• TABLE OF CONTENTS
HIDE
 Copyright
 Title Page
 Table of Contents
 Introduction
 Simulation, industrial dynamics,...
 Dynamo model of the frozen concentrated...
 Implications of study
 Validation of model, simulation...
 Reference






Group Title: Agricultural Economics report 1
Title: Grower-oriented supply and marketing policies for frozen concentrated orange juice
CITATION PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00027476/00001
 Material Information
Title: Grower-oriented supply and marketing policies for frozen concentrated orange juice
Physical Description: 46 leaves : ill. ; 28 cm.
Language: English
Creator: Raulerson, Richard C
Publisher: Dept. of Agricultural Economics, Florida Agricultural Experiment Stations, Institute of Food and Agricultural Sciences, University of Florida
Place of Publication: Gainesville
Publication Date: 1969
 Subjects
Subject: Frozen concentrated orange juice -- Florida   ( lcsh )
Frozen concentrated orange juice -- Marketing -- Florida   ( lcsh )
Genre: bibliography   ( marcgt )
statistics   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
non-fiction   ( marcgt )
 Notes
Statement of Responsibility: Richard C. Raulerson.
Bibliography: Bibliography: leaf 46.
General Note: Cover title.
General Note: Based on the author's thesis (M.S.), University of Florida, 1967.
Funding: Florida Historical Agriculture and Rural Life
Bibliography: Agricultural economics report - University of Florida Dept. of Agricultural Economics ; no. 1
 Record Information
Bibliographic ID: UF00027476
Volume ID: VID00001
Source Institution: Marston Science Library, George A. Smathers Libraries, University of Florida
Holding Location: Florida Agricultural Experiment Station, Florida Cooperative Extension Service, Florida Department of Agriculture and Consumer Services, and the Engineering and Industrial Experiment Station; Institute for Food and Agricultural Services (IFAS), University of Florida
Rights Management: All rights reserved, Board of Trustees of the University of Florida
Resource Identifier: aleph - 001615844
oclc - 03239001
notis - AHP0287
lccn - 77359963 //r90

Table of Contents
    Copyright
        Copyright
    Title Page
        Title
    Table of Contents
        Page i
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
    Simulation, industrial dynamics, and dynamo
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
    Dynamo model of the frozen concentrated orange juice industry
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
    Implications of study
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
    Validation of model, simulation results, and analysis of policies
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
    Reference
        Page 46
Full Text





HISTORIC NOTE


The publications in this collection do
not reflect current scientific knowledge
or recommendations. These texts
represent the historic publishing
record of the Institute for Food and
Agricultural Sciences and should be
used only to trace the historic work of
the Institute and its staff. Current IFAS
research may be found on the
Electronic Data Information Source
(EDIS)

site maintained by the Florida
Cooperative Extension Service.






Copyright 2005, Board of Trustees, University
of Florida





NJovember, 1969
p4,/


Ag. Econ. Report 1


Grower-Oriented Supply and Marketing

Policies for Frozen Concentrated

Orange Juice


Department of Agricultural Economics
Florida Agricultural Experiment Stations
Institute of Food and Agricultural Sciences
University of Florida, Gainesville


Richard C. Raulerson


I I













Table of Contents

Page


INTRODUCTION . . . . . . 1

The Problem .... . . . . . 1
Approach to the Problem . . . . ... .. 3

SIMULATION, INDUSTRIAL DYNAMICS, AND DYNAMO . . . 5

Information-feedback Loops and Simulation ...... ... 6
Industrial Dynamics and DYNAMO . . . . .. 10

Model Building . . . . . 10
Model Testing and Revision . . . ... 12

DYNAMO MODEL OF THE FROZEN CONCENTRATED ORANGE JUICE INDUSTRY .. 14

VALIDATION OF MODEL, SIMULATION RESULTS, AND ANALYSIS OF POLICIES 22

Validation of Model . . . . ... .... 22
Results Using Free Market Policy . . . ... 25
Analysis of Alternative Policies . . . .. 29

Method of Comparing . . . . ... 29
Comparison Using Weather Two . . . ... 29
Comparison Using Weather Six . . ... .... 35
Relative Costs of Alternative Policies . . ... 36

Consideration of a Physical Pool of Concentrate . .. .. 37

IMPLICATIONS OF STUDY . . . ... .. ....... 38

Implications of Study for the Industry . .. . 38
Implications for Further Research . . . ... 44















Grower-Oriented Supply and Marketing Policies for
Frozen Concentrated Orange Juice*

Richard C. Raulerson


INTRODUCTION

This study was concerned with the general problem of fluctuating orange

supplies and grower profits in the frozen concentrated orange juice (FCOJ)

industry. IMore specifically, a second-generation DYNAMO model was devel-

oped and used to simulate industry behavior under alternative supply and

market control policies. These policies were then compared as to their

ability to stabilize orange supplies and grower profits2 at acceptable

levels.


The Problem

The FCOJ industry has been characterized by large shifts in supply--

particularly during the 1957-58 and 1962-63 crop years. During these years,

low supplies caused by severe freezes led to high levels of grower profits.


*This report is based on an M.S. thesis by the author [8] under a grant
from Florida Citrus Mutual, Lakeland, Florida. Richard C. Raulerson is a
graduate student at the University of Minnesota, St. Paul, Miinnesota.

W. E. Jarmain, [4], developed the original version of the model.

2The term "grower profits" as used in this study refers to the net
returns above average estimated cost of production of $.85 per box.
3
"An acceptable level" will vary among individual growers. For the
purposes of this study a level of grower profits of $.15 per box was speci-
fied as being acceptable. This assumption could, of course, be varied in
the model.











These high profit levels induced large investments in orange groves which

raised supplies a few years later. These supplies of oranges in turn

created a period of low prices and lower grower profits.5

Data collected by the Market Research Corporation of America indicate

that during a period of restricted supplies of FCOJ a drop in sales occurs

as a result of individual housewives buying less orange juice and fewer

housewives purchasing FCOJ. In either case, further data indicate that

consumers turn to alternate products at those times [9]. Industry leaders

are concerned that some consumers, once they switch to an alternate product,

may not return to FCOJ when ample supplies are again available at lower

prices.

Reduced marketing efficiency resulting from orange supply fluctuations

raises the average retail price for FCOJ. Thus, the grower's share of the

consumer's dollar is reduced since the housewife pays relatively more for

marketing services and less for FCOJ.

Fluctuating grower returns increase the effort and cost of proper

grove management. Further, growers usually considered efficient may be

forced into selling their groves due to a period of lower than normal

grower returns.


4"High profit levels" refers to total industry profits without regard
to the distribution of these profits. It is recognized that grower profits
are not evenly distributed.

5Since the price elasticity of demand for oranges on-tree is inelastic,
total revenue for growers decreases as the quantity of oranges increases.










Approach to the Problem

The problem was studied with a DYNAMO simulation model. To validate

the model as a representation of the FCOJ industry, the separate industry

components were examined and then simulated model results were compared to

actual industry behavior.

After validation, the model was used to test the effectiveness of the

following policies in stabilizing orange supplies and grower profits:

1. Free market policy. The FCOJ industry was simulated with no

supply controls.

