DlB—Lancaster Fanning, Saturday, June 18,1983
Using production
THE U.S. DEPARTMENT OF
AGRICULTURE ESTIMATES
OF THE COST OF
PRODUCING MILK
Description of studies
and procedures
The most comprehensive recent
estimates of the cost of producing
milk have been published annually
by the U.S. Department of
Agriculture since 1974 (1, 2,3, 4,
and 5). The first report, pertaining
to the year 1974, was based on a
national survey of dairy farms.
The succeeding final reports for
1975, 1976, 1977, and 1978, the
preliminary 1979, and the
projected 1980 estimates used the
1974 study as the base structural
situation. Costs and returns were
updated using a set of estimating
procedures referred to as the firm
enterprise data system (FEDS)
(7). In 1980 a national survey was
conducted pertaining to the year
1979 to establish a new base for
estimates that are to be made in
succeeding years.
Recognition of the several dif
ferent procedures that might be
followed in estimating the cost of
producing nulk, and of the dif
ferent assumptions that might be
made with respect to several key
components of costs and returns,
are presented in the U.S. Depart
ment of Agriculture cost of
production (USDA-COP) series.
Several changes were made m
those procedures during the 1974
through 1979 period. In the report
for 1974, four basic estimates of the
cost of producing milk were
developed:
(1) Feed at cost of production;
land at acquisition cost
(2) Feed at cost of production;
land at current value for
agricultural purposes
(3) Feed at prices received by
farmers; land at acquisition cost
(4) Feed at prices received by
farmers; land at current value for
agricultural purposes.
The estimates of the cost of
producing 100 pounds of milk that
resulted from each of the four sets
of assumptions above are shown m
the following table. The range of
these estimates is $1.34, and
clearly illustrates the effect that a
simple choice of assumptions can
have on the computed cost of
producing milk.
Estimate* of the cost of producing milk,
1974 USDA survey (I)
At cost of At prices
production received
by farmers
Land
(dollars per hundredweight)
At acquisition cost $8 58
At current
agricultural value 59 05 59 92
Included in all the 1974 estimates
was the "milk equivalent” of dairy
cattle and calves that were sold
during the year. To arrive at such
a figure, receipts from the sale of
dairy cattle and calves were
divided by the price received by
the farmer for milk sold off the
farm. Then total hundredweights
of milk, for purposes of computing
the various costs per hun
dredweight, was the sum of milk
sold, milk consumed on the farm,
and the milk equivalent of dairy
cattle and calves that were sold. It
was deemed too difficult to
separate the costs associated with
creating the meat of those dairy
cattle and calves from other dairy
enterprise costs. The implication
of the proceduce used was that the
production of such meat was no
more nor less profitable than the
production of milk.
In the reports following the one
for 1974, the results of only one of
the four alternative combinations
of assumptions shown in the table
were provided. These are the ones
that incorporate feed costs at
prices received by farmers and
land at its current value for
agricultural purposes. As shown in
the table, this combination in 1974
yeilded the highest estimate of the
cost of producing milk. Feed prices
have fluctuated quite widely in
recent years, and there have been
times when cost of production
(COP) based on prices received by
farmers might have been lower
than COP based on cost of
producing feed. Nevertheless, the
normal expectation is to the
contrary, or else dairy farmers
would be expected to buy all their
feed. The rationale for dropping
the alternative of including
homegrown feeds at their cost of
production is that . . the dairy
enterprise must pay its own way,
Feed
$9 45
Blair J. Smith
costs to figure price supports
that it should not be supplemented
by other enterprises to make it
profitable’ ’ (3, page 3). The reports
do not seem to reveal the reason
for dropping the alternative of
including land at its acquisition
cost, however.
Another change
One other important change has
been made m the USDA-COP
series since 1974. This showed up
for the first tune in the 1979 report,
and concerns the way the value of
sales of cull cows and veal calves is
incorporated into cost estimates.
How it was handled earlier was
mentmed above. In the 1979 and
succeeding reports, the value of
such sales was not subtracted out
of the dairy enterprise costs until
after the per hundredweight costs
of the different items were com
puted. This had the effect of in
creasing the 1979 and succeeding
estimates of per hundredweight
costs above what they would have
been had the method of handling
the value of sales of cull cows and
veal calves prior to 1979 continues
to be used.
Critique of the USDA-COP studies
The general approach of the
USDA-COP studies is enterprise,
rather than whole farm. That is,
only the costs and returns that are
specifically assignable or at
tributable to the dairy herd are
included in the computations. It is
doubtful this can be done with a
high degree of accuracy with the
resources that are committed to
the effort on a national basis.
Furthermore, the notion that "...
the dairy enterprise must pay its
own way, that it should not be
supplemented by other enterprises
to make it profitable” might create
a potential for bias and loss of
realism that must be recognized.
The general reasons for possible
upward bias in estimates of costs
were noted earlier in this report.
The tendency under the USDA
enterprise approach may be to
‘load” the dairy enterprise with
such a heavy share of farm costs
that the apparent profits to any
other enterprises present are
overstated. These "profits” might
be so high, in fact, that one could
wonder why any milk is produced
at all.
