q
I
O
2
I
H
M
u.
M
8
u
M
2

 

vvvvvv

THESIS

This is to certify that the

thesis entitled

A S\‘M\L\‘\:0~Y\€OLLS PYGd'KC‘h-VC’ Monl
0C {he waited Sides Domestic
$07590.“ fvuiuShr)’

presented by

“amok. dent) Ckvx‘s-CQ V1;_QV\

has been accepted towards fulfillment
of the requirements for

M S ' degree in QQMM WOMWS

 

(/

Major profe I

Date 5/95/77

0-7639

 

A SIMULTANEOUS PREDICTIVE
MODEL OF THE UNITED STATES

DOMESTIC SOYBEAN INDUSTRY

BY

Thomas H. Christensen

A Thesis

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

Master of Science

Department of Agricultural Economics

1979

ABSTRACT

A SIMULTANEOUS PREDICTIVE MODEL OF THE
UNITED STATES DOMESTIC SOYBEAN INDUSTRY

By Thomas H. Christensen

This study estimates parameters related to soybean meal
and soybean oil demand for the United States in the marketing
years 1954-1977. Parameter estimates related to soybean
supply in the United States are also estimated for this same
period.

A twelve-equation simultaneous model for the United
States soybean industry is developed. The quantity of soy-
beans produced at the beginning of a marketing year is
first estimated from pre-determined (known) values. The
estimated supply of soybeans is then interacted with domestic
(and exogenous foreign) demand conditions to yield an« esti-
mate of domestic soybean utilization and soybean price.

The entire system of equations is then solved, using
forecasted values for all exogenous variables, in order to
provide estimates of the market conditions that will prevail
in the United States soybean industry through the market year

of 1982.

TABLE OF CONTENTS

ABSTWT . I O O O O O O O O O Q 0 O O O O O O 0

LIST OF TABLES . C C O O O O O 0 O C O O O 0

LIST OF

Chapter

I.

II.

III.

IV.

FIGUMS . O O O O O O O O O C O O O 0

INTRODUCTION . . . . . . . . . .

The Problem . . . . . . . .
Objectives of the Study . . . . .

THE UNITED STATES SOYBEAN INDUSTRY . .

Growth of the United States Soybean

Industry. . . . . . . . . . . . .
Soybean Meal. . . . . . . . . . . .
Soybean Oil . . . . . . . . . . . .
Direct Utilization of Soybeans. . .

Soybean Production: Alternative Crops

and Regionality . . . . . . . . .
Soybean Yields . . . . . . . . . .
Soybean Market Interaction . . . .
Soybean Supply . . . . . . . .
Soybean Demand . . . . . . . .

REVIEW OF THE LITERATURE . . . . . . .
THE MODEL . . . . . . . . . . . . .

Variables Employed . . . . . . . .
Endogenous Variables . . . . . . .
Predetermined Variables . . . . . .
Exogenous Variables . . . . . .
Notes on Some of the Variables
Demand Component . . . . . . .
Supply Component .
Assumptions . . . . . . . .

Model Rationale . . . . .

Page

iii

iv

MN

Gib

10
10
14
15

17

Chapter Page

V. RESULTS OF ESTIMATION . . . . . . . . . . 40
Supply Equations . . . . . . . . . . 40
Demand Equations . . . . . . . . . . 41
Results in General . . . . . . . . . 44
The Results: Supply Component . . . 44
The Results: Demand Component . . . 48
Analysis of Forecasting Ability . . . 53
Theil Coefficients . . . . . . . . . 53
Decomposition of Mean-Square Error . 55
Results of Estimation: Forecast . '

Ability . . . . . . . . . . . . . 56
Results of Simultaneous Forecast . . 57
Underlying Assumptions . . . . . . . 58

Exogenous Forecasts From National .
Model . . . . . . . . . . . . . . . 60
Conditional Forecasts . . . . . . . . 61
VI. CONCLUSION. . . . . . . . . . . . . . . . 70
Major Findings . . . . . . . . . . . . 70
Results of Forecasts . . . . . . . . . 72
Closing Remarks . . . . . . . . . . . 74
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . 75
APPENDIX A . . . . . . . . . . . . . . . . . . . 75
APPENDIX B . . . . . . . . . . . . . . . . . . . . 78
APPENDIX c f . . . . . . . . . . . . . . . . . . . 35
APPENDIX D . . . . . . . . . . . . . . . . . . . 94

APPENDIX E ‘ . . . . . . . . . . . . . . . . . . . 102

ii

LIST OF TABLES

Table Page

l.A Comparison of Estimated Supply
Elasticities. . . . . . . 46

2.A Comparison of Short-Run Demand
Elasticities for Soybean Oil and
Soybean Meal. . . . . . . 52

3.Analysis of the Forecasting Ability
of the Model. . . . . . . 54

LIST OF FIGURES

Figure
l.Soybean Production Areas, U.S..

2.U.S. Supply and Disposition of Soybeans
and Soybean Products. . . . .

3.Acres of Soybeans Planted - Northern
and Southern U.S. . . .

4.Quantity Demanded of Soybean Meal,
Soybean Oil, and Total Oil.

S.Soybean Price, Dollars Per Bushel,
1978 Actual = 6.50. . . .

6.Soybean Meal Production, 1000 Tons,
1978 Actual = 23, 754. . . .

7.Soybean Oil Production, Million Pounds,
1978 Actual = 10,800 . . .

8.Beginning Soybean Stocks, Million
Bushels, 1978 Actual = 161, 1979
Actual = 150. . . . .

9.Soybean Production, 1000 Bushels,
1978 Actual = 1,842,647..

10.Acres of Soybeans Harvested - Southern
U.S., 1000 Acres, 1978 Actual=lS,560.

ll.Acres of Soybeans Harvested - U.S.,
1000 Acres, 1978 Actual = 63,033.

12.Soybeans Processed, Million Bushels,
1978 Actual = 1000. .

l3.Soybean Price, Dollars Per Bushel, 1978

Actual = 6.50, Reduced Yield Scenario.

l4.Soybean Meal Production, 1000 Tons,
1978 Actual = 23,754, Reduced Yield
Scenario. -. - . - . . -.

iv

Page

12

16

47

49

62

63

64

65

67

68

69

78

79

Figure

15.

16.

l7.

18

19.

20.

21.

22.

23.

24.

25.

Soybean Oil Production, Million
Pounds, 1978 Actual = 10,800,
Reduced Yield Scenario.

Beginning Soybean Stocks, Million
Bushels, 1978 Actual = 161, 1979
Actual = 150, Reduced Yield Scenario.

Soybean Production, 1000 Bushels, 1978
Actual = 1,842,647, Reduced Yield
Scenario. . . . . .

.Acres of Soybeans Harvested - Southern

U.S., 1000 Acres, 1978 Actual =
15,560, Reduced Yield Scenario.

Acres of Soybeans Harvested - U.S.,
1000 Acres, 1978 Actual = 63,033,
Reduced Yield Scenario. .

Soybeans Processed, Million Bushels,
1978 Actual = 1000, Reduced Yield
Scenario. . . . . .

Soybean Price, Dollars Per Bushel, 1978
Actual = 6.50, Increased Yields
Scenario. . . . .

Soybean Meal Production, 1000 Tons,
1978 Actual = 23,754, Increased ‘
Yields Scenario. . . . .

Soybean Oil Production, Million Pounds,
1978 Actual = 10,800, Increased
Yields Scenario. . .

Beginning Soybean Stocks, Million
Bushels, 1978 Actual = 161, 1979
Actual = 150, Increased Yields
Scenario. . . . .

Soybean Production, 1000 Bushels, 1978
Actual = 1,842, 647, Increased Yields
Scenario. . . . .

Page

80

82

'83

84

85

86

87

88

89

9O

Figure Page

26.Acres of Soybeans Harvested - Southern
U.S., 1000 Acres, 1978 Actual =
15,560, Increased Yields Scenario. . 91

27.Acres of Soybeans Harvested - U.S.,
1000 Acres, 1978 Actual = 63,033,
Increased Yields Scenario. . . . 92

28.Soybeans Processed, Million Bushels,
1978 Actual = 1000, Increased Yields
Scenario. . . . . . . . 93

29.Soybean Price, Dollars Per Bushel,
1978 Actual = 6.50, Increased Export
Scenario. . . . . . . . 94

30.Soybean Meal Production, 1000 Tons,
1978 Actual = 23,754, Increased
Export Scenario. . . . . . . 95

31.Soybean Oil Production, Million Pounds
1978 Actual = 10,800, Increased
Export Scenario. . . . . . . 96

32.Beginning Soybean Stocks, Million
Bushels, 1978 Actual = 161, 1979
Actual = 150, Increased Export
Scenario. . . . . . . . 97

33.Soybean Production, 1000 Bushels, 1978
Actual = 1,842,647, Increased Export
Scenario. . . . . . . . 98

34.Acres of Soybeans Harvested - Southern
U.S., 1000 Acres, 1978 Actual =
15,560, Increased Exports Scenario. . 99

35.Acres of Soybeans Harvested - U.S.,
1000 Acres, 1978 Actual = 63,033,
Increased Export Scenario. . . . 100

36.Soybeans Processed, Million Bushels,

1978 Actual = 1000, Increased Export
Scenario. . . . . . . . 101

vi

Figure Page

37.Soybean Price, Dollars Per Bushel,
1978 Actual = 6.50, Reduced Exports
Scenario. . . . . . 102

38.Soybean Meal Production, 1000 Tons,
1978 Actual = 23,754, Reduced ;
Exports Scenario. . . . . 103

39.Soybean Oil Production, Million Pounds,
1978 Actual = 10,800, Reduced Exports
Scenario. . . . . . 104

40.Beginning Soybean Stocks, Million
Bushels, 1978 Actual = 161, 1979
Actual = 150, Reduced Exports
Scenario. . . . . . 105

4l.Soybean Production, 1000 Bushels,
1978 Actual = 1,842,647, Reduced
Exports Scenario. . . . . 106

42.Acres of Soybeans Harvested - Southern
U.S., 1000 Acres, 1978 Actual =
15,560, Reduced Exports Scenario. 107

43.Acres of Soybeans Harvested - U.S.,
1000 Acres, 1978 Actual = 63,033,
Reduced Exports Scenario. . . 108

44.Soybeans Processed, Million Bushels,

1978 Actual = 1000, Reduced Exports
Scenario. . . . . . 109

vii

CHAPTER I
INTRODUCTION

Soybeans have come from a position of relative obscurity
in the post-World War II era, with about 10 million acres
harvested, to approximately 65 million harvested acres in
1978. Soybeans are second only to corn in contribution to
the cash income of U.S. farmers. In the world today, soy-
beans are the most dominant of the oilseeds; soybean oil is
the most important of the edible oils; soybean meal is the
leading high-protein feed for livestock. Rapid expansion
in soybean production has been the result of technological
improvements and of an attractive economic position in re-
lation to other crops. Improved corn varieties, herbicides
and pesticides have increased corn yields, making it possible
to satisfy the demand for corn with a smaller proportion of
total cropland than in the immediate post-war period. Feed-
grain prices have been relatively low in most of the post-war
years, and government programs have been aimed at reducing
feed grain production. These conditions have opened the
door for a new crop in the Corn Belt: soybeans. In a like
manner, soybeans have moved on to acreage formerly devoted
to cotton in the South. Cotton production in the South has
become less attractive as the effects of government programs
aimed at reducing cotton acreage and competition from other

1

areas and fibers has been felt. The demand for soybeans has
been growing steadily over this same period. both domestically
and abroad. Soybean prices have remained bouyant throughout
the post-war period. In years where allotments have been
imposed on corn and cotton acreage, soybeans have been a1-
1owed on the feed grain base in all but 1971. A combination
of increased world demand for soybeans and soybean products,
acreage controls on competing crops in the U.S., and the
variability of foreign supply have all acted together to
maintain favorable soybean prices while production has

steadily grown in the United States.

The Problem

 

The wide fluctuations of both soybean and soybean pro-
ducts' prices that has been observed in recent years has
stimulated general interest in the ability of the agricul-
turalist to foresee changes in market conditions and the
impact that these changes will have, both in the United States
and abroad. The intent of this study is to examine the
economic forces which interact in the domestic market for
U.S. soybeans and their products in order to establish a
framework for analyzing the impact of actual and expected
market conditions. 'The need for such a framework has long

been recognized",1 and several other studies are devoted to

 

lDavid Kenyon and R.S. Evans (1975) "Short-Term Soybean
Acreage Projection Model Including Price and Policy Impacts"
Virginia Polytechnic Institute and State University Research
Bulletin (106):4.

. 2,3
this same end.

The Objectives

 

The main objective of this research is to provide a
framework for examining the interaction of the domestic
soybean market with the purpose of forecasting the impact of
expected market conditions. To accomplish this, an economic
model is constructed based upon the theories of the firm and
of the consumer as they apply to the production and consump-
tion of soybeans and their products.

The ultimate goal of this effort is to provide a de-
scription of a method of analyses of the impact of varying
sets of market expectations. It is hoped that the results
of this research will provide a practical technique for
interacting the analyst's expectations as to the level of
exports, income, yields and other relevant market information,
in order to provide an estimate of the resulting conditions

in the U.S. soybean industry.

 

21bid.

3James P. Houck, Mary E. Ryan, Abraham Subotnik (1972)
Soybeans and Their Products, University of Minneapolis Press.

 

CHAPTER II

THE UNITED STATES SOYBEAN INDUSTRY

The United States soybean industry is a complex entity.
As the complexity of a subject grows, so does the number of
interpretations. It is useful to review the vantage point
from which the analyst views the subject in order to better
understand how a particular interpretation has evolved.
The following is a discussion of the aspects of the soybean
industry that are considered important to this particular
study. It is by no means an exhaustive history, but rather

a sketchy outline.

Growth of the United States Soybean Industry

 

Soybean production has grown steadily over the past few
decades. Immediately after World War II, output had risen
from 78 million bushels in 1940 to 200 million bushels an-
nually in 1946 (spurred by a cutoff of foreign edible oil
sources). Things did not return to normal after the war;
U.S. soybean production continued to climb so that by 1978,
1,792 million bushels of soybeans were harvested. The value
of soybeans has grown in relation to other crops. It is now
second only to corn as a cash crop, with a farm value of over
$11 billion in 1978. Soybeans are the most important source

of edible oil and animal high protein feed, both nationally

and worldwide. The United States has become the most im-
portant producer and exporter of soybeans and soybean pro-
ducts, making the soybean a real American success story.

Soybeans have been raised in the United States for
decades (mostly in the Atlantic States), but the first real
boost towards their present prominence was given by World
War II. The cutoff of foreign oil sources by the Japanese
led to the introduction of soybeans to new acreage, notably
in the Corn Belt. Since the soybean is ideally suited to
the same soil and climatic conditions that favor corn and
other feed grains,4 their introduction to the Corn Belt was
opportune. Advances in corn technology brought great in-
creases in corn yields, and with greater yields came govern-
ment policies designed to reduce corn acreage and maintain
corn prices. By allowing the demand for feed grains to be
. filled by smaller areas of cropland, the advances in corn
technology opened acreage to a new crop, soybeans.

The demand for soybeans is derived from the demand of
the two joint soybean products: soybean oil and soybean
meal. Both of these products have enjoyed a continuing
growth in demand over the past three decades. Technological
changes in the soybean processing industry have improved ex-
traction efficiency. The reduced costs that have resulted
from this new technology have enhanced the competitive

position of soybean products and fueled the demand for them.

 

4James P. Houck, Mary E. Ryan, Abraham Subotnik (1972)
Soybeans and Their Products, University of Minneapolis Press.

 

The growth in demand for soybean products and the
attractiveness of soybeans as a crop alternative provides a
framework for further study. In analyzing the domestic
markets for soybean oil and soybean meal, the sources of the
demand that have encouraged the growth of this soybean in-
dustry will become clear.

Soybean meal and soybean oil account for most of the
utilization of soybeans. The relative contribution made by
each of these products to total value of a soybean crop has
changed dramatically over time. In the early 19505, the
value of the oil produced from a bushel of soybeans was about
equal to the value of the meal produced from the same bushel.
By 1978, almost two-thirds of the value of soybean production

came from the meal portion of soybeans.

Soybean Meal

 

Soybean meal is used primarily as a high-protein sup-
plement in livestock and poultry feeds. As disposable in-
come levels have risen in the United States, so has the de-
mand for animal products. Livestock production has become
increasingly concentrated as the efficient commercial pro-
ducer has come into prominence. This new breed of operator
has come to rely increasingly upon formula feeds and custom
mixes that incorporate high protein supplements as a major
constituent. As the use of prepared feeds continues to
expand, the demand for high-protein sources will grow pro-

portionally, with soybean meal being the most attractive of

the alternative protein supplements.

Soybeans, cottonseed, peanuts, sunflowerseed, rapeseed,
linseed, copra and palm kernels are all sources of oilseed
meal. Among these substitute sources of high-protein meal,
soybean meal is one of the highest in crude protein content
by weight. Soybeans have a higher percentage of meal to oil
than the other oilseeds. This makes soybeans an attractive
source of oilseed meal when the demand for edible oil is not
as strong as the demand for oilseed meal. The most important
competitive source of oilseed meal is cottonseed meal. Cot-
tonseed meal is a byproduct of the cotton industry and as
such, has never contributed more than 8 percent to the total
value of United States' cotton production. It is doubtful
that cotton production will ever be responsive to changing
oilseed market conditions.

There are several other sources of high protein avail-
able to the feed industry. Fish meal is one which is very
high in protein, but has the disadvantage of imparting a
"fishy taste" to the resulting animal product (notably,
milk). Synthetic urea is another important source of high—
protein feed. It has not gained a large share of the market
to date and, with current unfavorable natural gas market con—
ditions, further expansion of this source appears to be
questionable.

The status of soybean meal as the most important source
of high protein for livestock does not seem threatened at

present. The horizontal integration of the prepared feed

and soybean processing industries only serves to demonstrate
the widespread expectation of the continued importance of

soybean meal.