2. Allocation of FCOJ to two separate markets. This policy

assumed that two separable markets for FCOJ either existed or

could be established. The economic reasoning behind this

policy was that if different price elasticities of demand

existed in two different markets, then total revenue could

possibly be increased in the short run by shifting FCOJ from

the market with the smaller (in an absolute sense) elasticity

coefficient to the market with the higher elasticity coeffi-

cient. In this study the two markets were designated as

primary and secondary markets, respectively. The normal

retail market was considered to be the primary market, while

such possible markets as school lunch, vending machines, and

domestic food program were considered together as the secon-

dary market. All FCOJ was considered sold in the primary

market when grower profits were above $.15 per box. When

profits were below this level, an amount of FCOJ necessary to


Simulation in this study refers to a method of translating dynamic
processes into mathematical relationships and then observing, through the
use of a digital computer, the effects of these relationships on the
variables of the system over time.










raise profits to $.15 per box was shifted to the secondary

market. It was assumed that FCOJ was sold at cost and that

entry to and exit from the secondary market was possible at

any time. Further, the demand was assumed to be perfectly

elastic at the secondary market price.

3. Elimination of fully productive trees. This policy was

designed to stabilize profits above the $.15 per box level.

To accomplish this goal, a certain portion of the existing

productive trees were destroyed whenever grower profits

dropped below $.15 per box. When the policy was effective,

trees were removed at a rate which would eventually raise

grower profits to $1.00 per box. However, the policy would

never actually bring profits to that level since the policy

was no longer effective when grower profits rose above $.15

per box.

4. Curtailment of new tree plantings. This policy was formulated

so that no new orange trees could be planted whenever grower

profits were above $1.00 per box. The rationale behind this

policy was derived from the known tendency of orange growers

to overinvest in new tree plantings when grower profits were

at an unusually high level (e.g., following a freeze when

supplies were very low and prices were high). Due to the

nature of its control over supplies, this policy was perhaps

more forward looking than the other policies. Instead of

representing an attempt to correct a low grower profit situa-

tion, the policy was designed to help keep such a low profit

situation from developing.










5. Allocation of FCOJ to two separate markets used in conjunction

with elimination of fully productive trees.

6. Allocation of FCOJ to two separate markets used in conjunction

with curtailment of new tree plantings.

In order to test the effectiveness of the various policy decisions

under different weather conditions, each policy was simulated with two sets

of randomly selected weather conditions. To assure that industry behavior

was a result of policy decisions and not just weather, the same two sets of

weather conditions were used for all policy decisions.


SIMULATION, INDUSTRIAL DYNAMICS, AND DYNAMO

Economists have long been aware of dynamic problems8 but only recently

have they begun confronting these problems empirically with mathematical

programming and simulation.


The free market policy was simulated six times with different sets of
random weather conditions to see if similar patterns of industry behavior
occurred under different weather assumptions. Refer to page 19 and Table 1
for a discussion and presentation of the random weather effects used in
this study.

For example, both the model of economic stagnation as developed by
early classical economists and the later development of the cobweb model by
Ezekiel [2] employ dynamic concepts. Lindahl [5, pp. 38-39] described
period analysis, which captures the essential elements of simulation:

Starting from the plans and the external conditions valid at the
initial point of time, we first have to deduce the development
that will be the result of these data for a certain period forward
during which no relevant changes in the plans are assumed to
occur. Next we have to investigate how far the development during
this first period--involving as it must various surprises for the
economic subjects--will force them to revise their plans of action
for the future, the principles for such a revision being assumed
to be included in the data of the problem. And since on this
basis the development during the second period is determined in
the same manner as before, fresh deductions must be made con-
cerning the plans for the third period, and so on.










Forrester [3] credits the discovery of simulation to two major develop-

ments--the concept of servomechanisms or information-feedback loops which

is essential for the understanding of time-linked, interdependent variables,

and the electronic digital computer which has greatly reduced the cost of

the mathematical calculations.


Information-feedback Loops and Simulation

The emphasis in simulation is placed upon time-linked decision-making

processes which contain information-feedback loops. An information-

feedback loop exists whenever the environment leads to a decision that

results in action which affects the environment and thereby influences

future decisions. Critical to these information-feedback loops are the

various amplifications and delays present in a system.

For an example of a simple information-feedback loop consider a man

driving an automobile (Figure la). His function is to operate the vehicle

in such a manner as to guide it on a reasonable course down the highway.

If he knows his vehicle well, he can successfully operate it. His eye

senses the vehicle's direction and if it is wrong, he decides, after a

delay, to alter his course. After another slight delay his hand turns the

steering wheel. This action is amplified by the steering gear, the wheels

turn, and after another delay he senses the vehicle moving in a different

direction. The process in continually repeated until the car comes to a

halt. The path of the car might resemble that in Figure lb. Suppose,

however, that the driver had to steer the car by looking out the back

window and deciding his course of action by where he had been and not

where he was going (Figure Ic). The path of the car would probably be

quite erratic, and if he tried to negotiate a bad corner (which, of course,












Mechanical
delay

0


Decision
delay


Amplification


Figure la. Typical information-feedback loop.


------ Actual path
Desired path

^o-^ -"--^^^ "'^^


Figure lb. Path of car with driver looking ahead.


..r Actual path
.__ Desired path

-- -. \ c _


Figure Ic. Path of car with driver observing rear-view mirror.


.e











he cannot see approaching), he would probably have an accident. This may

not be the best way to drive a car, but managers and economists are faced

with a similar situation. They have to rely on past data and face an

uncertain future.

Citrus management personnel, while looking out the "back window" of

their "vehicle," the Florida citrus industry, tried to negotiate a "bad

corner" in the form of the 1962-63 freeze. They subsequently decided to

invest heavily in new orange tree plantings. Consequently, an accident is

imminent if these plantings yield supplies of oranges that cannot be dis-

posed of through existing marketing channels at acceptable prices.

Another example of an information-feedback loop is the intra-year

pricing of FCOJ shown in Figure 2. The object here is to arrive at an

equilibrium price. However, due to the delays in the system the actual

price fluctuates about the equilibrium price. The components of the

information-feedback loops are connected by material flows and information

flows. In the figure, flows of FCOJ are shown as solid lines. This flow

begins with growing and processing activities. The rate at which FCOJ

becomes part of processed inventory is symbolized as an hourglass-shaped

valve. Processed inventory moves into retail inventory at another rate.

Then consumption removes FCOJ from retail inventory at still another rate.

Information flows, upon which the pricing of FCOJ depends, are shown as

broken lines. An F.O.B. price is set by processors after a delay while

they consider the level of processed inventory and the rate at which it is

disappearing. This F.O.B. price dictates a certain retail price. After a

delay in discovering this retail price, consumers adjust their rate of

buying according to their demand schedule. This reduces retail inventory.























Delay
C- --
- -


Processor Consumer --
(1Delay Delay
behavior behavior



I i




F.O.B. Retail
price ... "--- price









Price
1 Actual price





Equilibrium
price
_- ---------------------------.-.. --------------




Time


Figure 2. Intra-year pricing of frozen concentrated orange juice.










Then after a pipeline delay of ordering and shipping, FCOJ moves from pro-

cessed to retail inventory at some given rate. At this time the processors

reevaluate the level of inventory and rate of disappearance to consider a

new F.O.B. price. Then the pricing process is repeated. The equilibrium

price, shown to be stable in Figure 2, will probably never be reached

because the pricing process is continuous. That is, while processors are

debating inventories and movements, consumers are still reacting to the old

price. By the time the new price reaches the consumer the inventories and

rates will have changed and a new price will be forthcoming.


Industrial Dynamics and DYNAMO
9
This study used the approach of industrial dynamics, which basically

relates the information-feedback characteristics of dynamic (time-varying)

organizations or organizational structures to their performance. The

approach shows how the structure, amplification of policy decisions, and

time delays in decisions and actions interact to influence the performance

variables. Industrial dynamics utilizes the DYNAMO computer program [7].


Model Building

The first step in building an industrial dynamics model is to identify

a problem. Next, the system exhibiting the problem must be investigated to

ascertain whether or not it contains information-feedback loops. If it does,

simulation may be appropriate, and the factors that appear to interact to

create the observed behavior of the system must be isolated. This is

accomplished by delineating the information-feedback loops within the system.


