distribution of expense (that is, the assessment of
a just and proper fraction of it as a part of the cost of
each item of our product) is not only one of the most dif-
ficult, but also one of the most controversial and most un-
satisfactory problems of works management or shop ac-
counting. This is because expense is not like the material
and labor components of a manufactured product, which
are absolute, concrete factors known quantities that are
permanent, fixed and absolute in value. The expense
component of any single item is really an elusive variable,
to which we give a value arbitrarily taken because it solves
some particular case or problem.
Let us illustrate the point again by means of a pocket-
knife. Let us suppose the simplest possible conditions
that we are making nothing but one kind, size, and style of
knife. Suppose our cost records show that the material used
in this knife is worth 20 cents, and the labor that made it an-
other 20 cents. Our prime or flat cost, as it is called, is 40
cents. We find, perhaps, that by the most careful and cor-
rect compilation and distribution we can make of all our fac-
tory expense (that is, our expenditures for things other than
material and direct labor), this knife should be burdened
with an expense charge of 10 cents that is, it should be
considered to have cost 20 cents for material, 20 cents for
labor, and 10 cents for expense, in order to return to us
our entire manufacturing expenditure. Let us suppose
that of this 10 cents expense burden i cent goes to pay
this knife's proportion of the president's salary, and I cent
goes toward the general manager's salary, and 3 cents go
for other office salaries, and i cent goes for rent, and i
cent for the coal bill, and i cent for general repairs and
2 cents for sundries.
Now suppose we had not made this particular individual
knife. Our cost facts as to material and labor would prove
their absolute truth by transposing the equation. We
should actually save 20 cents for material and 20 cents for
the labor. That 40 cents would remain unexpended and
we should have it in the treasury. We would save 40
cents in actual money by refraining from the manufacture
of this particular article. But our assumed expense fact
goes all to pieces. We do not, by not making this knife,
save i cent on the president's salary, or i cent on the gen-
eral manager's salary, nor do we reduce our rent, or lessen
our repairs, or cut down any of those other expense items
(except possibly the coal) by the figures we attributed to
the expense burden of this individual knife. What does
happen is that all the other knives we do make have to
bear between them just the same total expense as before,
or a little larger expense burden each.
But let us not leave this example without noticing an-
other point. We have remarked so far that a difference
of even one knife more or less in our total product makes
a corresponding actual difference in our total outlay for ma-
terial and labor, but practically no difference in our total
expense account; and we have deduced from this that a
scheme of expense distribution that is true for a certain
volume of output becomes untrue at any other volume of
output, whether larger or smaller.
It would be incorrect, however, to assume that the ex-
pense burden as a whole does not ever vary, or indeed that
it does not vary considerably, with varying volume of busi-
ness. The truth is that expense burden is made up of a
large number of elements, some of which go up and down
in general correspondence with the volume of business and
some of which do not. In other words, our total expense
is divisible into two classes constant and variable. The
former division (constant expense) includes all expense
items necessary, so to speak, to the mere existence of the
business, while the latter division (variable expense) includes
all items connected with the activity of the business.
For example : In the constant-expense section we should
include rent, or its equivalent in interest, insurance and
taxes, if we own our real estate and buildings. This clearly
remains uniform or unchanged, whether the factory be
busy or idle. Another such item is the salaries of general
officers ; they draw their pay the same in good times or in
bad. It is true that on a very great expansion of business
we might have to acquire more ground and put up more
buildings, or rent more space, or enlarge our organization
and add more salaried officers. Or in very dull times we
might give up some of the property we have been renting
and we might cut down official salaries; and so these so-
called constant expenses may change. But if they change
it is by occasional large steps of this kind. They remain
level for long periods, and there is a minimum below which
they can never go if the business is to continue to exist at
all.
On the other hand, expenses like advertising, selling,
correspondence, clerical assistance, drafting, power, trans-
portation, foremen, yard labor all these go up and
down on curves corresponding closely and quite sensitively
to the amount of business we are doing, and many of them
can be completely cut off if the plant is wholly shut down.