Soybean Oil

 

Soybean oil is produced at the same time that soybean
meal is made, but the markets for these two soybean products
are unconnected. Soybean oil competes with animal and vege-
table fats and oils, especially those fats and oils used in
food production. Edible fats and oils are used in processed
foods, cooking oils, salad oils, shortening and margarines.
Peanut oil, cottonseed oil, corn oil, and to a lesser extent,
palm oil, are some of the substitute oils that compete
directly with soybean oil as ingredients in products which
incorporate edible fats and oils; butter and lard substitute
for finished edible oil products in consumption.

Soybean oil has come to dominate the market for fats
and oils in the United States. There has been a 15 percent
increase in the per capita consumption of all fats and oils
products since 1963. During this same period the per capita
consumption of soybean oil has increased by 40 percent. Most
of this growth in soybean oil consumption has occurred be-
cause<xfchanging consumer preferences. Consumer tastes have
moved away from animal products, such as butter, towards
vegetable oil products, like margarine.

Products with desirable flavor and improved physical

characteristics have been made possible by new technology.

There is now widespread public acceptance of the health and
cost advantages of vegetable oil products, and as these
vegetable oil products improve, the substitution of vegetable
for animal products involves little loss of consumer satis-
faction.

There exist many substitutes for soybean oil, yet few
appear likely to gain the status that soybean oil enjoys in
the fats and oils market. Cottonseed oil is the byproduct
of a larger industry, and as such, it contributes only a
small proportion of the total value product of the industry.
The contribution of the cottonseed oil to the total value
of the cotton crOp was only 6.5 percent in 1976. Only a
cataclysmic change in the fats and oils market would have any
significant effect on the cotton industry. The supply of
cottonseed oil is exogenous to the fats and oils market. In
a like manner, lard and butter supplies are relatively un-
affected by fats and oils market conditions. Butter was the
source of only 7 percent of the total value of milk produc-
tion in 1976. Peanuts are another important source of oil,
but if current government policy continues, it is doubtful
that peanut oil production will expand much more rapidly
than it has in the past.

The most important source of competition for soybean oil
in the future is likely to be palm oil. Presently palm oil
is used primarily in shortening and processed foods, but it
is becoming a more common ingredient in margarine and cook-

ing oil. Palm oil is extremely competitive with soybean oil:

10

"U.S. imports of duty-free palm oil from Malaysia and
Indonesia are expanding sharply because of its price ad-

vantage over soybean oil and cottonseed oil."S

Direct Utilization of Soybeans

 

Some soybeans are not crushed but are used directly on
the farm as feed or held as seed. These usages combined have

not exceeded 6 percent of total production in the last decade.

Soybean Production: Alternative Crops and Regionality

 

Soybeans are a hardy and versatile legume. They require
less nitrogen fertilizer than most crops (although they may
require more herbicides). Most soybean varieties have a
short growing season and do not require any great capital
outlay for specialized machinery. These characteristics com-
bine to make soybeans an attractive crop alternative in much
of the United States. Much of the crop land formerly devoted
to corn or cotton has now become soybean acreage. The
switch from corn to soybeans has been precipitated by ad-
vances in corn varieties and the potential glut that higher
corn yields have brought. The acreage devoted to cotton

production is particularly sensitive to government policy

 

5George W. Kromer (Aug. 4, 1975). "Trends and Patterns
in Soybean Oil Use For Food And Industrial Products" U.S.
Department of Agriculture Economic Research Bulletin (611),12.

 

11

changes.6 Much of the acreage that was at one time in cot-

ton is currently in soybeans, reflecting both the results of
"the cotton program which has restricted acreage" and "the
increase in soybeans' competitive position relative to cotton."7

In reporting the acreage and production of soybeans
in the United States, the U.S. Department of Agriculture has
identified six major production regions: the‘Corn Belt,
Lakes states, Delta states, Atlantic states, Plains states,
and other states (see Figure 1b.).

The Corn Belt states are the most significant of all
regions with about half of the total U.S. soybean cropland
and production. Small grains and hay acreage have been re-
duced greatly in the past half century, so that corn is now
the dominant competitive crOp in the Corn Belt states.
Typically, a third of the total acreage of this region is in
soybeans, while most of the rest of the Corn Belt cropland
is devoted to corn.

In the Lakes states, the main crop alternatives are
corn and soft red wheat. Typically, 10 percent of national
soybean production occurs in this area, with the bulk of the

soybean cropland being found in areas adjacent to the Corn

Belt region. There has been little growth in Lakes states'

 

6David Kenyon and R.S. Evans (1975). "Short-Term Soybean
Acreage Projection Model Including Price and Policy Impacts"
Virginia Polytechnic Institute and State University Research
Bulletin (106), 6.

7W.A. Boutwell, H.A. Harris, D. Kenyon (1966) "Competition
Between Soybeans and Other CrOps in Major U.S. Regions" USDA,
Economics Research Bulletin (588).

 

12

Figure l. - - Soybean Production Areas - United States.

‘ )A /

 

 

 

 

 

 

 

Location of U.S. Soybean Production.
One dot equals 5,000 acres.
Source: Census of Agriculture, 1964

 

 

 

Production Regions for U.S. Soybeans.
The shaded areas include those
states in “other states” (see text).
Source: U.S. Department of Agriculture.

13

soybean acreage in recent years.

The Plains states contribute about five percent of
total U.S. soybean production on average. Most of the soy-
bean acreage in this area is found along the region border-
ing the Corn Belt and the Lakes states.

The Delta states have eXperienced a continued expansion
in soybean cropland during the last thirty years. Between
the years 1968 and 1978, the rate of increase of soybean
cropland expansion in the Delta states has reached an average
of 14 percent per year. Some of this expansion in soybean
acreage has been on newly Opened cropland, but much of the
land now in soybeans was previously devoted to corn, small
grains or cotton. Cotton acreage has been declining in the
Delta states as a result of government policy, the movement
of cotton production to the West, and the strong competitive
position of soybeans. Soybeans are now the major crOp of the
Delta states, with nearly two-thirds of this region's total
cropland now in soybeans.

The Atlantic states have experienced a steady growth
in soybean acreage. Here the main Options are cotton and
corn. Declines in cropland devoted to cotton and corn have
accompanied the increase in soybean acreage in the Atlantic
states .

The bulk of soybean production that occurs in the other
states region is found in the areas bordering the Corn Belt
region.

The preceeding analysis of the U.S.D.A.-defined regions

l4

identifies the competitive crOps and geographic location

of soybean production in each respective region. .In some
regions the important crop alternative is corn and geographi-
cally, most of this production is clustered around the Corn
Belt. These regions are: the Corn Belt states, the Lakes
states, the Plains states. Other states have crop alter-
natives of more importance than corn; cotton is often found

to be in a competitive position. These regions are found

in the more temperate areas to the South that were identified
as the Delta and Atlantic regions. Similar crop alterna—
tives and geographic locations suggest two major areas of
soybean production in the United States. One of these re-
gions would be in the North, with corn as the main competitive
crop and production centered around the Corn Belt region.

This region will hereafter be referred to as the Northern
region and consists of the states in the Corn Belt, Plains
states, Lakes states and other states. The other production
region, the Southern region, is made up of the Delta and
Atlantic states. This area has corn and cotton as competitive
crOps and located in the more temperate zones of the southern

United States.

Soybean Yields

 

Soybean yields along with the harvested acreage of soy-
beans will determine production for a given year. Soybean
yields in the Northern United States region were 23 bushels

per acre during the post-Korean War period and have risen to

15

an average of 29 bushels per acre in recent years. In the
Southern United States region, the average yield per acre
was about 20 bushels after the Korean War. In recent years
this average has risen to 22 bushels per acre. Enhanced
soybean yields have resulted from the introduction of new
varieties suited to local climatic conditions; improved weed,
insect and disease control; and better management practices.
Much variability in crop yields has resulted from changing
climatic conditions. Evidence of sunspot cycles being as-
sociated with drought cycles, and thereby affecting crop
yields has been demonstrated in other studies. "Many analysts
have suggested a twenty to twenty-two year cyclical pattern

for yields in the Great Plains and Corn Belt."8

It may be
possible to better understand this aspect of soybean yield

variability through the study of these sunspot cycles.
Soybean Market Interaction
Soybean Supply

In the beginning of the crop year, soybean production is
determined by the acres of soybeans harvested and the yield
per acre of soybeans (See Figure 2). A small portion of the
crop is held for seed, or residual uses on the farm, but most

of the production enters the market. As the new crop becomes

 

8J. Roy Black and Stanley R. Thompson (1978) "Some Evi-
dence on Weather-Crop Yield Interaction" American Journal of
Ag. Economics (60).

 

FIGURE 2 .

16

--U.S. Supply and Disposition Of
Soybeans and Soybean Products.

 

 

RETAINED
FOR SEED

 

 

 

 

RETAINED FOR
RESIDUAL USES

 

 

\ /

 

BRODUCTION] -.——— <——-

 

 

I
ISTOCKCARRYINJ ___.. [I'm A 3
I

 

 

 

 

 

 

| STOCK CARRY oufl
’ SUPPORT
PRICE

COMMERCIAL AND /
GOVERNMENT STOCKS

‘ I , SOYBEAN
STORAGE PRICE
DEMAND

 

 

 

/

 

 

U.S. DEMAND
FOR MEAL

 

 

\ I

 

 

 

[ MARKET DEMAND FOR SOYBEANSJ

I
W

[MARKETING AND PROCESSING MARGIq

I

 

 

LVALUE OF BEANS CRUSHEITI

SOYMEAL SOYOIL
PRICE \ PRICE

 

EXPORT DEMAND
FOR MEAL

 

 

 

/
lm m A ‘
NORTH SOUTH
\

 

 

 

 

 
 

SUPPORT
PRICE

 

 

EXPORT BEAN

DEMAND

 

 

 

EXPORT DEMAND
FOR OIL

 

 

 

 

U.S. DEMAND
FOR OIL

 

 

 

17

available, there are stocks Of soybeans remaining that are
also available in the market at a discount (reflecting a
loss Of quality over time). The carry-in stocks and produc-
tion, minus soybeans held for seed and residual uses, de-

termine the soybean supply for that crop year.
Soybean Demand.

There are several sources Of demand for soybeans. Pro-
cessors demand soybeans for crushing in order to satisfy
their markets for soybean Oil and meal. The revenue from
the sales of Oil and meal combine to create the value Of
the beans that are crushed, which will influence the total
quantity Of beans demanded by the processors. Foreign and
domestic markets exist for soybeans; over one-third Of all
U.S. soybean production has gone to foreign markets in re-
cent years. If exports Of domestically produced soybean
Oil and soybean meal are considered, half Of the United
States' soybeans and soybean products have been exported in
the past decade. Speculative demand for soybeans and their
products may influence the stock levels that are held domes-
tically. The level Of stock demand (if any), domestic
crushing demand and foreign demand for soybeans, are the
components Of the total demand for soybeans each crop year.

The price that each farmer receives for his soybeans
is determined by the interaction Of the supply of soybeans
at harvest time with the total market demand conditions.

The price received by farmers for their soybeans will interact

18

with other considerations (i.e. support price levels, pre-
vious planting practices, the market price for alternative
crops, etc.) to determine soybean acreage planted in the
coming season.

The price received by farmers for their soybeans is
determined by the interplay Of the demand for beans with a
predetermined level Of supply. The demand for beans,
whether foreign or domestic, is derived from the demand for
the two soybean products: meal and oil. With about 4 pounds
Of soybean meal yielded per pound of Oil produced, the
ratio Of Oil to meal production is physically determined.
However, the demands for the two soybean products are largely
unrelated. The quantities Of soybean meal and Oil demanded
must approximate to the crushing ratio for the market to
clear. The only way the quantities Of meal and Oil demanded
can be linked to this crushing ratio is through price ad-
justment. In recent years it is interesting to note that
soybean meal prices have been trending upwards while the
price of soybean Oil has Often been weak. The strength Of
demand for soybean meal has Often been strong in relation
to that Of soybean Oil, as attested by the shift in their

relative prices.

In.

ntl

MI
by!

“H
5U

‘O
0-

I
v._

I
O‘V.

am.‘
‘ 5'

V.
c“

CHAPTER III
REVIEW OF THE LITERATURE

In 1958, King9 published a study which focused upon
the relationship Of all high-protein feeds as a group, to
all feed grains. In addition, the relationship Of each
high-protein feed to all other high-protein feeds was
estimated in a series of two-equation relationships. A
demand equation for soybean meal was estimated and compared
to an equation for other high-protein feeds (it is important
to note that during the data period Of this study, soybean
meal was not as dominant in the high-protein feed market as
it is today). In the soybean meal equation the quantity Of
high-protein feeds as a group was dependent upon the price
Of high-protein feeds, the price Of feed grains, the price
Of livestock and associated products, and the number Of
grain consuming animal units. The market share Of soybean
meal was estimated from soybean meal price and other high
protein meal prices in a simultaneous solution with an
equation for other high-protein feeds.

The King study demonstrated: (a) that high-protein feeds

 

9Gordon A. King (1958). "The Demand and Price Structure
for Byproduct Feeds" U.S.D.A. Technical Bulletin (1183).

19

n"

~r

U.

.A

CA
iv

WV

(n
'( t

to

‘AI

.V.

\-

D“

If;—

5“.
It

20

are highly competitive with feed grains, and (b) there exists
qualitative differences in the various high—protein feed
groups which is reflected in a degree Of independence Of
demand for each group.

In another study conducted by the U.S.D.A.,lo soybean
meal price was estimated in varying specifications which in-
cluded the following variables: the total supply Of soybean
meal, the quantity Of soybean meal fed to livestock, the
total quantity Of other high protein meals produced, the
quantity Of other high protein meals fed, high-protein con-
suming animal units, the average prices for livestock and
livestock prices and the production Of formula feeds. The
specification that was finally arrived upon had soybean meal
price as a function Of the total supply Of soybean meal, the
average price received for livestock, and the production of
formula feeds.

Attempting to develop a predictive equation for soybean
meal price led Hieronymus to try a large number Of alter-
native specifications for soybean meal demand.11 The equa-
tions that were deemed best included soybean meal supply
(either as total production or as production minus exports),
high-protein consuming animal units and livestock prices.

The construction Of the high-protein consuming animal

 

loU.S.D.A. (1959) "Factors Influencing Soybean Meal
Prices" Feed Situation (July).

llT.A. Hieronymus (1961) "Forecasting Soybean Meal
Futures Prices" Commodity Yearbook, Commodity Research
Bureau Inc.

 

21

units series is of particular interest. In contrast to the
U.S.D.A. data series, which computes average consumption
rates based on the utilization of all high-protein feeds,
the rates used in this study were specific to soybean meal.
The data series is then adjusted for trends towards greater
soybean meal consumption over the period considered.

James P. Houck has been the pioneer in the study of the
entire soybean industry with all its complexity. In 1963,
Houck developed an eight-equation model to simulate the
market demand relationships for soybeans and their products.12
Houck envisioned the price determination of both soybeans
and soybean products as a simultaneous interrelated process.

The process of market interaction was simulated in the
Houck study by an eight-equation model representing the
various market demands and price links that exist in the soy-
bean industry. Five of the eight equations are viewed as
stochastic: (a) soybean meal demand; (b) soybean oil demand;
(c) crushing and handling margin; (d) export demand for
soybeans; and (e) storage demand for soybeans. The other
three equations were definitional for: the price of soybeans
received by farmers, the value of the two soybean products --
meal and oil.

The stochastic parameters were estimated via two-stage
least squares and ordinary least squares regression procedure

in the reduced form. Both methods yielded statistically

 

12James P. Houck (1963) "Demand and Price Analysis of the
U.S. Soybean Market" University of Minnesota Ag. Experimental
Sta. Technical Bulletin (244).

22

significant estimates of a similar magnitude. It was the
first successful treatment of the soybean industry in an
integrated form.

In the following years Houck updated his model to in-
clude estimates of soybean production in the United States

13 The

and export demand for soybeans and soybean products.
supply of soybeans was treated as the sum of the production
Of six regions in the U.S. In a like manner, the export
demands for soybean meal, soybean oil and soybeans are
estimated for each of either five or six regions and then
summed to determine total export demand. The market demand
model was respecified from the formulation originally put
forth,14 and while the crushing and handling margin was no
longer estimated or defined, individual equations for soybean
meal and soybean oil exports were estimated. Another stochas-
tic equation was introduced for the stocks of soybean Oil.
Meal and oil production, soybean prices, market clearing
equations for soybeans, soybean meal and soybean oil are all
treated as identities.

As an alternative specification, total world exports
of soybeans, meal and oil were made in three equations, with
highly satisfactory results. All parameters were estimated

by three alternative techniques: (a) ordinary least squares;

(b) two-stage least squares, and (c) three-stage least squares.

 

13James P. Houck, Mary E. Ryan, Abraham Subotnik, Op. cit.

14James P. Houck, op. cit.

23

"Of particular interest is the similarity between the LS,
ZSLS and BSLS estimates. The parameter estimates were very
stable, with little change seen when an alternative estima-
tion procedure was employed."15
Houck's treatment of the soybean industry has been
fundamental and has served as the foundation for much of the
predictive work conducted by the U.S.D.A. in following years.16
A recent study conducted by Kenyon and Evans incorporates an
effective soybean price (which considers both prices re—
ceived by farmers, the set-aside program and support price
levels) in predicting the harvested acreage in the six pro-
duction areas treated in Houck‘s most current study.17
Other economic investigations of the soybean industry
have utilized spatial equilibrium analysis18 and Operations

research.1'9

 

15James P. Houck, Mary E. Ryan, Abraham Subotnik, op. cit.

16Fats and Oils Situation, November, 1971, and July 1973,

17David E. Kenyon, R.S. Davis, op. cit.

18H. Nakamua, T.A. Hieronymus and G.C. Judge (1963).
Interregional Analyses of the Soybean Sector, University of
Illinois, Dept. of Ag. Econ. AERR-67,

19Earl C. Hedlund (1952) Transportation Economics of the
Soybean Processing Industry. University of Illinois Press.

 

CHAPTER IV

THE MODEL

 

The model of the United States domestic soybean industry
presented in this study is a simplified one, both from
economic and statistical viewpoints. Economic relation-
ships are formulated by using a generalized version of the
market for soybeans and soybean products in the United States
(see Figure II,p.16) and a number of simplifying assump-
tions. These simplifying assumptions are necessary in order
to reduce the analytical complexity, inherent in the inter-
related markets for soybeans, soybean meal and soybean oil,
to a manageable level. Variable selection for each of the
economic relations, represented as a stochastic equation,
is based upon a priori considerations from economic theory
and upon the results of previous investigations.