Refer to Forrester [3] for a complete description of the approach of
industrial dynamics.










A flow diagram may help by giving a clear view of the important cause and

effect relationships in the system. It also gives a compact overall view

of the system and points the way for data collection.

At this time decision policies that describe how the decision-makers

utilize available information are formulated. This step involves the

transformation of the components of information-feedback loops into vari-

ables and the available information into coefficients.1

The DYNAMO simulation model is constructed by writing equations that

link all the variables in the manner consistent with the flow diagrams.

Any delays inherent in the information-feedback loops must be included in

the mathematical model since delays are a principle source of unexpected or

undesirable system behavior. Simulated system behavior over time is then

obtained with the aid of a digital computer to execute the necessary calcu-

lations. The simulated behavior is then compared with the known historical

behavior of the system. This comparison is made to determine the validity

of the simulation model as a representation of the real world. No model will

be perfect. Some divergence between the real world and model behavior must

be expected because of the improbability of being able to include all the

relevant variables and specify the exact magnitude of their effects.


1Since it is assumed that the ideal simulation model should contain
all the significant variables in the system, a lack of data concerning any
variable will of necessity require the assignment of a somewhat arbitrary
coefficient to,that variable. This action may not be fatal to the model.
It is more important that the variable should not be left out of the model.
There is probably some range of values into which a coefficient must fall
and eliminating the variable from the model is equivalent to letting the
coefficient be zero, one value which is known to be wrong. Actually, the
model may be fairly insensitive to a range of values for specific coeffi-
cients and model results will not be adversely affected by a "ball park"
estimate, whereas an estimate of zero would possibly yield poor results.










Model Testing and Revision

In determining the validity of a simulation model, primary interest is

in the nature of the system being simulated, and the nature of the system

is revealed by the cyclical nature of the variables over time. Specifically,

the period and amplitude of the cycles encountered are important. Since

the simulated behavior is generated over time, one would possibly be in-

clined to compare simulated and real behavior at specific points in time.

Such a procedure could cause the researcher to discard an excellent model.

Application of the time-comparison criteria to the case in Figure 3 in

order to choose between alternative simulation models reveals that simula-

tion I is closer to the actual behavior than is simulation II. However, it

is obvious that simulation I gives no indication of the true behavior of

the variable. On the other hand the period and amplitude of the cycles

exhibited by simulation II are nearly identical to the period and amplitude

of the cycles actually observed. In other words, simulation II is quite

successful in revealing the nature of the system (assuming of course that

similar results were obtained for the other system variables). Another

test of model validity is the confidence placed in the individual components

of the simulation model. This "test" is particularly valuable when



/ \ (Actual Behavior
K \ \ "/
S\% Simulation I

?, 9 <

Simulation II




0
Time


Figure 3. Actual and simulated behavior of a hypothetical variable
over time.









historical data are not available or when the structure of the system being

studied has recently undergone such drastic changes that the historical

data are of little value.

If the simulation model produces unacceptable results, then each

component must be carefully examined to see where possible flaws exist. If

a coefficient has been assigned an arbitrary value from a probable range of

values then two possible courses of action exist. If the expense of

obtaining an accurate value is small, then this estimate should be obtained.

Or, if this endeavor is quite expensive, then the model may be subjected to

sensitivity analysis. This process involves further simulation runs with

one or more arbitrary but plausible values for the coefficient. If the

model is quite sensitive to different values, then the researcher must

undergo the expense of obtaining a good estimate or accept that portion of

the model as it is. The latter decision may not endanger the research

effort as will be shown later. If all coefficients are acceptable and

simulation results are still poor, then the information-feedback loops must

be examined and revised as necessary. The process of revising the model

continues until the model is an acceptable representation of the real world.

Such an approach must be used with care, in order to avoid introducing

extraneous variables or coefficients into the model to achieve purported sys-

tem behavior. While it is true that desirable results must occur as a

result of a correct model, it must also be observed that desirable results

can occur as a result of unwittingly arranged system components. For
example, 23 = 8, but 2 x 4, 4 + 4, and 16 2 also = 8.
example, 2 = 8, but 2 x 4, 4 + 4, and 16 -4 2 also = 8.









When the model has been validated model experiments are performed.

The process of experimenting involves altering the simulation model in
11
order to discover changes that will improve actual system behavior. In

the case where simulation results do not conform to actual system behavior,

the model may still be useful. Suppose, for example, that meaningful

coefficients for a particular segment of the model cannot be obtained due

to great expense or other reasons. Assume further that sensitivity analysis

has shown the model to be quite responsive to different values for those
12
coefficients. If the remainder of the model appears adequate,2 it may be

possible to redesign the sensitive portion to produce simulated behavior

which, when compared to actual system behavior, may reveal that the

simulated results more closely fulfill the criteria by which the system is

appraised. The logical conclusion of this procedure would be to follow

through by redesigning the corresponding portion of the actual system.


DYNAMO MODEL OF THE FROZEN CONCENTRATED
ORANGE JUICE INDUSTRY

The FCOJ industry and the DYNAMO model of the industry are composed of

several interrelated sectors and are characterized by dependence upon the

consumer, by the influences of weather, and by forces subject to control by

the industry. These combine to determine the level of orange supplies and

grower profits over time.

The DYNAMO simulation model represents an attempt to define--in a

mathematical framework--the interrelationships between the relevant industry

variables. To facilitate exposition, the relationships presented here will


11It is intuitively obvious that changes will be made in the model only
if these changes can be implemented in the real system.
12
1Adequacy implies that reasonable confidence is placed in the remaining
individual components of the model.









13
not be in the form required by the DYNAMO program.3 Instead, the discussion

will abstract from the mathematical formulation of the model and will focus

on some of the major forces which were thought to be operating in the
14
industry at the time of the study.4 Additionally, the model is extended

to include the supply management policies previously outlined so that some

additional forces subject to control by the industry are included.

The block diagram shown in Figure 4 points out the gross interrelation-

ships which control supplies. Long-run supply in the form of tree numbers

is affected exogenously by weather and endogenously by the level of grower

profits. This long-run supply influences the general level of F.O.B.

(processor) prices. Then grower profits are a reflection of the existing

F.O.B. price. This completes the long-run supply information-feedback loop

for the industry. Of special importance in keeping the long-run supply

continually out of equilibrium are the effects of weather and the delay

between the planting of an orange tree and its eventual yield.

For any given season the initial short-run supply determines the F.O.B.

price. Then the retail price is determined from the F.O.B. price.15

Through the consumer demand schedule, the retail price determines the level

13
1Refer to Raulerson [8] for a complete listing of the model equations
as written in the DYNAMO format.
14
14966-67 season.

15The demand schedule at the consumer level is actually the independent
relation upon which all pricing structures in the FCOJ industry (and the
DYNAMO model of the industry) are based. However, it is common industry
practice to think of grower and retail level prices as dependent on the
price received by the processor. Examination of the model equations in
Raulerson [8] reveals that the F.O.B. price is based upon disappearance
levels which in turn are based on short-run supplies remaining. Then the
remaining supply depends on retail sales which are of course, a function of
the consumer demand. Thus, the economic relationships actually involved
were not violated.
























Grower
profit


F.O.B.
price


Retail
sales


Block diagram of major components of the frozen concentrated
orange juice industry.


Figure 4.









of retail sales. Retail sales and the effects of weather combine to influ-

ence short-run supply. This completes the short-run supply information-

feedback loop which is kept out of equilibrium largely because of delays

in information and the effects of weather.

