ElB—Lancaster Farming, Saturday, June 4,1983
Should production determine price supports?
(Editor’s Note: Should costs of
production be used as a basis for
establishing milk price supports?
Would that system be better than
the present parity But
how do you mee - costs of
production? Thes< and other
questions have bee cplored in a
detailed “Critique - Cost of Milk
Production Esttuuss and Their*
Use in setting Price Supports”
prepared by Blair J. Smith,
associate professor of agricultural
economics at Penn State. It will be
reproduced in Lancaster Fanning
beginning in today’s issue and
continuing throughout June.)
Part I
INTRODUCTION
The Agricultural Adjustment
Act of 1933 is the basic law under
which manufacturing grade milk
prices are supported through
government offers to purchase
certain manufactured dairy
products at announced prices. The
Agricultural Act of 1949 requires
that manufacturing grade milk
prices be supported at 75 to 90
percent of parity. The latter act
has been amended a number of
times by Congress, but, in the
absence of such amendments to
the contrary, the required level of
support always returns to or
remains at 75 to 90 percent.
The concept of a parity price for
milk is one that will give milk the
same purchasing power, in terms
of the things that farmers
generally buy to produce farm
products, as milk had with respect
to the goods and services farmers
bought in the five-year period 1910
through 1914. Theoretically, if
manufacturing grade milk is
currently selling at 85 percent of
parity, then milk has 85 percent of
the purchasing power that it had
from 1910 through 1914.
The federal government at
tempts to assure that the an
nounced support price will be
achieved in the market by offering
to buy butter, milk powder, and
cheese at specified prices. Prices
offered for these products are at
levels that are thought to permit
manufacturers to cover their costs
while paying farmers a price at
least equal to the desired level of
support. The market price that is
used as the gauge of the success of
the price support program is what
is referred to as the Mmnesota-
Wisconsm manufacturing grade
milk price (M-W). It is generally
the monthly average price paid to
farmers for manufacturing grade
milk sold in these two states.
Approximately 15 percent of all
milk produced in the United States
is still of manufacturing (Grade B)
quality, and approximately 50
percent of that Grade B milk is
produced in Minnesota and
Wisconsin.
The “basic” price m all Federal
Milk Marketing Orders (FMMOs)
is the M-W. This is the minimum
price that buyers must pay
producers for milk that is used to
produce such hard manufactured
products as butter, powder, and
cheese. All milk in all FMMOs is of
Grade A quality, eligible for use in
fresh, fluid form. Eighty percent of
all Grade A milk produced in the
United States, 68 percent of all
milk, is priced under one or
another FMMO. Milk in FMMOs
that is sold in fresh, fluid form
must return to producers the basic
price plus a fixed fluid or Class I
differential. This differential
varies across markets but not
across time. Thus, ail the variation
in required minimum Grade A
producer prices over time is due to
variation in the M-W.
The concept of parity as the
basis for determining support
prices, the price support program
itself, the use of the M-W as the
basic price in FMMOs, and the
FMMOs themselves have been a
continue to be under critical
review by a number of persons and
agencies.
The focus of this bulletin is on
one of the most commonly
suggested alternatives to parity,
cost of production. Other alter
natives are; (1) A parity price
based on the tlungs dairy fanners
buy rather than on those that all
farmers generally buy; (2) direct
payments to dairy farmers when
market prices fall below desired
levels: (3) subsidies to consumers
when controlled prices are higher
than they are willing to pay; and
(4( production or marketing
quotas to hold marketings down to
the level that will generate the
price that is desired m the market.
Interest in cost of production as a
possible basis for pricing milk at
the national level was evidenced in
the Agriculture and Consumer
Protection Act of 1973, which
directed the Secretary of
Agriculture to “... conduct a cost of
production study of wheat, feed
grains, cotton, and dairy com
modities ...” The notion was that
the estimates of such costs might
serve as "targets” for national
farm price and income policies and
programs.
The plan of this publication is to
develop a theoretical framework
within which costs of producing
milk might be measured with
complete precision. Then, the
practical problems associated with
trying to work within that
theoretical framework are
discussed. A critique of cost of
production studies generally is
offered, followed by a specific
critique of the senes of estimates
of the cost of producing milk that
are annually conducted by the U.S.
Department of Agriculture.
Finally, the appropriateness of
using cost of production as the
basis for pricing milk, in con
junction with one or another policy
instruments, is examined.