«The structure of the model is presented in terms of
the economic relationships between the variables. These
relationships are then explained, along with the rationale
for each respective structural equation. Beginning with a
presentation of the variables that are employed, followed
with a discussion of model structure, a description of the
economic model is developed. A discussion of the under-
lying assumptions and model rationale complete the discussion

of the economic model.

24

.__._ __._1__ ._ _ _

25

THE MODEL - DESCRIPTION

 

The following model of the United States soybean industry
is treated as a two-part system. The domestic supply of
soybeans is recursively determined: it is assumed that all
major production decisions are made on the basis of infor-
mation available before harvest. The domestic demand for
soybeans is treated as a simultaneous system, where the
derived demands for the joint soybean products -- meal and
oil -- determines the level of crushing demand. In this

simultaneous system, the current pggipdwpgigegvand‘quantities

w—

 

of soybeans and their products interact with each other and

~‘ ‘W‘wm 2M1 ‘TII. ‘mWAW

 

with other market conditions to determine the level at which

wrwhv-‘qu'ar-fw. ow W‘WW 0-1“ («N ”W

thaefistrfl} 3.135135;

The statistical technique employed to determine estimates
of the parameters for both the supply and demand components
Of the model was the ordinary least squares procedure. Simul—
taneous estimation procedures were not employed in the
estimation of the parameters for the demand equations' vari-
ables. Houck used several simultaneous estimation tech-
niques in his study of the soybean industry and noted, "The
similarity between the LS, ZSLS and 3SLS estimates" with "the
magnitude of the coefficients" being "generally similar form
method to method."20 In light of these results, the use of

ordinary least squares estimation procedures seems adequate.

 

20James P. Houck, Mary E. Ryan, Abraham Subotnik, op. cit.

26

Variables Employed

 

The model makes use of 19 endogenous, ll predetermined

and 12 exogenous variables. These variables were either re-

ported on a market year basis (year beginning October 1) or

converted from a calendar to a market year via adjustment

weights. Mnenomic variable names are used for ease in inter-

pretation of the total model structure. The sample period

is from 1954 to 1977.

Endogenous Variables

The following are treated as endogenous variables in the

model:
AHSBTS=

(1a)

AHSBTN=

(la)

APSBTS=
(la)
APSBTN=
(la)
SOYBYTS=
(la)
SOYBYTN=

(1a)

 

acres of soybeans harvested in southern U.S.
region, thousand acres, southern U.S. includes:
Ark., Miss., Ala., N. Car., S. Car., Virg.,
Mar., Del.

acres of soybeans harvested in northern U.S.
region, thousand acres, northern U.S. includes
all production areas not found in the southern
U.S. region.

acres of soybeans planted in the southern U.S.,
thousand acres.

acres of soybeans planted in the northern U.S.,
thousand acres.

average soybean yield in the southern U.S.,
bushels per acre.

average yield in the northern U.S, bushels per

acre.

SOILPT=

(1d)

SOILSTKT1=

(loo

ISBSEDT=
(lb)

SBRIST=

(lb)

OILSPREAD=

(1f)
CRUSHT=
(lb)
SOYBQT=
(1a.)
SBSTKT1=
(1b.)
SOYBPT=
(1a.)
SOYMQTD=
(lc.)
SOYMPT=

(1d.)

TOTOIL=
(le.)
OTHEROIL=

(le.)

27

average price of domestic crude soybean oil,
cents per pound.

ending stocks of soybean oil (year ending
September 30), million pounds.

soybeans used for seed, million bushels.

residual and feed usage of soybeans, million
bushels.

farm-retail price spread of soybean oil,
cents per pound.

soybeans processed in United States, million
bushels.

Soybeanmproduction in the U.S., thousand
bushels.

ending stocks of soybeans (year ending Sep-
tember 30), million bushels.

average soybean price received by farmers,
dollars per bushel.

quantity of soybean meal demanded domesti-
cally, excluding exports, thousand tons. ‘
average price of soybean meal, bulk Decatur,
at 44 per cent protein, dollars per hundred-
weight.

total U.S. consumption of all oils and fats,
million pounds.

total U.S. consumption of all oils, excluding

soybeanoil, million pounds.

SOILQTD=

(1c.)

SOYBEXP=
(1c.)
I somor=
(lcrl
SMEXT=
(1c.)
SOILEXT=
(1c.)
SOYBEXPT=

(1a.)

28

quantity Of soybean oil consumed in the U.S.,

excludes exports and changes in stocks,
illion pounds.

quantity of soybeans exported during the

market year, million bushels.

soybean meal produced in the U.S., thousand
tons.

Soybean meal exports from U.S., thousand tons.

soybean oil exports from U.S., million pounds.

U.S. export price of soybeans, dollars per

bushel.

Predetermined Variables

The predetermined variables in the model consist of all

those variables which have either lagged values or values

that would be known at the beginning of a marketing year.

These are:
DSOYBPLI=
(1a.)
DCORNPLI=
(19.)
DWHTPLI=
.(lh.)
DCOTPLI=

(11.)

deflated average soybean price received by
farmers, lagged one year, dollars per bushel.
deflated average corn price received by
farmers, lagged one year, dollars per bushel.
deflated average wheat price received by
farmers, lagged one year, dollars per bushel.
deflated average cotton price received by

farmers, lagged one year, cents per pound.

29

DPVISBT= deflated effective support rate for soybeans,
dollars per bushel.
LAPSBTNLI= log Of acres planted soybeans, northern U.S.,
(1a.) lagged one year, thousand acres.

'LAPSBTSLI= log of acres planted soybeans, southern U.S.,

(1a.) lagged one year, thousand acres.
TIME= linear trend, 1954=54...1980=80,
TSQR= curvilinear trend, l954=2916(542)...l980=
6400(802).

DUMWEATHER=dummy variable with value of one for years
with predicted drought cycles, (1952-58,
1974-80).21

Exogenous Variables

These variables are also predetermined in the sense that
they are not generated internally in the model, yet these
are variables which would not be known before a market year
began. These are:

CPI= index of prices paid by consumer, U.S. Bureau

of Labor statistics.

 

INTT: short term prime interest rate, bankers ac-
(2) ceptances, 90 day notes, 1967:100.

CORNPT= average price of corn received by farmers,
(lg) dollars per bushel.

21

J. Roy Black and Stanley R. Thompson. (1976), op. cit.

LPBEEF=

HPCAUT=

CONSUMRAT=
(lj)
EDIBLEI=
(1k)
POPT=
(3)
DPCIT=
(4)

RAINS=

GMSOYMEAL=

30

log of average price received by farmers for
various categories of beef cattle, weighted
by various feed rates, adjusted to market
year, dollars per hundred weight. The log
function is used to simulate the decreasing
rate of adjustment to price that has occurred
over time.22
high-protein consuming animal units, U.S.D.A.
weights applied to major livestock production
categories, adjusted to market year.

ratio of per capita margarine consumption to

per capita lard and butter consumption, pounds.

index of edible oil price, 1967:100.

resident population of the U.S., July 1,
millions persons.

disposable per capita income, dollars per
person.

deviation from average rainfull in Lake
Charles, La., inches.

the gross margin for soybean meal, calculated
from annual livestock data, soybean meal price

and adjusted for market year, cents per pound.

Notes Qn_Some of the Variables

The sources of the data correspond to a numbered code

 

22

T.A. Hieronymus (1961) Op. cit.

31

svhich appears directly below each variable name. These are
listed in Appendix It. The following will help clarify fur-
'ther questions.

1. The prefix DEF- refers to any price that has been
deflated by the consumer price index, CPI. In the case of
A calendar year data that has been adjusted to market year,
the same adjustment has been made to the CPI.

2. The suffix -Ll refers to a lag of one period.

3. The prefix CHG- refers to the change in a quantity.

4. Beef prices and numbers were generated from an
allocation program which attempts to distinguish fed-beef
from non-fed beef using published data series. A suitable
substitute for PBEEF and HPCAUT can be obtained from data
series published by the U.S.D.A. in "Livestock Statistics".23

5. "The effective support rate is equal to the loan
rate discounted by the factor by which set asides impose
upon program participation plus deficiency payments dis-
counted by the national program allocation factor."24

A 6. Export quantities of soybeans, soybean meal and

soybean oil are not determined within this model. This
study has been a collaboration with another study of export

markets for U.S. soybeans and soybean products.25

 

23U.S.D.A. (1976) "Livestock Statistics" Statistical
Bulletin (531).

24Eric Wailes (1979) "The M.S.U. Agriculture Model -
Domestic Policy Component", Michigan State University Depart-
ment of Agricultural Economics, unpublished monograph.

25William Tierney (1979), Michigan State University De-
partment of Agricultural Economics, unpublished research.

 

32

The structure of the model is presented below, with
the causal relationships being identified as estimated
equations and the identities as definitional equations.
Mnenomic names identical to those previously identified are
used.. The endogenous variables in each structural equation
are presented first followed by a colon and then all pre-

determined or exogenous variables follow.

Demand Component

A. Estimated Equations

1. Average price received for soybeans, SOYBPT, OIL-
SPREAD, SOYMPT, SOILPT.

2. Quantity of Soybean Meal Consumed in U.S., SOYMQTD,
SOYMPT: CORNPT, LPBEEF, HPCAUT.

3. Total Oil Consumption in the U.S., TOTOIL: CONSUMRAT,
EDIBLEI, POPT.

4. Quantity of Soybean Oil Consumed in U.S., SOILQTD,
DEFSOILPT, OTHEROIL: DEFDPCIT.

5. Soybean Seed Demand in U.S., SBSEDT, APSBT: DFSOYBPLI.

6. Residual and Feed Use of Soybeans; U.S., SBRIST:
CORNPT, GMSOYMEAL, HPCAUT.

7. Residual Soybean Stocks in U.S., SBSTKTl, SOYBQTD:
SOYBEXP, SBSTKT.

8. Speculative Soybean Stocks in the U.S., SBSTKTl,
SOYBPT: SOYBEXPT, INTT, SBSTKT.

9. Residual Soybean Oil Stocks in the U.S., SOILSTKTl,

SOILQTD, SOYMQTD: SOILEXT, SOILSTKT.

33

10. Speculative Soybean 011 Stocks in the U.S.,

SOILSTKTl, SOILPT, SOYMQTD: INTT, SOILSTKT.

B. Definitional Equations

1. Soybeans Processed in the U.S., CRUSHT=,02(SOYMQTD)
+.05(SOILQTD).

2. Soybean Meal Produced in the U.S., SOYMQT=SOYMQTD+
SMEXT.

3. 011 Other Than Soybean Oil Consumed in the U.S.,
OTHEROIL=TOTOIL-SOILQTD. 3

4. Soybean Oil Produced in the U.S., SOILQT=SOILQTD+

SOILEXT+CHGSOILSTK.

Supply Component

 

A. Estimated equations

1. Acres of Soybeans Planted in Northern U.S., APSBTN:
DSOYBPLI, DCORNPLl, DWHTPLl, LAPSBTNLI.

2. Acres of Soybeans Planted in Southern U.S., APSBTS:
DSOYBPLI, DCOTPLl, DCORNPLl, DPVlSBT, LAPSBTSLI.

3. Average Soybean Yield in Northern U.S. , SOYBYTN:
TIME, TSQR, DUMWEATHER, DEFSOYBPT.

4. Average Yield; Southern U.S., SOYBYTS: TIME, TSQR,

RAINS, DEFSOYBPT.

Assumptions

 

The structural equations presented above make a certain
number of simplifying assumptions. The soybean industry is

treated as a competitive industry which implies that certain

34

conditions hold true.

1. Soybeans and soybean products are treated as homo-
geneous commodities. In reality soybeans vary in grade, and
the price received for different grades reflects its relative
quality.’ Soybean meal varies in quality, with some grades
having a smaller proportion of fiber to protein than do other
grades. Soybean oil is fairly uniform, with only minimum
quality requirements. The fact that the prices used for soy—
beans and soybean meal are average prices helps to adjust
for varying qualities, but the quantities reported of both
soybeans and soybean oil do not reflect grade differences.
The inaccuracy introduced by product heterogeneity could
conceivably become the source of unexplained variation.

2. In order to have a competitive market for soybeans
and soybean products the number of buyers and sellers must
be large enough so that no one individual has the market
,power to influence price. The large processors which in-
creasingly characterize the soybean industry fail to meet
this criterion. Any collusive activity on the part of pro-
cessors could violate the assumptions underlying the model
and introduce more “noise" to the system.

3. It is assumed that the processing of soybeans yields
meal and oil in a fixed proportion. This assumption seems
reasonable since there has been less than one percent varia-
tion of this yield ratio (about 4 to l) in the past twenty
years.

4. The trading of soybeans and their products is assumed

35

to occur uniformly throughout the marketing year. This is

not what actually occurs, so that there may be sOme inaccu-
racy in the adjustment of calendar year data to an October 1
marketing year basis when this adjustment is made by weighting
each month equally.

5. Soybean meal stocks barely exist, but soybeans store
fairly well for up to a year and soybean oil will retain
quality for long periods of time. This creates an opportunity
for speculative holdings, especially of soybean oil. To
test whether or not there is empirical evidence to support
the hypothesis that there is significant speculative storage
of soybeans and soybean Oil, the structural equations for
stocks of each commodity have two alternative specifications.
One specification is definitional in that it treats soybean
Oil and soybean stocks as the residual of the difference
between supply and demand. The alternative specification
has these stocks as a function of prices, both foreign and
domestic. The F-statistics will then be tested for signi-
ficance, and compared to one another to determine whether
a specification which treats stocks as speculative is prefer-

able to a residual treatment of stocks.

Model Rationale

 

Production of soybeans in the United States involves
both decisions on the part of the farm manager and influences
beyond his control, such as weather and improved technology.
In order to reflect this, the production of soybeans is

estimated in a two-step manner with the acres planted by

36

farmers as a function of economic criteria and the yield
realized on this acreage as a function of weather and trend
(trend variables represent various technological improve-
ments). The real price of soybeans at harvest is the economic
criteria that is used by the farm manager to decide how

much of his acreage is worthy of harvest. This decision will
in turn influence the average yield realized. On this basis,
the deflated price Of soybeans is included in the soybean ,
yield equations. Acres harvested of soybeans is treated as

a constant function of acres of soybeans planted. An alter-
native specification would be to estimate acres harvested

of soybeans as a function of deflated soybean price and acres
planted.

The production of soybeans is regionalized into the
Northern and Southern regions on the basis of similar crop
alternatives and weather within each region. The deflated
prices of the major crop alternatives and the soybean policy
variable are included in the acres planted equations for
each of the two regions. A weather variable is included in
the yield equation for the Northern U.S., whereas observed
average rainfall is used in the Southern U.S. for the lack of
an appropriate weather variable to account for climatic in-
fluences. Once acres planted are estimated for each region,
acres harvested are defined. Acres harvested are then mul-
tiplied by the yield estimates to provide an estimate of soy—
bean production for each region.

The demand component of the model estimates the domestic

37

demand for soybean oil and soybean meal and then defines the
quantity of soybeans demanded for processing from the quanti-
ties of each of these two joint products that is demanded.
The price received by farmers for soybeans is treated as a
function of the price of soybean meal, soybean oil and of the
farm-retail price spread for soybean oil (intended as a
measure of the average marketing margin realized by pro-
cessors). The quantity of total demand for soybeans can be
defined from the internal estimates of production, soybeans
processed, seed, feed and residual utilization, stocks and
estimates of soybean exports Obtained from an export market

study.26

In a like manner total quantities of soybean oil
and meal demanded can be defined from export, stocks and
production data.

At the heart of this demand component are the demand
equations for soybean meal and soybean oil. Soybean oil
accounted for 43 percent of total fats and Oils consumption
in 1977. Other fats and Oils are very competitive with soy-
bean oil and are good substitutes in most cases. The market
share of the total demand for fats and oils that soybean oil
will realize in any given year is expressed as a function of
the market share of competing Oils, deflated income and
deflated soybean Oil price. Total oil demand is also esti-
mated; based upon trends in consumption (CONSUMRAT), an index

of price levels for all edible Oils (EDIBLEI) and population

(POPT). When the system of demand equations is simultaneously

 

26Such as: William Tierney (1979) op. cit.

38

solved the relative market shares of soybean Oil and com-
peting oils are estimated, with the competing oils' market
share being defined as the difference between total and soy-
bean oil quantities demanded.

The position of soybean meal in the domestic market is
much more dominant than that of soybean oil in the fats and
oils market. Soybean meal accounts for about 90 percent of
current Oilseed meal consumption in the U.S. Most of the
supplies of competitive high protein feeds are constrained,
as was discussed above. It seems appropriate to estimate
soybean meal consumption directly since it is, in essence,
the total oilseed meal market. The major feed substitute for
soybean meal is corn, the dominant feedgrain. Typical ad-
justments in feeding rations involve a substitution of corn
for soybean meal or visa-versa, at least within their range
of substitutibility. The price of corn is included along with
the price of soybean meal to reflect this product substituti-
bility. The quantity of soybean meal fed will also vary
directly with the population of animals on feed. Since this
population is heterogeneous, each group is adjusted with
constants reflecting different rates of soybean meal utili-
zation for each livestock category. The adjusted populations
for each category are then summed to yield a measure of the
total population Of high-protein consuming animal units.

A change in the price received for the finished live-
stock product will influence the amount of soybean meal that

will be fed. Yet how this adjustment is made is an important

39

consideration. In short-cycle production processes (e.g.
pork, poultry) the number of animals will be changed in
response to the change in product price. An adjustment of
this sort would be reflected as a change in the high—protein
consuming animal units number. In longer production cycles,
liquidation is not as feasible of an alternative, so changes
in feeding rations will occur. For this reason, the price
of beef was included as an explanatory variable in the soy-
bean meal demand equation. The price of beef was entered
into the soybean meal demand equation as a natural log
function in order to reflect a growing inflexibility in
feeding rate adjustments to price. This price inelasticity
has been observed to be growing over the estimation period,
possibly because of the increased use of pre-mixed and
custom-mixed feeds in beef production.27
The stocks of soybeans and soybean oil are specified
in two alternative forms in order to test the hypothesis
,that demand for stocks may be speculative rather than
residual in nature. If the results indicate that the demand
for either soybeans or soybean oil is not speculative, that

quantity will be treated as a residual in the final model.