Figure 5 summarizes the major forces thought to be operating in the

industry. The figure represents the forces as individual decisions, collec-

tive decisions, prices, and other forces. Implicit in the figure are

delays of various durations which include delays due to time-consuming

physical and decision processes. The symbols used have the following inter-

pretations:

X represents a flow variable. It results from decisions made
by the individual grower, processor, or consumer, and can be

thought of as a valve which restricts or increases a given

flow.

Q represents variables affecting decisions.

| -- represents a stock variable. It is the cumulative result of

inflows and outflows.

represents a collective decision variable. It represents an

S industry policy and is not affected by individuals acting

alone.

Represents exogenous variables.

--represents physical flows.

represents flows of information.

Since the relationships depicted involve a system of interdependent

information-feedback loops, the decision as to where to enter the system in

order to examine it is rather arbitrary. The production process begins

with the planting of trees and provides a reasonable starting point.






















Planting
Decisions


Weather
^---0---

/ Cultural
Practices Decisions


Processing
Decisions


Retail Consumer
Ordering Buying
Decisions Decisions


Legend:


V Supply Control Policies


Figure 5. Flow chart showing the major
in the study.


forces operating in the FCOJ industry, including the hypothesized supply policies considered










Following Figure 5, individual planting decisions depend on the grower

profit per box. As defined earlier, grower profit per box equals the on-

tree price less production costs. The on-tree price is derived from the

F.O.B. price. The model assumes that individuals increase or restrict the

rate at which they plant trees as grower profits rise or fall. Historically,

it has been evident that these decisions in aggregate produce sizeable

fluctuations in the number of trees. More specifically, growers tend to

overinvest (collectively) in new orange groves when profits are unusually

high. For this reason the augmented model includes a collective policy to

force the individual to restrict new plantings when profits are unusually

high.16

Over time, the planting rate adds to the stock of trees. This stock

can be reduced instantly by the effect of weather, freezes in particular.

The model allowed for random freezes by sampling from 29 seasons (1937-38

through 1964-65) of weather data [10].

A freeze can affect orange supplies in one or more of three ways

according to the severity of a particular freeze encountered. Some trees

can be killed or so severely damaged that they have to be butt-cut. Some

trees can be damaged enough so that they have to be hatracked; that is, all

but the primary or secondary scaffold branches have to be pruned. These

trees come back into production in four to five years. The third way in-

cludes those trees which only suffer yield losses. Table 1 presents six,

20-year sets of randomly selected weather effects showing the fraction of

trees lost, crop lost, and trees hatracked for each random year selected.


1Refer back to pp. 3-5 for a more complete description of the policies.









Table 1. Six 20-year sets of weather effects in terms of fraction of trees lost, fraction of crop lost,
and fraction of trees hatracked

Weather Condition

Season One Two Three Four Five Six

Lost Yield Hat Lost Yield Hat Lost Yield Hat Lost Yield Hat Lost Yield Hat Lost Yield Hat


1961-62
1962-63 .11 .15 .18 .11 .15 .18 .11 .15 .18 .11 .15 .18 .11 .15 .18 .11 .15 .18
1963-64 --- .3 --- --- .3 --- --- .3 --- --- .3 --- --- .3 --- --- .3 ---
1964-65 --- .2 --- --- .2 --- --- .2 --- --- .2 --- --- .2 --- --- .2 ---
1965-66
1966-67
1967-68
1968-69 --- --- --- --- --- --- .08 .15 .1 --- --- --- --- --- --- .11 .15 .18
1969-70 --- --- --- --- --- --- --- .15 --- --- --- --- --- --- --- --- .2 ---
1970-71 --- .1 -- --- --- --- --- .04 -- --- --- --- --- --- --- --- .1 --
1971-72
1972-73
1973-74 --- --- --- --- --- --- --- --- --- --- .13 ---
1974-75 --- .1 --- --- --- .08 .15 .1 .06 .12 .06 --- .13 ---
1975-76 --- .1 -- --- --- -- --- .15 --- --- .05 ---
1976-77 .11 .15 .18 --- --- --- --- .04 --- .08 .15 .1
1977-78 .08 .35 .1 --- --- --- --- --- --- .08 .3 .1
1978-79 --- .25 -- --- --- --- --- --- -- --- .19 --- .06 .12 .06 -- --
1979-80 --- .17 --- --- --- --- --------- --- .04 --- --- .05 ---
1980-81



Actual freeze effects are shown for seasons 1961-62 through 1966-67; random freeze effects are shown
for seasons 1967-68 through 1980-81.










The stock of trees is also reduced as trees become unproductive.

Although this is a natural process, the rate at which trees become unpro-

ductive is influenced by the flow of year-to-year cultural practices.

Individual decisions concerning cultural practices depend on grower profits.

The crop size for any given year is based on the number of trees and their

yield. The yield is affected by the year-to-year cultural practices, and

the influence of weather.

Processed inventory accumulates as the result of processing rate

decisions. Since mature oranges can be stored on the tree, it may be

thought that the on-tree price and other factors will influence the pro-

cessing rate. This may be true to some degree, but historically the pro-

cessing rate has been quite consistent from year to year. Evidently, the

gains from attempting to operate processing facilities at peak capacity

outweigh other forces which influence processing decisions. Thus, the

schedule of processing rates during a season were fixed in the model.

Processed inventory decreases as a result of disappearance decisions based

mainly on the F.O.B. price, and hence the amount that retailers will take.

Additionally, a collective action that would divert FCOJ into the secondary

market was included. This action was presumed to take place whenever the

grower profit per box reached the minimum allowable level of $.15 per box.

The policy was designed to reduce short-run supplies and to raise the price

structure and grower profits.

All disappearing processed inventories not channeled into the secondary

market accumulate as retail inventory. Retail inventory is depleted as

consumer buying decisions regulate the outflow of FCOJ into final consumption.










storage costs of a pooling program is the other piece of information needed

to determine the economic feasibility of a market pool. The costs of such

a pool include costs of storage, procurement, handling and administration.
20
The estimated cost amounted to over $59 million for the 20-year simula-

tion period (Table 7), and the estimated procurement cost was over $157

million21 (Table 8). The estimated gain in grower profit was over $171

million for the same period (Table 9). These figures suggest an estimated

net loss (or net investment in stored FCOJ) of about $46 million as

compared to the free market over the same time period (Table 10). This

estimated loss (or investment) does not include handling and administrative

costs which would be substantial.22


IMPLICATIONS OF STUDY

Implications of Study for the Industry

Before the FCOJ industry adopts any supply-oriented policy it should

resolve some basic conflicts of interest regarding the desired level of

orange supplies and the length of time the proposed policy should be in

effect. Furthermore, it should be recognized that any policy will

undoubtedly change the industry environment which created the need for the

policy. Due to this latter situation the industry should be cautioned to

continually evaluate both the tangible benefits of any policy and the

structural changes in the industry resulting from such a policy.


20Storage cost was based on $7.176 per year per 52-gallon drum at 580
brix.
21
2Procurement cost was assumed to be $4.45 per case which was the
approximate value per case of raw fruit plus processing charges.
22
2The administrative costs of the industry advertising program amounted
to about 11 percent of the total advertising revenue at the time of this
study.