The loss of realism that arises
out of the assumption that the
dairy enterprise must “pay its own
way” is that although most milk is
produced on multiple-enterprise
farms, those other enterprises are
usually meant to complement, not
compete with, the dairy en
terprise. The benefits of such
complementarity should be shared
among enterprises, not ascribed to
all of them except milk production,
if one is interested in explaining
why there is as much milk
produced as there is, one must look
at milk production in the setting in
which it is actually earned out, not
at an abstraction from that setting.
In other words, the focus of cost of
production estimates should be on
what it costs to produce milk as it
produced, not on what it would cost
if it were all produced on single
enterprise dairy farms.
What about those farms that
produce some milk but on which
the dairy enterprise is not the
major one? A criterion for level of
specialization can be established
that will screen out such farms
from inclusion in cost of production
surveys. This would have little
effect on the representativeness of
the farm sample because so much
of the milk in the United States is
produced on very highly
specialized dairy farms. For the
farms that meet the test of
specialized dairy farms. For the
farms that meet the test of
specialization, the whole farm
approach to estimating cost of
production which incorporates the
porportioned treatment of sideline
revenues is recommended. This is
probably the least expensive and
most accurate way of estimating
the cost of producing milk. It does
lack the detail that the enterprise
approach demands. But when the
resources necessary to ap
propriately carry out the en
terprise approach are inadequate,
the accuracy of such detail must be
questioned.
In the opinion of this writer, the
USDA-COP studies have other
shortcomings in addition to the one
of their basic approach. The more
critical or bothersome of these are
discussed immediately below.
Home grown feeds
Home grown feeds, except
pasture, that are fed to the dairy
herd are charged in at the price
they could have received had they
been sold in the market. The costs
of producing all feeds are meant to
be separated from other farm
costs and deducted from total
costs. Thus, the revenues that are
obtained from any feeds that are
sold are not further accounted for.
An alternative to the approach
taken by the USDA is to implicitly
charge home grown feeds at the
cost of producing those feeds. This
is accomplished by taking the the
"whole farm” approach in com
puting cost of production and
handling sales of surplus home
grown feeds as proportioned
sideline revenue.
Again, for emphasis, the
preference of this writer is to
estimate the cost of producing milk
as it is actually produced in any
geographic area of interest. The
J
use of synthetic or engineered
milking enterprises using some
assumed set of “benchmark”
practices may not adequately
portray the real world. Any change
that takes place in the dairy in
dustry must begin with the current
base that has been established
over many years. Change is an
evolutionary process, and
wholesale leaps from the actual to
some assumed ideal rarely take
place.
Incidentally, few workers have
ever suggested that the cost of
raising dairy herd replacements
be separated out from the costs
associated with the milking herd.
Yet, to carry the notion that the
milking herd should pay its own
way to the ultimate, costs and
returns to the herd replacement
activity should be separated from
those of the milk producing ac
tivity.
Land costs
land costs are included m the
USDA studies based on the farmer
respondent’s estimate of the value
of that land for agricultural pur
poses. The alternative of valuing
land at cost when acquired was
deleted from the estimates early in
the COP senes. Valuing land at its
opportunity costs in its highest
valued alternative use,
presumably for urban or industrial
development, has never been in
corporated in the USDA-COP
estimates.
Placing a proper value on land so
that an appropriate cost can be
assigned to the dairy enterprise
has already been asserted to be
both difficult and important. The
only time one really knows what
land is worth for any purpose is
when it is sold and there are
very few whole dairy farms that
change hands in any given area in
any given year. More commonly,
one existing dairy farmer will buy
an adjoining farm or parcel of
land, and integrate that land into
his present operation without
actually creating a new or ad
ditional dairy enterprise.
Unless a farm has been in the
same family for several
generations, valuing land at cost of
acquisition is at least based on a
known figure. Most other methods
of valuing land are merely
estimates or judgements of the
value of the land if it were to be
sold for agricultural or other
purposes. An exception to these
latter methods would be the use of
net rental rates for establishing
land costs. Whatever rent far
mland brings in agricultural uses
could be used as the cost of land. A
fairly active rental market is
required to establish good
estimates, but the use of farmland
rents to establish land costs has a
great deal of merit, in light of the
alternatives available.
Economists are advocates of the
opportunity cost concept of valuing
equity assets. That is, the cost of
these assets to an enterprise is
what those assets could earn in
their best alternative employment.
Supposedly, the owner of such
assets will not keep them in any
use that pays less than they could
earn in another use at least not
for very long. Yet, if the op
portunity cost concept did prevail,
given the ways costs and returns to
dairying have usually been
figured, there would be little or no
milk produced in the United States
today at all. With the primary
exception of the USDA-COP
studies for the years 1977 and later,
the computed average costs of
producing milk have almost
always been greater than the
computed average prices received
for milk, and often by rather large
amounts. What this probably
means, most realistically, is that
opportunity costs haven’t been
figured correctly. Other factors
enter in, but we have trouble
(Turn to Page Dig)