 

27T.A. Hieronymus (1961) op. cit.

CHAPTER V
RESULTS OF ESTIMATION

The estimated equations are presented below, with the
parameter estimates preceeding each mnenomic variable name.
The t-statistic estimates are not reported here, but may be
calculated from the standard error estimates which appear
below each variable name in parenthesis. The corrected coef-
ficient- of determination (R2) appears below each equation.
Also the Durbin-Watson statistic (d) associated with each
equation appears as a test for serial correlation. The
Durbin-Watson statistic (d) is not reported in those equa—
tions where a geometrically-lagged dependent variable is
specified. Instead, Durbin's h statistic is used to test
for autocorrelation. This statistic (h) is distributed in

a standard normal distribution.

Supply Equations

 

1. Acres Planted: Soybeans, northern U.S.

 

APSBTN= -206437. + 7690.25 DSOYBPLI -12229.4
(22390.5) (1044.73) (3672.6)
DCORNPLI - 1470.90 DWHTPLI + 22750.2

(1640.3) (2329.7)
LAPSBTNLI

R 2=.96 Durbin’s h statistic = -.00006

41

2. Acres Planted: Soybeans, southern U.S.
APSBTS= -72838.5 + 1378.40 DSOYBPLl ~33.87
(12651.3) (600.36) (39.34)
DCOTPLl - 1119.11 DCORNPLl + 1092.81
(1285.76) (856.39)
DPVISBT + 8666.23 LAPSBTSLI
(1217.31)
R 2=.94 Durbin's h statistic =.000005
3. Soybean Yield, southern U.S.
SOYBYTS= -57.79 + 2.33 TIME - .016TSQR
(35.72) (1.09) (.008)
+.l44 RAINS - 1.42 DEFSOYBPT
(.05) (.69)
R 2=.61 d=l.33
4. Soybean Yield, northern U.S.
SOYBYTN= 136.63 - 3.67 TIME + .032 TSQR
(58.43) (1.77) (.014)
-2.11 DUMWEATHER - 2.31 DEFSOYBPT
(1.15) (.59)

R 2=.83 d=2.22

Demand Equations:

 

l. Soybean Meal, Quantity Demanded
SOYMQTD= -27108.30 - 344.52 SOYMPT
(2525.89) (117.80)
+474.22CORNPT + 4664.00 LPBEEF + .355HPCAUT

(331.05) (1363.94) (.045)

42

R 2=.94 d=l.82

2. Total Oil, Quantity Demanded
TOTOIL=2515.37 + 1864.01 CONSUMRAT - 11.03 EDIBLEI
(8666.43) (1416.47) (4.92)
+86.08 POPT
(49.63)
R'2=.87 d=l.63
3. Soybean Oil, Quantity Demanded
SOILQTD = -151.91 + 4.04 DEFDPCIT

(1657.66) (.298)

-45.77 DEFSOILPT - .53 OTHEROIL
(19.78) (.12)
R’2=.96 a=1.7o

4. Soybean Price, Received by Farmers
SOYBPT= -.92 + .05 OILSPREAD + .13 SOYMPT
(.33 (.014) (.06)
+.l4 SOILPT
(.016)
R'2=.96 d=2.84
5. Soybeans Used for Seed
SBSEDT=7.62 + .0011 APSBT-2.66 DSOYBPT
(2.82) (.00007) (2.36)
fi'2=.92 d=2.44
6. Soybeans Fed and Residual Usage
SBRIST=52.49 + 6.99 CORNPT -.28
(23.12) (4.88) (.37)
GMSOYMEAL + .0007 HPCAUT

(.0003)

43

R‘2=.38 d=1.79

7. A) Soybean Ending Stocks, Speculative Demand
SBSTKT1=13.66 - 245.77 SOYBPT + 214.67 SOYBEXPT
(71.89) (213.71) (189.97)
+214.67 INTT + .25 SBSTKT
(146.03) (.27)
E 2:.40 F(4,10)=2.00 Durbin's h statistic=.03
7. B) Soybeans Ending Stocks, Residual Formulation
SBSTKT1= -.26l.54 + .0008 SOYBQTD
(33.27) (.00007)
--l.54 SOYBEXP + .78 SBSTKT
(.15) (.08)

th

.93 F(3,11)=59.91 Durbin's h statistic =.61
8. A) Soybean Oil Ending Stocks, Speculative Demand
SOILSTKT1= -.364.58 - 2.20 SOILPT - 11.67 INTT

(337.58) (9.57) (38.91)

+.087 SOYMQTD + .045 SOILSTKT

(.028) (.24)

R 2:.40 F(4,10)=3.36 Durbin's Lh statistic =o.44

8. B) Soybean Oil Ending Stocks, Residual Formulation
SOILSTKTI = r914.40 - .34 SOILQTD -.45 SOILEXT

(315.31) (.14) (.16)

+.32 SOYMQTD + .46 SOILSTKT

(.09) (.21)

R 2=.67 F(4,10)=8.19 Durbin's h statistic =l.7

44

THE RESULTS IN GENERAL

 

The results are generally satisfactory with the ex-
pected coefficient signs and acceptable levels of signifi—
cance for most parameter estimates. Most of the estimated
equations explain an acceptable level of the variation in

the observed dependent variable.

The Results: Supply7Component

 

Acres Planted: Soybeans, northern U.S.

In light of the results of estimation, the parameter
estimates for this equation seem reasonable. The own-price
elasticity estimate for soybeans is in line with the results

of other studies (see Table 1 ). 28'29

Cross-price
elasticity estimates for corn and wheat are also reasonable
and correspond to those cited in Table' 1.. There is a

high level Of variation associated with the parameter estimate
for the price of wheat. Wheat (especially soft-red wheat) is
of some importance as a crOp alternative in the Corn Belt
states, but it is even more important in the Lakes States,
where 10 percent of national soybean production typically
occurs. While the effect of wheat price is not a powerful
one, it is reasonable in magnitude. It's inclusion in this

equation is desirable, despite the low level of significance.

associated with wheat price's parameter estimate.

 

28James P. Houck, Mary E. Ryan, Abraham Subotnik (1972)
op. cit.

29David E. Kenyon, R.S. Evans (1975) op. cit.

45

Durbin's h statistic30 is not significant at the 95 percent
confidence level; the test for serial correlation is in-

significant.

Acres Planted: Soybeans, southern U.S.

All of the parameter estimates vary with the dependent
variable in a manner which corresponds to the expectations
one would develOp from economic theory. The variation of
the parameter estimates associated with corn and cotton
price are both rather high. When compared to the cross—
price elasticity estimates generated in other studies (see
Table 1 p547, the cross-price elasticity estimate for cotton
price derived in this estimation is low. Durbin's h statistic
is again insignificant at the 95 percent confidence level;
the test for serial correlation is insignificant. The
fitted and actual values for soybean acres planted are pre-

sented in Figure 3a.
Soybean Yield Equations

At first glance these estimated equations may seem dis-
appointing. The soybean yield equation for the northern U.S.
may overestimate yields in the future, as the positive in-
fluence of the square of time overshadows the negative effect
of time variable. The estimated yields for the southern re-

gion should exhibit more stability as the signs on time and

 

30Samuel B. Richmond (1964) Statistical Analysis,
Ronald Press Company, New York.

46

Table l. - - A Comparison of Estimated Supply Elasticities.

Region/Study Effective Price Effective Price
Short-Run Long-Run
Own-Price Cross-Price Own-Price Cross-Price

Soybeans Corn Cotton Wheat Soybeans Corn Cotton Wheat

 

NORTH

Christensen '
(1979) .70 -.47 — -.08 6.26 -4.68 — -5.02

Houck, Ryan,
Subotnik
(1972) .68 -.57 - — — _ _ _

Kenyon,
Evans
(1975) .71 -.51 — - 5.47 -3.63 - -

SOUTH

Christensen
(1979) .39 -.13 -.09 - 2.63 -2.01 -l.71 -

Houck, Ryan,
Subotnik
(197 2) .96 - - - - - - -

Kenyon,
Evans
(1975) .30 - o.38 - 4.00 - -5.05 -

 

 

 

 

 

 

 

 

 

 

47

'53le

th

Sou.

‘P

 

m0h0< 20‘ HA 3..

.3
a!

GI

r9
’3

r7
’0

 

p3

 

no

1.

2".le

AC

 

a... 41 3 «J 6‘
nuuo< :o-~_:

 

=4
.l

a.
«I

4..
«I

.l
and

= v
-.°

48

time-squared variables are reversed. The weather variable
(DUMWEATHER) performs well. The tracking record in the
southern U.S. region is very poor, even in light of the great
variability of soybean yields in this region. In contrast,
the estimate of average soybean yield for the northern U.S.
tracks well, correctly following eleven out of nineteen
turns. In the last five years the estimates of soybean

yields in the northern U.S. were estimated within 10 percent

of their actual value, with every turn being caught correctly.

The Results: Demand Component

 

Soybean Meal: Quantity Demanded

The results of estimation are satisfactory for this
equation. While the coefficient associated with the price
of corn is not significant at the 95 percent confidence level,
all of the estimates behave in accordance with economic
theory. The Durbin-Watson test for serial-correlation is in-
significant at the 95 percent confidence level. The
equation tracks well, missing only two out of fourteen
turns (see Figure 4a.). The residual terms are stable,
except in the marketing years of 1971 and 1975, when the

residual term is larger than one standard error .

Total 011: Quantity Demanded
While all of the estimated coefficients have the signs
indicated by economic theory, the parameter estimate for

the consumer taste variable (CONSUMRAT) is not significant

49

 

 

 

 

 

 

 

C 7.
we. 3
. T. l 7
n .r/ 7
a . Ifl
U ’1!
Q 9 L
. ‘ .HJ
o... SJ
u— .4
Q 1 I3
sc -/
n e
a .C
e n
.C a
V. mu.
0 e .3
3 D .i
.
.
A.“ o J
E mw
«Fm -c
“a. w. L.1 e
L .I d
m U
2.. G n
a z r m
n: e
L - PL «7 C
.O l
3 7 5 .3 a...
. 39:5: :. a. a a a: a
.I
L
.3
L...
L
sfl‘ ell.
e a
.3 2 .L
.I 9. LC
1 C
7. 3“
L
.
. .
L7 u
.
r .
u
I J. (Jr! new“
I. 9
4L

 

accchOAAHaz access to...1r

50

at the 95 percent confidence level. The Durbin-Watson test

for autocorrelation is inconclusive at this same confidence
level. The estimate of the dependent variable tracks poorly

in the first third of the observation period, missing the first
five turns. Thereafter, all ten of the next turns are

correctly caught. The residual terms become successively small-
er over the final two-thirds of the sample period (see Figure

4c.)..

Soybean Oil: Quantity Demanded

This equation is satisfactory. It has the expected
signs, a high level of confidence associated with each para-
meter estimate, and a good fit with the dependent variable's
observations (see Figure 4b). Only two out of fifteen turns
are missed, both early in the sample period. The residual
terms are stable over the entire sample period beyond the
forth observation. The Durbin-Watson test for autocorrelation

is inconclusive.

Soybean Price Recieved By Farmers

This equation is nearly definitional, which explains
many of it's desirable characteristics. The Durbin-Watson
test for autocorrelation was significant at the 95 percent

confidence level.

Soybeans: Ending Stocks
Two alternative specifications were considered: one
based upon the assumption of speculative demand and one which

treats soybean stocks as a residual. The F-statistic for

51

the speculative stocks specification was not significant at
the 95 percent confidence level; the residual specification was.
stocks are treated as a residual in the final specification.
Durbin's h statistic is significant at the 95 percent con-
fidence level; the test for serial correlation is significant

in the soybean ending stock residual formulation.

Soybean Oil: Ending Stocks

Two formulations were tested for significance here, one
for a speculative demand assumption and one for a residual
stocks assumption. The F-statistic associated with the
speculative demand specification was insignificant, while
the residual stock specification's F-statistic was signi--
ficant at the 95 percent confidence level. Soybean oil
stocks are treated as a residual in the final specification.
It is interesting to note that the quantity of soybean meal
demanded is statistically significant in both specifications.
This reinforces the assumption that high levels of soybean
meal demand, relative to soybean oil demand, may result in
a buildup of soybean oil stocks. Durbin's h-statistic is
significant in both specifications; the test for serial-

correlation is significant.

Soybeans Used for Seed

Parameter estimates for both total acres of soybeans
planted in the U.S. and for the deflated price of soybeans
are in accordance with the expectations of economic theory.
A large part of the variation of the dependent variable is

accounted for.

52

Table 2. - - A Comparison of Short-Run Demand Elasticities for Soybean Oil and Soybean Meal.

 

Price/Equation-Study Soybean Corn Beef Deflated Deflated
Meal Price Price Soybean Disposable
Price Oil Price Per Capita
Income

Soybean Meal Demand

Christensen (1979) - 0.18 0.06 0.46
Hieronymus (1961) - 0.74
Houck, Ryan, Subotnik - 0.18 -
(1972)
Paulino (1966) - 0.76

Soybean Oil Demand
Christensen (1979) - - - - 0.11 2.14
Hieronymus (I96I) - -

Houck, Ryan, Subornik - - - - 0.28
(1972)

Paulino (I966) - - - - 1.30 .

 

 

 

 

 

 

 

 

53

Soybeans: Residual Usage

This specification is notably weak, with only a third
of the variation of the dependent variable around it's
mean being explained. Luckily, residual usage of soybeans
has never accounted for more than 2 percent of soybean

production historically.

Analysis Of Forcasting Ability

 

As a prelude to the actual forecast, an examination
of statistics which measure the ability of an equation
to track Observed data is made. A final test of the

forecasting ability of the entire system of equations will

be made with data from 1978, a year beyond the estimation
sample period. Evaluation of the accuracy of the fitted
values, in both the sample period and in 1978, will provide

additional criteria for evaluation.

Theil Coefficients

 

Theil coefficients are a means of residual analysis which
tests the ability of an equation or system of equations to fit
the actual data. There are four Theil coefficients which
measure predictive ability (see Table 3).

The first Theil coefficient, U1, compares the actual and
fitted values. It is standardized to fall with the range of 0

and l, where 0 corresponds to a perfect forecast.

The second Theil coefficient compares the estimated value

to the value observed in the previous period. This serves to

54

who. Ado. oao. ooo.H ooooo.ooooo. mmh. mmv. ovo. Nmo. cusOmIOHOfi> cmmn>om
mom. mmo. Hoo. moo. ooooo.ooooo. mmh. mmv. one. NNo. :uHOZIUdOM> :mmnxom
mmh. ohm. Hoo. _ooo.H ooooo.ooooo. Nam. one. mmo. hmo. susomlmcmmnhom DODEOHD mmuo<
mam. moo. Hoo. ooo.H ooooo.ooooo. 55m. How. mom. mHo. :uHOZImCDOnhom Umucmam OOHO<

mom. oao. oHo. ooo.H ooooo.ooooo. mam. omv. vvm. oHo. UCOEOD HMO HmuOB

woo. woo. voo. ooo.H ooooo.ooooo. Hon. ooN. mom. HNo. DQOEOD HAO SOOQ>Om
vmo. moo. moo. ammo. hoooo.ooooo. omo. mom. mow. omo. UCOEOD How: cmmn>om
DmD mmD EmD DMD i mHD SHD VD mD ND HD

coflufimoosoomo Houum Ohmsomlcmmz mucOflOMwmmoo Adore coflumsom

.HOOOE ecu mo >DHHMQ4 mowummomHOh 0:» mo mam>acc<Il.m mqm<a

55

give a point of reference, where the model under scrutiny
is compared to a naive model: one which assumes that this
year's value is identical to last year's value. When the
average of the residual terms of the tested and naive model
are identical, the value of U2 will be 1. If the tested model
performs well, then the value of U2 will be less than one,
whereas if the tested model performs very poorly, (i.e. worse
than the naive model), then value of U2 will be greater than one.
The third Theil coefficient, U3, compares the observed
change to the fitted change of a value. U3 is standardized
to fall within 0 and 1, where a perfect forecast would have
a value of 0.
The fourth Theil coefficient compares the fitted changes
to the change observed in the previous period. The value of
U4 that indicates an equality between the tested and naive

models varies between samples.

Decomposition of Mean-Square Error

 

Mean-square error (MSE) is decomposed in one of two ways:
bias, variance and covariance components or bias, regression
and disturbance components. The bias component of MSE, here-
after denoted Um, measures the deviation between the actual
and fitted average changes in value. If Um is large3l, the

fitted values do little to explain this deviation. On the

basis of past evaluation, there is little information to be

31(Um+US+Uc = Um+Ur+Ud= l). .1

56

gained from the examination of the variance and covariance
proportions of the mean-square error term (denoted Us and Uc
respectively).32 In the same analysis, it was demonstrated

that the regression proportion (Ur) is consistent: in a

perfect forecast both Um and Ur will tend towards zero.
Results of Estimation: Forecast Ability

The results of estimation in terms of Theil coefficients
and mean-square error decomposition are presented in Table 3.
In general, the Theil coefficients indicate there has been
a substantial improvement over the naive model in all three
demand equations, with a good fit to actual values. Changes in
the actual data are not explained as well, as both U3 and U4
are higher than U1 and U2 in all the.demand equations. The
Theil coefficients of the supply equations show less difference
from a naive model than do the demand equations. With a lagged
dependent variable in the acreage planted equations, and
trend in the yield equations, this is no great surprise.

The changes in actual data are not explained well in
either of the yield or acreage planted, Southern U.S.,
equations. This may be the result of the use of trend and
geometrically lagged endogenous variables to estimate relation-

ships over a long period of growth.iriobserved values.
The decomposition of mean square prOportions attributes

very little of the error to bias of the regression in any of

 

32C.W.J. Granger and P. Newbod (1973) "Some Comments on the

Evaluation of Economic Forecasts" Applied Economics (p.35-47).