Table 7. Simulated storage costs for a frozen concentrated orange juice
pool

Retail sales Cost data
Season
FM S Additions Years of Cumulative cost of storage
to poola storage compounded @ 6 percent


(Million cases)c


(Million dollars)


1961-62
1962-63
1963-64
1964-65
1965-66
1966-67
1967-68
1968-69
1969-70
1970-71
1971-72
1972-73
1973-74
1974-75
1975-76
1976-77
1977-78
1978-79
1979-80
1980-81


Total


35.7
22.8
21.2
27.4
38.4
43.0
44.4
46.0
47.2
48.4
49.3
49.8
50.0
50.0
49.8
49.6
49.0
48.4
48.0
47.6


35.7
22.8
21.2
27.4
38.4
42.6
43.0
44.0
45.0
45.9
46.5
47.2
47.6
48.1
48.5
49.0
49.5
50.1
50.2
49.6


.4
1.4
2.0
2.2
2.5
2.8
2.6
2.4
1.9
1.3
.6


2.1
8.8
15.7
24.1
32.6
40.0
47.8
53.2
56.8
58.9
59.6
59.6
59.6
59.6
59.6

59.6


866.0 852.3 20.1


Frozen concentrate was assumed to be placed in the pool whenever
grower profits in the free market fell below 15 cents per box. There were
no removals from the pool because of the decision to remove concentrate
only when profits in the free market were at least one dollar per box, and
profits never reached this level during the latter years.
b
Storage cost for a 52-gallon drum of FCOJ @ 58 brix = $7.176 per
year; 52 gallons @ 580 brix = 71.05196 gallons at 450 brix or 31.578649
cases; $7.176 31.578649 = $.2272421 per case per year. Storage cost used
here does not include handling charges.


c h case contains 48 six-ounce cans of concentrate.
Each case contains 48 six-ounce cans of concentrate.










Table 8. Simulated procurement costs for
juice pool


a frozen concentrated orange


Season FCOJ added to poola Cumulative procurement cost plus
interest compounded @ 6 percentb


(Million cases)e









.4

1.4
2.0
2.2
2.5
2.8
2.6
2.4
1.9
1.3
.6


1961-62
1962-63
1963-64
1964-65
1965-66
1966-67

1967-68
1968-69
1969-70
1970-71
1971-72

1972-73
1973-74
1974-75
1975-76
1976-77
1977-78
1978-79
1979-80
1980-81

Total


(Million dollars)









4.3

18.4
33.5
53.2
74.4
96.7

116.2
133.2
146.0

154.2
157.7
157.7
157.7
157.7
157.7

157.7


FCOJ was never removed from the pool since the grower profit was
always below $1.00 per box except during part of the first five seasons
1961-66 when there was no FCOJ in the pool to remove.
b
Procurement cost = $4.45 per case.


Each case contains 48 six-ounce cans of concentrate
Each case contains 48 six-ounce cans of concentrate.


20.1









Table 9. Gain in grower profits from a frozen concentrated orange juice
pool


Grower profits
nGrower profits Years of Cumulative pooling gain
Season
FM SM Difference storage compounded @ 6 percent
FM SM Difference


(Million dollars)









3.8
19.7
43.5
71.1
101.5
132.8
161.6
186.1
205.6
219.5
227.2
225.4
211.5
190.7
171.4


Total 431.6 509.7


1961-62
1962-63
1963-64
1964-65
1965-66
1966-67

1967-68
1968-69
1969-70
1970-71
1971-72

1972-73
1973-74
1974-75
1975-76
1976-77
1977-78
1978-79
1979-80
1980-81


35.8
53.3
92.4
107.2
71.2
1.6
- 5.8
- 9.5
-12.2
-14.1
-14.6
--12.7

- 9.0
- 3.9
2.3
9.4
18.6
30.8
41.0
49.8


35.8
53.3
92.4
107.2
71.2
3.2
1.2
1.7
1.5
1.9
2.9
4.3
6.4
9.1
12.1
15.1
17.2
19.1
22.5
31.6


+ 1.6
+ 7.0
+11.2
+13.7
+16.0
+17.5
+17.0
+15.4
+13.0
+ 9.8
+ 5.7
- 1.4
-11.7

-18.5
-18.2


78.1


171.4











Table 10. Net value of a frozen concentrated orange juice pool


Costs


Gains


Storage

Handling

Procurement

Administrative

Total cost


(Million dollars)

59.6

Not estimated

157.7

Not estimated

217.4


Grower profit








Total gain


(Million dollars)

171.4








171.4


Net loss or net investment in stored
FCOJ at end of 20 years:

Present value of loss or investment
in FCOJ:


aAssuming a 6 percent rate of interest.


46.0


14.3a










The price elasticity of demand is more elastic for the processor

than for the grower sector of the FCOJ industry at any given level of crop

size. This situation is common in agriculture and is due to the addition

of processing and marketing charges to the value of raw fruit. These

different price elasticities imply that a crop size which maximizes net

revenue for the grower may produce less than the maximum net revenue for

the processor.

One solution to any conflict which may exist or arise between growers

and processors concerning optimum crop size is vertical integration in the

FCOJ industry. Historically, vertical integration has been occurring in

the industry. The continuation of a free market policy will probably pro-

vide greater impetus to more integration than the other policies considered

in this study.

The study shows that supply policies designed to control supplies on

a year-to-year basis may create additional supply problems in the future.

If so, there may be a general conflict between short and long long-run

policy goals. The implication for the industry is that careful study

should be given to any policy (including policies not considered by this

study) to determine if such a conflict exists.

Even though a supply policy is successful in stabilizing grower profits

and supplies of oranges at an acceptable level, industry decision-makers

should be aware of possible unforseen dangers. It was pointed out earlier

that a policy may change the industry environment in such a way that

producers will respond in an unpredicted or unpredictable manner. For

example, when the Israeli government established a program of guaranteed

forward prices to help the poultry industry, farmers increased egg produc-

tion an estimated 81.3 percent above normal production for that price









because the price was no longer uncertain [5, p. 98]. If such a situation

developed in the FCOJ industry, then the results of this study would tend

to be optimistic for all policies (except free market) tested.

A final word of caution--before the FCOJ industry adopts any policy it

must recognize the irreversibility of the industry supply function. The

industry is more responsive to price increases than to decreases. This

situation is attributable to a large degree to the fixity of investments in

groves. The result is that any policy which improves prices in the short-

run and does not attempt to control tree numbers may aggravate the long-run

supply situation.


Implications for Further Research

A DYNAMO model should be a continuous representation of a real world

system or an extension of that system. This requirement necessitates the

inclusion of all relevant system variables in the model. Further, all

variables should have accurate parameters or constants associated with them.

This requirement of an accurate description creates a problem when stimulating

the future rather than the past. A system will generally undergo changes

as a result of recent behavior patterns observed by the decision-makers

within the system. These changes cannot be fully predicted and allowed for

in the model.

Some relief from the requirement of perfect industry description is

possible because a number of system variables may not actually exert much

influence on the behavior of the system, especially in the long run. In

addition, there may be a considerable error allowance in some of the para-

meters before the system is inaccurately represented. It is possible to

discover which variables and parameters have only a small influence on

total system behavior by subjecting the model to sensitivity analysis.










It should be pointed out that the requirement of an accurate system

description implies that future DYNAMO models should be "better" since more

relevant data will probably be collected and analyzed over time. Recogni-

tion should be given to the fact that DYNAMO models are not designed to

provide "optimum" industry behavior, but to aid in the improvement of a

system. The key here is the differentiation between optimization and

improvement. A model resembling the real world may be capable of providing

ideas for improvements.

The effectiveness of DYNAMO models is directly related to the quality

of available data needed to choose the parameters for the model. It is

true, as indicated above, that the choice of parameters for certain

mathematical relationships may not seriously affect overall model results.

However, in general, accurate data are required in order to construct

accurate models and the researcher should not expect sensitivity analysis

to rescue a model constructed with low quality data.










The consumer buying decisions are based on their demand curve which is a

function of the retail prices, income, population and industry advertising.

Retail price is derived from the F.O.b. price and the industry adver-

tising is a collective decision with its level directly tied to a per-box

tax on the fruit processed.