ESTIMATING THE COST OF
PRODUCING MILK
Laymen probably are inclined to
believe that estimating the cost of
producing milk ought to be simple
that an accurate and objective
figure shouldn’t be difficult to
come by. Yet knowledgeable
students of the matter recognize
that one can “crank out” costs that
vary over a wide range even
Blair J. Smith
from the same data set! Fur
thermore, there are many dif
ferent ways of developing a data
set, and each of these also may
result in a different estimate of the
cost of production as well.
More specifically, the primary
sources of variation in estimates of
the cost of producing milk result
from:
(1) The selection of a basic data
set
(2) Which items are included on
both the costs and the returns side
of the dairy farm business ledger.
(3) Which accounting
procedures are used to establish
values for the items that ere in
cluded in the computations.
Each of these potential sources
of vanaltion in estimates of the
cost of producing milk will be
discussed, in turn, immediately
below.
CHOICE OF A BASIC DATA SET
Basic data for estimating costs
of production have traditionally
been generated in two ways. One is
the farm survey method, the other
the farm budget method. The
former has the appeal of reality,
but is very expensive. The second
is a much cheaper method, but it is
difficult to know how well the
budgeted farm follows the changes
that take place on real dairy farms
over time. The farm survey
method risks -bias due to the
sampling procedure used, and the
farm budget method risks bias to
the extent that the modeled farm is
not truly typical or representative
»f all dairv fnrm>-
It is common to see a com
bination of the two approaches
used, however. Dairy farms are
surveyed periodically to establish
a base period, and budgeting is
used to develop estimates of costs
between surveys. This com
binational approach is the one
currently being used by the U.S.
Department of Agriculture in its
annual estimates of the cost of
producing milk. (1,2,3,4, and 5).
Since the survey of some sort
usually is involved in most
estimates of the cost of producing
milk, the attention of the
remainder of this article will be
focused on the survey method of
obtaining basic data. Some
workers will argue that the farm
budget or economic engineering
approach is potentially superior to
the survey method, in both theory
and practice. This author agrees
on the matter of theory, but
disagrees on the matter of prac
tice. The pros and cons of the farm
budget method will, therefore, not
be developed in this report.
In selecting a sample of milk
producers to be surveyed, a purely
random procedure is seldom
followed. Some further criteria are
ordinarily used. For example,
producers of fluid grade milk
(Grade A) ought to be
distinguished from producers of
manufacturing grade milk (Grade
B) if the cost of producing milk for
fluid use is of interest. One sup
poses, other things being equal,
that if costs more to produce Grade
A than Grade B milk. If Grade B
producers are included in the
sampe, then the cost of producing
Grade A milk, will be understand.
Another selection standard
might be size of dairy farm in
terms of numbers of milk cows.
Many surveys systematically
exclude farms with herds below a
certain size. Then there is the
definition of the dairy farm itself.
What will be used to differentiate it
from a non-dairy farm? Usually
some test of the important of the
dairy enterprise, relative to all
other enterprises, is applied. That
is', some cut-off level of
specialization is imposed on the
sampling procedure. Finally, there
is the matter of the actual
collection of the data. Will it be by
personal interview or mail
questionnaire? One supposes the
former method has the potential,
at least, of being more accurate
than the latter. Important in either
method is the design of the
questionnaire, and the knowledge
and industry of the field
enumerators and editors of the
questionnaire.
WHICH COSTS AND WHICH
RETURNS?
This question may seem trivial
to the uninitiated. All costs and all
returns, of course! Yet, an
examination of existing reports of
the cost of producing milk reveals
that workers have substantially
different notions about which costs
and which returns are relevant.
Why would any returns be relevant
in the computation of costs?
Because certain returns are offsets
to certain costs, as will be seen in
the later section dealing with the
theoretical classification of
returns.
It should be noted here that if
there is a bias m estimates of the
cost of producing milk, it is more
likely to be toward the higher
rather than the lower side. This is
because costs tend to be overstated
relative to returns. Or perhaps
more accurately, returns tend to
be understated relative to costs.
The reasons for these tendencies
may be summarized m two
statements, given the political,
regulatory, and economic climate
in which milk is produced in the
United States:
(11 The only case that comes to
mind wherein it might be ad
vantageous to the farmer to report
lower rather than higher costs of
producing milk is when farm
business financing is being sought
and
(2) Many of the off-farm people
who have been involved in
estimating the cost of producing
milk are sympathetic to dairy
farmers. Consequently, the data
gathering and interpreting
methods they use often have cost
escalating biases built into them.
Of some costs and returns items
there is little debate as to whether
they ought to be included in any
estimate of the cost of producing
milk. (Here, for the moment, the
problem of allocating certain costs
among enterprises on multiple
enterprise farms is put aside).
Items such as purchased feeds,
hired labor, fertilizer and lime.
(Turn to Page E2O)