 

57

the demand equations. Such is not the case in the supply
equations, where the proportions of the mean-square error
attributed to bias or regression error are substantial.

The most important indication of the analysis of
Theil coefficients and the decompostion of mean-square error
proportions is an inflexibility and significant degree of
error associated with the four supply functions. While the
demand equations explain a good deal of the change in the
quantities of soybean meal and oil demanded, the soybean
acreage planted for the southern U.S. seems to be more rigid.
Since this supply equation was estimated ovdr a long sample
period, and since this was a period of acreage expansion
and average yield increase, one might expect this rigidity

to be in a downward direction.

Results Of Simultaneous Solution

 

The entire soybean model was solved via the Gauss-Seidel
simultaneous solution technique. The Gauss-Seidel simultaneous
solution technique solves a system by first calculating a
value for each endogenous variable from actual values for
the right-hand side endogenous variables and the exogenous
variables. In each of the following iterations a value
is calculated for each right-hand side endogenous variable
which equals seventy-five percent of the value that was
-calcu1ated for each right-hand side endogenous variable in
the previous iteration plus twenty-five percent of the value

calculated for each right-hand side endogenous variable during

58

the current iteration. This process is continued until
the change in the estimate for each endogenous variable
is less than one percent in the final iteration.

Where Yln= b0+b1Xl+b2Y2n

Xl= exogenous variable 1

Yln= endogenous variable 1, nth iteration

Y2n= endogenous variable 2, nth iteration

Y1,l= b0+b1Xl+b2Y2

Y1,2= b0+b1Xl+b2((.75*Y2,1)+(.25*Y2,2))

Yln= b0+b1Xl+b2((.75*Y2,n-l)+(.25*Y2,n))
This solution provides information which will be used to
analyze the capability of the entire model to predict
future market conditions. The most important criteria in
this assessment is the accuracy of the model estimates when
the model is in simulation mode (i.e. internal estimates are
used to generate the final set of estimated values). The
stability of the model can be tested by comparing the results
of simulation mode solution to the actual data. A further
test of the model is made when forecasts of known values are
made, but these values were not included in the sample period
over which each equation was estimated. This occurs in the

year 1978.

Underlying Assumptions of the Forecast

 

To understand the forecast made in this solution, the

59

assumptions about the exogenous variables must be known.
These assumptions are:

l. The level of soybean oil exports will rise at same
rate as the 1974-1978 average of 4 percent annually.

2. The trends of consumer preference, reflected in
the ratio of margarine to butter and lard consumption vari—
able: CONSUMRAT, will continue to increase at the 1974 to
1978 average annual rate of 4 percent annually.

3. Soybean Oil stocks are set at the 1974-1978 average
rate of 4 percent increase.

4. The consumer price index is assumed to increase to
194.8 in 1978, 208.3 in 1979, 220.9 in 1980, 232.4 in 1981,
and 245.9 in 1982.33

5. Disposable Personal Income is set at 1452.9 in 1978,
1598.5 in 1979, 1740.0 in 1980, 1884.2 in 1981, and 2048.8
in 1982.34

6. Soybean exports are assumed to expand at the 1974-78
average rate of 7 percent annually.

7. Rainfall in the southern United States will be the
same as during 1977 and 1978 (four inches above average).

8. The level of resident U.S. population will be 219

millions in 1978, 220 millions in 1979, 222 millions in 1980,

 

33Michael K. Evans (1978) Long-Term Macroeconomic

Forecasts, Chase Econometric Associates, Inc.

34Ibid.

 

 

60

224 millions in 1981 and 226 millions in 1982.

35

Exogenous Forecasts from the National Model

All other exogenous data is provided by the M.S.U.
Agricultural Model solved in a simulation mode.

estimates are summarized below:

High Protein 1978
Consuming Animal
Units (HPCAUT)
Beef Price —
dollars per cwt.
(LPBEEF) 40.0
Policy Variable

for Soybeans -

dollars per bushel
(PVISBT) 2.50
Cotton Price -

cents per pound

(COTPT) 60.1
Corn Price -

dollars per bushel

(CORNPT) 3.21

The Forecast

 

1979

43.5

51.1

1980

38.5

2.25

1981

37.

50.

These

.25

.88

1982

98,765 100,330 104,156 106,240 106,175

41.5

57.5

The results Of the simultaneous solution of the entire

 

35

Bureau of the Census (1976) Statistical Abstract of

the United States, Washington D.C.:

 

 

Government Printing Office.

61

system of equations are presented graphically in the Figures

5 through 12, inclusive. It is important to note that while
no actual values are reported for the years from 1978 to

1982, the actual value for 1978 is plotted on the graph in
each figure and reported above each table in each of the

figures.

Conditional Forecasts

 

Additional forecasts are presented in Appendices 3.
C, D, E, pages 78 through 109 inclusive.

In Appendix .8, all of the original underlying assump-
tions are maintained while the national average soybean
yield is reduced by 12 percent for the entire period (1972
to 1982).

In Appendix. C, the national average soybean yield is
increased by 12 percent, while all other exogenous factors
are held at the levels assumed in the original forecast.

In Appendix D, soybean meal exports and soybean ex-
ports are increased by 10 percent annually, while soybean
oil exports are increased by 5 percent annually. All other
exogenous factors are held at the levels assumed in the
original forecast.

In Appendix .E, soybean meal exports and soybean ex-
ports are reduced by 10 percent annually, while soybean oil
exports are reduced by 5 percent annually. All other exo-

genous factors are held at the original forecast levels.

62

FIGURE 5.--Soybean Price, Dollars Per Bushel, 1978
Actual = 6.50

    

THIS Is A Rut OF THE MODEL IN A PROJECTICN MODE-FROM 197: TO
7.39 . f ’ ’ ' ' ' I ' ' ' ' * ' ' ' ' ' ‘ ’ ’ ' ' ' ‘ ’
7.60 . 3 E
7.30 - '- ,I’

7.01 : E I
6.72 : Z /
6.42 - Z .l
6.13 3 I
s... : 4

.54 - '
5.25 I Z
4.95 : :
‘tobc : E
«.37 I ~

;§-o;:"o;;"o;R-o;s'o;§'o’o‘o‘o‘o'o'o’o'o’o" ’0'.

ACTUAL (+) AYE ESTITATCS (*) VALUES F0? l-SGYEFT

YF’R LEVELS

(+) (t)
1972 4.370 8.739
1973 5.682 5.030
157+ 6.54; 6.123
1975 4.920 5.0‘6
197% 5.513 6.565
1977 5.88? b.C14
FCRECLST
1978 5.888 6.279
1979 5.883 6.491
lie 5.8t3 7.529
1‘"; 3058:. 80,336

\0
If
IV’
0

IIIIIIIIIIIIIIIIIIIIIIIIIIIJ

63

23.754.

FIGURE 6.--Soybean Meal Production, 1000 Tons,
1978 Actual =

.‘~
— h a
PL
F1.
FL

TI

Dy

Rt?

15

THIS

++ ++Q

5..
rJ
+

.L
a.
7
2

O

rag
.T
.L
:1.
.b
2

.3.
n.
.9.
PL
.3
c.
2

O

5

a

*

E
5
7o
2

o

r...
3
+
.L
9.
9..

o

OZUDE+L.5
.1965*L

A.

.3
r. U
+
E
7
8

CI.

.3
.mL
+
E
7.
{D
l
o

 

(‘F

' (.) VALUES F

FSTIVATEC

(+) AND

ACIUAL

*

(a

_L

V)
75.9
Ll‘

$90329
700.9441“
010.888
0 0 o o o 0
11:44.18
61350.02
21cc 80.2....
70.15;...nffia
141.111n‘

».L....I.a....\-...,Jn.v
PavelarU “5an
Wupaalofiuuhqc .U
C . O O C 0
90.4214 8.1
...L.7PU.C....C7
7L§L141c
(DC)..D: amen/L
51.14:...CIIJL

77iv~55¢d8
1.6601
69.858
0 O O O O
ficr—DCCITs
“ILQSTPIL
REVILFC.
“.harnualrn.
22.2.2»)!—

oilfLfiufiunu
Grdflbeau
“cu—UCAUAU
O I O O 0
11.111
77.777.
311.743....v
rzn._2..cn¢
0:43.. 9. 9..

T

CD

A

C
2.39:..57E5c. 0.1.2
77. 77.7.7.3...7.7rt...t€
u). u,.C.u..o,Gi.Uc. C; q, 019
1111.1..I4F1‘J11J1a

64

FIGURE 7.--Soybean Oil Production, Million Pounds,
1978 Actual = 10,800.

    

18 L RU? OF THE “DUEL IR A SROJECTIUN ”nDE-FPCM 1973 T3 1982.
75+3§ ‘ ' f -
Mar:- - - I -
15£+25 - - ’ -
- - I -
EE‘JD " ' J "
ea+fz - - r' ‘
- - - I -
35*u3‘ “ ' I "
- +/ + + + -
.992E*C4 " f’ -
.949E+04 - -/ -
- 7 -
07E+UQ - r ‘
OSE+34 - - -
2E+74 - ’ -
35*3~ - ‘ -
TE+CQ - - -

NI!
(73.
(L

I
u).

I
I
I
|

HI

[n
In
.4
H

1:
I)
a
{n

ACTUAL (+) AH?

Yrfiw LEVELS

(*) (*)
1972 7ECI.CU3 9769.015
1973 ECQSoQGJ 9503.453
197a 7375.303 796°.280
1975 3638.603 9C56ofilé
1978 3575.333 3585.768
1q77 18298.333 1L276.Ulo
FUSECAIT
1978 112E8.COC 13875.233
1979 132%”.036 11275.59?
1983 17258oC30 liEl9o983
1°91 13208.333 11152.784
lQPZ loZb8.JCJ 12672.073

65

s
l
e
h
S o
uO
BS
1
n
O:
i
11
la
iu
Mt
C
[A
5
k9
C7
09
tl
S
nl
a6
el
b
Y:
0
$1
a
gu
nt
iC
nA
n
i8
97
e9
Bl
.
_
8
E
R
U
G
I
F

 

a;
”C
9
1... . _ . . . . . . - . . . . . - . . . _ — . . . . _ c
.
PU. o T-
T: . ¢W
. o T.
.J - C...
7.. o u.
0, . S.
4L. o .
. é
um. . 0
an .
D. . o
F . Q
. . o C
E . F,
H... o
m; — C»
u... . o r. ) 3.64133 39931
. U .r 77.....5385... 7.1.232,“
?. o L ( 823357 59974
as» . A, 0.000. 000.0
7.. o V 1841.58 310.98
T . 7.38195 0:39.90,
.L. o s) «L1._n.¢..3 73),. . l
r... . t .
Iu— o I\
3 .
Du - \k and um...
3r \. \ . a. i
. 9|“ 0 ~75 S
A \. \.. . A L
\\ —_ .
.x. L
I V) fi..r..:J.de_HJ \J..U.U.Jh.
T. FT, 649.404b 051030
C, L‘ Q,rOQ))nJ/2nl «UfJflvau
f,— 000009 on...
199547.. 7:25.51on
x...“ 7 P3734... 1.. V...‘ ..
u\ l 1 6&1 1 1.“ no... «.1. 1
.L
\J
+
(\ T:
To
L RM
.A C
U or 11:34 5 67L... :0 C2019.
o.. T p 7777771fin77 u....,..«.p_
HU- . . . . . _ . . . . . . . . . . . . . . . . . - . C r.” 0.9QIQ’.go.nUQ/.7QJQ.Q,
R RH VI. 1.1..1111Flollolil
uh
S
v... o o o o o o in. 4 9.. 7.. o o o
(D Aug 7 3 8 I». O O O O Q] 4 d
8 7v 0) Q no t). 1.. Q, .4. nu Q. 04 b O;
I 3 2 2 an 1 1 6 n4 2 ,b 1 1.. 1
H . . . . .
YI

66

S
l
e
h
S
u
B
0
07,
04.
16
2
4
8
’
1

FIGURE 9.--Soybean Production,
1978 Actual

L

PROJECTIL

U;A

MOTEL

THE

Rum CF

A.
H

«\v

I

THIS

 
   

.
.
.
u
.
.
.
.
-
.
.
.
.
I,
. If.
I
.
.
.
. . . . . . _
-
.
.
.
-
.—
...d :5 .H
..J U pJ
+ * +
.L E .L..
1a b [u
flJ 9 rm.
«l— 1 1
O O O

_3
pr .
+
tr.
3
«.0
1

O

"3
an,
O

E
,3
7a
1.

o

:1.

CL

A
n—

.J
1,
KO
1

O

pflJ
. U
.7
PL
or.

O
1

.135€+0:

. . . -
R .. r:
+ +
P: Pr—
8 .1
n/. 3.—
1.. l

o o

sevfiar

’
‘ -

(*)

:—

5+
L‘

2.x

ck

lfiQID—I-IDS
nu737J7wU
7.955214
0 o o o o 0
540/474
0510/3415
acconc HE.
9.2412744 A.
..iq.i.4L11-..

«U :6 F. n...“ N... ..U
U:.?..;;U.U.3
1‘. »O Nu 3.. ,D C...
o o o o o 0
(018157»
Fed." I“ .L. 91.1
7.41488
3. .3 I. ,7? n47.

114 .1LQ-4L

to BEE“
51217.
6.....5 57b
0 o o o a
91037
86565
.DACQAQJCJ
77.7Q .2.

7. «L .I‘ 1...]..—

U.) 336
Snfiufiuumk.
.5273 C e:

14 Laz¢l 7‘
{Urn/D .D rC
77777
1 12‘ 40.1 1.‘

2 7 . .4. 5 .57 E 8 9 U1 2
777.777.?77c...Air.
G 99992. 399—3023,.
«LlllllFlllll

I

Southern U.S.

q

Cl

15,560.

0,-Acres of Soybeans Harveste

.L

1

1000 Acres,l978 Actual =

FIGURE

0
9L

8 .
G1
1... . . . . . . . . . . . . . . . . . . . . . . . . . o
.
n-.. o
T. .
. .0
Ad .
7. 0
Q, .
1. o
.
M.. o
O .
R. o
F. .
. . o
E .
DJ. 0
3 .
V. o
.
an. .
O .
I. o
T. .
C. o
E .
.d. o
O .
F+ .7.
o /
+ I.
A
+
+
+

Morg'urw

E
H
TI
F
C
........
0..............._..........
R
5 :J 5 5 5 3 5 ,3 5 5 ...... 5 .2
a. .:J C C C .U 3 U 0 .U .U 3 C V
+ + + + + .v + + + .v + + +
SEEEEEEEEEEEEE
14185295307418
4 4 3 3 3 2 2 2 2 l 1.. .1 no
51111111111111
I 0 O O 0 O O O O O O O O o
H .
T

ll-AHSdTS

FCQ

VALUES

(*)

 

(+) AND

ACTUAL

*)

.L

V)
Cr...
Ll.\

16512.4.
311858
7671258
0 o o o o 0
2.36432
93039.9.
796.345
”Jllslndnd
111111

.H... flu nUn .1. #0.. r0
Dab-U300
UOGaUr-JO
O O O O O 0
0681423
c.3998?»
1714Q,S
12239.74
11L1\$11

CAST.

21.35.30
5.49.22
989:7...
O O O O O
770......8
0431a.
cogs/0c;
3nc37J7.
1111...].

OPUfiUprrln.
03030
SAJflwugfl-b
O O O O 0
33333
9ozc56.
auscdrds
«44.4.4.4.
19.1.1.1

fit. 1..“ C467V.BQI.U1AL
7 7777.79.77 H. ,5 t
:1 90.999549019 Q52
.1 111.1qu11. .111

68

1000

FIGURE ll.——Acres of Soybeans Harvested - U.S.,

63,033.

1978 Actual

Acres,

PU,

J

MODE-FROM 197

A PROJECTIDN

OF THE MODEL IN

RUh

A

THIS IS

.543E+55

0529E‘K':

j
.3
+
F.
a?
G
.5,

I

.478E+25

.467E+35

.1

0457E*11‘

 

CT

3 l2-AHS*

O

VALUES F

1

YEAR

67é:034
O I I a o .
77.284958
549.253
U29.:\1..1
(a. 4 30,59. 0.
450/455

pUnanjn,.nU
.Unleanlufil.

Q 5Q.:JQ.RJ

952/25
.8849}?
0455?.
o . o . c
7nUEq7p3
21.—.917
65200,
“52577
55555

Unaniflbrl.
fithn—DD
.Lnbh..afluxd
I I I O O

cauJG...:.D

254.557E8OL .12
777777;.77885
o.99999099999
1111.11F11111.

FIGURE

THIS IS A
.117E+04
.113E+s4
.1095+:4
.13§E+3h
.131E‘3k
976.
936.
897.
858.
819.
780.
7&0.
701.

69

1978 Actual = 1000.

Rt} CF THE

.4
1
X,»

~030uxoomr
'D'T *th "-4w4-4 ‘14
.n bran:

~13 ac
ASU’JV'I \HIV
('7
3
J)
-I

p,‘pAH|-H1’1H,.ap.np—ap4w

'D‘f“
IUD-‘()

I c

OLEL I\ A PFO
: r
- I
- I
- .v
- /I‘

: I
I

' I

-/

     

+ + + + +

721-“33
5"1-?"
771.30]
835.16?
7“0.00?
*27.QQC
9.570003
Q27oi3.
927.?3“
927.9CC
9‘7-LC‘

;..’.a._- k'

A
i
v

l3?.913
947.4169
72“.E79
111.&28
77$.?5‘
$39.52?
393.E56
513-578
358.c85
739.955
1°?.979

12.—-Soybeans firocessed, Million Bushels,

IllllllIllllr—

\C

G,
a
n.)
0

CHAPTER VI

CONCLUSION

The estimated model performs well over the sample
period. The fitted values of the dependent variables
track the actual values well, especially in current years, a
period of greater variation (see Figures III and IV ). Tne
estimated economic parameters seem reasonable, as they com-
pare favorable with the findings of other studies (see

Tables 1 and 2 ) .

Major Findings

 

1. Comparison of the elasticity estimate for soybean
meal price from the soybean meal demand equation in this
model, to estimates from other models, (with older sample
periods) indicates that soybean meal demand is becoming more
inelastic over time.