This completes the physical flow of FCOJ which begins with the initial

decisions to plant trees and ends with FCOJ flowing into final consumption.

Figure 5 shows that the main interest in the study was with grower profits,

and that all policy decisions added to the model were based on the goal of

stabilizing grower profits at acceptable levels.


VALIDATION OF MODEL, SIMULATION RESULTS,
AND ANALYSIS OF POLICIES

Validation of Model

The FCOJ industry was simulated for the period 1961-62 through 1966-67

using the free market policy and the actual weather during that period in

order to provide a comparison of model and actual industry behavior.

Figures 6 and 7 show this comparison with respect to the pack of FCOJ and

grower profits per gallon.

These variables are representative of the other variables in the model

insofar as model behavior vs. actual industry behavior is concerned. The

model and the actual system exhibited the same modes of behavior for the

six year period. The fact that the magnitude of these variables differ

somewhat for the simulated and actual system is only of secondary importance.

17 Footnote 15.
See Footnote 15.







2.50


2.25


2.00


1.75


1.50


1.25


1.00


0.75


0.50


0.25


0.00


-0.25


1.19


1.21


S. .. n


1961-62


1962-63


Figure 6. Simulated and actual
through 1966-67.
aEstimated.


1.90


2.44
W"f


----4


1963-64


Simulated


SActual


1.64


1.21


1964-65


0.77




B/


0.71


1965-66


0.02
1 C=


1-\J
-1.0a


1966-67


Season
grower profits of frozen concentrated orange juice, 1961-62


Source of actual grower profits: Florida Canners Association Statistical Summary Season of
1965-66, Winter Haven, Florida.


0.61
0.46








100

90

80

70

60

50

40

30


1962-63


Simulated and actual
through 1966-67.


1963-64


1964-65


1965-66


1966-67


Season
retail pack of frozen concentrated orange juice, 1961-62


aBased on 480 Brix.
bEstimated.
Estimated.


Source: Florida Canners Association Statistical Summary Season of 1965-66, Winter Haven, Florida.


7-


E Simulated

I Actual













I/


7-


10 -


77-





h///


7


















7


7-i~






7~~


1961-62


Figure 7.










More important is the fact that the cycles of the variables have the same

periods, while the amplitudes (absolute change from a high point on a curve

to a low point) are similar.


Results Using Free Market Policy

Simulation of the free market policy with no freezes after the 1966-67

season is presented in Figure 8.18 This run, like all runs, was started

with the initial conditions corresponding to the start of the 1961-62

season. Crop size (for FCOJ only) was 55.9 million boxes, F.O.B. price per

dozen six-ounce cans was $1.44, productive trees numbered 13.9 million and

grower profits averaged $.73 per box.

The behavior of only three variables is traced in Figure 8. Violent

shifts in these variables in the early years were a consequence of the

1962-63 freeze.

Beginning with the 1964-65 season grower profits fell continuously and

reached zero in the 1966-67 season. During this same period crop size rose

to 69.7 million boxes, the F.O.B. price fell to about $1.15, and the number

of productive trees rose to over 18 million. Grower profits did not become

positive again until the 1975-76 season after which they followed a general

upward trend with slight seasonal variation.

Grower profits and retail sales for the simulation runs using a free

market policy and one of the six randomly selected sets of weather are

summarized in Tables 2 and 3, respectively. Each of the runs with the

other five randomly selected sets of weather encountered at least one

freeze in addition to the 1962-63 freeze (see Table 1).


18Refer to Raulerson [8] for a graphical presentation of the results of
the other simulation runs.
















































1963- 1965-
64 66


1967- 1969- 1971-
68 70 72


1973- 1975-
74 76


Season

Figure 8. Average Grower Profit, F.O.B. price and Crop Size for a 20-year simulation with
weather assumption One and a free market policy, 1961-62 through 1980-81


4.0


1961-
62


80


75


70 n

10
65 .
N


60

0
o



50 m


45


40


35


1979-
80


1977-
78










Table 2. Yearly and total grower profits for six computer simulations of
the free market policy using random weather (million dollars).

Weather Condition
Year
One Two Three Four Five Six


1961-62

1962-63

1963-64

1964-65

1965-66

1966-67

1967-68

1968-69

1969-70

1970-71

1971-72

1972-73

1973-74

1974-75

1975-76

1976-77

1977-78

1978-79

1979-80

1980-81

Total


35.8

53.3a

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

4.2a

.8

-20.4

-15.5

- 7.0a

19.6a

41.9a

67.4a

85.6

138.8

136.9a

786.3


35.8

53.3a

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

-14.1

-14.6

-12.7

- 9.0

- 3.9

2.3

9.4

18.6

30.8

41.0

49.8

431.6


35.8

53.3a

92.4

107.2

71.2

1.6

- 5.8

35.0 a

76.4

55.4

2.4

-12.1

- 9.9

40.6a

90.0

78.6

36.4

25.1

28.3

33.0

834.9


35.8

53.3a

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

-14.1

-14.6

-12.7

15.2a

52.7a

63.9

65.9a

88.6a

137.3

130.6

101.5

948.3


35.8

53.3a

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

-14.1

-14.6

-12.7

- 9.0

19.9a

25.7

2.1

10.7

50.6a

91.0

79.8

563.4


35.8

53.3a

92.4

107.2

71.2

1.6

- 5.8

40.3a

95.1

92.2

38.5

- .5

- 4.4

- 2.1

1.5

5.1

8.5

12.8

17.8

24.0

684.5


indicates the occurrence of a freeze during the season.
Indicates the occurrence of a freeze during the season.









Table 3. Yearly and total retail sales for six computer simulations of the
free market policy using random weather (million cases)a

Weather Condition
Year
One Two Three Four Five Six


1961-62

1962-63

1963-64

1964-65

1965-66

1966-67

1967-68

1968-69

1969-70

1970-71

1971-72

1972-73

1973-74

1974-75

1975-76

1976-77

1977-78

1978-79

1979-80

1980-81

Total


35.7

22.8b

21.2

27.4

38.4

43.0

44.4

46.0

47.2

44.4b

49.4

50.9

50.6

50.3b

46.0b

45.7b

31.8b

20.8

24.0

29.0b

769.0


35.7

22.8b

21.2

27.4

38.4

43.0

44.4

46.0

47.2

48.4

49.3

49.8

50.0

50.0

49.8

49.6

49.0

48.4

48.0

47.6

866.0


35.7

22.8b

21.2

27.4

38.4

43.0

44.4

33.5b

35.2

43.0

48.1

49.3

49.7

36.5b

37.1

44.0

48.2

48.9

49.1

49.4

804.9


35.7

22.8b

21.2

27.4

38.4

43.0

44.4

46.0

47.2

48.4

49.3

49.8

44.7b

40.2b

44.1

35.0b

24.1b

28.6

36.3

40.5

767.1


35.7

22.8b

21.2

27.4

38.4

43.0

44.4

46.0

47.2

48.4

49.3

49.8

50.0

44.8b

49.9

50.9

49.9

39.4b

41.2

44.6

844.3


35.7

22.8b

21.2

27.4

38.4

43.0

44.4

29.4b

29.6

37.2

45.2

47.9

48.8

49.4

49.8

50.2

50.5

50.8

50.9

50.8

823.4


aEach case contains 48 six-ounce cans of concentrate.

blndicates the occurrence of a freeze during the season.









The runs generally support the historical evidence that a period of

high grower profits stimulates new orange tree plantings which result in

large quantities of oranges and low grower profits a few years later.


Analysis of Alternative Policies

Method of Comparing.--Supply control policies were compared on the

basis of their ability to stabilize supplies and grower profits at accept-

able levels. Two simulations of each of the six policies were run in order

to observe their effectiveness under two sets of random weather conditions.