So the second great point to keep in mind is that while
the ratio of expense to productive labor and materials (or
in other words, the proportion of our total cost chargeable
to expense) is variable and is constantly varying in a way
that from an accounting point of view is very troublesome,
this variation is caused by the fact that a certain very
large part of our expense account is constant, or nearly so,
however our total volume of business may vary. It sounds
like a paradox, but the proportion of expense varies be-
cause the total of expense does not. This fixed necessary
outlay stands little changed from month to month, while
the gross income against which this is balanced fluctuates
now up and now down.
The result is that as business becomes more active the
expense ratio drops even though the expense total may
rise, while as business shrinks the expense ratio rises even
though the expense total may fall. This is the reason why
in dull times dividends on industrial and railways stocks are
so frequently reduced or passed. Business may be (say)
50 per cent of normal; purchases are cut down, hours are
shortened, employees are discharged, trains are laid off,
purchases of material are suspended, actual operations and
expenditures for actual production are cut down to one-
half but profits do not remain at half the normal. They
vanish entirely and a deficit appears instead because the ir-
reducible constant expense eats all and more than the gross
profits earned by the 50 per cent activity.
To come back now to our imaginary knife factory; we see
that while we may be certain enough what our whole ex-
pense account amounts to, the assumption that the indi-
vidual expense burden chargeable to each individual knife
is 10 cents is an assumption only. It is a convenient ap-
proximation to truth which holds good under average con-
ditions, but begins to depart from truth as soon as and as fast
as conditions depart from average. That is the first diffi-
culty in distributing expense burden.
But suppose, further, we are making not only pocket
knives, but also carving knives and safety razors. We can
tell exactly how much material and how much direct labor
each pocket knife, and each carving knife and each safety
razor takes. We can tell exactly how much our total ex-
pense is. But how shall we tell just how much of this
total expense is occasioned by the manufacture of a carving
knife, of a safety razor, or of a pocket knife, or whether
there is more general expense occasioned by the manufac-
ture of one of these articles than by another? Does forg-
ing a carving blade consume more power and use more coal
than forging a pocket-knife blade, or does timekeeping and
clerical labor run higher in the safety-razor shop than it does
in the pocket-knife department? Should, therefore, each
carving knife or each safety razor (for these and other
similar reasons) bear a larger burden of expense than each
pocket knife? If so, how much?
May be the carving-knife account does not show satis-
factory profits, and we think of giving up that branch of
the business. But are the apparent profits small because
we are charging it with more than its true share of ex-
pense, and thus relieving the pocket knives and the safety
razors of some of the burden they ought to bear? If we
drop the manufacture of carving knives, will our expense
account drop by the amount of burden we have been charging
up to the carving-knife department, or shall we still find the
same old expense totals bearing now wholly on pocket
knives and safety razors and shall we be worse off rather than
better? Would it be sound policy, instead of abandoning
any line, to add still another that would bring a reasonable
profit over the flat cost of materials and labor, in the ex-
pectation that in fact no increase of expense would be oc-
casioned, and we should be just that much ahead on our
total profit and loss account?
Here we see the second difficulty in the expense distri-
bution, which is to apportion the total properly among
the several or many lines of product in a varied manu-
facturing business, so that the calculated costs of each (on
which we base our selling prices) may be as near as possible
to truth. Then whatever line may expand or contract we
shall be safe from disastrous disappointment in the total
of our profits.
In order to see more clearly how the proportion of ex-
pense justly chargeable to various lines of products may vary
that is, how various components of expense are created
in unequal proportion by various classes of manufactured
goods, and hence should be borne with corresponding in-
equality by these various classes and to see also some of
the considerations affecting the distribution of expense, let us
imagine that we are making a hasty tour through a machine-
shop and let us see in part how and where the burden is
created. We will assume that the shop makes its own
castings and we will begin with the foundry. The material
(pig iron) and the labor of molders, helpers, core-makers,
etc., on each and every job and piece, can be pretty closely
recorded, so that our material and direct-labor costs are
reasonably exact. But here are some men who are not en-
gaged in making any special casting into which their work
goes and to which it can be charged; they are wheeling sand,
shaking out flasks, charging the cupola. Here is coke go-
ing into the cupola to be burned, and power being used for
the cupola hoist and for furnishing the blast. Without
searching any further, we find already an aggregated out-
lay an expense burden which we can not attach to any
one piece of material or to any one job, but which we must
distribute somehow among all the jobs done that day or
on that melt.
We see however, further, that there is another cupola
on which men are busy making repairs. Evidently there
is an outlay for refractory linings, labor, and incidentals,
which must somehow be loaded on to the foundry product
and repaid by its sale. We must keep our cupola in re-
pair; it costs money to repair it, and we must manage to
get our money back. But this expense was incurred through
wear and tear attending the melting of iron for all the
castings made in a week or a month, perhaps. Our total
of this repair bill, then, must be distributed over the jobs
of that whole period if each is to bear its fair burden.
We see, further, that other men are at work removing
dust from the rafters, repairing the roof, and white-wash-
ing or painting the whole foundry building. They are
remedying the deterioration or decay of possibly a year.
Again, money spent in general expenses, to be recovered
in the sale of product. Here is another item of burden to
be spread over a still wider section of our output.
Here, in all these cases, we have burden limited in dis-
tribution according to time.
We pass to the machine-shop, and we see a large overhead
crane transporting a heavy engine bed to the planer. The
crane itself represents invested capital which is disappearing
constantly year by year as the machine in which it is in-
vested wears out. Furthermore, it costs money to run that
crane money for interest on the investment required for
its installation, for power to run the crane, for the man who
operates it. Some of this cost accumulates night and day,
whether the crane is running or is idle; some accrues only
when it is in operation. But it accrues, and we must charge
it against our product somehow and get it returned to us
with profit. Evidently, though, it would be unfair to levy
any of it against our lighter lines of manufacture, which do
not need crane service and never use it. On the other
hand, here is a little industrial railway used for moving
light and medium-weight pieces around the shop. This is
an expense item of similar sort, but here the burden is not
chargeable against heavy product.
Here we have burden limited in distribution by weight or
character of product.
We enter the lathe department and find a foreman in
charge. His wages are paid him every week and enter
into the total of our manufacturing costs, but they do not
appear on the job tickets for any of the individual items
of work handled on the lathes. His wages, also, then,
must be taken care of in the manufacturing-expense burden;
but they are incurred in connection only with the lathe
work, and in justice no fraction of them should be attached
to any of our manufactured product which has not had lathe
work done upon it.
Here we have burden limited in distribution by the char-
acter of operation.
As we pass through the shops, we notice here and there
a timekeeper at work, securing data as to the times when
jobs had been begun or finished, and here as we approach
the offices is a room where several clerks are entering the
time records and computing premiums or bonuses. Evi-
dently this is a necessary auxiliary to our productive system,
although it is itself unproductive. The cost of the employ-
ment of these clerks and of attendant expenses must go into
our burden; what particular fraction of it is theoretically
attached to any particular machine we manufacture and
sell, obviously should depend upon the complexity of that
machine the number of parts, and hence of operations
and times, which had to be recorded, and the demands its
computations and calculations make upon the time and
services of the time clerks. Here we have burden varying
according to the complexity of the product. Next, if we
look into the sales office (as we should do) we shall see
a probably large and expensive force of men, with the aid
of considerable outlay for office assistants, advertising, and
publicity work. The total of this expense of this com-
mercial burden must be taken care of, and if we look
into it we shall probably find that the necessity for these
expenses varies very widely between different lines of our
manufacture. Standard product disposed of through
dealers probably almost sells itself. Special business, or
new business for which the market must be created, prob-
ably costs a great deal to work up. Here we have burden
varying according to commercial conditions.
It will be apparent from the view we have had so far
that no absolute, mathematically correct and invariably
true distribution of expense can be made. We must ac-
cept some reasonably fair distribution that will serve within
allowable limits of error under ordinary fluctuations in busi-
ness, and we must give separate and careful attention to
extraordinary conditions that may make our methods and
figures, temporarily at least, inaccurate. The methods
generally used are more or less rough-and-ready approxima-
tions, convenient to use, sometimes as misleading as they
are convenient, but often quite good enough for practical
purposes, especially as the experienced industrial manager
has a sort of sixth sense, or specially trained common-sense,
by which he corrects the occasional false readings of his
cost system.
These methods will be outlined in the following chapter. 1
1 A very thorough discussion of this subject will be found in " The
Distribution of Expense Burden," by A. Hamilton Church ; The Engineer-
ing Magazine.
Ud. 3.4.2026
Pub. 18.9.2023