2. Stocks of soybeans and soybean oil are the result
of quantities demanded being in less than quantities supplied,
not the result of speculation.

3. Feed grains (i.e. corn) are in direct competition
with soybean meal in the livestock industry. This is evi-
denced by a positive cross-price elasticity estimate for

corn in the soybean meal demand equation. Corn is a substitute

70

71

for soybean meal in the production of livestock and live-
stock products.

4. The logrithmic specification of beef price in the
soybean meal demand equation works well. This would sup-
port the hypothesis that there may be a declining cross-
price elasticity between beef price and soybean meal price
as feeding practices become more fixed.

5. The soybean acreage planted equations, taken to-
gether, performed well. The average absolute error of
2,200,000 acres annually for the 1966-1974 period represents
a 4.7 percent error. In comparison, another, more dis-
aggregated, model appears to perform better over the same
observation period (3.6 percent improvement). Kenyon and
Evans'study had an average absolute error of only 510,000
annually for this same sample period.36

6. The weather variable, based upon sunspot activity,
is statistically significant at the 95% confidence level.
This variable has promise for future efforts to forecast
crOp yields in the midwestern United States.

7. The own price elasticity for soybean oil was
estimated to be 2.14, a value which would indicate that
soybean oil is a superior good in consumption, or at least
a constituent of such goods.

The only obvious improvement that can be made on the
basis of the results would be the disaggregation of the

supply of soybeans into the six regions suggested by the

 

36D. Kenyon and R. Evans, 1974 op.cit.

72

U.S.D.A. The gains that would be realized by disaggre—
gating are tangible, but not enormous. There is a trade-
off between simplicity and incremental accuracy gains when

disaggregation is considered.

Results of the Forecast

 

The estimated quantities of soybeans produced and pro-
cessed domestically bOth correspond closely to the observed
values throughout the entire sample period. In the year
1978, the production estimate is 4 percent too low for the
quantity of soybeans produced (compared to an average
absolute error of 7.5% over the estimation period). The
estimated quantity of soybeans processed in 1978 is 27
percent too low, which compares favorably with a 3 percent
average absolute error for the estimation period

A closer examination of the source of the error in
the production estimate for 1978 reveals that the estimate
for total acreage harvested for soybeans in the U.S. is
5% too low. This pattern repeats itself throughout the
sample period, with a negative 3 percent average error for
the entire U.S. and a negative 8 percent average error for
the southern U.S. region. With an average of 25 percent of
all production occurring in the southern U.S. during the
1972-1978 period, a substantial portion of the total error
in the soybean production estimate could stem from a con-
sistent underestimation of soybean acreage harvested in the

southern U.S. region. The estimated yields are apparently

73

too high in many years, which would be the only plausibl
explanation for a positive error for estimated soybean
production in years where the error for soybean acreage
harvested is negative. The results of simultaneous solution
re-affirm the results of statistical analysis: the soybean
acreage planted in the southern U.S. performs poorly and

the yield equations have an upward bias. The improvement

of these equations will be the focus of future efforts.

The source of error in the estimate of soybeans pro-
cessed in the U.S. is easily decomposed. While the average
absolute error for the soybean meal quantity produced in
the U.S. has an absolute error of about 3 percent associated
with it, the average error is about 1 percent too low. The
bias is in the Opposite direction in the soybean oil pro-
duced equation; with an average absolute error of almost 5
percent and an average error of 2.5 percent. The average
error associated with the quantity of soybeans processed
in the U.S. estimate is less than 1 percent. There seems
to be no systematic under or over-estimation of this quantity.
In light of favorable performance in all respects, the demand
component of the model: soybean meal demand, soybean oil
demand and quantity of soybeans processed, is considered
satisfactory.

Beginning soybean stocks are overestimated consistently
from 1975 on, but this is due, in a large part, to the large
positive error associated with soybean production in 1974.

Since each year's estimate of beginning stocks will depend,

74

in part, on the previous year's estimate, the error begun

in 1975 will be repeated in each of the following years.
Soybean price does not respond well to year-to-year

changes in total supply and total demand. A plausible method

of adjustment would be to respecify soybean price as being,

in part, a function of soybean stocks (as a measure of

total market conditions). This alternative will be explored.

Closing Remarks

 

While the model put forth in this study has some short-
comings, it does provide a workable framework for the examina-
tion of the entire U.S. soybean industry. In this sense the
initial objective of this research has been satisfactorily
fulfilled. Applications of this study are illustrated in
Appendices B, C, D, E, where various assumption alter-

natives are incorporated into conditional forecasts.

BIBLIOGRAPHY

Black, J. Roy and Thompson, Stanley R., 1978. "Some Evidence
on Weather-Crop-Yield Interaction" American Journal of Agri-
cultural Economics, Vol. 60, No. 3, August 1978.

 

 

Harris, H.M. and Kenyon, D., 1974. Competition Between Soy-
beans and Other Crops in Major U.S. Regions, Economic Research
Service, U.S. Department of Agriculture, E.R.S.-588.

 

 

Hedlund, Earl C., 1952. Transportation Economics of the Soy-
bean Processing Industry. Urbana, Illinois: University of
Illinois Press.

 

 

Hieronymus, T.A., 1961. fForecasting Soybean Meal Futures
Prices" Commodity Yearbook, Commodity Research Bureau, Inc.,
New York.

 

Houck, James P., 1963. Demand and Price Analysis of the U.S.
Soybean Market. University of Minnesota Agricultural Experi—
ment Station, Technical Bulletin 244.

 

 

Houck, James P., Ryan, Mary E., and Subotnik, Abraham, 1972.
Soybeans and Their Products. University of Minnesota Press,
Minneapolis.

 

Kenyon, David E. and Evans, R.S., 1975. Short-Term Soybean
Acreage Projection Model Including Price and Policy Impacts.
Virginia Polytechnic Institute and State University, Research
Bulletin 106.

 

 

King, Gordon A., 1958. The Demand and Price Structure for
Byproduct Feeds. U.S. Department of Agriculture, Technical
Bulletin 1183.

 

 

Kromer, George W., 1975. Trends and Patterns in Soybean Oil
Use for Food and Industrial Products. Economic Research
Service, U.S. Department of Agriculture, E.R.S.-6ll.

 

 

Lamm, R. McFall, Jr., 1977. "Demand and Supply of Vegetable
Oil Products in the United States: A Short-Run Supply
Analysis", Southern Journal of Agricultural Economics, July,
1977.

 

75

76

Matthews, Jimmy L., Womack, Abner W. and Hoffman, Robert G.,
1971. "Formulation of Market Forecasts for the U.S. Soybean
Economy With An Econometric Model:, Fats and Oils Situation,
November, 1971. Ecomomic Research Service, U.S. Department.
of Agriculture, FOS—260.

 

Matthews, Jimmy L., 1973. "Conditional Market Forecasts
and Implications for the U.S. Soybean Economy:, Fats and
Oils Situation, July, 1973. Economic Research Service, U.S.
Department of Agriculture, FOS-267.

 

Nakamura, H., Hieronymus, T.A. and Judge, G.C., 1963. Inter-
regional Analyses of the Soybean Sector. University of
Illinois, Department of Agricultural Economics, AERR-67.

Paulino, Leonardo A., 1966. "A Recursive Model of the United
States Domestic Soybean Market", Thesis for the Degree of
Ph.D., Michigan State University.

Paxton, Kenneth W., 1978. Cotton and Soybean Production,
Costs and Returns. Louisiana State University Agricultural
Experiment Station, Department of Agricultural Economics,
Research Report No. 528.

 

 

Richmond, Samuel B., 1964. Statistical Analysis. Ronald Press
Company, New York.

 

U.S. Department of Agriculture, 1959. "Factors Influencing
Soybean Prices", Feed Situation, July, 1959.

 

Wailes, Eric, 1979. "The M.S.U. Agricultural Model - Domestic
Policy Component“, unpublished monograph. Michigan State
University, Department of Agricultural Economics.

77

APPENDIX A.

SOURCES FOR DATA

. Agricultural Statistics, l978.S.R.S., U.S.D.A.

a. Table 178, page 129.
b. Table 180, page 130.
c. Table 182, page 131.
d. Table 185, page 133.
e. Table 194, page 138.
f. Table 188, page 143.
g. Table 39, page 30.

h. Table 82, page 61.

i. Table 48, page 38.

j. Table 201, page 144.

k. Table 203, page 146.

Survey of Current Business, various issues.

 

Statistical Abstract, 1978. Table 2, page 6.

 

Business Statistics, 1978.

 

78

APPENDIX B
REDUCED YIELD SCENARIO

FIGURE l3.--Soybean Price, Dollars Per Bushel, 1978
Actual = 6850, Reduced Yield Scenario-

THIS IS A RUN OF THE MODEL IN A PROJECTION 8085-8808 1970 TO 1982.
9.77 - ' - - - - ' - ' - -,’ - - ' - - ' - - - - ' - - -
9.32
8.87
8.42
7.97
7.52
7.87
8.82
8.17
5.72
5.27
4.82
4.37

I
;

f

I

/’

‘s

‘\

\~

    

+++++

I I I I I I I I I I I I I I I I I I I I I I I I I
V .
0 I I I I I I I I I I I I I I I I I I I I I I I I I

75 78 78 85 82

-.-
70
K.
_ --‘--

ACTUAL (+) AND ESTIMATED (.) VALUES FOR l-SOYBPT

YEAR LEVELS

(+7 (*7
1972 4.570 5.289
1973 5.883 5.352
1974 8.848 8.255
1975 4.920 8.825
1978 8.818 8.522
1977 5.880 7.122
FORECAST
1978 5.888 7.702
1979 5.880 8.517
1980 5.680 9.259
1981 5.888 9.507
1982 5.880 9.998

79

FIGURE l4.--Soybean Meal Production, 1000 Tons,.1978 Actual
= 23,754, Reduced Yield Scenario.

    

THIS IS A RUN 0F THE_MO0EL IN A PROJECTION MODE-FROM 1970 To 198
.284E+CS 3 ' - ' ' ' 3 - - ' -,’.- - ' ' ' ' - - - ' - - - 3
.274E+05 E E l/ E
.264E+§5 - - f -
.255£+95 : I /I :
.245£+35 E E f’ E
.235E+05 - - -
.225£+85 : 'I Z

- + + + + + -

.216E+55 - -
.238£+85 : 3
.I98C+C5 : 3 I
.187Ef35 : : :
.177E+05 I I :
.167E+05 : 3 I
. ;é-o;;-o;é-.;é-.§E-og£-o-.-.-.-.-.-.-.-.-.-.-.-.-.

ACTUAL (+) 4ND ESTIMATED (i) VALUES FOR 2-587887

YEAR LEVELS

(+) (*)
1972 16758.300 17271.137
1973 19674.000 19211.519
1974 16732-003 16845.044
1975 20754.000 18411.726
1976 18488.033 17998.794
1977 22371.00? 22026.166
FORECAST
1978 22371.000 24938.526
1979 22371.303 25538.809
1980 22371.033 25:84.078
1981 22371.000 27808.477
1982 22371.003 28879.92?

7“)
o

80

FIGURE 15.--Soybean Oil Production, Million Pounds,1978 Ac-
tual = 10,800, Reduced Yield Scenario.

0
CO
I‘~ )

IC‘

THIS IS'A RUN OF THE MOSEL IN A PROJECTION MODE—F80. 1970 T
.124E+05 ' - - - - - ' - ’ - - ' ‘ - - - - - - - - - - -
.120E+05
.118E+05
.112E+35
.107E+05
.103E+35
.99OE+04
.948E+04
.906E+04
.884E+04
.822E+04
.780E+34
.737E+04

IIIIIIIIII
‘

    
   

IIII III! III! IIIII llll IIII
‘\
0 III! III! III! III! Illl llll IIH

ACTUAL (+) AND ESTIIATED (f) VALUES FCF Q-SDILCT

YEAR LEVELS

(+) (k)
1972 7501.000 :181.700
1973 8995.000 9740.237
1974 7375.000 8482.387
1975 9830.000 8778.905
I978 8578.00; H622.558
1977 13288.000 1=278.238
FanCAST
1978 10288.000 13001.859
I979 10288.000 12301.818
1980 10288.05; 13834.01:
1981 12288.000 11138.520
1982 10288.088 12828.137

81

FIGURE l6.—-Beginning Soybean Stocks, Million Bushels,
1978 Actual = 161, 1979 Actual = 150,Reduced
Yield Scenario.

THIS IS A RUN OF THE

MODEL IN A PROJECTION MUSE-“For 197: TO

  
 

0IIIIIIIIIIII'lllllllIOOIII"

1950 "
97-3 ' + + + + +
-9907 " ‘J \ ‘
. ' \ '
-1980 ’ l
- V*4
-2960 ' '\
-395. “ ’ \
-4930 " ' \
-592. - - \
-6900 " ' ‘
7 - - §
-788. - - \
' ' \
-887. " "
- - 'L
’9850 " " ‘
72 7a o7é 075 085 .?§_ o\o o o o o o o’o'o‘o‘o-o'
_ -- h--
ACTUAL (+) AND ESTIMA Eu (*) VALUES FL? F-SESTKT
YELP LFVELS
(+) (t)
1G72 710953 710866
1:373 5906“] ”96.462
157% 176.881 ’59.62C
1975 195.040 -4C.531
1976 244.603 "215.131
1977 105.303 '164.469
FCRECAST
1979 153.503 -Q5&.2CE
1350 103.003 -713.022
1931 103.00: -POB.533

\1‘
‘V:

82

1978 Actual =

1000 Bushels,

1,842,647, Reduced Yield Scenario.

FIGURE l7.--Soybean Production,

TO 1982.

I973

MCDE-FFOM

A PROJFCTION

OF THE MODEL IN

IS A RLH;

THIS

.

.

-

.

_

.

.

.

.

_

.

i
I

../l
p: ,3
no r...
+ +
F. E
3 :4
«U 9
2 1

. o

.3
«Iv

.188E+

P—J
AU
.7
.L
0
8
.l.

O

017SE*GS
.lESE+

C).
n)

5
0-.

.3
‘0
§
t;
.3
3
1

o

.128E+05
.121E+0

:
a

 

7-sovu'

VALUES FOP

ACTUAL (+) ANU

(*)

.2
V)
E...

Ll‘

YEAR

95044.0
92669.5
nUQu/zl-Iql.
O O I O I O
89.20arxc1...
92.9352
3%,..3469
715224.:
111L1L1A-1.

"U C C C G 0
052955
36325:.
O O O O O 0
518157
874671
7.4.1.486
252527
111.111

“.10.;“4/
9.8381.
37g77flz,
O I O O 0
5712.0
4.0462
2233.:
7.685... .
11122

G. «U nu flu 0....
£335.52.
52.2355
0 o o o 0
77.777
101.111...
£706,sz
77777
11110‘

234 c..b7E8c.,012
777777R77 889
9999a..909990..o..
llllll—flllll

83

FIGURE 18.--Acres of Soybeans Harvested - Southern U.S.,
1000 Acres, 1978 Actual = 15,560, Reduced
Yield Scenario.

OJECTIC". MDOE'FF'C-M 1°70 TO 1982.

   

.149E+?5 - I -
.146£+05 - + + t/{ . I
.IQZEHJE ‘ B I :
.139£+05 : I‘ I I
- V -
.1355+05 - -l -
.132E+05 - 1 :
.129L+03 - 3 :
.1255+05 - - I
0142£+05 ' " :
.118E+35 - - :
.1195..3 - - :
.111£+n= - I - :
.1uaE+:3 - 1 - I
7: 7" 07:3 07); 083 .t15 o o o o‘l‘o-o-o'o‘o‘o’o‘o‘o
_ -~——.
ACTUAL (+) AND STIHLTLO (.) VALUES Fca 11-AHSATS
YELQ LEVELS
(+) (.)
1972 11190.00: 1:779.350
1973 12736.0CL 12533.92:
1974 12198.03: 11819.82}
1975 13483.330 11344.003
1973 12982.00; 13151.31?
1977 14593.0c0 12997.574
FDFECAST
1975 14593.00: 14413.71?
1979 14593.303 13959.995
1953 14593.03: l“b3;.791
1981 14593.0C0 lfifinq.893
1992 14593.230 131C9.*12

84

FIGURE l9.--Acres of Soybeans Harvested - U.S., 1000 Acres,

THIS IS A

.648E+35
.632E+05
.616E+35
.600E+85
.SBQE+GS
.568E+JS
.552E+85
.5366+&5
.520E+GS
.505E+95

.489E+&5.

.47SE+05
.457E+35

HHHHHTIHo-‘HHHH

1978 Actual = 63,033,

RUE

  

72 74 76
ACTUAL (+) AND
YEzk

972 4559
973 5566
974 5134
975 5357
S76 4935
v77 5761
CRECAST

978 5761
979 5761
988 5761
981 S761
9&2 5761

  

NI'VNI‘ON NCI‘J‘IH‘HJI

OF THE MODEL IN A PRO

"‘1"
1‘

F“
+
v

(30000:)

0 o o o n

CJUCJLJD

-ICIOOO DC)CVCJC)Q
C')CJ(‘) (DC‘HDCJCJC

DC!

JECTICN M

A
fi
U

«5378.781
58217.724
38622.371
«9c25.549
57139.759
55399.49:
54372.911
58101.537
63571.753
55577.049
55529.31:

(\fi'“
but.

Reduced Yield Scenario.

'FROH 1973 TO 1982.

85

FIGURE 20.--Soybeans Processed, Million Bushels, 1978 Actual
= 1000, Reduced Yield Scenario.

     

THIS IS A RUN OF THE MODEL 1% A pROJECTION MODE-FROM 1970 TO 1982.
.117E+04 - - I, -
.115E*84 - - 1 -
.139E+C4 - - I)! I
.105954 - - / Z
.1019»: - - l -

- _ j _
97". - - / :
935. - _

— + + + * ¢ -
896- ‘ .-
8579 ‘ - :
8180 " - :
779. ‘ - —
7#0. - - :
701. " " _ -"

ACTUAL (+) AND ESTIWATEC (*) VLLUES FCR 18-CRUSFT

YEuR LEVELS

(+) (I)
1972 721.800 738.935
1973 821.33. 851.“74
1974 731.330 703.056
1973 865.103 798.746
1976 790.590 774.968
1977 927.303 935.691
FORECBST
1978 927.008 985.235
1979 927.003 1311.305
1989 927.&SJ 1356.712
1951 927.305 1097.9Eé
1982 927.300 1188.382

86

APPENDIX C
INCREASED YIELDS SCENARIO

FIGURE 21.--Soybean Price, Dollars Per Bushel, 1978
Actual = 6.50, Increased Yields Scenario.