Weather conditions Two and Six were used for this purpose (Table 1).

Policy comparisons were made on the level and relative variability of

grower profits. The comparison used assumes that industry utility (from

the growers' point-of-view) is an increasing function of total grower

profits. Variability of grower profits and supplies of oranges is not

desirable due to reduced marketing efficiency, reduced grove management

efficiency, and possible adverse shifts in the demand for FCOJ that may

result when consumers must continually reappraise their consumption patterns.

As a consequence, the basis for comparing policies assumes that utility is

a declining function of variability of profits.

Comparison Using Weather Two.--The free market (FM) policy appeared to

be the least desirable of the six studied. It produced the lowest total
19
grower profits and had the highest rel-variance9 (Table 4). However, at

the end of the simulation it ranked third in the level of grower profits.

This indicates that a free market is capable of correcting a low grower


19The rel-variance is a statistic giving the ratio of the variance to
the mean squared. A "small" (vs. "large") rel-variance indicates lesser
fluctuations in the value of a variable relative to its average value.










Table 4. Total grower profits, rel-variances, grower profits for the 20th
year, and total retail sales for 12, 20-year simulations of the
frozen concentrated orange juice industry, 1961-62 through 1980-81


Weather Condition Two Weather Condition Six


Policy


Policy


SMPR
PR
SMTR
TR
SM
FM


SMPR
PR
SMTR
TR
SM
FM


Total
grower profits
(Million dollars)

773.9
771.0
699.8
694.5
509.7
431.6
Rel-variances
0.56
0.58
0.61
0.72
1.46
2.83
20th year
grower profits
(Million dollars)
69.9
66.5
49.8
44.6
44.3
31.6
Total
retail sales
(Million cases)
886.0
852.3
827.4
826.9
816.8
816.5


FM
SM
TR
SMTR
PR
SMPR


Total
grower profits
(Million dollars)

1159.1
1152.6
905.5
899.4
722.3
684.5
Rel-variances
0.23
0.24
0.45
0.46
1.01
1.18
20th year
grower profits
(Million dollars)
69.3
69.1
26.9
26.3
24.0
15.1
Total
retail sales
(Million cases)b
823.4
816.1
790.5
789.7
754.3
752.2


aFM = free market, SM = secondary market, TR = tree removal, PR =
planting restriction, SMTR = secondary market combined with tree removal,
SMPR = secondary market combined with planting restriction.

Each case contains 48 six-ounce cans of concentrate.


PR
SMPR
TR
SMTR
FM
SM


SMPR
PR
SMTR
TR
SM
FM


SMTR
TR
SMPR
PR
SM
FM


PR
SMPR
FM
TR
SMTR
SM


FM
SM
TR
SMTR
PR
SMPR









profit situation without the aid of supply-oriented policies. But the time

required to correct this situation may be quite long. The hypothesized

utility function suggests that some type of policy program would be pre-

ferred. The time required for profits to recover in a free market environ-

ment can be seen in Figure 8. Note that the period of high profits

following the freeze lasted only four years, while it took 12 years for

profits to come back to pre-freeze levels once they declined. This situa-

tion is common among several important tree crops (see M. J. Bateman [1]).

The difference in investment between a nursery tree and a fully productive

tree probably contributes to the longer period of depressed prices and

grower profits. Growers seem to be more willing to plant new trees during

periods of high profits than to destroy productive trees during periods of

low profits. These differences in growers' response substantiate the pro-

position that large fluctuations in crop size and grower profits tend to

lower long-run profits because of the stimulus to planting during periods

of high grower profits.

The policy which combined the secondary market and planting restric-

tions ranked among the most desirable policies. It generated the highest

total grower profits and had a low rel-variance (Table 4). Note that four

policies: secondary market-tree removal (SMTR); tree removal (TR);

secondary market-planting restrictions (SMPR); and planting restrictions

(PR), had rel-variances that were similar (Table 4) and will be considered

as equal for this discussion. The SMTR and TR policies were inferior to

the SMPR and PR policies, due to the relatively large difference in grower

profits.

The SMPR and PR policies produced similar total grower profits.

However, the behavior of industry variables for the two policies was some-

what different over the 20-year simulation period. Yearly grower profits










were higher for SMPR during seasons 1966-67 through 1972-73 (Table 5).

This occurred because the secondary market part of the policy limited

supplies earlier than a planting policy by itself. However, for the

balance of the simulation, PR had higher yearly profits. Since no imme-

diate supply removal policy was used with PR to raise the grower profits,

the planting restriction was in effect longer. In addition, the lower

profits forced the removal of more productive trees. This resulted in

lower supplies and higher profits later. An examination of the profits for

the latter years of the simulation reveals that the planting restriction

only policy was gaining on the dual policy at the rate of over three

million dollars per year (Table 5). Since the total grower profits were

773.9 and 771.0 million dollars for SMPR and PR, respectively, then grower

profits for the single policy would probably be greater had the simulation

been extended one more year. It would seem in comparing these two policies

that SMPR is more beneficial in the short-run, while PR is more beneficial

in the long-run.

The SMTR and TR policies had rel-variances similar to SMPR and PR but

produced lower grower profits. Total grower profits for SMTR were almost

identical to those for TR (Table 5). Unlike SMPR and PR which exhibited

different industry behavior, major industry variables reacted almost

identically for SMTR and TR. Table 5 shows a difference in grower profits

of over three million dollars in only one season (1973-74). This behavior

occurred because the removal of trees occurs in the short-run but has long-

run implications. In effect, the tree removal and secondary ideas both are

designed to bring grower profits up to $.15 per box. Thus, the addition of

the secondary market policy added very little to the short-run effects of

the tree removal policy. In later years of the simulation the grower










Table 5. Yearly and total grower profits for 16 different 20-year computer
simulations of the retail sector of the frozen concentrated orange
juice industry in Florida (million dollars)


Policy FM FM FM FM FM FM SM SM

Weather Condition
Year
One Two Three Four Five Six Two Six


1961-62

1962-63

1963-64

1964-65

1965-66

1966-67

1967-68

1968-69

1969-70

1970-71

1971-72

1972-73

1973-74

1974-75

1975-76

1976-77

1977-78

1978-79

1979-80

1980-81


35.8

53.3

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

4.2

.8

-20.4

-15.5

- 7.0

19.6

41.9

67.4

85.6

138.8

136.9


35.8

53.3

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

-14.1

-14.6

-12.7

- 9.0

- 3.9

2.3

9.4

18.6

30.8

41.0

49.8


35.8

53.3

92.4

107.2

71.2

1.6

- 5.8

35.0

76.4

55.4

2.4

-12.1

- 9.9

40.6

90.0

78.6

36.4

25.1

28.3

33.0


35.8

53.3

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

-14.1

-14.6

-12.7

15.2

52.7

63.9

65.9

88.6

137.3

130.6

101.5


35.8

53.3

92.4

107.2

71.2

1.6

- 5.8

- 9.5

-12.2

-14.1

-14.6

-12.7

- 9.0

19.9

25.7

2.1

10.7

50.6

91.0

79.8


35.8

53.3

92.4

107.2

71.2

1.6

- 5.8

40.3

95.1

92.2

38.5

- .5

- 4.4

-2.1

1.5

5.1

8.5

12.8

17.8

24.0


35.8

53.3

92.4

107.2

71.2

3.2

1.2

1.7

1.5

1.9

2.9

4.3

6.4

9.1

12.1

15.1

17.2

19.1

22.5

31.6


35.8

53.3

92.4

107.2

71.2

3.2

1.2

43.6

97.9

98.9

32.4

4.1

6.4

7.6

8.5

9.2

10.0

11.4

12.9

15.1


786.3 431.6 834.9 948.3 563.4 684.5 509.7 722.3


Total











Table 5. Extension.