  

 

 

6.85 :-----:----”------'------
6.6. E I ,’
6.1.3 - 1 J
6.23- : :/
6.02 I.
5031 : + + + + +
5.61 3 3
5.40 I I
.5.23 3 3
«.99 I 3
4.76 I I
4.55 3 3
4.37 3 :
‘0'o-o‘o‘o‘o ' ‘o‘s‘.'s‘.“9‘a‘9'v°o'.‘o'.‘.‘o

72 7a 76 75

ACTUAL (+) AND ESTI ATE (’) VALUiS F0? 1-307397

Y=AR LEVELS

(+1 (.)
1972 9.373 n.57:
«1332 2'2? 15"”?
1975 .392: 51337
1976 6.310 5.9..
1977 5.¢dC 5.47.
5872“” 5 .

ea. b 217
1979 5:660 6:816
1981 5.662 6.532
1931 5.563 6.77-
1952 5.682 b.94“

FIGURE 22.--Soybean Meal Production,
23,754,

.2815+05 '
.2725‘05 '
.262E+85 °
'.253E*UE °
.ZQ3E+05 -
.23AE’05 '
.22h5705 '
.215E+U5 '
.ZOSE+05 ‘
91965‘05 ‘
.186E005 5
61775005 '
.1675‘05 -

72

YFAQ

1972

1973

197k

1975

1976

1977

F093

1973

1979

1930

1981

1932

'7“

CAST

'76

 
 

78

(IA u..l‘:|, l‘..
0".)C1r “111' ('1
841 H" ’ ML“ 2

AI—\ I":
. ».~.r'.r r

I’ Far-0 717)

NNNNN NHNHHH
NNNNIV NmmeO‘
“(.4wa OIF‘JNO‘N
V—nrflu -umumruo
“HFH*H bemtTvru

87

ACTUAL 1+) ANC ESTZLATT

E0
I

+ + O + +

1000 Tons,
Increased Yields Scenario.

JCCTIUN MODE-F20“

VALUES

(‘)
-732". 137
-993309-3‘.
17c3;.92r
l0639n‘18‘.
L5129113”
22329-253
393530 351‘
25735.279
26322.655
3722*.29-
:d‘bu90977

-.-.-.-,-.-.-.-.- o o

1978 Actual

O III III III IIII III III III III IH

LL;

\L
N
c

88

FIGURE 23. --Soybean Oil Production, Million Pounds, 1978
Actual = 10, 800, Increased Yields Scenario.

I-

    

THIS IS-“ R3”-cf [H5 fcfgf EN-‘_PLOJEETEOE YUQEIFSOT E93"-T§
.1205oas - - ,
.1165+95 : 3 I
.1125+us I I I
.1oae+us E E I}
.1cue+esz - ”if + +
.1ooa+us - -’
.9eee+uu 3 -
.9285+ou : :
.390E+cu : :
.asze+ou 3 3
.814E+0u E l E
.7765+Gh - -
.737E+cu : 3
i ;.2-.;z-.;é-.;g-.3:-.;£-.-.-.-.-.-.-.-.-.-.-.-.-.-

YCAR LavaL:

(+1 (9)
1972 7501.;;: 5C98.159
1973 8999.913 9913.523
197a 7375.;cc c23E-19b
1975 9633.11: 5975.39-
1976 8576.;[2 9512.12;
1977 10285.2;5 1.276.255
Feiacasr
1975 13265.):5 ;;c77.2:9
1979 10255.:ac 1:;37.192
1930 10285.;C; 112:7.gez
1951 13288.;L; ;::6-.:-7
1992 13289.L13 12:42.3:3

[“0

u)
u
o

I1IIIIIIIIIIIIIIIIIIIIIIIIIII”

89

FIGURE 24.-~Beginning Soybean Stocks, Million Bushels,
1978 Actual = 161, 1979 Actual = 150,
Increased Yields Scenario.

THIS IS A RUN OF THE MODEL IN A PFUJE CTIUH UDE‘FEOH 197; TO

.126E+04:-----E"; $1.: ------- --- -
.116E+0t¢ : TI,"
o106£+0k : t
961. - I-
861. I II:
760. I 7/ 2
sea. 1: ’ E
560. 2 [I 2
use. : I. :
360. I I I
260. I 1 2
150. I W
59.6 - + + * + +
-'-°-°-"'-' 9" 5‘9gé-.-.-.-.-.-.-.-.-.‘.-.-.-.-.

YCAR LSVELS

(*I (‘)
1972 71.963 7lod7k
1973 59.655 193.245
197% 17Cobb§ fidkgd9C
1975 1550L9; 19;.su3
1976 2h~.BCC 639.27u
1972 10301.5.‘C 15330 331
FORCCAST .
1978 103.8t8 1173.348
1979 103ot38 1296.932
198? 103.12? 131;.178
1961 103.;fiC 137:0“Z4
1982 153o¢L3 1361-309

[‘0
o

u)
u

IIIIIIIIIIIIIIIIIIIIIIIIIIH

1952.

-5.
'J

p
U

97«

4
b

1978 Actual

Increased Yields Scenario.
FQOM

(.
C-
-

'T 10H mt

J

1000 Bushels,

90
~ A PROJEI

HOCEL

1,842,647,

FIGURE 25.-~Soybean Production,

THIS IS A RUN 0F TH

 

-
a T7
flu
. C.
v.
. 0
3
. .
7
.
.
‘vx
. 0
F
.
no
- HC \I 22¢..413Fh. 8.8 .1446
U # 731~U63 9.» 225
. n. ( 65.1.69? 6135.6
.44. I I I I I I I I I I I
n V. shiaefiu? 333.93
53966... 3.35.9.0
. I 251.932 5.2266
.. “.75p=6po 9999:.
. (x ......«..... ....4.05:.
7 .D
I _E
. III T S
A» L
. o’J — M. .2
—I V) nUflufi.P.-fi11 Open—nu... .
. \ T. 5+ «U52...» LE 555.25
3 LI 36 $265 55555
c r —.k. 090000 .0000
OI 6184.297 $37777.
.........fif C 079671 11111
N 7141B.86 66666
. AM 252527 77777
111111 11111
. I
.9
. ( T.
S
. L A
In C
. . U R 23h567589312
T A 777777977888
. c g — . o o c o . n o c . c . . ... n n c . . n o C 92 9999901099999
A VI 111111F11111
5 5 5 5 5 5 5 5 5 5 5 5 5
O .U 0 0 0 .U B 0 0 O G 0 .0
s. + .9 + + + + 1. + + .9 + +
E E E E E F: E E E E E E F.
3 5 8 fl... 3 5 7. o 2 a“ 7 9 1
1 3 9 9 8 7 6 6 5 h 3 2 2
2 Z 1 1 1 1. 1. 1 1. 1 4... 1 1
O 6 6 O 6 O C C O 6 C O O

FIGURE 26.--Acres of Soybeans Harvested - Southern U.S.,

.1th+85
.1h1E*05
.1355105
.135805
.13ZE+05
.129E+85
.istwS
.1238405
.1205 +05
.1175 +125
.1145 +05
.1113 +05
.1083 +85

1000 Acres,
Yields Scenario.

*
ACTUAL (+)

<
n

D
m

CAST

rerererere'nrerer‘rer‘re
qu3d3d3d0C3¢3u3¢3J3¢3d3
UDODCD‘4~QJn‘d~i~4~a~d‘J
"OF‘CDUOCDIn‘QCF\nJ?CMrU

    
 

L1,).LIIIIIII

13'

[fl
(1
.211

“o
76

ANC

VC

.nmmxflmo

O 0 O O O Q
I .t‘»qj?)l' ~Cn3 “I”

f: l”.fif.l‘.‘b(,) upfr;

r mac: 1 1r) (:3

HHHHP HHHPPH
Fttt-P tNuNNP
\nuuanIxnmt+qu
OOJhOIO
CMCHCACdCd (“POOJCDCDCD
O I O o I
(fifth/1.3!.-
(.JoDOCJLJ

Cal. “‘3‘ IC‘~

9].

1978 Actual = 15,560,

HI

I"

U)

(“)

l'i‘l'l‘091s

I‘I’I‘I‘I'

ION" I ' F‘ (.0

MNMNW

4‘ “INK-F 4 1 on \L-V
4"Vw‘30' \DNImC)“;

‘90-; Nmm

JALUcS F”

CLI‘CN-Duotu
C...

(‘)

.0
G‘I-xa
U'I’OJ

JH‘IJ

I‘WDU'I
‘JO‘NOVX' I‘.) ['1 '

'JII‘O

incutxt‘OJ
CVNNPUI

..
UK

Increased

‘TIIS IS A RUN OF THE MOCEL IN A PROJECTION NUDE-FROM 197C TC 1962.

FIGURE

1978 Actual

THIS LS A
.5795005
.S695+05
.5585005
.5“8£+05
.5366955
.5285905
.51SE+GS
.508E*95
.h975+05
.H875‘GS
.Q77EODS
.h675+85
.QSTE+GS

R

Z

[IIIIIIIOIIIIIIOIIIIOIIOIIC

ACTUAL (+) ANG

.<
[I

D
I)

CAST

HHHHPflPHPHHH
\DsOUDOsOOOkO‘OOQO
mmmVNmVVVNVV
NHOQ¢MmerwN

 
 

ll

 

73 8

{III
:1:

“I"
4W"
V<

U‘IUIUIU’IU’I Ultwnuun‘
VNVV“ ‘10“?‘Wm
O‘O‘O‘O‘O‘ O‘NU’IOJO‘O‘
PHHHH Hfllqtmrn
NNNNN N‘DQH NM
0 o o O o O
C If l!-;l )r‘A.)
t13(3|‘)t3|3‘3

O O o O o
rwfltnr‘u‘l

o
‘u

1
I

—1

['0

"I

'—

U;

92

63,033,

I

Cl

(‘)

‘mnr4uw

I Yuma “N401: 0‘
mm: NP: ""

‘fl“‘UlU.U1

VAtUdS FCQ

N3 N'DN p'C-mO-“JTU‘ A
HHLH—‘IP Ntnfl‘bflm *
'Lcfiu«4-VN v

VW‘J’TN'C

27.--Acres of Soybeans Harvested - U.S.,
Increased Yields Scenario.

—’-’-.-.-.u.- 9.- ¢ .

1000 Acres,

0 O O

lZ-An

-
\
u

ET

aIIIIIIIIIIIIIIIIIIIlltllilo»

\U

A]:

N
o

93

FIGURE 28.—-Soybeans Processed, Million Bushels, 1978 Actual =
1000, Increased Yields Scenario.

     

THIS IS A sun or THE MOEEL :N A PROJECTION nuns-5209 197: TO 1952
.11ae+cu 3 3 3 3 3 3 3 3 3 3 3” 3 3 3 ..... 3 3 3 3 3 3
.111£+nu 3 3 a 3
.107E+09 3 3 X 3
.103E+au 3 3 ,v” E
995. - - x -
959. 3 3 3
922. Z + + + + + Z
855. 3 3
658. : : 3
312. 3 3 3
775. 3 3 3
73a. 3 3 3
731. 3- ‘ _ _ - _3 ________ _ - _ - - ______ 3

O .00 g
72 7B 76 78 SC 52

ACTUAL <+> ANC ESTIMATE? (h) VALUE: Fa; 1a—c5usur

Yinv LiVELS
(6) (‘)
1972 721.3;2 736.76C
1973 821.3L3 599.53;
1979 7c1.3.3 737.22:
1975 665.;c2 636.267
1976 790.Lt§ 277.332
1977 927.;LL 9+;.t59
FORECAST
1978 92z.zgc )88.931
1979 92/.-.0 1L14.S:3
1380 92;.CQ3 LL32.U:5
tyU: 1L 4": l
193% 357.:35 1152.656

FIGURE 29.--Soybean Price, Dollars Per Bushel,

8.73
8.36
8.00
7.64
7.28
6391
6.55
6.19
5.82
5.k6
5.18
“.73
h.37

 
   

9

4

APPENDIX-D

INCREASED EXPORT SCENARIO

Actual

THIS is A RUN OF THE

M0

VI
0".

€'-—---
ACTUAL (*) AND

HHPHHWHHHHHH
\DQ“)\O\OOLOLO\D\O\O\O
wmmflwnwuwwww
mpmommqmmrum

I“

O
p
(I)
_g

o o o O o o
gamma; mammo-ou Ar’

mmmmm morom;
a o-Uma O‘HN r mfl

6.50,

ES

3 I
- f
I I
- I

- .I

: f

- I
-.J

9”.
m
r’

('irJC'n man V<

fvul

H)LJC3€

1978

Increased Export Scenario.

L IN A PgOJacrznn

II I. up an I' .. -‘ u. .n

h
\’

.__QI£..‘.JN A

r‘mUlNCJCJ
t J 1‘04"; :3

cum-4.1.” confirm F
.L‘,‘
'ooqd-NH

O C I C O
mrhwo
La\n."

‘1(_)
-

olllfillllllllliliillIIII'AIH

u.
0;

F.)
o

95

FIGURE 30.——Soybean Meal Production, 1000 Tons, 1978 Actual
= 23,754, Increased Export Scenario.

THIS IS A QUN O: THi k A “SL-‘:§CW .37 '3 9‘

.3115‘05
.299E+25
.287E+CS
.27SE+85
.263E+05
.2515‘05
.2395+C5
.227E‘05
.215E+CS
.203E+CE
.191E+05
.179E+05
.1675‘05

 

0 III l0. IlOI l8' l0. IIII '8'0 IIP

'o‘o'o‘o‘o‘ ‘9'0‘9‘0‘0‘9'u‘o‘9‘9‘u'c‘c'0‘.‘o‘.‘.‘
72 7k 76 78 fit 62
ACTUAL (*) AND ESTIMATL‘ (‘) VALUdS ;t€ Z-SCYVGT
YLAR Liqu:
1+) (‘)
1972 16703.16? L7251.f76
1973 1967+.iLC L9133.;.Q
1974 16782.;ti :695h.-~E
1975 237Eu.2£: ;o£e~.c~3
1970 1655602L: LCC310¢13
1977 22371.;25 £222c.849
FCSECAST
197d 22371..L5 35131.~Zr
1979 22371.;13 26391.*55
1983 22371.3;3 26132.52?
1931 22371.;23 19919171;
1982 22371.L2? 3:71.. 27

Ix

96

FIGURE 31.--Soybean Oil Production, Million Pounds, 1978

.1305*05
.1255+c5
.121E+C5
.116£+GS
.1115*GS
.107E+85
.1325+GS
.9725+cu
.9255+0u
.878E+8u
.331E+Ch
.78uE+0u
.7BTE+CL

HPHHHHHHHHPH
0.0.09.1“) 0.9.00 .0 4’)

THIS IS A RUN CF THE

.<
"I
p
n

 

O
b
.1

thwm‘i'JJDN‘IVVV ‘A
(I)

NHLhD'JvI Human LAN

Actual = 10,800, Increased Export Scenario.

HOEEL IN A DLUJccrznu "nos-5209 :9
- I

AND 'C I“ TLC (‘) VALUiS

(*) (‘)
75c:..;: 5.63.915
3995...: 9533..53
737505E: 7‘J9‘5025L
9630.8;2 9Lbécdlt
8578.;32 0582.768

1CZC’30.:: 3...:7602‘10
1328C02L: ;;;9‘éc‘561
123253.312: LLSZCKFIP.‘
15263.;.t 11166.83L
10256.;CC L159%.259
15288....3 1323-9. 59.3

0OIIDIOIOIIIIIIIDIOIIAI'AIIO"

97

150,

1979 Actual

Increased Export Scenario.

FIGURE 32.--Beginning Soybean Stocks, Million Bushels,
1978 Actual = 161,

o
A.,/—
.3

Q)
1.......................... .
.
SJ. 0
T. o
o o
... .
7. o
C. o
a... o
c
M. o
0 .
3.. 0
F .
.. 9
:— c
D.
U
H-
N.
H;
74‘-
v.
.In.-
.5.
J.
D
L..—
D-
.3.
N
...
L.
.L
0
C
H.
C...
H

N
~Haw:-.........-...o...—.......

An

S

T. O 0 O 1 1 9 O 0 0 0 O 0 O
1 6 1 O o o 9 1* 9 la. 9 h 9

s 2 Q 7 6 .1 3 2 1L 7 =4 2 c 7

I 3 2 1. 9 2 5 1 2 2 3 1.9 5 5

m u c c g c p c .

 

ACTUAL (+) ANC

(3)

«.0

S.

V)

.L...
L(

1. (D b 4-35;?»
77930.20
R 2....ddls7
O O O I 9 O
. ..~.\.u.1.ph..3
VIC/31.3.9
a .3753

a “5.3.. RUE
6 .70.?» .
Cibe. 6.1-
9 0 O O 0 O
4.9..U543
7S78“fi

14‘21

CAST

A..an.u:.u.2
741.....nhvz
3....-. u 6......
. . . . '
A55..33
gall—«.77.
1324.33.

n a -

ref-7.3:).
a-..LF...u-IL
.9... .85....TL
. Q C C 0
33333
GUCGU.U
4‘14L11

23... 557....89.u1.2
77777732/7 338
ggoaqggoggggo,
11.411113. 11111

98

FIGURE 33.—-Soybean Production, 1000 Bushels, 1978 Actual =
1,842,647, Increased Eprtg Scenario.

 

TrIs IS A RUN OF THE nccsL 1N A PROJECTIHH «uni-Faca 1971 TO 1:22.
.2115+os : - - - - - 3 - - ° "; - - - ' ° - - - ' ° - - - 3
.204£+os : 3 7’ :
.196£+us I I I I
.189£+e5 3 3 I 3
.181£+05 3 3 I I
.17wE+05 E - R.+/+ +-+ E
.166E‘05 I ’ ‘1’ :
.159E+as - -
.151s*os I I
.iuue+us : 3
.1sea+cs : 3
.1295+cs I I
.1215+cs : :

§5¥é?'53¥5'53035§5‘3?’I:§:Z‘"""""""""""
72 7a 76 78 a: :2

ACTUAL (+> ANE EEfEJZTEI (., VALUES Pu; 7-soveqr

YEAF LEVELS

(+) (*)
1972 1:736.3;3 1§026.761
1973 13971.65; 1231“.372
197k 121H6.;2~ Loagt.rsb
1975 15b61.2£5 Le239.63?
1976 12875.5L3 L§§47.27b
1977 17617.95: ;b>;+.573
FOQECAST
1978 17617.558 17691.52h
1979 17617.SSC L6271.01?
1985 17617.55; 1:3oi.b27
1981 17617.55; 2C532.783
1932 17617.55L 2;933.913

99

FIGURE 34.--Acres of Soybeans Harvested - Southern U.S.,
1000 Acres, 1978 Actual = 15,560, Increased
Exports Scenario.

H

to

THIS IS A QUN OF THE MC L'L IV A PLOJELTION MOD£‘F€OM 197;

- c + f + ++. no- -o-u . o c

0'. I... "lil OOIII IOIIIIIIOIIF‘

 

01““5‘05 ‘ ’

01“15*05 ‘ I

.135£+35 - //

.1355+as — /

.1325+es - f ’
- \ I

01295*05 " v

0126E+05 ‘

0123E+05 ‘

01295*05 -

0117E+05 ’

ollhE+U5 °

01115705 ‘

.108E+35 '

7d .53 .65. . O O Q 9 § 0 O O
ACTUAL (H AH: 3755.? (L) mu -3 7:2: 1-:.+~.::':
YEA= LiVELS .
(+) (3)

1972 1119).-.: ;;29;.73_
1973 127360aLC 11:33.61?
197k 121€8.2c5 ;1§3€.Zlq
1975 13483QL$3 L193Lo£§1
1976 lZQIZOELC 121330593
1977 1&SQ3cdaC 12:32.95;
FORECAST ‘ -
1975 1H593oCCC 13337.392
1979 1&593oLLC L2842c736
1935 1Q59301L3 L35-70311
1931 1&593.;;c ;3o;s.79b
1982 14593.32: 1+312.2:1

w
J

ru
0

100

FIGURE 35.-—Acres of Soybeans Harvested — U.S., 1000 Acres,
1978 Actual = 63,033, Increased Export Scenario.

THIS IS A FUN 0: THE «0:51 IN A PLUJLtrzuw Moos-=:0w
.5955+cs - - ' - - - - - - ’ - - ' ' - D ' ' -
.5868705
.57us+os
.562E+CS

.5515‘95

I"
I)
IN
‘1
-I
It.)

.485E+C5
.R6QE+85

.057E+CS

.5395+95 }
.527E+os 3
.51ss+cs 3
.saue+os I
.u925+05 I

‘OOIIIIIIOIlllllllll'lllllir‘

 

ACTUAL 1+) ANC 'E?f75?:3 1*) VALUES Fri 12ozn>37
vra. -LVFLS

1+) (1)
1972 456E3 as: 623’ 5~
1373 55567111‘ 9»:»:1$1$
137g a13u1.c13 geeze.:53
1977 53579.;1: v052-.C7E
12;; 33355..;_ 5:;52.319
. 9 612 ;L: a ; h'a
£27964“ 5, ° "
. 61°.-1: 55'77 :5a
1979 57612.:;: 5:33:1«27
195s §Ie12.e:: 57-32.o1~
1331 97612.11? 99365.53h
1582 57512.::: o;3»7.73;

101

FIGURE

Actual 1000,

   
   

THIS IS A ?UN OF THE MOCEL LN L
oiZlE‘Cl‘ ' '
OIIBE‘CQ - O i

- - I
o1125+fih - - /
- - ’
01C7€+C§ " - /
.1czs+eu - - /
977. - C
931. ‘ f
- + + 4* +
885. -
339. - -
793. - .
7A7. - c
711. - -
'o'o".-o' -l-D-0-0-1
72 7A 76 78 :9
JCIUAL (+) ANC £STIVA’
YEA? LaliLS
(*)
1972 721.6»:
1973 621.3L3
197k 751.3L3
1975 865.;LC
19‘s 79c.gc:
1977 927.4;5
FG?ECLST
1978 927.555
1979 927.;LT
igfl; 927.-il
1931 9§7o2tE
1982 9 7.2L:

36.-—Soybeans Processed, Million Bushels,

1978
Increased Export Scenario.

'4
|(‘

+

£5 (‘) VALUES F0? 18‘CEUSHT

A
I
‘1

Val“ #11; (AV! ‘N-l v4
O-muuum no.3," 1,; NU]

NJ-‘Iv‘u‘l unicorn; G-r
mama-UN NUINVU‘H
rrmwr: u ammo-DA

°IOIIIIIIIIIIIIIOIoIIIOII.“-

\U
u'

N
o

102

APPENDIX E

REDUCED EXPORTS SCENARIO

FIGURE 37.--Soybean Price, Dollars Per Bushel, 1978
Actual = 6.50, Reduced Exports Scenario.

THIS IS 1 Run OF THE MODEL 1' A PROJECTICF HCDE-FPOM 1971 TO 1982.
7,37 2 - ' ' - - I ' ' - - ' ' - ' ' ' - ' ' - ' ' ' ' I
7.12 : : / :
6.87 : Z [I :
6.52 3 3 / I
5.37 : I ,’ E
6.12 - - -
5.87 I + + + + + :
5.62 : I
5.37 3 3
5.12 I 'Z
4.87 I :
4.52 E 3
4.3 - :

-.::§-.;é-.-.-.-.-.-.-.-.-.-.-.-.-.-.

 

YF~R LEVELS

(+) (.)
1972 4.376 4.7"0
1973 5.680 5.000
l9?“ 6.643 $.123
1975 5.910 6.066
1977 5.383 6.314
§2§§CLST

' 5.88? .1

197: 5.88” 2.é%§
1990 5.951 7.514
1361 5.é83 7.286
19r2 5.88) 7.491

1982 .

TC

A
y.

1978 Actual

197

MCDE ‘F'R 3‘1

1000 Tons,

PROJECTION

103
A

I i
\I'

I

= 23,754, Reduced Exports Scenario.

CF THE VOCEL

FIGURE 38.--Soybean Meal Production,
A :Ln

THIS IS

 

T!
Q
.1...
V:
C
S
.
9..
F.
C
F
c...
E \: fincruqénéo“ 77U49
U .. 7.0.4.419 23448
L l. 31489.56 5380.1
.n» o o o o . o o o o o a
V. 1.3.4418. 5784b
635.402 3367...“.
23967.19. GHQ/p.94.
7 Q, 68 E 4... Q3174 C
11.1....34 7.659749.
+
.7 «b
L
+ .1.
V ) .U. u 0...... r. .nt. .1» “H . 7. 3.4,
+ .:+ C..u»u3r..:U 905...”.3
Ll. 533.6...an 7.0.3.05.
+ o o o o o o o o o o 0
87424.51 111.11.
373597 777.77
767743 2.33.533
6.3.5:! Rid Grin/.92..
1 .15 110/. n... rd 56.9.4111— p13;
)
+
. l. 7..
a...
. .L A
.A C
. U R 234.....67E8c2U12
T. .n 777777P77SRE
......................._.. C .L c..90;99c1h990.99
_ .H. Y. 1111.....1F1111I.
c... r... .9 r? .U .3 a. 5 “3 a 5 3 E
.b 2. .b no .1”. .2... H. ....... no 3 .U a .2.
+ 4 + + + + + + + + + + +
EE.LEE.L.L.L.LE.LEE
1..:493.In..h..526937
432211099577..c
27.22222111111
o o o o o o o o o o o o o

FIGURE 39.-—Soybean Oil Production, Million Pounds, 1978
Actual = 10,800, Reduced Exports Scenario.

THIS IS A RUE OF THE MC EL IN_ A CRC OJECTIE: CE
.IIEE+;: H’
.11«E+L: I

30!‘ 1970 T0

.111E*G§
.1E7E+ub
.123€+55
.9SSE*QQ
.959E+C’
.9225+34
.885E+C4
.848E+34
.811E+94
.774E+5Q
.737E+;4

‘\
I I I I I I l I I I I I I I I l I I I I I I I I I I
I I I I I I I I I I I I I I I I I I I I I I I I I I P‘

 

(+) (*I
1°72 7501.030 tbe.C15
1973 8999.035 9603.¢53
1974 757$.L0: 7969.986
1575 £630.30} 3356.c15
1%76 5378.302 8585.758
1377 1:2"3.035 1 276.316
FORECAST
197b 1L2F8.00J liCSéo8£3
197? 13288.38; 17112.98?
19EU 1725&.OCS 19457.735
1°51 122?E.QOQ 1:723.592
1982 17288.80; 11¢?l-33‘

If)
N
I

TnIS

.I'LJE‘I‘E‘I

FIGURE

18

GE .
61 .
C?

J... .
45 .
37 .
295.
dlbo

A

9

‘.
I.

IIIIIIIIIIIIIIIIIIIIIIIIIIC

“wqwxhrq

.D‘D\0\D\I")\D\P'~I D DMD

'1‘ (1‘. 1's a} ‘1'1'

f'~)>4(‘: 41mm 4 mm 1: (mo

HHbJrJHTP—‘HP'F—‘Hy.fi

0‘

(‘1
I)

105

161,

1979 Actual

Reduced Exports Scenario.

TFE

   

In
.4

IIIII}IIIIIIIIIIIIIIIIII

q L'Txl

(1‘ 4p IJ‘IP'Dr-h

1):}.

’:(‘f'

I
(r1011)! (,nLu
o o o o o

HHHHH HRH-J”

{.3

”CiiL Id
--

O I O O O O
uIL"L.-GJU"~D
00 #35451)

(..IL:I(IL."L) L'L'JC'CJO'..-

(--(‘.- I“! :._:IC'
(41)?) (fif‘)

I)

1)

‘ul’

[*1

ICW

"o’o‘o'o'o‘o".
W2

5’ (*) VALUIS
(*)

71.874
58.276
184.994
313.631
295.593
358.703
303.578
384.645
471.351
(355-0149
l:#2.c98

40.--Beginning Soybean Stocks, Million Bushels,
1978 Actual

150,

TICh MDCE-FPCM 1973 TO

1992.

106

11.
a
u
t
C
A
8
7
9 o
1.0
i
.r
S a
1.n
e e
huC
88
u
B S
t
0?;
0.0
DVD.
1.x
E
I:H
nau
0 e
.l G
+.u
mam
AuR
O
r I
Pal
4.
nrO
a I
6A4
h~4
YB
O I
31
_
_
l
4
U
G
I
F

1970 TO 1982.

OF THE MODEL IN

RUN

A

THIS IS

   

n‘4-‘3‘5

'0-

.3
..Y..
¢
.7...
I...
G)
1

o

2.
N...

.187E+
.186E+

Q o
o. .,

'1

.17“E+?
.1675+35

.161E*'J‘3

-
”d
,.

olafiE*

5
3
+
7..
1;
q
1

O

.r.-.
\J
+
.b
3 u
3
.1

o

.
. T
_ Au.
0 .....l
. V:
o ,\J
. S
o .
. 7.
o
u
o
_ O.
o ,J.
. F
o
. C)
o F. \l
- U r
0 L l.\
. A»
a V
. C
O O \I
— ..L t
Gnu l\
. .l.
o 4... ad
- V. n). .:
3b «I S
_ .H .L
. .L .3 —.. .L
. FL 9... .1 V.)
.0 vi pa+
- 1.... .5 IL!\
o M. ,b qr.
. I7
0 _.u h...
. r...
0 A. R. A
. v17
.A-H )
. nu. +
0 “v I.\
. .7.
0E .L
. TI A
9..» .2 U P.
. K47 T. {A
. — . . C ..r_
.A Vo
Cu. ".3
3 ....J
+ +
.L 9:
9.. 1..
9. 2
1 1L
0 0

12.07.63
“07337.0
7455.614
0 O O O O 0
8931...!“
9.19541
AVG/Ranger?
0.2.13“ [4 r5
110A1Llol

._u.b..u.:..._.,b.U
052325
.J 037.52.
0 O O O o 0
51.5157.
.L7h». C71
770.148»?
achCbbndgl
14A..LqL.IL}

.U7.n.9nu
84.0.55
915912o
o o o o 0
5.5597:
833.10.
€17.59.
7. 7 7. 9
111.11.

«by. W n .mUfi a . U
Sad—3.5:,»
5:1..WVC;_H;
. . . C .
77777
.19‘11...
656 0,9
77777
.l“.ol.lo.~

9.34 R... .075 89 51.2,.
7777.77 2:7 789...»;
0.99.9 0.9 09 9Q. G.C..
1 «1.11Q81F19L111‘

107

FIGURE 42.--Acres of Soybeans Harvested - Southern U.S.,
1000 Acres, 1978 Actual = 15,560, Reduced
Exports Scenario.

   
 

THIS IS A FUN OF THE NODFL IN A RDGJECTISN MGDE-FnO“ 1973 TO
- - - - - — + + + + + - - — - - - - — - - — - -
.144E+55 -
0141E*£5 '
.138E+¢5 - -
.1355+u§ - .‘ I,
- - -¢
4320:: - - 1‘ ,r
- -1 \I
.1295+05 - -( ‘
.1"~s€+=5 - J
4 L. - J
0123E*C5 ‘ '
olZCE‘OS ’ ‘
.117E+US e -
011aE*:}S " ‘
0111E*35 - I '-
.1c3€:+&5 - l -
7§ 07‘: 07:: .75 ORE 0R5- O O O O O O O O O O O O 6
fl —- - --
ACTUAL (+) AML tSTlflATEu (r) anUES F39 ll-AHSSTS

YE'P LEVELS

(+) (t)
1972 11190.C9‘ 1 792.751
1973 12736.C03 11935.616
1974~ 1;198.COU 115069715
1975 13453.JCC 11934.391
197E 12992.35C 12483.55}
1977 14595.020 1L592.884
FQFECAST
197$ 14593.UCQ 13537.952
1979 19593.33} lbq51.158
1993 14593.003 13384.627
19511 14593.353 13482.3(."
1982 14593.000 11793.155

D
L13

n
O

0llllIllllll'lllllIIIIIIIIIH

1000 Acres,

Reduced Exoorts Scenario.

-

108
63,033,

FIGURE 43.——Acres of Soybeans Harvested - U.S.,
1978 Actual

TO 1982.

. o
. T
. o W”
n.» . Q.
7. . H
9 — A
l. a _
_ 7..
M. o 1
no .
on. o
F . R
_ . a 0...
E . F
n... I
O . s
M. . E ) 20.3595 93.381
. U .2 00137.15 83:329..
.w. . L ( 575C774. .uuryciE;
n... _ A o I u o g u o o o c c
I. . V. 723428 77832
T . 542253 25442
C. . \I n.....¢:.:.11 54.2.55
r... . t 5.. 451...“.20 3.22.727
d. . ( 4:44.955 55.355
0 .
a. + a 3. —D
D I _ u wrt
Q.
— L
FL
V.) U.JF.:¥.4«IJ 3nd..u.U.J

Eo. QCOQCO .UCQCU
Ll. CCOCSE 0003C
I O O 0 o a O O O l O
Iu71oéf2 6422):...—
o 64.7.24. 117.11
663.335 56 £56
5513.97 77777
u .3c_5hc, 5:..SRID

 

T.
r...
A
C
D. 234.337,...89312
T L 777777F7.79P;.H,
L. _ . . . . . . . . . . _ . . . . _ . . . . . . . . C .L 9999 015699.99...)
on A v. 1.1.1111F11111
.2 :1 .3 z .3 5 a r 5 3 5 5 5 5
A J 3 3 J J O 3 .1. J a a). n. .3
+ + + + + + + + + + + + +
S .L .r. .L E E E E r. E ..r_ E E F.
I 1 .1 .J 3 .J .7 9 9 8 8 8 7 7
C 7 a c. 9 n... l U 9 .0 7 .,D 5
s .3 r... p: :4 G. 5 r3 5 a. a 4. h. 14
I C o O I O O I O O 0 O I O
H
T

‘3
1..

TO 193

'3
J

C?

1

MGDE-CROM

CTION

L“

Reduced Exports Scenario.
350d-

1
F

109
1000,
A

Actual
MODEL IN

OF THE

FIGURE 44.—-Soybeans Processed, Million Bushels, 1978
Rbt

THIS IS A

. a TI
. H
. o S
. .U
. o R
. C
. o .
. D.
. . 1..
.
. o
. - «fl
. c "U
_ F
_ o
_ S
. . .L ) 399.530 523.53.
_ U i 1.:7ficrsac 77347
. . L C 4689 :1: 14531
— A o o o o o o o n o o c
. o v 57417vC4 r2793???
— 7.749.173 .erQunTb
. n \: _fu(wl.cd.lc,. Quiuxr?
. t 11 ..
_ a (\
.
i + 0%... — mu
. A. .L
. l + . .T S
— —\» L
. / § 0 L r... Ct.
— .H —.I. V) 3...».0 .u.urL JIUVJHJTU
. d + . .l 5+ 0.....JQLF; .UCCCC
. _.b L( 91.71718... .JCJS...
. r + .b r: .o.... a...-
.7 111537 77777

  

....... _. ___ . _ _ _ _ — . _ _ . . D can/7.)! 0.2 "(ARI/:92
_ .ww 787879 c.3999
. .5 A
.7
c I \l
_ +
a I!“ l\ T
.7 S
_ o L A
_ A FL
_ 05. U D. 2174:757C_QHC.C19..
. .7 T. .u 77777.7—fi.l7.h.nr.h
..___..._.....-.......—... C _,r. 99990/909cagc,9
IA Y. 1.11s111F1A1.111
u .4.
U .3
o O
E .L
r? at o o o c o o o o a I c
nuu2.3n~5..u789.UU1
1 1 9 b 3 0 7 4. 1. 8 ,b 3 0
o o a; O 9 9 8 Ho 8 7 7 7 7