SMPR SMPR TR TR SMTR SMTR PR PR

Weather Condition

Two Six Two Six Two Six Two Six


35.8

53.3

92.4

107.2

71.2

4.0

2.8

4.8

6.8

9.1

11.3

13.5

17.4

26.3

36.1

44.5

51.3

57.4

62.2

66.5


35.8

53.3

92.4

107.2

71.2

4.0

2.8

44.3

98.2

103.3

63.7

34.7

36.8

42.9

50.0

56.5

61.2

64.6

67.1

69.1


35.8

53.3

92.4

107.2

71.2

19.9

25.7

13.6

13.2

14.2

11.4

16.4

19.0

17.3

18.7

22.1

26.9

32.8

38.8

44.6


35.8

53.3

92.4

107.2

71.2

19.9

25.7

50.2

99.4

98.1

48.6

22.7

26.2

20.0

18.6

19.8

20.0

20.3

23.1

26.9


35.8

53.3

92.4

107.2

71.2

21.1

27.1

13.5

15.4

16.6

11.8

15.4

15.9

17.1

20.2

23.1

27.2

32.6

38.6

44.3


35.8

53.3

92.4

107.2

71.2

21.1

27.1

50.3

99.0

99.0

48.1

23.3

27.4

19.6

18.8

20.6

18.7

21.3

25.0

26.3


35.8

53.3

92.4

107.2

71.2

2.3

- 3.8

- 4.3

- 2.0

2.1

7.1

14.3

22.6

32.5

40.8

48.4

55.1

60.9

65.2

69.9


35.8

53.3

92.4

107.2

71.2

2.3

- 3.8

41.6

96.3

99.6

66.0

36.8

38.1

44.0

50.8

57.1

62.2

65.0

67.4

69.3


905.5 771.0 1152.6


773.9 1159.1 694.5


899.4 699.8









profits are above $.15 per box so that the secondary market part of the

policy adds nothing. Thus, it appears that a secondary market policy used

in tandem with a tree removal policy has little advantage over a tree

removal policy used by itself.

The secondary market policy used by itself had the lowest total grower

profits, and the highest rel-variance of all policies except the free

market. At the end of the simulation it produced lower grower profits than

any other policy (Table 5). Its influence on industry variables was to

raise profits in the short-run while limiting long-run gains. This result

can be explained because this type of policy only attacks the symptoms of

the problem and allows the cause to go unchecked. A secondary market

policy adjusts supplies on a year-to-year basis. However, only oranges are

removed. In order to affect future supplies the trees must be controlled

in some manner. Model experimentation shows that curtailing short-run

supply actually encourages the retention of productive trees. Hence, long-

run supply problems are actually magnified.

Comparison Using Weather Six.--The relative desirability of the six

policies studied under weather assumption Two was almost unchanged for

weather condition Six, while the absolute and percentage differences in

grower profits between the leading policies showed some increase under

weather assumption Six.

Under weather condition Two the total grower profits for the policies

involving the removal of productive trees were about 90 percent of the

total grower profits for the policies that restricted new plantings.

However, this figure dropped to about 78 percent for weather condition Six

(Table 6). This implies that the more exogenous fluctuations in supply










(freezes) that occur, the more valuable the planting policy becomes. This

is reasonable considering the argument presented earlier that growers are

more responsive to high prices than to low prices. Consequently, inter-

mittent periods of high grower profits induce larger supplies unless a

restriction on planting is imposed.


Table 6. Simulated total grower profits for six policy decisions under two
weather conditions expressed as percent of total grower profits
of leading policy


Weather condition Two Weather condition Six

Total grower Total grower
Po y profits as Polprofits as
percent of percent of
SMPR SMPR


SMPR 100.0 SMPR 100.0
PR 99.6 PR 99.4
SMTR 90.4 SMTR 78.1
TR 89.7 TR 77.6
SM 65.9 SM 62.3
FM 55.8 FM 59.1




Relative Costs of Alternative Policies

The policies discussed have been judged only on their capability to

stabilize grower profits and supplies of oranges at acceptable levels.

Implementing any policy program will necessitate some cost outlays. The

two dual policies--planting restrictions plus secondary market, and tree

removal plus secondary market--would certainly cost more than their single

policy counterparts--planting restrictions and tree removal. Since the

single policies performed as well as the dual policies, the single policies










will probably be preferred due to the cost consideration. This actually

leaves four policies from which to choose--free market policy, secondary

market, tree removal, and planting restrictions. Of these the free market

policy is obviously the least expensive to employ. However, it produced

the highest rel-variance and the lowest total grower profits. Of the three

single policies remaining, the planting restriction policy would probably

be the least expensive to administer. This policy also adds the most to

grower profits.

Of the four single policies, the secondary market idea would probably

be the most expensive to administer because of the record supervision

required to assure that supplies would be channeled into the correct market.

This policy also ranks third among the four single policies in terms of

total grower profits and variability of profits.

The tree-removal policy probably falls between the planting restric-

tions and secondary market policies in terms of cost. This is also where

it ranked in terms of grower profits. Thus, it appears that a consideration

of the cost of alternative supply policies probably will not alter the

relative desirability of these policies.


Consideration of a Physical Pool of Concentrate

The results of the secondary market policy using weather condition Two

discussed above were used to evaluate the feasibility of a physical pool of

FCOJ. Assume that FCOJ was placed into an FCOJ pool rather than a secon-

dary market when supplies of oranges were large enough to lower grower

profits to some chosen level (e.g., 15 cents per box), and would be removed

from the pool when supplies were small enough to raise grower profits to

some chosen level (e.g., $1.00 per box). A comparison of grower profits

under a free market vs. grower profits under the secondary market policy

provides an estimate of the gain in profits from pooling. The estimated









References


1. Bateman, M. J., "Aggregate and Regional Supply Functions for Ghanian
Cocoa, 1946-62," Journal of Farm Economics, Vol. 47, No. 2,
ilay 1965.

2. Ezekiel, M., "The Cobweb Theorem," Quarterly Journal of Economics,
Vol. 52, 1937-38.

3. Forrester, J. W., Industrial Dynamics. Cambridge, Massachusetts: The
Massachusetts Institute of Technology Press, 1965.

4. Jarmain, W. E., "Dynamics of the Florida Frozen Orange Concentrate
Industry," (unpublished Master's thesis), Massachusetts Insti-
tute of Technology, September 1962.

5. Lindahl, E., Studies in the Theory of Money and Capital. New York:
Rinehart and Co., 1939.

6. Mundlak, Y., An Economic Analysis of Established Family Farms in
Israel, 1953-58, The Folk Project for Economic Research in
Israel. Jerusalem, Israel: The Jerusalem Post Press, July
1964.

7. Pugh, A. L., DYNAMO User's Manual. Second Edition, Cambridge,
Massachusetts: The Massachusetts Institute of Technology
Press, May 1963.

8. Raulerson, Richard C., "A Study of Supply-Oriented Marketing Policies
for Frozen Concentrated Orange Juice: An Application of
DYNAMO Simulation," (unpublished Master's thesis), University
of Florida, 1967.

9. U. S. Department of Agriculture, in cooperation with Florida Citrus
Commission, Consumer Purchases of Citrus Fruit Juices, Drinks,
and Other Products. Washington, D. C.: Government Printing
Office, various months, 1962-63.

10. U. S. Department of Commerce Weather Bureau, Winter Minimum Tempera-
tures in Peninsular Florida, Summary of 20 Seasons, 1937-57,
and Annual Reports, 1958-65. Lakeland, Florida: Federal-
State Frost Warning Service.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs