PATTERNS OF ENERGY USE, ENEDGY COST INCREASES AND TaEIR IMPACTS ON
CROP PRODUCTION ON THE BIG ISLAND OF HAWAII:
A LINEAR PROGRAMMING APPROACH
A DISSERTATION SUBHITTED Ta THE GRADUATE DIVISION OF THE
UNIVERSITY OF HAWAII IN PARTIAL PULFILLIlENT
oF THE REQUIREMENTS FOR THE DEGREE OF
OOCTOR OF PHILOSOPHY
IN AGRICULTURAL AND RESOURCE ECONmlICS
AUGUST 1982
By
Egnonto N. Koffi-Tessio
Dissertation Committee:
Chennat Gopalaltr1shnaD. Chairman
James B. Karsh
Heinz Sp1e1.JDann
Salvatore Comitini
Frank S. Scott, Jr.

---

11
w. certify th&t. we turve read this dissertat.1otl and. that 1.11 our
op1J110a. it 1a sal:1sfactory in seope. and quality &8 a disserta.tion
for the. de.gree of Doctor of Pbi.losophy 1'0. Agric:ultural &Dd lluouree.
zeeeeer.cs ,
DISSERIA1'ION COMMI:rr:EE
/
" ....-F·~ 1
._~A. , .1, , , - - . _
,
>........
/
1
----
~~
Q.
.
.J la u (
..
,.IL r77
(

---

111
ACKNOWLEDGMENTS
This maDuscr1pt could Dot have been accompllshed without the
continuous encouragement and support of many individuals.
FiraC, 1 vould like to express my deep sense of gratitude and
apprec1at1on to Dr. Chennst Gopalakrlshnan, my academlc advlser and
leader of the project ent1tled "Energy in Western Agriculture:
Requirements, Adjustments and Alternatives" that provlded the
flnancial support for chis research.
The scholarly demeanor. the
h1gh standards of professional competence and the klndnes8 that he
d1splayed have great!y contr1buted to enhance the quality of chis
rese&rch effort.
Second, l am also 1ndebted to Dr. Frank S. Scott, Jr., the
motivating force beh1nd my admission to the University of Hawaii, for
his geuerous support throughout my graduate program.
Special thank8 and recognition are a180 due to Dr. James B.
Maxsh, Dx. Heinz Spielmaon and Dr. Salvatore Comitini, dissertation
commit tee membexs, fox theix valuable criticisms and suggestions.
Third, l vould also llke to extend my s1ncere thanks to the
Afxican American Institute fox its financial support that made
possible the pursuit of my gxaduate studiee at the Univexsity of
Bavsii.
To all the members of my family, eepecially my fathex, Julien N.
Xoffi-Tessio, and my mothex, Florencia, l expxees my s1ncere
gxatitude.
To my deax brother, Amedee Miga, who suddenly lost hie

---

iv
health iD the course of this study. this pieee of work, as an
expression of my deep concern and love, 18 dedicated.
Special thanks are also due to Kitty and Michael Dabney for
typing this dissertation.
Finally and .bave aIl. l am greatly Indebted
to sr wife. Sophia,
my daughter, Karina, and my new-born son. Serge, for their contlnued
support, inspiration and encouragement throughout my scademie
struggles.
Honolulu, Bawaii
Egnonto N. Koffi-Tessio
June 1982

---

v
ABSTRACT
In recent years. drastic changes have occurred in input priees.
output priees and in the institutionsl structure w1thin wh1ch
agr1eultural producers operate.
These changes sre largely the
upshot of sharp 1ncreases ln energy priees that are directly or
indirectly translated ioto h1gher production costa for the farmera.
The main objective of this study 19 ta examine the
interrelat10nship betveen the energy sector and the production of
three agr1cultural crops (sugar. macadamia Dut and coffee) by small
growers on the Big Island of Hawai!.
Specifieslly, it attempts:
(a) ta explore the patterns of energy use in agriculture; (b) ta
determ1ne the relative effic1ency of fuel use by farm size among the
three agr1cultural crope; and (c) ta investigate the impacts of
higher energy costs on farmers' net revenues under three output price
and three energy cost 8cenarios.
To meet these objectives, a linear programming model wss
developed.
The objective function waa to ma:dmize net revenues
subject ta resource availability, production. marketing and non
negativity constraints.
The application of the model to sugar, macadamia nuts and coffee
yielded the follow1ng results.
With respect to sugar, indirect
ecergy (fertilizer and herbicide) use appears to be an increasing
function of farm size.
Direct energy (gasoline, diesel and
electricity) does not lead to a specifie conclusion.
In the case of
I118.cadamia nuta, both direct and indirect energy use, vith the
•

---

vi
exception of gasoline and electrlclty, appears to he a decreasing
function of farm size.
Wlth respect to eoffee, the resules inclicate
chat direct energy U5e 15 a decreasing function of farm size.
However, the relatioDship between fertilizer use and farm size 15
not conclusive.
Findings a1so reveal that sugar, with only 10% of
energy cast, appears ta he more vulnerable ta higher energy costs
chan macaclamia nuts and eoifee with 16% and 18% of energy cast,
respectively.
In addition, hlgher energy costs tend ta have
differential impacts depending upan the output priee.
Some of the major conclusions emerging fram this study are:
(a) higher energy costs have not significantly impacted on farmers'
net revenues. but do have a differential impact depending on the
resource endowments of each cropgrower; (b) low output priees tend to
reinforce the impacts of higher energy costs, vhereas high prices
tend to cegate th~; (c) farmers are faced vith many constraints that
do not permit factor substitution.
In terms of poliey formulation, it vas observed that poliey
makers se~ to be overly eoneerned vith the problems facing grovers at
the macro level, vithout takins into aecount the constraints that
grovers face st the micro level.
These micro factors play a dominant
raIe in the context of resource allocation.
They must. therefore, be
ineorporated into a comprehensive energy and agricultural poliey at
the county and state level.

---

vii
TABLE OF CONTENTS
ACKNOIILEDGHENTS
111
ABSTRACT
v
LIST OF TABLES
ix
LIST OF FIGURES
xiv
CHAPTER 1.
INTRODUCTION
1
Overvdev
1
Background:
Problem in Perspective
2
Problem Statement
7
Objectives
10
Hypotheses
10
Study Ares.
11
Structure of the Study
21
QiAPTER II.
REVIEW OF LlTERATURE
22
CHAPTER Ill.
ANALYT1CAL FJWŒWORK
31
Procedures and Data Sources
3l
Procedures .
31
Data Sources
34
Sugar .
.
34
Macadam1a Nut
57
Coffee and Macadamia Nut Interplanting
74
The Linear Programming Model
89
Notation . .
94
Assumptions
95
Assumption of Proportionality
95
Assumpt10n of Additivity
95
Assumption of D1visibility
95
Assumption of Certainty
95
General Formulation of the Model
96
Objective Function
. . .
96
Resouree Availability Constraints
96
Production Constraints
97
Marketing Constraints . .
.
• .
.
97

---

viii
Non-Negativity Constraints
97
Case 1
98
Case 2
98
Case 3
98
CHAPT ER IV.
STUDY RESULTS AND THE IR POLICY IMPLICATIONS
100
Study Results
100
Sugar .
.
.
100
The Current Output Priee Scenario
101
The High Output Price Scenario
103
The Low Output Price Scenario
106
Macadamia Nut a
106
The Current Output Priee Scenario
107
The High Output Price Scenario
110
The Law Output Priee Scenario
112
Coffee and Macadamia Nut Interplanting
114
The Current Output Priee Scenario
115
The High Output Price Scenario
117
The Law Output Priee Scenario
118
Policy Implications
121
CHAPTER V.
SUMMARY AND CONCLUSIONS
127
Summary
. . . .
. . . .
127
Conclusions
.
132
Hodel Application for Further Research
133
APPENDIX .
135
LITERATURE CITED
168

---

1x
LIST OF TABLES
Table
Page
1
Free Wor1d Energy Consumption and Fuel
Share, 1978 and 1995 .
.
. .
. .
. .
.
3
2
Energy Consumption, Imports and Costs,
1979-1981
• • • • • • . • • • . • • •
4
3
Fossi1 Fuel Priees in U.S., 1960-1978
8
4
Sugarcane: Acreage, Production and Value of Sales
in Hawaii, 19B1
.
14
5
Macadamia Nut: Acreage. Production and Value
in Hawaii, 1981
.
. .
. . .
.
.
.
16
6
Coffee: Acreage, Production and Value of Sales
in Hawaii, 1981
17
7
Population and Samp1e of Farms Interviewed by
Sd ae , 1981
. . . . .
33
8
Farm 'Wage Rate by Method of Pay and Type of
Work, 1981 .
. . . . .
36
9
Sugar: Labor Input per Acre by Farm Size
Group, 1981
. . . .
37
10
Sugar: Ferti1izer Use per Acre by Farm Sdz e
Group. 1981
. . .
. . • .
. . . . .
42
11
Sugar: Ferti1izer Recammendations by HCPC, 1981
43
12
Sugar: Ferti1izer, Unit Priee. 1981
45
13
Suger: Herbicide Use per Acre by Farm Size
Group. 1981
46
14
Sugar: Herbicides, Unit Priee, 1981
48
15
Sugar: Coat of Direct Energy Inputs per
Acre, 1981 . . . . . . . . . . . . . .
49
16
Sugar: Direct Energy Inputs, HCPC, 1981
50
17
Sugar: Direct Energy Inputs, Mauna Kea
Company, Inc .• 1981
. . . .
51

---

x
Table
Page
18
Sugar: Direct Energy Inputs per Acre by
Type and Opere tdon , 1981 .
. . .
.
.
52
19
Sugar: Direct Energy, Unit Priee, 1981
s4
20
Sugar: Average Yie1d, Raw Sugar and Halasses
by st ee , 1981
. . .
. . . . . . . . .
56
21
Sugar: Prices of Sugar and Molasses, 1972-1982
58
22
Sugar: Production Cost per Acre by Farm
Size, 1981 .
.
59
23
Macadamia Nut: Labor Input per Acre by
Sd ae , 1981 .
. · . .
. . .
62
24
Macadamia Nut: Insurance, Interest, Land Cost,
Depreciation and Capital Cost per Acre, 1981
65
25
Macadamia Nut: Fertilizer Inputs per Acre
by Size, 1981
· . . . . . .
67
26
Macadamia Nu t : Unit Price of Ferti1i2er by
Type, 1981 • • · . . . .
. . . . . .
69
27
Macadamia Nu t: Herbicide Use per Acre by Type
and Size, 1981 · . . . . . . . .
70
28
Macadamia Nut: Unit Price of Herbicide by
Type, 1981 . .
71
29
Macadamia Nut: Direct Energy Inputs per
Acre by Size, 1981 .
.
. .
. . . . . . .
. . . . .
73
30
Macadamia Nut: Unit Price of Direct Energy
Inputs by Type, 1981 .
.
.
.
75
31
Macadamia Nut: Production Input Cost per Acre
by Type and Size, 1981 .
. . .
.
.
.
76
3Z
Macadamia Nut: Price (in Shells). 1972-1982
77
33
Macadamia Nut: Yield per Acre by Si2e. 1981
78
34
Macadamia Nut: Cost per Acre by Type and
Si2e, 1981 .
. · . . . . . . . . . .
79

---

xi
Tllble
Page
35
Coffee: Labor Input per Aere by
Size. 1981 . • • . . . • . •
82
36
Coffee: Fertilizer Inputs per Acre by
Size and Type. 1981
. • . . . .
•
85
37
Coffee: Fertilizer Unit Priee and Type. 1981
86
38
Coffee: Herbicide Input per Acre and Type.
1981 • • . •
. • • • • . •
· . . . .
87
39
Coffee: Unit Priee of Herbicide by Type.
1981 . • • .
88
40
Coffee: Direct Energy Inputs per Acre by Size
and Type. 1981
90
41
Coffee: Yield per Acre by Size and Type,
1981
91
42
Coffee: Priee. 1972-1982
92
43
Caffee: Cast per Acre by SiZe and Type. 1981
93
APPENDIX
1
Definition of Linear Programming .
136
2
Sugar: Optimal Levels of Activity, Current
Output Priee and Re 0
· . . . . . . . . · . . . . 137
3
Sugar: Optimal Levels of Aetivity. Current
Output Priee and EC 50 . · . . . . . . . . · . . . .
138
4
Sugar: Optimal Levels of Aetivity. Current
Output Priee and Ee 100
· . . . .
139
5
Sugar: Optimal Activity Levels. High Output Priee
and Base Period Energy Cost (EC 0) .
140
6
Sugar: Optimal Activity Levels, High Output Priee
and EC 50
. . . . . . . . . . .
141
7
Sugar: Optimal Activity Levels, High Priee
and EC 100 • • . . . . . . . . . .
142
8
Sugar: Optimal Levels of Aetivity. Low Output
Priee and EC 0, or EC 10 or EC 100
143

---

xii
Table
Page
9
Macadamia Nut: Optimal Levels of Activity,
Current Output Priee and Ee a
• · · · · · · · ·
144
la
Macadamia Nut: Optimal Levels of Acttvity,
Current Out put Priee and Ee 50 · · · · · · · · · ·
145
11
Macadamia Nut: Optimal Levels of Activity,
Current Output Priee and zc 100
· · · · · · · · ·
146
12
Macadamia Nut: Optimal Levels of Activity,
High Output Priee and zc a · · · · · · · · · · · ·
147
13
Macadamia Nut: Optimal Leve La of Activity,
High Output Priee and EC 50
· · · · · · · · · · ·
148
14
Macadamia Nut: Optimal Levels of Activity.
High Output Priee and Ee 100 · · · · · · · · · · ·
149
15
Macadamia Nut: Optimal Levels of Activity,
Lov Output Priee and Ee a
· · · · ·
150
16
Macadamia Nut; Optimal Activity Levels. Law
Output Priee and Ee 0
. . · · · · · · · ·
151
17
Macadamia Nut: Optimal Levels of Activity,
Law OUtput Priee and Ee 100
· · · ·
152
18
Coffee: Optimal Levels of Actlvlty,
Current Output Priee and Ee a
· · · · · · ·
153
19
Coffee: Optimal Levels of Acttvlty.
Current Output Priee and EC 50 · · · · · · · · · ·
154
20
Coffee: Optimal Levels of Activity,
Current Output Priee and EC 100
· · · · · · ·
155
21
Coffee: Optimal Levels of Aecivity.
High Output Priee and Ee 0 · · · · · · · · · · · ·
156
22
Coffee: Optimal Levels of Activity ~
Righ Output Priee and EC 50
· · · · • · · · · · •
157
23
Coffee: Optimal Levels of Aetivity 1
High Oucput Priee and EC 100 · · · · · · · · · · ·
158
24
Coffee: Optimal Levels of Activity 1
Law Output Priee and EC a
· · · · · · · · · · · ·
159

---

xiii
Table
Page
25
Coffee: Optimal Levels of Activity,
Law Output Price and EC 50 • .
• . . . . . . . . .
160
26
Coffee: Optimal Levels of Activity,
Low Output Price and EC 100
.
.
. .
161
27
Energy Cost Scenarios, Crop Priee Scenario
and Their Impacts on Net Revenues
162
28
Loan Pro gram: Amount Allotted per Aere and Total
Budget . . • • . . .
163
29
Resource Constraints (Sugar)
164
30
Loan Program: Amount Allotted per Acre, 1981
165
31
Resource Constraints (Hacadamia Nut)
166
32
Resource Constraints (Coffee)
167

---

xiv
LIS! OF FIGURES
Figure
Page
1
Ene t-gy Flow Madel for Agriculture
6
2
The Big Island of Hawaii . .
.
. .
12

---

1
ClIAPTER l
INTRODUCTION
Overvlew
The energy problem 18 probably one of the mast persistent issues
chat has aggravated the economic difficulties of bath developed and
developing nations ln recent years.
It has resulted largely from
our {ailure ln the past to identlfy and address some energy realitles
and ta see clearly our energy future (54).
Hlstory will record
that it vas the Organization of the Petroleum Exporting Countries
(OPEC) chat brought into sharp focus the seriousnes5 of the energy
problem and the depletable and non-renewable nature of the 011
reeource ,
Tc be sure, we have reached a turning point ln energy
aval1ability:
the pach of low energy costs and of perce1ved
abundance of 011 has been reversed ta one of a continuous rising trend
in energy priees.
[Although the first quarter of 1982 seems to
indicate a declining trend in energy priees, that i9 hardly any
basis to warrant the conclusion that energy priees will continue to
fall in the months or years ahead.
For instance, Kenneth T. Derr,
president of Chevron U.S.A., lue., has poinced out that the second
quarter petroleum inventory for 1982 has already registered a
deeline.
He eautioned that "reeent deeline in crude oil and
petroleum priees may end eoon and priees may rise later this year"
(56).]
But the speed with whieh energy costs will rise in the
future 19 largely dependent OP the rate at whieh conventiopal energy

---

2
resources become
scarce and more dlfficult to find. on the
technologiea! change that lowers the cast of non-conventional energy
sources. on the behavior of OPEC cartel. and the dames tic energy
pelieies of various countr1es (59).
Background:
Problem in Perspective
The United States 15 still the world's largest consumer of
energy.
In 1978, the energy used in the U.S. economy wss estimated
st 78.8 quadrillion BTU's.
With 5% of the world's population. the
U.S. accounted for about 32% of the world energy consumption.
At
the same rime, the entire Sina-Soviet black vith 28% of the world
population consumed about 31% of the world energy.
Table 1 gives an
intercountry co~arison of energy consumption and fuel shares for
the Free World with some projections for 1995.
In 1981 it was reported that U.S. net energy imports (total
imports less exports) of about 9.5 quadrillion BTU decreased by 22%
as compared to the 1980 leve1.
Similarly, energy consumption dropped
by 2.4% as compared to consumption during 1980.
At the same time,
U.S. energy import costs increased from $244.871 million in 1980
to $261,008 million in 1981, an increase of about 7% (Table 2).
It is clear from the above that although Americans have eut
their use of imported oil. they still have to face higher energy
costa.
Renee, for American consumers in general, energy will remain
a severe problem as we manage to live with the rea1ities of the
1980 1s.
More importantly. it may constitute the major constraint

---

Free World Energy Consumption and Fuel Sharee , 1978 and 1995
197 8
1 9 9 5
Total Energy
Total Energy
Consuaed
Fuel Shares
Consurned
Fuel Shares
Quadrillion
Cosi
011
Css
Dther
Quadrillion
Cosi
on
Css
Dther
Region or Country
8TU
(%)
BTU
(%)
s
U.S.A.
78.8
18
49
25
8
94.7
37
32
17
14
Canada
9.0
6
42
22
30
11.8
3
33
21
43
Japan
14.9
13
73
5
9
28.2
16
51
19
14
Western Europe
54.7
19
56
14
11
63.7
20
43
17
20
Australia/New Zealand
3.5
40
42
10
8
4.6
32
38
20
10
Total DECD
160.9
18
53
19
10
203.0
27
38
18
17
Total Non DECO
30.7
20
66
10
4
74.3
23
54
13
10
(OPEC)b
(6.2)
(0)
(71)
(24)
(5 )
(16.8)
(1)
(68)
(30)
(1)
c
Total Free World
191.6
18
55
18
9
277.3
26
42
16
16
s.
lncludes Puerto Rico and Virgin Islands
b.
Included in total DECD
c.
Total of DECO and non DECD
Source:
U.S. Department of Energy, Energy Information Administration, Annual Report ta Congress.
1980. Vol. 3.
e.>

---

Table 2
Energy cousoepr ton, Imports and ccars , 1979-1981
Total Energy Consumed
Total Energy Imports
Total Energy Coste
Years
(Quadrillion BTD)
(Quadrillion BTU)
(Mi 11ion Do 11ars)
1979
78.9
19.6
206,256
1980
75.9
15.9
244,871
1981
73.9
13.9
261,008
Source:
U.S. Department of Energy. Energy Information Administration. Honthly Energy ~eview,
Harch 1982.
'"

---

5
to the expansion of the agricultural sector in O.S. and the rest
of the world in the years to come.
The agricultural sector. in general, encompasses various
activities ranging from on-farm production, marketing, and processing
to consumption activities which require either direct energy such
as diesel fuel, gas, and electricity, or indirect energy such as
1
pesticides, fertilizers. and herbicides.
In a recent study.
Gopalakrishnan bas addressed "the complex methodological issues
mvo lved in the accurate estimation of energy r equf.reœent.s" (21).
lt was pointed out in this atudy that a uniform definition of the
term "production" and enë r gy data disaggregation are essential to
de termine the energy requirements of different products on a
comparable basis.
10r these purposes, an energy flow model
(Figure 1) showing the linkages of various activities has been
developed in this study to deal with the estimation of direct and
indirect energy inputa in the agricultural sector (21).
Energy inputs of farming have increased enormously during the
past sa years (58).
The decrease in farm labor use has been offset
in part by the growth of support industries for the farmer.
These
changes on the farm have led to a variety of other changes in the
O.S. food system.
For instance, in the past 50 years. canned.
frozen. and other processed foods have become the major items of the
American diet.
At present. the food processing industry is the
fourth largest energy consumer of the Standard Industrial Classifi-
cation grouping (54).
Transportation associated with the food

---

G~RE(T E\\'[RG'( 1
1
1 IffOiRECT EI>f.Im'
1
,•,•
fERT,
III
~
~ P{ST. TRN~;}i 1~RN5P. r "...
"I\\Ol,
""""~­
"·AU.
TI'"
, t\\
li.
~
L
T
Il1lNlSI'· r
~Wl:[(T E:\\EI!C, 'lOI"
lJo'DIIl[(f E:\\UC'f He;,:!:
FROIlOCT FLtl";S
STA~S (»' PIm.CfUJI:
.~
~L_~
~
~~~
~
._,_••_.
•__!tt
~
lY__._
v
_
YL
.
~'!1_~
.XILL
ArAl"nD f~':
TIIE u.s. FroD AND fltU SltTOll., rt\\1.RC'f US! ,,-'1:1) WTl..DOl n·906, V.S. OOV[R~K;fft l'aIMTlf:f. Or.flCE •
.....SIIlNCfO:4. D.C., I!lH.
Figure 1.
Energy Flow Model for Agriculture (21)
'"

---

7
system has grown apace. and the proliferation of appliances stl11
continues in homes, institutions. and stores.
Even farmers purchase
mast of the Ir food from markets in town (68).
Thus, energy inputs
have become 80 Integral to modern agriculture that Increases in
energy casts are likely to have severe impacts on food production and
agricultural Income (57).
From 1973 through 1978, direct energy cost in American
agriculture rose as follows:
gaso11ne 173%; diesel fuel 280%; fuel
011 89%; LP gas 144%; naturai gas 242%; and electricity 707. (16).
Consequently, mast farmers are faced vith higher energy bills which
are automatically trsDslated ln ta higher costs of production and
higher priees for cODsumers.
Table 3 shows the trends in foesil fuel
priees.
Problem Statement
Agriculture constitutes a eignificant sec ter of the State
ecenomy.
ln 1980~ ite total farm value reached $989.4 million~ the
highest within the decade.
Sugarcane~ pineapples~ and macadamia nuts
continue to be the leading agricultural crops in the State.
From
1979 te 1981, the farm value of sugarcane showed an 11% decrease due
to the substantial fall in sugar priees.
On the other hand, the
farm value of pineapple surged, vith a record $76.6 million 8S
compared with the 1979 level of about $69 million.
Slmilarly~
returns fram divereified agriculture registered an Il percent
increase from the previous year.
With the exception of cattle~
receipts from nursery products ($27.4 million) and macadamla nuts

---

8
,
_ 0
-- o o
•
o
••
o
•
N
N
N
•
• • o
g •
o,
•
0
N
"e. ,
0 0 0
..- ,..., .., N Q r'l .....
•••
.-- .., ~ go ,..., ...
,
...
o
•
•
•
•
-
:8
•
-•
N
o
N
• • •
-•
•
•
-• •
-•s
••••
N
0
•
•
• ••
• • •
•
-
o
•
•
'0 Ô
-
o
ci
o
8
• • • ·• -
• -
•
-
-
c
e
-"g "
N
N
•
•
"e
"•
-.
•
•
•
•
0
-
: :
0
-
•
o _
•
,
•,
-
o
•
• • •
-•
·
•N -
•
N
- ~
-•"
-
•
,
o
c
-•
•
N
N
o
•
•
-
o
•
•
"
--
o
"c
o
o
. 0
o
•
0
o
-
•
0
s
•
· .;
.0.
o
-•
•
- -
--. -.. --. o -•
-
N
g
• •
•.
"
o
•
•
-
_
0
0
.,.,
0
...,
""
o
•
•
-
•
-. .
-,
- •o,
o
o
_
N
• • •
•
o
· -
• •
.
~
o
"
N
•
•
•
•
-
~
•
- -,
•
>
"
•
•
-
•
-
o
-w•
8
-~. "
-t
"
•
•
•
8 -
•
• -
• -
• ;:8
8
-
8 8
"
~
•
• :
-.
-
•
- ~
~
-8.
•
~ -
"
-••
•o !
•
ô
1
"
·"•o
!
•
"
& • -
•, -
•
• •, !
-
", • -
o
-•~-•",
"
•
!
-
~
•
•
"
"
"
"
" 8 5
"
: : ~
•
• "
-
o
•
i ~
•
•
~
•
,
• •
8
-•
8
,; .ci
~

---

9
($28 million) represent a significant share of diversified
agriculture, edging out vegetables and melons which tallied a record
of $19 million in 1981.
The State's de?endence on imported ail exposes the agricultural
sector ta the full impacts of rising ail priees and the gro~ing risk
of supply disruptions.
ln Hawaii, the direct energy inputs used in
the agricultura! sector are bssically gasoline, diesel, natural gas,
liquefied petroleum and electriclty.
The indirect energy inputs
consiet of items su ch as nitrogenous fertilizers and pesticides.
,,
During the decade of 1970-1981, electricity and gasoline priees in
the StaCe have increased by 200% and 184%. respectively (11).
These
incre8aes translate not only into direct higher energy bills, but
alao into indirect increases in the priees of energy-based inputs
that farmera must purchase.
Since the increases in the priees of these energy resources are
largely determined at the regional, national and international levels
(exogenously determinedJ, a study of their impact is essential ta
6uggest possible adjusements or directions for the future.
Exogenous forces or factors may constitute a serious threat ta the
continuous economic development of Hawaii.
In addition, since the implementation of various policies at
the State level is partly dependent on the economic activities at
the national and international levels, the assessment of these
external forcea and the magnitude of their impacts is essential ta

---

10
the formulation of meaningful policies for the State as weIl as for
the Big Island of Hawaii.
Objectives
The basic purpose of the present study is to de termine the
impacts of increased energy costs on the production of agricultural
crops in the county of Hawaii.
The specifie objectives are:
1.
to identify the patterns of energy use in agriculture;
2.
ta de termine the relative efficiency of fuel use by farm
size among different agrlcultural crops; and
3.
ta explore the impacts of energy cost changes on farmers'
net revenues.
Hypotheses
The general hypothesis ta be tested is that the agricultural
sector ls sensitive ta energy cast increases.
The specifie
hypotheses ta be tested are:
1.
The larger the farm size, the more energy efficient it
tends ta be.
2.
The more energy intensive the production of an agri-
cultural cr op is in relation ta other crops, the more
vulnerable it ls to energy cast increases.
3.
The lover the output priee of a crop is in relation ta
that of other crops, the greater is the impact of higher
energy costs on the farmer's net revenues.

---

11
St:Ucly Area
Introduction
ln arder [0 have reliable estima tes of and meaningful insignts
Into the patterns of energy use, energy cost increases. and their
impact on Hawaiian agriculture, the Big Island of Hawaii (Figure 2)
has been selected as a case study.
At least four reasons can be mentioned for the choice of the
Big Island:
First, the CouDty of Hawaii vith 64% of the State land
area has about 569,364 acres of farmland. whlch represent S8i. of
the agricultural land in the State.
Second, vith the exception of
pineapple, the major proportion of crope in Hawaii are grown on the
Big Island.
In terms of cultivated ecreege , the proportions of crops
~rown in 1981wereas follows:
sugarcane (42%), caffee (100%).
macadamia nut (97%). fruits (69%), vegetables aod melons (44%).
Third, the Big Island of Hawaii has a variety of climates rangiog
trom tropical rain forests ta deserts and a variety of sail types.
The average rainfall is about 90 inches, which is hlgher than the
State average.
Fourth, the agricultural income is second ooly ta
tourism. which 15 the leading incame-generating sector of the
county (19).
The Blg Island is the youngest in the Hawaiian Archipelago and
the largest county of the State, covering an area of 4,038 square
miles.
Different geologlc and climatic conditions on the island have
resulted in the classification of 70 different soil series and 12
miscellaneous land types combined into 14 sail groupings.
The Big

---

12
UPOI.U POl""
"'~T
,,
,
,
,
,
,
r
,
,
,
,
,
,
"-
\\.~ 4
'
-
"t'
, ,
,........, ,
...
,
110 ... ,
,
,
".
, '- ,,
MilO
' - - - - -
-,
,
'.
'.
.' -
.
' .
s UrH
". ,
,
"",".11 VO~1I
,
"''l'JOlloU,l''.1I1l
,
"CIü, ,,
""
,,,
-.'.
Dl.trlcl I~"d.~y
.. " . . . . .
• .... U
"'ole _ _ •
·."S '_r. Io.. "d'r,
Figure 2.
The Big Island of Havai!

---

13
Island aIs a has a large variety of cltmates.
lt has almost elght
times Oahu's volume of ground vater and almast oine times Oahu's
volume of surface water (19).
More than 20% of the Big Island's 93,700 inhabitants are
employed in the agricultural secter.
The per cap1ta personal income
of the county 15 currently estimated at about $8,586 (19).
Tourlsm.
the leadlng income generator of the county, has been experiencing a
deep slump in recent years.
This, coupled vith uncertain sugar
priees. continues ta affect the econamy of the Big Island (67).
hsriculture and Energy
Agriculture plays an important raIe in the economic development
of the Big Island of Hawaii.
The island's energy supply sources are
varied and range from imported ail ta indigenoue energy sources.
AGRICULTURE
The principal agricultural crops on the Big Island of Hawaii
are sugarcane, coffee and macadamia nuts.
A detailed discussion of
each of these crops is presented belaw.
Sugar
The Hawaiian sugar industry consists 'of 330 farms which control
or lease about 216,000 acres.
The industry is the third largest
in the State and its contribution ta the farm sector is
approximate1y $385 million.
The Big Island is the largest growing
area in the State (Table 4).
Sugar, the leading agricultura1
commodity of the county, is largely grown along the Hamakua Coast
and Kau district.
The graving and processing of sugar on the Island

---

14
Table 4
Sugarcane Acreage, Production and Value of Sales ln Hawaii. 1981
Cane Acreage
Raw Sugar
Value of Sales
Island
(acres)
(tons)
(1000 dollars)
Reve11
90 t 489
384.234
151,572
Kaua!
45,801
236,118
83,267
Maui
47.147
254,374
100,478
Oahu
32,662
172,815
68,262
Total
216,099
1,047.541
413.768
Source:
Telephone Interview vith HSPA. 1982.

---

15
1s damlnated by Kau Sugar Co. s Hila Coast Processing Co. (HCPC).
Davies Hamakua Sugar Co. and Puna Sugar Co.
In 1981. the total
contribution of the industry ta the county's economy vas approximately
$220 million.
Hovever, 1981 vas an exceptionally bad year for the
sugar industry.
The unusual1y low sugar priee plunged the sugar
industry ioto a severe crls1s.
The current and expected lasses are
sa large that the sugar industry 1s considering a variety of measures.
includlng reduction ln acreage, lncreases ln efficiency and
reduction in labor casts.
Macadamia Nuts
The macadam1a nut industry consists of 464 farros vhich control
or lease about 12.510 acres (26).
lt 15 an important agricultural
cr op vith an annual farm value of approximately $28 million (Table 5).
Virtually all the crop is grown on the Big Island of Hawaii.
However,
some acreage 15 being
added on Maui, although there will be no
harvest there for another five years.
The industry has a very promis-
ing future.
The eurrent and the expected priee of nuts is good. and
growers are expecting a larger erop in 1982.
Producers agree,
h~ever, that additional promotion 15 needed in the face of inereasing
supp1y.
Coffee
Coffee la also an important induatry in Havaii.
In 1981, its
farm value vas estimated at $4.5 million.
The induatry at present
consiats of 625 farms whleh control or lease about 1800 acres
(Table6).
Virtual1y a1l coffee is grown on the Big Island.
In recent

---

16
Table 5
Hacadamia Nut:
Acreage, Production and Value in Hawaii, 1981
Production
Value
Island
Acreage
(l,OOO pounds )
(1,000 dollars)
Hawaii
12,510
35,800
27,566
Kauat/Haut/Oahu
1,190
200
154
Total
13.700
36,000
27,720
Source:
Hawaii Agrlcultural Reporting Service, 1982.

---

17
Table 6
Coifee:
Acreage. Production and Value in Hawaii. 1981
Production
Value
Island
Acreage
(1.000 pounds)
(l,DaO dollars)
Hawaii
1.800
2,240
4,480
Total
1.800
2,240
4,480
Source:
Hawaii Agric:ultural Reporting Service, 1982.

---

lB
years. caifee production has experienced a continuolls decrease.
Ta
combat this decl1ne in the industry. efforts vere made to market
Kona coilee as a gourmet item at priees substantially above those of
grocery-store grades (19).
This, comblned vith the interplanting of
coffee and macadamia nuts, promises a bright future for the industry
in the years to come.
ENERGY
The State of Hawaii 15 highly dependent on foreign sources for
its energy needs.
With 92% of its eoergy derived from imported ail,
of which 64% cames from foreign sources, Hawaii remains one of the
mast vulnerable states to the full impacts of rising 011 priees and
the growing risk of supply d19ruptions (12).
The degree of these
impacts varies, however. from county to county depending on its
resource endowments.
The Big Island has an exceptionally varied
source of energy which consists of biomass, geothermal power. ocean
thermal energy conversion. wind power and hydroelectric power (28).
Biomass
Blomass i5 an important alternate form of energy that continues
to contribute markedly ta the State's quest for energy self-
sufficiency (22. 23. 24).
The Big Island of Hawaii has a varied
aource of biomass.
The biomass sources that hold out promise as
important sources of energy on the Big Island are sugarcane.
macadamla nut shells. coffee pu1~, euca1yptus and leucaena.
Estimates
of the total contribution of these biomass crops have varied
somewhat.
However, recent studies indicate that the Island of Hawaii

---

19
currently generates about 45% of its electricity tram biomass sources.
Rawal1 ts biom8S5 resources have the potential of supplying 15% of
the State's total energy by 2005 (12).
Geothermal Power
Geothermal power 15 getting increaslng attention as a source of
electric power generation.
In 1981, the first generatcr began
operation vith a promise ta supply 3000 KWH ta the Statets utility
grld.
The plant, located in the Puna district. i5 a joint effort of
the Federal government, the State. the County of Hawaii and the
Hawaiian Electric Company (HECQ).
Recently, the plant has been hit by
a series of malfunctions and equipment failures.
These have resulted
in the reduction of output and increased rate ta an average of 62
cents per month per residentia1 customer (30).
Ocean Thermal Energy Conversion (OTEC)
The Ocean Thermal Energy Conversion (OTEe) system i5 another
promising energy source on the Big Island.
In 1980. the V.S.
Oepartment of Energy (DOE) issued a Pro gram Opportunity Notice for a
c10sed cycle OTEC pilot plant of at least 40 megawatts (28).
Changes
in Administration, reorganization of the ODE. and drastlc cutbacks of
energy research funds placed the project on ho Id for over a year.
However, it was recent1y announced tbat funding for the first phase.
conceptual design. will be forthcomlng for rwo Havaii-based projects.
The OTEC potentia1, its techno1ogy deve1opment. engineering
problems. economic6. environmenta1 effects, legal issues. politica1
concerna, socio1og1cal concerna and po1iey implications and

---

20
rec~endationB have already been assessed by the Hawaii Natural
Energy Institute (HNEI) in 1981.
Wînd Power
The Big Island appears ta have one of the best wind regimes in
the world.
Ils total energy patential 15 equal ta many times the
county's needs.
The Deparonent of Heteorology, University of Hawaii.
in conjunct1on with the Hawaii Natural Energy Institute (RNEI).
has been leading ~ resource assessment program for the past decade.
This has resulted in the establishment of a Wind Energy Application
Network (WEAN) Program designed ta assess the wind power parential.
Yind Farms Ltd. has plans to establish 8 large vind machines,
produclng 500 kilowatts each, at an area on Parker Ranch just west
of Kahua Ranch on the Big Island.
Hawaii Electric Light Co.
(HELCO)
has agreed to purchase an equivalent of 4 megawatts of electricity
from Wind Fsrms Ltd.
The g~owing interest in wind farm development
and its energy potential continues ta e c erecc many ma.lnland flrms
ta Hawaii (2B).
Hyd~oelectric Power
Hydroelectric pOWer is a1so an important source of energy on the
Big Island.
The lsland's ~aiDY northern and eastern areas provide
sites for severa1 hyd~oe1ectric facilities.
Most of the facilities,
howeve~, do not have Any storage capacity and therefore operate
depending on river flow.
CODsequently, their full potential is
reached only under ideal conditions.
For instance, the hyd~oelectric
plant on Wailuku River~ which was expected to pro duce up to 3.4

---

21
megawatts of electrlclty~ vas unable ta utl!lze lts full potentiel due
ta drought conditions this year (2B).
This clearly Indicates that
gaod environmental conditions are necessary for the full realizatlon
of hydroelectric generating capacity.
Although it appears that the
expansion of hydroelectrlc capacity on the Big Island 15 feasible~ the
economlc5 of such an undertaking are unlikely ta be favorable in
comparison ~ith a number of alternative strategies (12).
The total
contribution of hydroelectricity ta the County's utility grid is
currently about 0.9%.
Although the Big Island 15 richly endowed vith indigenous sources
of energy. their full development 15 not necessarily attractive due ta
cast considerations.
Consequently. in the very short term, energy
resourees will remain the critical inputs in the expansion of
agricultural output.
Structure of the Study
The study 15 organlzed luto five chapters.
The first chapter
presents an overview of the problem and study area, and states the
objectives and the hypotheses of the study. The second chapter i5
devoted to the review of earller flndlngs as they relate ta the study.
The third chapter diseusses the analytical framework.
Specifically,
it examines the procedures of data collection, and the application of
a linear programming model ta sugar. macsdamianuts and eoffee.
The
fourth chapter analyzes the study results and the poliey
recommendatlons. and the fifth chapter presents the summary and
conclusions of the study.

---

22
CRAPTER II
REVIEW OF LITERATURE
In the early seventles and follawing the 1973 OPEC 011 embargo,
several studies have emerged relating agricultural production to
energy use.
Instead of attempting an exhaustive survey of aIl
these studies, representative studles have been chosen for revlew.
Mirst (33) provides some of the first estima tes of food-
related energy requirements in the United States.
He used data from
the 1963 O.s. input-output tables ta determine the quantities of
energy consumed in the agricultural. processing, transportation,
wholesaling and retail1ng, and household sectors for personal
coneumption of food.
The study concluded that the energy used by
the O.S. food cycle constituted about 12% of the national energy
budget.
Processed fruits and vegetables were identified to be
particularly energy-intensive with regard to both their calorie intake
and their protein content.
Flour and cereals, fresh vegetables, and
dairy products. on the other hand, were shawn to require relatively
small energy inputs per unit of food nutrient.
Following the 1973 OPEC oil embargo, many studies appeared
purporting to show that U.S. agriculture 15 an efficient user of
energy.
A common argument running through these studies 15 that the
use of energy-based inputs may be less in the future than in the past
and may constitute a severe threat to agricultural output, with
long-run implications for productivity.

---

23
Perelman (57) suggested that if efficiency 1s measured in terms
of energy input (energy requirements) in production. then U.S.
agriculture comes out very poorly.
Measuring efficiency in terms of
conservation of energy. Perelman concludes that U.S. agriculture
appears ta result in a net energy drain.
Pimentel (58) and Stelnhart (68) conclude in separate studies
chat food production Casts are higher ln the V.S. than in other
countries wlth less energy-!ntensive agricultural production
technology; furthermore, the S8me study concludes that knovn
petroleum reserves would be rapidly exhausted if U.S. agricultural
technology were employed ta produce a high-protein dlet for the entire
world population.
The ail criais also provided an
impetus for a eeries of
mathematical programming etudies of the national and regional impacts
of both increased energy costs and energy shortages on agriculture.
An exhaustive liat of such studies i8 not provided here.
However,
representative studies are reviewed to illustrate the efforts in
this area.
Dvoskin and Heady (15) analyzed United States agricultural
production under lim1ted energy supplies, high energy costs, and
expanding agricultural exports.
High energy costs as weIl as energy
shortages were found to have a slgnificant impact on both regional
crop production and regional income distribution.
An energy crisis
in the form of reduced energy supplies or higher energy costs or
both would have a severe long-run impact on lrrigated farming in the

---

24
vestern United States.
The study cODcluded chat higher energy costs
might actually prevent farmers from applying water ta their irrigated
crope.
Also Dvoskin and Heady concluded that the real hope for
irrigated farming in the long run lies in increased agricultural
exports and ample energy supplies ta agriculture.
Higher exports
promise farmers higher returns for their output and these more than
offset high energy priees; moreover the 5tudy showed chat a major
part of higher exports must come from irrigated farming and Increased
fertl11zation. bath of which are energy intensive operations.
Adams, King, and Johnston (1), in 1977, analyzed some of the
impacts of Increases in energy costs and reductions in energy supplies
on the product mix of field crops and vegetables in California.
A
quadratic programming model including risk is used ta evaluate the
effects of increased energy COst5 and reductions in fertilizer and
fuel supplies.
The model includes a demand matrix of nine field
crops and 28 seasonal vegetables.
The stucly attempted ta isolate
the velfare implications of energy changes on producers and
Consumers.
The major findings of the empirical investigation suggest
that alternative energy policies have st rang diIferential impacts.
For example. the impact of increased energy costs vas found ta fall
primarily on producers. whereas the impact of reduced fuel supplies
was found ta fall primarily on consumera.
The study raised seme
key questions about the impact on agriculture of any proposed
energy program.

---

25
Another study by Casey, Lacewell , and Jones (6) provided an
analys1s of the regional effects on agric~ltural output and
producer net returns for varylng levels of fuel restrictions in the
Southern Hlgh Plains of Texas.
Fuel shortages were found ta have
different effects on agrlcultural output and producer net returns
dependlng on the nature of the shortase (in seasou, ae harvest, or
for irrigation).
Diesel fuel shortages up to ahout 15%, during the growing
season and/or harvest, have littie effect on output and net returns,
given that the farmers adopt a reduced tillage strategy durlng the
growing season.
In contraBt, output and assoclated net returns were
found to be much more aensitive Co irrigation fuel (naturai gas)
ehortages than to die~el fuel shortages, both in season and at
harvest.
This 15 explained by the depenclence of agricultural
production on irrigation and the 1nab1lity [0 make adjustments that
would main tain yields with less irrigation water.
Ta supplement
estimates of mlnLmum output and net return reduct10ns expeeted at
varions fuel levels, the authors suggest addltionai research ta
quant ify production shifts and associated net returns that oceur with
inereasing fuel costs.
Mapp and Dobbins (49) examioed the impact of increaaing natural
;a6 priees on the patteru of 1rr1gated crop production, farm net
.neome and the quantity of water pumped through time for
epresentative farms in the Oklahoma Panhandle.
Inereasing natural
as priees were found ta have several potential effects.
First, chey

---

26
lncreased the cost of pumping irrigation water. and ether things
being equal. reduced the level of net returns associated with
irrigated crop production.
Second. shlfts from hlgh ta moderate
levels of irrigation occurred due ta changes in the water table and
pumplng casts.
Third. increasing Datura! gas priees prompted a
shift to dry cr opland production.
About a two-thirds reductlon ln
net returns accompanled lncreasing natura! gas priees and the shift ta
dry cropland production.
In addition. the following studies deserve
particular attention in the context of the proposed investigation.
Merlin (50) provided some of the latest findings ln the area.
Uslng a statlc I1near programmlng mode!, the Buthor analyzed the
effects of increases in energy priees on net revenues from crop
production.
When aIl activities. except energy priees, are fixed at
their 1977 levels. net revenues declined to $2.3 million with each 2S
percent increase in the averall cost of energy.
When energy priees
reach 206.1 percent above base levels. total production costs equal
gross return and net revenue i8 zero.
The study concluded that the
impact of rising energy priees is more severe at greater pumping
depth than for shallow irrigation wells.
Commoner et al. (10) analyzed the energy requirements for
producing fourteen different field crops in Lwenty-nine different
8ituations.
They found that along vith energy price increases
during 1970. the cost of other crop production inputs rose Just as
rapidly.
The comparative energy input costs of different crops are
measure.d as "Energy Vulnerabllity Index."
This index compares the

---

27
increase in energy input costa ta 1) the change in priee rece1ved for
the crops, and 2) the change in total variable production C05tS.
Skold (65) presents severa! adjusement possibilit1es that
farmera uslng pump irrigation systems should consider when faced with
higher energy priees.
He concluded that few producers are able ta
pas5 these increas1ng costs on ta consumera because of the nature
of agricultural markets.
Likewise, there are Ifmited opportunlties
ta 5ubst1tute other inputs for higher-priced energy inputs.
Conservation measures can help ta preserve pump irrigators but the
impact of higher energy priees la greater for pump 1rrigators than
for other producers.
Young (70) evaluated irrigation costs of representative wells
on the Texas High Plains. with inereasing energy priees along with
the break-even irrigation costs for seleeted erops with alternate
commodity priees.
Pumping costs vere estimated for a range of naturel
gas and eleetrieity priees.
He added distribution eosts ta pumping
costs ta de termine total irrigation eosts.
A vide range of total
costs was evident.
He also eompared the estimated break-even
irrigation coet.a wt tb three sets of commodity priees.
His "law
priees" are the approxima te target or support level priees for 1978;
"In tezmedf.at e forces" are set at approximately 75% of parity; and
"high priees" are approximately 100% of parity.
The break-even cast
for irrigeted vheat increases fram. $2.37 per acre-inch with "law
priees" ta $5.52 per acre-inch with "bf.gb priees."

---

28
Short (64) used a recursive, reglonal, I1near programming model
to evaluate the effects in 1990 and 2000 of the fal11ng water table.
rising energy priees and varylng exports.
The model represents
production alternatives w1th more than 2.500 rotations, each with a
different relat1onsh1p between y1elds, resource use and costs.
Production 15 constra1ned principally by available land suhdivided
according ta productlvity and production costs into 216 categories
in the Oglala Zone and 204 categories ln the rest of the nation.
The
model assumes competitive equilibr1um; it determines priees for
land, water and endogenous crops. while ether resources receive
market rates of returns.
The study concluded that both incressed
energy pr1ees and deereased exports reduee farm ineome per sere
at t rdbut ab ï.e te irrigation.
The effect of 8 doubling of energy priees
18 to 1nerease c rop priees. Incr eaee the priees of land, tnduce the
conversion of land irrisated with groundwater ta dryland, and reduee
water and energy use.
Lltterman (48) lnvestigates the relatlonship of
energy to
non-energy 1nputs, speciflcally, capital, labor and 1ntermediate
mater1als 1n twenty manufacturing sectors from 1947 to 1976.
Ta
accomplish this objective, ~o models are used. a nonlinear statlc
model and a dynamdc 11near model.
The funet10nal form of the Dan-
linear model 15 a generallzed Box-Cox cast funetlon that allows
estimation of elaatlclties of lnput dewand, economies of scale and
bias ta teehnical change wltbout a priori restriction.
The form of
the dynamic linear model 1s a vector autoregression wtth an
-

---

29
exchangeability prior.
The importaDt tesules obtained fram the
Bax-Cox cost function show that capital and energy are substitutes
and labor and energy are complements in paper products~ primary
metals and agriculture.
On the other hand, the important conclusions
froœ the dynami~ linear model Indlcate that in mast secters, capital
increases in response to energy priee increases, but this capital
increase 15 uot su~talned. indlcating that capital purcnases are
geared more toward one-time conservation measures rather chan
extensive changes in the production process.
Bellock (5) developed a structural œodel to simulate the V.S.
potato indusrry vith special emphasls on examinlng the Interregional
effecta of changlng energy costs.
The model estimates national
demande. identifies five production regions and four product forms.
Covering a saœple period from 1961 through 1978, the model is
employed to aimulate the probable impacts of changing energy prices
on total production, the mix of production forms, and regional
patterns of production.
The results of the estimation suggest that risk and energy C05ts
do not significantly influence planting decisions and that supplies
are generally highly inela5tic with respect ta expected returns.
The
supplies of the specific product forma from any g1ven reg10n are
found ta be ltnked to energy costa.
Particulerly, higher energy
costs encourage the production of processed potatoes in the North-
west and d1scourage it eLsevhere ,
acvever , the simulations do not
reveal any significant impact of energy costs on the total production
-

---

30
of each region.
It 16 argued, therefore. that the failure of the
mode! and the simulations ta detect an energy cost-regionsl
production link may be due ta the existence of thresholds. below
which energy C05tS do not impact on planttng decisions.
These and many other etudies have contributed ta the und er-
standing of the relationshlps between the agricultural sectoT and
the energy sectoT, and the potential impact of energy priee increases
on the agricultural sector.

---

31
CHAPTER III
ANALYTICAL FRAIlEl/ORK
This chapter e%amines each of the basic inputs used in the
production of Bugar. macadamia nuts and coifee.
This includes a
description of the various inputs, how they are obtained and the
manner ln which they are used in the linear programming model.
Procedures and Data Sources
Procedures
The basic data used in chis snaiyais cames both from primary and
secondary sources.
Comprehensive Buger data from secondary sources. sufficient ta
meet the objectives of the study were available.
Consequently. the
input and output coefficients for Bugar used in this study are based
largely on secondary sources.
This has been supplemented. where
necessary, with primary data.
On the other band. the macadamia nut and coifee data used vere
obtained fram eurveys of macadamia nut and coffee gravere on the
Big Island of Rawaii.
The methodology used far data collection i9
stratified random sampling vith proportional allocation.
Thus each
crop is stratified by farm size.
Sugar farma are divided into four
size categories A, B, C. and D correeponding ta less than 10, 10-49,
50-159 and 160 acres and over farm, respectively.
Simi1arly, the
macadamia nut farma are subdivided into five farm sizes At B. C. D,
and E corresponding to Ieee than 5. 5-9. 10-19. 20-49 and 50-499

---

32
acres, respectlvely.
Collee farms, on the other hand, are
disaggregated into three farm size categories corresponding tO the
ftrat three categories of macadamia Dut larme.
This classification
of farma by size and the use of stratified random sampl1ng enable us
to aseees the technology differences among different farms and their
attendant economies of Beale.
In arder tO make meaningful policy
recommendations, tt 16 crucial to take into consideration these
differences.
This makes stratified random sampling procedure more
attractive than simple random aampling.
The best sample size vas chosen by minimizlng the variance using
the follawing fo~la adapted from Cochran (9)
52
(C - CO)i~fWhCh/;C;)
• •
(1)
52
i~{WhSh;C;)
where
C
• total cost; C
- fixed cast
o
Wh • proportion of stratum h in the total population
Sh • variance of each atratum
C
• cost per unit in stratum h
h
N
total population of macadamia nutsand coffee gravera.
h•
Based on the above formulai the aemple size of macadamia nuts
and coffee gravera tnterviewed vas computed and the results are
presented in Iable 7.

---

33
Table 7
Population and Semple of F~rms Interviewed by Size. 1981
5ize Group
Population of Farms
Sample 5ize
h
Nh
"h
A
(0-4)
330
37
B
(5-9)
64
7
C
(10-19)
36
4
D
(20-499)
29
21
E
(500+)
5
1
Total
465
52
Source:
Hawaii Agriculture! Reporting Service. 1981.

---

34
Data Sources
Sugar. macadamia nuts and eoffee data obtalned from primary and
secondary sources are presented ln the following sections.
SUGAR
Data used ta st1mulate the production of sugar by the independent
grawers ve re obtained fram the cast study conduc ted by Holderness 1
Vieth, Scott and Briones in 1981 (34).
This vas checked against a
similar study done by Holderness, Vieth and Scott in 1979 (35).
Production Input
Analysts
The production of 8ugarcane by the independent growers on the
Island of Hawaii le governed by the fermerts ability to pey for his
labor. rent, machinery or equlpment. energy, herbicides, and
fertilizers.
These production input expenses constitute the major
coet components.
Labor.
Various farm operations, i.e., land preparation. seed
planting, harvesting are performed by one or a combioation of the
following eypes of labor.
Family labor.
The growing of sugarcane by independent
grawers is mostly a family operation involving the cultivation on an
average of 23 acres.
As such. the growers cultivate the1r cane on
a part-time bas1s while they vork primarily for the large sugar
planeations.
In some cases. the field work is done by the farmer and
his family members on veekenda and after heure.
FamUy labor 1s
regarded by the farmer as unpaid labor.
Hawever. in this study. 1t

---

35
18 assumed that family labor 18 valued st $5.00. the average vage
paid to the hlred workers ln 1961 (Table 8).
Horeover, the ftndings from the Btudy suggest chat a 10-49 acre
Bugar farm. in geners!. 18 more labor-intensive than the other
farma.
This farm uses about 49.92 man hours per acre for his faœl1y
labor.
This la comparable to a less chan 10 acre farm Chat uses
about 45.32 man hours per acre.
However, the 10-49 acre farm la about
one and one-haIt cimes and two and one-hall times more labor-intensive
chan the 50-159 and 160 acre and over farms for bis fsmily labor.
S1mllarlYI the family labor 16 st least three times higher than the
less than 10 acre farm and la as high as chat of the 160 and over
acre farm (Table 9).
Oversl!, the 10-49 acre farm 16 more labor-
intensive chan the less than 10, 50-159, and 160 acre and over farm,
respectively.
This seems ta suggest that the small size farms are
more labor-intensive than the large cnes.
The latter can afford
capital-intensive technologies and therefore use less labor,
vhereas the former rely heavily on their own and family labor for
their regular farm operations.
Hired labor.
Another category of labor that is used in the
production of sugarcane i8 the hired labor.
It consists of labor
that comes from off-farm.
Traditionally. the independent gravers
rely heavily on tbeir own and family labor ta verk on their fields.
This traditional source of labor has greatly changed due ta the
acarcity of family labor and the change in the size of their farm
operations.
In fact, family labor appeared to be relatively scarce

---

36
Table 8
Farm \\lage Rates by Kethod of Pay and Type of \\lork
April 12-18, 1981 vith Compatisons1
Hethod of Pay and. Type of \\lork Performed
Dollars Per Hour
AlI hlred farm workers
6.00
Paid by other than plece-rate
5.99
Paid by haut ooly
5.73
2
Paid by houri by cash vages ooly
4.45
Field workers
5.17
Llvestock workers
5.08
Machine operators
6.83
3
scpervtscr s
8.85
~erqu1Bltes Buch as room and board and housing are provlded te
Bome workers ln aIl categories.
2Includes revised estimates for some states.
3Includes ooly hourly workers not rece1ving perquls1tes.
Source:
Agr1cultural Reportlng Service, 1981.

---

37
Table 9
Sugar:
Labor Input Per Acre by Farm Size Group, 1981
(Mau Houra Per Acre)
Fartn Size
A
8
C
D
Less than
160
Type
10
10-49
50-159
and Over
Fam.11y
45.32
49.92
24.38
8.00
Hired
9.60
30.60
27.66
24.43
Total
54.92
BO.SO
52.04
32.43
Source:
(32) •

---

38
amang Hila Coast independent grcwexe ,
Most of these gravera are old
and their ch11dren have lietle des1re ta work on the farme.
In
addition, the search for efficiency has Induced some independent
growers ta adopt capital-intensive technologies chat become more
coat-effective on larger farms.
CU8tom or contract work.
This constituees an important
80urce of labor.
Becauae seme farma do not have the financia!
abiIlty ta purchase their ovn machinery, ta prepare their land, ta
plant, and fertilize. they enter into a contractuel arrangement with
the plantations chat provide mase of the custom or contract work
needed.
Vlewed in chis perspective, the custom or contract labor can
be regarded as a substituee for hlred and family labor.
Land.
The land on which most of the independent producers grow
eugarcane i9 acquired through leases either from their affiliated
plantations or large land holding estates or owned in fee simple.
Wbile the former is a common practice. the latter 1s also a fairly
common type of ownership.
In this study, the land costs as used in the production expenses
include rent for the lease operator and land charge for the home-
steader who ho1ds lands in fee simple.
Homesteads are usually
defined as a portion of the holding. limited bath as ta total aree
and value. owned and occupied by familie.s as their home.
The land cast by farm 9ize group obtaiued in this study varies
slguificantly fram a low $51 per acre ou farm A to a h1gh $110 pe.r

---

39
acre on farm B.
Although land costs on farms Band D are qulte
sim11ar, they are twlce and one and one-ha If times higher than on
farms A and C, respectively.
OYerall, the average land cost la about
$85 per acre (Table 22),
Capital.
The capital input used in this study includes ooly
farm machlnery such as tractora, sprayers and trucks used on the farm.
There are many ways in which the capital input ls measured.
Flrst,
if rentaI rates of varlous farm equipment and capital expenditures are
readily 8vailable. then the latter can be deflated by the former to
convert the capital spending aggregate lnto equipment machinery
bours.
The renta! rate of machlnery 18 then used as the priee of
capital.
Although the procedure 15 desirable when one capital input 18
coneldered, it becomes less satisfactory when different maehinery
inputs are eoncerued beeause of the variations in the rentaI
priees of machinery.
Io overcome this diffieulty, capital expendi-
tures are 1nstead deflated by an index of a11 the renta1 rates ta
obtain a measure of the rea1 quantity of farm maehinery used.
The
same veighted average of the machinery rentaI priees vas taken as
the priee of capital ~th the weights determined by the share of
eacb type of machinery in capital spending.
A1though this approach
1& auperior ta the former procedure, it neg1ects aubstitution ~thin
the capital aggregate, auch as the ehoice between airp1anes and
tractora in apply1ng fertilizers or insecticides.

---

40
Slnce the renta! rates of various types of farm equipment are
not readily available, we are unable ta use th!s procedure te
esttmate the priee of capital and the physical quantity of capital.
Instead, the study relies on the procedure suggested by Chr1stensen
and Jorgeoson (8) te construct the
•
(I-k)[qk
r
+
k
U
-
(l-qk
) J
t -1
t
kt
t-I
where Pk 18 the service priee of machinery and equ1pment; k ls the
investment tax credit; qkt-l 18 1977 value of the tracter priee
index where q 1978
• 1; r
16 the Lnte.res t rate. charged for
k t
eacbdnery and equipment; ~t 1s the replacement rate for fam
equ1pment.
The cost of capital was then calculated by multiplying the value
of capital etock by the service priee of machinery and equipment.
Based on the Aboye formula, the service priee of mach1nery and
equipment was valued at 15% ••
The analysi& shows that the capital cast ranges from a low $141
per acre on farm D ta $263 per acre on farm B.
The capital costs on
farms A and C are respectively $166 and $201 per acre (Table 22) •
• The data used ta eat1mate the service priee of capital are only
available for 1918.
Although this tends ta understate the real price
of capital, it is a more acourate figure of capital priee than the
current intere5t rate.

---

41
lt ~y be po1nted out that 1n the case where the mach1nery ts
owned and financed through loans. the cost of capital lncludes the
replacement rate of capital, the 1nterest rate charged on the 108n
and the taxes and 1nsurance pa1d.
This procedure 15 a1so an
acceptable procedure and 15 often used 1n many cost etudies.
Christensen and Jorgenson's procedure 1a a1so used to measure capital
for macadamla nuts and coftee.
Patterns of indirect energy use.
Fert11izer.
Fertil1zer inputs are cons1dered as indirect
energy inputs.
They constitute an important part of the production
axpenses.
The Bugar growers use different types of fert111zer that
are a comhination of different doses of nitrogen. phosphate and
potaah.
The d1fferent k1nda of fertillzer used by the lndependent
growers are aummarlzed in Table 10.
Fertllizer recommendatlons for sugar growlng by 8mall
Inde pendent producers are usually made by large plantations.
As
such, lt ia intereat1ng to compare the amount of fertillzer used by
lndependent gravers wlth the guidelines suggested by the plantations
(Table 11).
The application of fertilizer is usually done by hand, machine,
or a comblnation of the two.
Assuming that other variables are not
held constant, the findings suggest that farm B uses more fertllizer
per acre than any other farm size group in the study.
Specifically,
farm D uses at least 3 times leae fertillzer per acre than farms A
and C and. appro:d.mately 4 times less fertllizer than farm B.
In

---

42
Table 10
Sugar:
Fert11izer Ose Per Acre by Farm S1ze. 1981
(Pounds per Acre)
Farm S1ze
A
8
C
D
Less than
160
Type
10
10-49
50-159
and Over
A-l
953
1579
1066
548
K-I04
731
968
458
548
",,"28
448
520
A-5
788
732
A-4
838
1178
1043
Source:
(34) •

---

43
Table Il
Sugar:
Ferti!izer Recommendations by Rila Coast Processing Company
1981 (Pounds per Acre)
ta t.oon Crap
High
Law
M 1B1
700
M 1B1
700
A 2B
350
A 1
300
A 2B
450
A 1
375
A 2B
400
A 1
375
A 4
350
A 4
350
A 4
300
A 4
300
'Lent Crap
M 1B1
950
M 1B1
950
A 2B
300
A 1
250
A 2B
450
A 1
375
A 2B
400
A 1
350
A 4
350
A 4
350
lcurce e
Hilo Coast Processing Company. 1981.

---

44
most instances, the pattern of fertilizer use per acre 15 higher chan
the amount recammended by the plantations.
For example. the average
amaunt of A-l recommended by the plantation 15 about 270 pounds per
acre for bath ratoon and plant crop.
This amaunt 15 about 3, 6. 4,
and 2 cimes smaller chan chat used by A, B. C. and D farms.
respectively.
The results are summarized in Table 10.
The patterns
observed do not provide any bas1s to accept the first hypothesls.
Based on the unit priee of different types of fertl1izers used
by the independent gr avers (Table 12), fertilizer inputs constitute
the major component of the energy cost.
Specifically, fertil1zer
expenses represent about 71t. 88%, 78% and 67% of the energy casts on
farms A, B. C. and D, respectively or an average of 76% of energy
costs.
Consequently, farm B has the highest expenses of fertilizer
per acre compared to other farms.
Similarly, farm D i9 a more
efficient user of fertilizer. since it has the least cost of
fertilizer per acre.
Herbicide.
Herbicide inputs are also considered indirect
energy inputs.
Independent sugar grovers use different forms of
herbicides that make it difficult ta obtain an aggregate figure of
herbicide use.
However, disaggregated figures exist and can be seen
in Table 13.
For instance. all four types of farm use less than one
gallon of surfactant and roundup per acre.
Assuming that other
variables are not held constant. the rate of use of dowpon per acre
on B farm is about two times smaller than the rate of use on A and B
farma. and about one and one-half times smaller than on farm C.

---

45
Table 12
Sugar:
Fertilizer:
Unit Priee, 1981
Type
Unit
$/Unit
A-l
lb.
.15
H-104
lb.
.14
H-28
lb.
.12
A-5
lb.
.16
A-4
lb.
.16
Source:
Telephone interview with C. Brewer Chemical.

---

Table 13
Sugar:
Herbicide Use per Acre by Farm Size, 1981
(Pounda or Gallons per Acre)
Farm Size
A
B
C
D
Type
less than 10
10-49
50-159
160 and Over
Dowpon
lbs.
4.26
2.26
3.80
4.00
karmex
lbs.
5.13
5.80
7.00
8.00
Atrazine
lbs.
4.19
5.13
9.65
8.00
Roundup
gal.
0.21
0.11
0.22
Surfactant
gal.
0.53
0.06
0.68
1.00
TCA
lbs.
Ametryne
lbs.
2.62
1.54
2.92
Sticker
gal.
-
0.21
0.10
Paraquet
gal.
-
0.06
DCMU
lbs.
-
4.04
Source:
(31) •
..'"

---

47
Similarly. the rate of use of Karmex per acre 15 almost the same on
farms A and B, whereas farms C and D use about 7 and 8 pounds per
acre.
However, the results do Dot show that the larger the farm.
the le6s herbicide 1t uses.
The first hypothesis is therefore
rejected on that basis.
In arder ta obtain comparable values by farm size, dollar values
of different herbicide inputs vere computed.
Based on the costs of
the herbicides (Table 14). the results suggest that farm A has the
highest herbicide expenses per acre compared ta the ether farm size
groups.
Specifically. herbicide costs are about 117., 5%, 147. and
10% of the energy costs on A. B. C, and D farm. respectively.
The
average cast of fertilizer 15 about $25 per acre.
Patterns of direct energy use.
The direct energy inputs uBed
to grow and procees eugarcane are diesel, electricity, gasoline and
residual ail.
These inputs sre becoming more and more critical as the
cast of these inputs is constantly increasing.
In 1981, the Bila Coast Processing Company (BCPC) hervested and
processed about 113,573 tons of sugar grown on 10,803 acres.
Of
this, about 24,436 tons of suger vere provided by the United Cane
Planter Cooperative, a cooperative of independent growers.
The
total acre age vas estimated at 2603 acres.
Similarly, Mauna Kea
Sugar grew about 8200 acres and harvested about 89,137 tons of sugar.
Based on the total amount of energy used to produce and process
sugarcane in 1981 (Tables 15, 16, 17), the energy inputs per acre
vere derived and presented in Table 18.
Since the energy use figures

---

48
Table 14
Sugar:
Herbicide:
Unit Priee. 1981
Type
Unit
$/Unit
Dalapon (Do"'P on)
lb.
1.65
Diuron (Karmex)
lb.
3.25
Atrazine
(Aatrex)
lb.
2.18
Roundup
gal.
69.50
Surfactant
gal.
6.35
TCA
lb.
1.10
Ametryne
lb.
3.43
Sticker
gal.
7.00
Paraquat
gal.
30.00
DCMU
lb.
1.20
Velpar
lb.
20.55
Lo Drift
gal.
17.70
2, 4-D
gal.
11.25
Sencor
lb.
9.88
Source:
Telephone interview vith C. Brewer ChemicaI, 1982.

---

49
Table 15
Suger:
Cost of Direct Energy Inputs per Acre, 19B1
(dollars)
Diesel
163.02
Electricity
10.97
Gasoline
26.94
Residual 011
54.06
Source:
Telephone interview vith C. Brewer & Co. Ltd., 1982.

---

50
Table 16
Sugar:
Direct Energy Inputs Used for Processing, BepC, 1981
Type
Amount
Cost (dollars)
Diesel
1.381,600 gal.
1,349,700
Electricity
675,500 kwh
75,400
Gasol1ne
133,900 gal.
174,000
Residual oil
50,440 barrels
1,511,600
Source:
Telephooe interview w1th C. Brewer & Co. Ltd., 1982.

---

51
Table 17
Sugar:
Costs and Direct Energy Inpu~s Used for Grow1ng
Mauna Kea Sugar Company, Ine., 1981
Type
Amount
Cce t
(dollars)
Diesel
315,400 gal.
216,600
Elec tr1city
305,200 kwh
40,100
Gaso11ne
68,300 gal.
90,000
Residusl 011
733.6 barrels
31,600
Source:
Telephone 1n'terv1ew w1th C. Brewer & Co. Ltd., 1982.

---

52
Table 18
Sugar:
Direct Energy Inputs per Acre by Type and Operation
Harvesting
and
Unit
Growing
Processing
Total
Diesel
gal.
38.46
127.89
166.35
Electricity
kwh
37.22
62.53
99.75
Gasoline
gaL
8.33
12.39
20.72
Residud oil
gal.
3.76
49.04
52.S0
Source:
Telephone interview ",ith C. Brewer & Co. Ltd . • 1982.

---

53
by farm size could DOt be obtained fram Bepc, an average figure was
used ta estimate the energy inputs by type and operation.
The total
figure was then calculated and used for aIl farm s1%es.
The findings suggest that the energy input per acre used ta
harvest and process sugarcane 1a higher than that used for growing.
Specifical1y. the amount of diesel used ta process and harvest
sugarcane 15 about 3 tlmes higher than that used for growing.
Similarly, the amount of electricity. gasoline. and residual 011
used in harvesting and processing la about 2, one and one-hall and
13 times higher than that used for growing in 1981 (Table 18).
Based on the unit priee of the different forms of energy input,
the cost of direct energy inputs per acre 15 calculated and
6ummar1zed in Table 15.
The cast per unit is also presented 1n
Table 19.
Seedcane.
Seedcane 1s a short cutting of the sugarcane stalk
that 1s planted in furrows ta establiah new cane plants.
For the
independent growers, seedcane is an inFut that must be Furchased or
produced.
Based on the cast Fer ton of seedcane of about $21, the
aeedcane expenses Fer acre are at least tvice as small on farm A as
on farm C and almost ident1cal on farms Band D.
The average cast is
about $75 per acre (Table 22).
Management and overhead expenses.
Management cost is an
important part of the general expenses.
The latter inc1udes general
farm overhead and management under the budgeting procedure and
general and administrative expenaea under coat accounting procedure.

---

54
Table 19
Sugar:
Direct Energy:
Unit Priee. 1981
Type
Unit
$!Unit
Gasol1ne
gal.
1.30
Diesel
gal.
.98
Electricity
kwh
.11
Residual ail
barrels
43.00
Source:
(26. 27).

---

ss
ln fact, these costs include management Bnd executive staff office
Ixpen5es, legal fees, professional fees, and association dues for the
'Iudgeting approach.
Based on the National Eeonomics Division and
St8ti5tics Service of V.S.D.A. (52), a management fee of 10 percent
~f total costs (excluding land charge) is assumed in this analysis.
rue detalls are sunanarized in Table 22.
The farm overhead expenseS
on the other hand, include property
J
i~suranee, finaneial and legal fee5, business and legal time, and
loci&l security (Table 22).
The above approach used to impute value
to management and overhead expenses i5 assumed to be the same for
-nf f ee and macadamia nuts ,
Output and output priees.
The average yield of rav sugar and
Dolasses by farm size group used in this study vas obtained from the
l1rect survey of independent sugar grawers.
The results by farm
.tze are shawn in Table 18.
To ob tain the revenue, three output price scenarios vere
:onsidered.
The first price scenario, the current output price
Icensrio, assumes a break-even price of $440 for raw sugar and $66
for molasses.
The second output price scenario, the high ouput
)rice scenario, assumes a 40% increase from the break-even or current
)utput price scenario.
The prices of rav auga r and molasses are set
It $616 and $93, respective1y.
The low output price scenario, on
:he other hand, assumes a 40% decrease from the current output price.
l'he prdcee of raw sugar and molasses are set at $264 and $40,
~espectively.
ln addition, a ten-year time series price data for

---

56
Table 20
Sugar:
Average Yield, Ra~ 5ugar and Molasses by FaTm Size, 1981
Farm S1ze
Ra.. Suga't' (96 0 )
Holasses
Group
(tons/acre)
(tons laere)
A
9.50
2.94
B
10.08
2.57
C
13.43
2.42
D
10.09
2.47
Source:
(33) •

---

57
raw augar and molasses 1s presented to show the trend of raw sugar
and molasses priees (Table 21).
Production cost.
The production cast 18 a very important
componeot of this analyste.
The var10us cast components that make
up the production cast are summarized in Table 22.
These casts are
considered as base per10d casts in this analysis.
HACADAHIA NUT
Macadam!a nut cult1vat1on 1s a long-term 1nvestment that requires
a relatively long per10d between plant1ng and bear1ng.
Depending upon
particular env1ronmental conditions
such as the 8011, the temperature
9
and the amount of mo1sture and var1ety. macadam!a trees come 1nto
bearing st 5 ta 6 years of age.
Because of these long wa1ting
periods, banks and other agricultural production credit associations
are reluctant to provide the necessary loans that farmers need in
their first years.
The development of new macadam1a orchards requires land clearing,
preparation, pur chase of nursery stocK and continuous app11cation of
herbi~ides. fertilizer before and after plant1ng.
These operations
constitute major expenses that have an important bearing on the
decision to invest in ma~adamia nut cult1vat10n.
The growing of macadam1a nuts and the performance of these farm
operatioDS involves the direct or indirect use of labor inputs,
indirect energy inputs such as fertilizer. herb1cide and direct
energy inputs auch aa gasoline. diesel and electricity.
An analyais
of theae production inputs is presented in the follawing sections.

---

58
Table 21
Suger:
Priees of Suger and MOlasses
1972-1982
t
Year
Raw Suger 9&0 ($/ton)
Molasses CS/ton)
1972
158
2&.10
1973
180
&0.40
1974
&91
58.00
1975
320
38.20
197&
234
41.80
1977
212
27.10
1976
2&2
50.&0
1979
304
72.10
1980
554
87.90
1981
395
53.00
1982
355
58.00
Source:
2&) •

---

Table 22
Sugar:
Production Cost per Acre by Fsrm Si~e, 1981
(Dollara)
Farm SiEe
A
B
C
D
Type
Lesa chen 10
1D-49
50-159
160 and Over
Labor
187.51
156.53
122.82
91.87
Contrsct work
274.15
410.05
778.18
435.74
Seedcane and procurement
55.65
60.46
123.90
60.69
Land cost (charge and rent)
51.43
110.46
72.99
106.32
Capital coat;
166.44
262.69
201.45
140.61
Marketing processing cost
2521.28
2640.50
2281. 90
2724.79
Total energy cast
501.00
577.00
450.00
400.00
FertUizer
212.06
304.08
166.13
126.33
Herbicide
33.83
17.65
29.22
18.90
Gasoline
26.94
26.94
26.94
26.94
Diesel
163.02
163.02
163.02
163.02
Electricity
10.97
10.97
10.97
10.97
Residusl ail
54.06
54.06
54.06
54.06
Repairs
25.36
103.00
108.80
956.40
Ferm overhesd expense
34.34
133.50
269.84
1138.00
Management
99.40
158.49
189.04
301. 78
Production cast (including
marketing)
1093.00
1743.00
2079.00
3319.65
Total cost
3816.00
4613.00
4599.00
6256.00
Source:
(32) •
....
'"

---

60
Data used vere pr1marily obtained from personal interviews Yith
the gravers.
These data have been checked agaiDst the studies by
Keeler and Huang (40). Keeler and Pukunaga (41). Hamilton and
Fukunaga (25), and Scott and Karutani (62).
These studles are uBeful
sources that al10w us ta check the re1i.billty of the data collected.
Production Input
Analyste
Labor, land. fertllizer. herbicide inputs, gasol1ne. diesel
and equipment are the major inputs that are requlred ta cultivate
macadamia nuts.
Labor.
The growlng of IIl.8cadamla nuts by sma!l lDdepe.ndent
gravers la mostly a family operation tbat involves extensive
ut111zat1on of f.ml1y labor and seasonsl hlred Labar.
Most grewers
are part-time. whereas seme ethers spend st least 40 hours a week on
thelr farma.
In some cases, the field work is done on weekends by
family members.
Fam11y labor 1s usually cons1dered as unpaid Labor
by the farmer.
Bowever, in this study, family labor ia considered
as a substitute for h1red labor and is valued at its opportunity
cost.
The wage rate assumed here 1s $5.00 per hour, vhich 1s the
vage rate paid by the farmers when add1tional labor bas to be hired
to carry out farm operations. generally hsrvest1ng.
The cr1t1cal shortage of labor usually occurs between August
through Ja1DJ.ary.
October and lIovember are usually considered peak
ecnthe , although ecee nuts ee tut-e every month of the' year ,
Dur1ng
tbese periode, the 8carc1ty of labor is very pronounced.

---

61
This explains the high rate of spoilage observed in some areas
of the Big Island.
Macadamia nuts have to be p1cked off the ground
and husked ~th1n 2 or 3 days to reduce the rate of spoilage.
In
meny instances, family labor 1s insufficient and has to be
aupplemented by "outside" Lebor ,
The findings appear to suggest a negative correlation between
farm size and the use of family labor, i.e., between 6mall-s1ze
farms and family labor and large-size farms and hired labor.
For
example, farm. A uses about one and one-half, 2, and 17 t1m.es more
family labor than farms B, C, D, and E, respectively.
S1m.ilarly,
farm E uses about 9 and 19 times more hired labor per acre than
farma B and A, reapectively (Table 23).
Labor input ls certainly a critical factor for the independent
growers.
The reason is tbat alternative employment opportunities
outside agriculture, auch as in tourigm and construction, exist and
are highly pa1d.
Bowever, in recent years, labor-saving mechanical
harvesters have been developed for large growers or groups of
cooperating gravera as weIl as smell growers.
Although these
harvesting devices _ay be attractive to large growers, they are too
expensive to attract small growers.
Based on the rate of $5.00 per hour charged for labor, the
labor cost constltutes an average of 60% of the production cost.
In reality, _ost of the farmera do not tmpute BOY coat to the1r ovn
and family labor.
Bovever, a realistic assessment of the1r

---

Table 23
Macadamia Mut:
tabor Input per Acre by Size and Type, 1981
(Man-Hours per Acre)
Fam Sir.e
A
B
C
D
B
Type
LeaB than 5
5-9
lD-19
20-49
5D-500
FSIIi1y
340
247
216
162
20
Hired
12
25
118
200
225
-
Total
352
272
336
362
245
Source:
Persona1 interview with gravers. 1982.
'"
N

---

63
production cost must !nclude the opportun1ty cost of th~ scarce
resource (Table 34).
Land.
Mac:adamia nut acreege has increased signif1cantly in recent
yesrs.
From 1971 to 1980, the total and bearing sc:reage of mac:adamia
nuts increased 96 and 33 percent, respect1vely (26).
Similarly. the
number of mac.adamla Dut farma increased from 295 in 1971 to 465 in
1980, which represents a 59% lncrease.
Of these 465 farma, about
83% consiat of Ieee than 10 acres snd produc:e a smal1 portion of
m&csdamias harvested.
The large portion of the total output cames
from the small percent of large gravera.
The land on wh1ch macadamia Dut 1e grown 18 e1ther abandoned
suger or cofee farma aCQu1red through lesses fOrlll pr1vate and public
institutions or owned by the growers in fee simple.
Although the
former ls found ln Most instances. the latter ls a very common type
of ownership that is found in Kona.
Soil. natural vind protectlon, elevetion. rainfall end
accessibillty for harvesting and cultural operations are important
factors to be consldered ln the cultivation of macadamia nuts.
Although the crops have proven best adapted to mlld. frost-free,
subtropical climates vith at least SO lnches of annuel rainfall
weIl distrlbuted throughout the year, macadamia trees can tolerate
and survive aild frosts and drought conditions.
tn Hawaii,
macaduda trees srow best be raeen 700
and 1800 foot elevations and
vhere there is sood, natural vind protection or adequate, planted
vuu!breal<s (25).

---

64
In addition. macadamis trees appear to grow 8uccessfully on a
variety of Baval1an so118 rang log from loose voleante lava 8011e to
well-dralned. lateritic elays.
In most instances, the relat1vely
low amounte of nitragen, phosphorus and potassium in the s011 have to
be supplemented by fertl11zers in arder to increase yields.
Lesses
vary in cost depending on land productiv1ty and location.
For
instance, farme in central Kona close to the main highway tend to
have greater rental cost than those that are note
Land cost as
used in this study includes land rent and real property taxe
The
findlngs suggest an average land cost of $149 per acre, ranging trom
a low of $110 per acre on farm C to a h1gh of S171 per acre on
fsrm D (Table 24).
Capital.
Capital 1& a aomewhat difficult input to quant if y in
production economic theory.
Empirically, depreciation and interest
are often uaed as proIies ef capital cost.
Since seme of the macada-
mia nuts
gravera do not allow for deprecistion of their farm
equipment, the value of capital is multiplied by the service price of
capital developed earlier to ebtatn the capital cost used in the
analysis.
Farm equipment used to grov macadamia nut ia varied.
Depending 00 eacb farm situation. the type of e~uipment used includea
a combination of husker. drier, trucks or farm trailer. power sprayer.
knapsack sprayers. hlower and mechanical harvester.
Other
miscellaneous materiala include band tools. pruning shears. sickles.
picks and sbovels.

---

Table 24
Macad8.llia Mut:
Insueence , Incereat , Land ccer , Depreciation. and Capital Cost per Acre by Size. 1981
(Dollars)
Farm Size
A
ft
C
D
ft
Coat
Lesa thaD 5
5-9
10-19
20-49
5G-499
Insurance
105
91
144
115
124
Interest
133
89
142
150
160
Land ccee
(l'en t and tu)
116
170
110
171
162
Depreciation
100
422
156
225
190
Capital
246
268
150
170
157
Source:
Personal interview vith gravers. 1982.
0-
u

---

66
Based on the above accounting procedure, the fiodings Buggest a
negatlve correlation between farm size and capital cost per acre.
Specifically, capital cest appears to be a decresslng function of the
farm size. 1ndicat1ng economles of Besle.
For exemple, the capital
cost on farm C 18 about one and one-half times smaller than the
capital cost on farms A and B, respect1vely. and almost the same
on farm E (Table 24),
Patterns of indirect energy use.
This section explores the
patterns of indirect energy inputs (fert111zers and herbicides) used
to gray macadamla nuts.
Fert111zer tnputs.
They constitute a major part of the
production expenses.
The types of fertilizer often used by the
macadamla gravera lnterv1eved are comb1nat1ons of d1fferent amounts
of phosphorouB. potasaium
and nitrogen, name1y 16-16-16, 14-14-14
.nd 10-15-20.
Although most growers uae a cambioation of the above, based on
location and the particular soi1 characteristics, some gravera tend
to concentrate on a particu1ar type of ferti1izer.
For exemple, the
A farm uses 800 pounds per acre of 16~16-16, 200 pounds of 10-15-20
and none of the other ferti1izer inputs, whereas the E farm uses a
cambioation of 260 pounda of 16-16-16, 30 pounds of 1'-1'-1' and
143 pounds of 10-15-20 per acre (table 25).
In soy case, the
resu1ts suggest that the 1arger the farm size, the 1ess ferti1izer
it uses per acre.
The rate of ferti1izer uae is found to be
dependent on soi1 types, farm location and on the particu1ar needs of

---

Table 25
Macadamia Nut:
Fertilizer Inputa per Acre by Size, 1981
Fam Size
A
B
c
c
E
Type
Unit
Leu than 5
5-9
10-19
20-49
50-500
16-16-16
lb.
800
620
617
HO
260
10-5-20
lb.
200
800
640
143
143
Source:
Personal interview vith gravera, 1982.
'"
~

---

68
each farmer.
The results obta1ned do Dot lead ta the rejectlon of the
firet bypothes1s.
7ertl11zer input expensee constitute an average
of 42% of the energy cast.
Overall. the fert111zer input expenses
appear ta he bigher chan the expenses for other energy inputs
coosldered in this study (Table 34).
Herbicide.
Weed control 18 perhaps the mast e%pensive and
one of the most important factors in nursery management.
Pre- and
post-planting weed control 18 often done e1ther by power sprayers or
knapaack sprsyers.
F81110g to control weeds from the initial
plant1ug can greatly retard the growth of macadam!a crees and result
in 1ncreased cast of weed control.
The different types of herbicides
used are Paraquat, Roundup, and in some instances Atrazine, Karmex
and 2, 4-D.
The f1ndings appear to suggest a significant variation in the
rate of herbicide application per acre by size depending on
particular needs of the farmer.
For instance, although farms B
and D appear to use the same amount of Paraquat per acre, farm E uses
about 2 and S t1mes less Paraquat per acre than farma C and A,
respective1y.
The 1arger farm appears to be more efficient than
other farma, a1though in seme cases the results are mixed (Table 27).
Based on the cost per unit (Table 28), the herbicide expenses
repreaent about 16% of energy expeDses.
In addition, the herbicide
appears to be a decress1ng function of the farm size, exhibiting a
stroug econamy of seale.
This provides factual evidence to accept
the first hypothesis.

---

69
Table 26
Macadamia Nut:
Unit Priee of Fertil1zer by Type, 1981
Type
Unit
$!Unit
16-16-16
80 lb. bas
17.50
14-14-14
50 lb. bas
37.00
10-5-20
80 lb. bag
13.50
Source:
Direct interview w1th C. Brewer Chemicals , 1982.

---

Table 27
Hacadamis Nut:
perbicide Use per Acre by Type and Size, 1981
(Gallon or Pounds per Acre)
A
B
C
0
E
Unit
Leu than .5
5-9
10-19
20-49
50-499
Paraquet
gaL
1.17
0.43
0.78
0.50
0.35
Roundup
gal.
0.47
0.53
0.32
0.11
Warfar1n
gal.
-
-
0.43
0.11
0.49
2. 4-D
geL,
0.38
-
0.13
Atrazine
lb.
-
-
0.33
D1uron (Karma)
lb.
-
-
0.43
0.11
Source:
Persona! interview with gravera, 1982.
~
o

---

71
Table 28
Macadamia Nut:
Unit Priee of Herbicide by Type. 1981
Type
Unit
S/Unit
Paraquat
gal.
62.50
Round.up
gal.
76.00
Warfarin
gal.
11.25
2, 4-D
gal.
15.20
Atrazine
lb.
3.50
Diuron (Karmex)
lb.
16.00
Source:
Telephone interview with Brewer Chemical. 1982.

---

72
Patterns of direct energy inputs.
Direct energy inputs do not
contatitute a major portion of the production expenses.
The growing
of macad~1a nut by the 1ndependent gravera 18 not a heav11y
mechan1%ed operation.
Consequently, the direct energy used does not
contribute signifieantly to the cast of production.
Gaso11ne 18
mainly used to operate the trucks. power sprayers and jeep tra11ers
used on farm.
Diesel 1a used for operat1ng tractora and electr1c1ty
1e a1so consumed wb11e husk1ng the nuts.
The finci1ngs 1nd1cate the follow1ng.
Gaso11ne coasumpt1on on
farm D 1a about 2. 4. 3, and 3 and one-half times smal1er than on
farma A, B. C. and E. respectively.
Farm B, on the other hand,
appears to use more gasoline per acre than any other farm cons1dered
in the Btudy (Table 29).
Similarly, the aame farm consumes more
diesel than any other farm considered.
The electricity relationship observed shows that farm D uses
more electricity than any other farms (Table 29).
The conclusions emerging fram this analysis are:
firet, the
amount of dieael fuel consumed appears to be a decreasing function of
farm si~e (assuming that other factors are not held constant).
The
larger the farm size, the less diesel it uses per acre.
Second, the
rate of use of gasoline and electricity, on the other hand, exhibits
• U curve on which the minimum ia reached on D and C fars,
respectively.
That i., the rate of gasoline and electricity use
appears to be first a decreasing function of farm size and then
atarts increasing from farma D and C, respectively.

---

Table 29
Macadamia Mut:
Direct Energy Inputs per Acre by Size and Type. 1981
A
B
C
C
E
Unit
Lesa than 5
5-9
10-19
20-49
50-499
Gasoline
gal.
19.17
35.00
30.40
11.00
26.93
Diesel
gal.
-
29.17
20.00
15.00
3.62
Elec tri city
kwh
6.67
2.50
21.00
39.25
Source:
Personal interview with gravers, 1982.
<j

---

74
Based on the unit priee of dtff~rent types of direct energy.
direct energy costs represent an average of 47% of the total energy
expenses (Tables 30, 31. 34).
Output and output priees.
The aver ege yield of macadam.1a nuts
(in shel1) by size used in this Btudy was obtained trom personal
interviews vith small growers.
The results are 8ummari~ed in
Table 33.
Ta arrive at the revenue, three output priee scenarios vere
considered.
The firet priee scenario, the current output priee
scenario, 8ssuaes a b~eak-even priee of 90 cents a pound.
The second
output priee scenario, tbe high output priee scenario, assumes a 40%
increase from the break-even priee.
The priee of macadamia nuts 18
set at S1.26 a pound.
The low output priee scenario. on the other
band, assumes a 40% decrease fram the break-even priee re9ulting in
an output priee of 54 cents a pound.
A ten-year time series data
of maeadamia nut priees is .lso presented in Table 32.
Production cost.
ln this study. production cost i9 used in
eombination vith grosa revenue to der ive net revenues.
A summary of
the production cee ee 18 preaented in Table 34.
GOFFEE AND HACADAMIA MUT INTERPLAN'IING
Coffee and macadamia pUts are becaming inereasingly interplanted
on the Big Island.
In liant of tbis. in this section, eoftee and
macadam1.a DUt are be.1ng treated as joint pr-oduct e and it is assumed
tnat the amounts of inputs per acre uS8d for both crops are
identical.

---

75
Table 30
Hacadamia Mut:
Unit Priee of Direct Energy Inputs by Type, 1981
Type
Unit.
PLiee (dollars)
Gasoline
88 l.
1.&7
Diesel
gal.
l.00
Electrlclty
J<>,h
12.74
Source:
Direct interview vith growers. 1982.

---

Table 31
Macadamia Nut:
Production Input Cast per Acre by Type and Farm Size Group. 19B1
(Dollars)
A
B
C
D
E
Type
Lees than 5
5-9
1().-19
2().-49
5().-499
Labor
1760
1360
1680
1819
1210
Land cast (including
tex)
116
170
110
171
162
'ert11izer
210
272
246
101
81
Herbicide
115
67
88
42
28
Diesel
-
29
20
15
4
Casoline
32
5B,45
51
lB
45
Electric:ity
85
32
26B
497
Total energy cast
442
458
672
673
158
Total coat
2318
1988
2462
2654
1530
Source:
Personal interview vith growers. 1982.
~
'"

---

77
Table 32
Macadamia Nut.:
Priee 1972-1982
Fana Priee
Yen
(cents per pound)
1972
23.3
1973
25.5
1974
32.0
1975
31.6
1976
36.9
1977
40.8
1978
53.8
1979
62.9
1980
72.4
1981
77 .0
1982
90.0
Source:
(26).

---

Table 33
MScadam18 MUt:
Yield per Acre by Size, 1981
(Pounds per Acre)
A
B
c
D
E
Type
Lesa than 5
5-9
10-19
20-49
50-499
Yleld
2649
4012
4115
4520
5759
Source:
Persona1 interview with growers, 1982.
~
œ

---

Table 34
Macadamia Nut:
Total Cast per Acre by Type and Psrm Size, 1981
(Dollars)
P8 rtIlI S il.: e
A
B
C
D
E
Type
Leas thaa 5
5-9
10-19
2D-49
5D-499
Lam coa t and taxe.
116
170
110
171
162
Labor
1760
1360
1680
1810
1210
Indirect energy inputs
325
339
333
143
109
Fertil1zer
210
272
245
101
81
Herbdcdde .
115
67
88
42
28
Direct energy iaputs
117
119
339
530
49
Casoline
32
58
51
18
45
Diesel
-
29
20
15
4
Electricity
85
32
268
497
Capital cost
246
268
150
170
157
Interest on opersting
capital
133
89
142
150
160
Total energy cast
442
458
672
673
158
Management
258
218
265
286
168
Production cast (excluding
management)
2697
2345
2754
2974
1847
Production cast
2955
2563
3018
3260
2015
Source:
Personal interview with growers. 1982.
~
~

---

80
Grovlng other marketable crope between the tree rows durlng
the tirst years 15 not necessarily an irrational decls1on.
First. it
reduces the rlsk of sudden change in fanm priees and. therefore
lncomes, if the farmer vere ta cultlvate ooly one crop.
Second. not
ooly does intercropping reduce great fluctuations of farm income.
but when intelligently carrled out, lt may even be beneflcial ta the
trees bec8use of improved 5011 fertility and weed control.
However,
such crops must Dot be planted 50 close ta the trees as ta interfere
wlth thelr development.
Adequate spacing of 35 ta 45 feet between
rows 15 recommended for eoiiee and macadam la nut lnterplanting (25).
Data used here vere obtained primarl1y from persona! interviews
with eoifee growers in Kona.
These data vere che~ked against the
atudies by Keeler, Ivane and Matsumoto (42), Fukunaga (20) and
Baker (2).
Among the coffee growers interviewed, a large per~entage
intercrops macadamia Dut vith coffee.
The ~ont1nuous increases 1n
coffee and macadamia nut priees in re~ent years have made inter-
planting an attra~tive proposition to the grovers.
Consequently, both coffee and ma~adamia nut growers do not knov
the number of a~res. hours. the amount of fertilizer or herbicide,
and the amount of direct energy inputs used ta grow coffee or
macadamia nut separately.
ID some instances, however, some fa~er5
who grow coffee exclusively vere interviewed and their information was
used Beparately in this study.
In those instances where intercropping
is pra~ti~ed, it ia 8Bsumed in thia study that the amount of inputs
per a~re used for ~offee and macadamia Dut is identical.
However, in

---

B1
the analysis of the production inputs, efforts are made ta describe
the various inputs vith special reference ta the typical problem
that the eoifee industry 16 fseing.
Production Input
Analysis
Labar.
The labor problem in the eoifee industry 15 almost the
same as that of macadaJJ11a nu t s ,
Essentially, i t centers almast
exclusively on the harvesting operation.
Most of the fermer e must hire
labor ta meet st least 300 hours required on farm A, 400 hours
required on farm B and 200 hours required on farm C ta pick one acre
of cherry eDifee (Table 35).
The harvesting period 16 usually in September and extends through
March or April.
The labor 15 critical durlng the se periods.
The
labor problem la complicated by the fact that eoifee does not ripen
st one time.
The orehard must be harvested many times in arder to
obtain 75 bags of parehment and 55 bags of eherry whieh are the
average yield of parehment and eherry per aere.
The skill of the
pieker and the nature of the field are two major faetors that
determiue the pieking rate of the worker.
Some farmers reported that
a good pieker ean piek as many as 4 bags per day at a rate of 4 man
hours per bag.
The eoffee pieker is pa id about 514 per bag of eherry
eaffee pieked.
Therefore the wage rate assumed here 15 about $3.50
per hour.
Land.
eoffee aereage has experieneed a deeline in reeent years
to the benefit of maeadamia nuts.
From 1971 to 1981, the in erop
and bearing aereage has deereased by 51 and 37 pereent , respeetively.

---

82
Table 35
eoifee:
Labor Input per Acre by Farm Size, 1981
(Man-Nours per Acre)
A
B
C
Lesa than 5
5-9
10-19
Fami1y
306
448
249
Hlred
12
25
118
Total
318
473
367
Source:
Persona! interview with grcve rs , 1982.

---

83
Stm11srly. the number of farms has decreased fram 750 in 1971 to 650
in 1981. a 13% decline.
Bishop Estate 16 one of the lessors of land on which eoifee 15
grown.
The land tenure system varies somewhat. ranging
trom lease-
bold to ownership in tee simple.
Land cost varies greatly trom
location ta location.
The reasons for chis variation, mentioned in
the discussion of macadamia nucs. are applicable for coi tee as weIl.
The procedure used ta impute value to land 15 iclentical ta that
used for macadamia nuts.
Specifically. land cost will 1nclude land
rent and rea! property taxes.
The findings suggest that the average
cost of land 15 about the seme as that of macadamia nu cs (Table 43)
ainee macarlamis nuts and eDitee are 1nterplanted.
Capital.
The machinery used in eoffee farm9 in Kona ls alm09t
identical to that used for macadamia nuts.
Because of the steep
slapes an which most of the farms are situated. farmers continue ta
use jeeps in conjunction with power sprayers.
Some farmers have also
mechanical driers in Kona.
Other structures commonly found on Kona
caifee farma are warehouses for storage and water tanks to wash the
coffee.
The accounting procedure used to measure capital i9 the same
as that used for macadamla nuts.
Due to the intercropping of maeadamia nut9 with coffee. the
capital costs are assumed to be identical for coffee and macadamia
nut farms for the sizes considered in this analys1s.
Fertll1zer.
Fert1lizer 18 a prime determlnant of yleld
and quallty in caffee production.
lt constitutes an important part

---

84
of the production expenses.
Farms A, Rand C use an average of at
least 350 Founds of coffee cherry (10-5-20) 1n arder to grow one acre
of coftee.
S1ml1arly. st least 500 pounds of ferti!izer Mac 8
(10-10-10) are needed to grow one acre of coffee.
These results
suggest that the rate of use of coffee cherry is a decreasing funct10n
of the farm size and that the use of the other fertil1zer (MaC 8)
eppeat-s to reach a max1lllum on farm B.
Although the former ls con-
sistent vith the tirst hypothesls. the latter does not provlde soy
basis to warrant a conclusion (Table 36).
The unit priee of
fertilizer by type ls also presented 1n Table 37.
Herbicide.
Weed control la a continuou5 farm operation that
le becoming more and more expens1ve as the coat of the herbicide
inputs is constantly increasing.
The herbicide inputs used are
varied.
Farmers uae Paraquat. Roundup. 2. 4-D, Atrazine and Diuron
(Karmex).
The findings appear to suggest a significant variation in
the rate of application per acre by size.
For example. farm A uses
about 4 and 2 times more Paraquat per acre than farms Band C,
respectively.
A similar conclusion can be reached for Roundup
(Table 38).
In any case, the results appear to show that larger
farms use less herbicide per acre than the small ones.
The unit
price of herbicide by type used is also presented in Table 39.
Patterns of direct energy use.
Coffee grcrwing is not a highly
mechanized operation.
It ie rather a highly labor-intensive enter-
prise.
Gaeoline ie the frequently used form of energy as the farmers
drive their jeep to fertilize. spray and prune coffee plants.
The

---

85
Table 36
Coifee:
Fertil1zer Inputs per Acre by S!ze and Type
(Pounds per Acre)
A
•
C
Type
Lesa than 5
5-9
10-19
Cof fee cherry
405
454
230
(10-5-20)
Mac 8
200
800
640
(l0-10-10)
Source;
Personal interview vith grawers. 1982.

---

86
Table 37
Coffee:
Unit Priee of Fertilizer by Type, 1981
Type
Unit
$/Un1<
Coffee cherry
(10-5-20)
lb.
.16
Mac 8
(10-10-10)
lb.
.17
Source:
Telephone interview with c. Brewer Cheeu ca l , 1982.

---

87
Table 38
Coffee:
Herbicide Input per Acre by 5ize. 1981
A
B
C
Type
Unit
less than 5
5-9
10-19
Paraquat
gal.
3.00
.71
1.14
Roundup
gal.
l.42
.58
.44
2, 4-D
gal.
.38
.39
Atrazine
lb.
l.00
l.00
1.00
Dluron (Ks:rmex)
lb.
.40
.11
Source:
Personal interview with growers. 1982.

---

88
Table 39
Coifee:
Unit Priee of Herbicide Input by Type, 1981
Type
Unit
I/Unlt
Paraquat
gal.
62.50
Roundup
gal.
)6.50
2, 4-D
gal.
15.20
Atrazlne
lb.
3.50
DluroD (Xermex)
lb.
16.00
Scurce s
Telephone intervie~ vith C. Brever Chemical, 19B2.

---

89
findlngs auggest that farm A uses about 1 and one-ha If and 2 times
more gasoline than farms C and B, respectively.
Consequently, the
cast of gasol!ne used per acre 15 higher on farm A than on any other
farm !ncluded in the study.
The same pattern of electricity use can
be found as we compare the three farms.
Also. only farm B reported
having used about 21 gallons of diesel in 1981 (Table 40).
In this
case, the results do not exhibit any economy of scale.
Outputs and output priees.
The output of coffee that 15
considered in this analysis includes cofiee cherry and parchment
eoffee.
The average yields obtained are summar!zed in Table 41.
Tc
arrive at the revenue, three output priee scenarios are Identlfied:
high. current and low output priee.
The current priee is set at
$2 a pound for parchment and $1 a pound for coffee cherry and
ref1ects the break-even priee.
Whi1e the high priee i5 set at $2.80
and $1.40 a pound, respective1y, the 10w priee is about $1.20 and
$0.60 a pound and represents a 40% increase and decrease from the
break-even priee or current priee scenario, respective1y.
A ten-year
trend of coffee priees is also presented in Table 42.
Production costs.
Production costs, exc1uding management cost.
are summarized in Table 43.
These costs are weighted against cne
revenue to arrive at the net revenue per acre by farm size.
The Linear Programming Madel
Rapid changes in input priees. great fluctuations in farmers'
income resultlng from cyclical changes in crop priees and sud den

---

90
Table 40
eoffee:
Direct Energy Inputs per Acre by Size and Type. 1981
A
B
C
less than 5
5-10
10-l9
Gasoline (gallon)
36.20
18.02
26.25
Diesel (gallon)
20.84
Electricity (kwh)
4.59
3.22
5.00
source:
Personal in terview wi th growers, 1982.

---

91
Table 41
Caffee:
Yield per Acre by Size and Type, 1981
(Pounds per Acre)
A
B
C
Type
Lesa thap 5
5-9
10-19
Parchment
1,566
14,590
1,746
Cherry
6,967
2,968
8,250
Source:
Personal interview with gr cwers , 1982.

---

92
Table 42
eoffee Priee, 1972-1982
Year
$/pound
1972
.35
1973
.50
1974
.56
1975
.46
1976
.75
1977
1.85
1978
1.38
1979
1.26
1980
1.43
1981
1.60
1982
2.00
Source:
(26) •

---

93
Table 43
eaffee:
Total Cost pel' Acre by Type and Farm Size, 1981
(Dollars)
A
B
C
Type
Less than 5
5-9
10-19
Land cost + taxes
116
170
110
Labor
1113
1655
1284
Indirect energy inputs
405
307
261
Fertilizer
99
209
146
Herbicide
306
98
115
Direct enet'gy inputs
124
92
107
Gssoline
60
30
43
Diesel
21
Electricity
64
41
64
Capital ccs t
246
268
150
Interest on operating capital
133
89
142
Management
202
241
194
Production cost (excluding
management)
2137
2581
2054
Energy cos t
529
399
366
Production coat
2399
2822
2246
Source:
Persona! interview with growers. 1982.

---

94
priee changes in their energy-hased inputs are investlgated here with
the help of a 11near programming model.
In the followlng section, the
notation, the assumpt10ns and the general formulation of the model
with reference to specifie cases are presented.
Notation
The following notation ls used to formulate the model as
applled to sugar, macadamia nuts and eoifee.
X
15 the number of acres of produced crap q on farm
"~
1 of type j
19 the average yield of processed type l of crap q
per Bcre on farm i of type j
19 the priee per ton or pound of processed type 1
of crap q
la the unit cost per acre of resource k used to produce
crap q on farm 1 of type j
d
15 the processing cost per acre of crap q on farm 1
"ij
of type j
~ij
18 the amount per acre of reSDurce k used to produce
crop q on farm 1 of type j
K
1B the total amount of processed type l of crop q
l q
produced

---

95
B
is the minimum amount of processed type 1 of crop q
1q
sold
~qiJ
15 the total emcunt; of resource k allotted ta produce
crop q on hm i
of type J
where q • l, 2, 3· 1 • 1, 2, 3, 4. 5
•
Assumptlons
The linear programming model ls based on the following
assumptions:
Assumption of proportionality.
Linearity 15 assumed in both
the objective function and the constralnts formulation.
This implies
chat, in the objective function, each activity taken separately ls
directly proportional ta the leve1 of that acttvity.
In the
constraints functions, chis implies a constant return ta scale.
Assumptfon of additivity.
This implies that the total
&mOunt of aIl activities he equal ta the sum of each activity taken
separately.
ABsumptton of divisibility,
This impl1es chat factors can
be used and commodities can he produced in fractional quantities.
Assumption of certainty.
This implies that the coefficients
of the model are fixed and known with certainty.
Consequently,
output and input priees and reeource coefficients are aseumed fixed,
i.e •• non stochastic for each scenario.
The current output priee, as 88sumed in this study, corresponds
to the break-even priee.
The high and low output priees, on the

---

96
other band, are assumed to represent, respectively. ~O% increase and
decrease irom the break-even priee.
In addition, it 1~ assumed that the inputs used ta gray macadamia
nut and cotiee. vith the exception of fertl1izer inputs which can be
eaaily disassociated. are Identlcal in an interplanting situation.
Finally. it 15 also assumed that the study cavets only small
Besle grovers of 8ugar, macadamia nut and coffee on the Big Island of
Havai!.
General Formulation of the Model
OEJECTI\\~ FONCTION
The objective function is to maxtmi~e net revenues or profit
derived from the production of field crope:
Max
[ [ (p
r
j
qI
i
subject to the following constraints.
RESOURCE AVAlLABILITY CONSTRAINTS
[
[
[
j
i
k

---

97
This constraint states that the amount of resource k allotted ta
produce crop q on farm i of type j cannat exceed the total resource
available.
PRODUCTION CONSTRAINTS
[
[
<
j
i
This cODstraint states that the amount of processed type 1 of
crop q produced on farm i
of type j cannat exceed the total processed
type 1 of crop q aval1able.
MARKETING CONSTRAINTS
[
[
>
J i
This constralnt states that the &mount of processed type 1 of
crop q produced on farm 1 of type j must be at least equal ta the
total amount sold.
NON-NEGATIVITY CQNSTRAINTS
>
o
>
o
This constra1nt states that the activity levels must be e1ther
zero and/or positive.

---

98
Case l
Wben ~ • 1, 1 .. 1, 2, we have respective!y raw 8ugsr and
molasses
where
1 .. 1, 2.
-
• 37
j " l , 2 , 3 , 4
k .. 1. 2.
- • 8
Case 2
Wben q .. 1, 1 .. 3, we have macadamia nuts (in shel1)
where
1 - 1 , 2 , 3 .
- • 52
j
.. 1, 2, 3. 4, 5
k - l . 2 , 3 ,
-
• 8
Case 3
Wben q .. 3. l .. 4, 5, we have respectively eoffee (parchment)
and coffee (cherry)
where
1 .. 1, 2, - - - - 52
j - l , 2 . 3
k - l . 2 . 3 ,
- 8
The cODstraints used in the linear programmlng are baaed on the
maximum amouut chat the loan institutions are currently willing ta
provide to the small gravera.
Although chis amoune varies vith the
expected priee of the crOPt ies allocation among the various
production inputs reflects past records chat grawera have established

---

99
vith the bank or loan institutions.
Once the 108D ls aprpoved, any
transaction by the grawer must he carried out through the cooperative
of independent gravera ta make sure that the amount 15 spent for the
production activities speclfled.
Consequently. the grower has no
input substitution posslbilities as he faces incresses in the
production inputs priees.
Since almost every crcp grower ls a
prospective bank borrower. the 10ao institution does play a key ra le
in the succesa of the suger, macadamia nut and eoffee lndustry.
Based on interviews with 10an officers, the monetary and
physicsl resource constraints used in the I1near programming model
were generated.
The constralnts for suger are presented in
Appendix Tables 28 and 29.
Similar constraints can be aeen in
Appendix Tables 30. 31, and 32.

---

100
CHAPTER IV
STUDY RESULTS AND THEIR POLIer IMPLICATIONS
This chapter 15 organlzed inta tvo parts.
The first part
presents the study results.
The second part discusses their paliey
implica tians.
Study Results
This section presents the results obtained fram the study of the
impacts of higher energy cost on the production of agricultural crops
on the Big Island.
The relatlonships between energy costs and the
production of crops have been examined under three output priee and
three energy cost scenarios.
The three output priee scenarios are
current, high and low output priee scenarios.
The current output
priee scenario corresponds ta the break-even priee, whereas the high
and low output priees correspond ta a 40% increase and decr8ase fram
the break-even price.
ihe energy cast scenarios are EC 0, EC 50,
Bnd EC 100 indicating the base period, 50% increase and 100% increase
in energy cast, respective1y.
For bath output priee and energy cast
scenarios, the year 1981 i6 considered as the base period, since data
used vere for that yeer.
Each output priee scenario is examined
separate1y under the three energy cast scenarios.
Suger
In this section, 1n vhich energy cast accounts for about 10%
of the cast of growing sugar, the impacts of higher energy cast are
examined under three output pr1ce
and three energy cast scenarios.

---

101
THE CURRENT OUTPUT PRIeE SCENARIO
Under the current output priee scenario, the raw sugar and
molasses priees are set at ~440 and $66 per ton, respectively.
The
reBults of the firat energy cast scenario (Ee 0). i.e •• the hase
period energy cast scenario in which energy casts constitute an
average of 10% of the total cost of production are presented in
Appendix Table 2.
The results show that given the various
constraiots that face aIl independent gravers and the current cast
scenario, farms A and C appear ta he the optimal sizes ta grow
Bugarcane.
In addition. aIl the dual values, vith the exception of family
and hired labor. show zero shadow priee.
The shadow priee of
labor of about $14 for family labor snd ~92 for hired labor reveals
that only the use of labor can add to net revenues.
Cansequently, a
reallocatlon of the Input mix in favor of labar may lead to a
greater revenue.
The second energy cost scenario 15 EC 50. 1.e.~ a 50% Increase
ln energy cast.
The results shov that a 50% increase in energy cast
has not changed the number of acres of eugar grown and the amount of
raw sugar and molasses produced.
Hovever. some changes were observed
to result fram a 50% increase ln energy cost.
The first change ls
an 18% reduction in the farmer's net revenues.
This decrease ln
net revenues le amaller than the SOI increase in energy cost.
This
result suggeets that the farmer's net Income ls Inelastlc witb respect
to the changes ln energy costs (Appendix Table ) .

---

102
The second change that occurred in the EC 50 scenario la a
reductlon ln the shadow priee of family and hired labor.
This result
suggests that sinee energy cost has Increased and the shadow priee
of family and h!red labor has decreased. a rational farmer, 1.e.,
the farmer who has the ultimate goal of maximizing profit or net
revenue, can still utl11ze more family and hlred labor st the expense
of the more costly energy inputs.
Alternatively, the farmer cao
use the same technology but 15 compelled to conserve the use of
energy resources ln the production pracees.
The 50% increase in
energy cost has decreased the shadow priee of family and hired labor
by 91~ and 3~, respectively.
The third energy cast scenario (EC 100), 1.e.,
a 100% increase
in energy cost, presents s somewhat different result for the
independent growers.
The findings show that a lOO~ increase in energy
cost from the base period will result in a 33~ decrease in net
revenues for the independent sugar growers.
As a result. only farm C
ie found to produce eugarcane. and only hired labor appears to have
a poetitive shadow price.
This result suggests that if the farmer's
objective is eolely to maximize net returns. the use of additional
units of hired labor slone can increase his net returns (Appendix
Table 4).
These results imply that a reallocation of the input mix
by eubstituting the binding reeource (labor) for the unused resources
may reduce cost and therefore may lead to greater profit for the
growers.

---

103
The above discussion has led to the conclusion that higher energy
costs have not severely impacted on farmera' net revenues.
Whereas
a 50% fncrease in energy cost has resulted in an 18% decrease in net
revenues. a 100% increase in energy cost has reduced revenues by
33%.
These results suggest that the net revenue 15 insensitive to
energy costs under the break-even or current output priee scenario
(Appendix Table 24).
THE HICH OUTPUT PRIeE SCENARIO
Vnder the hlgh output priee scenario, the raw sugsr and
molasses priees are set at $616 and $93 per ton, respectively.
The first scenario 15 the base period energy cost scenario.
According to this. the energy cost averages about 10% of the total
cost of production.
The resuIts of the base period scenario are
summarized in Appendix Table 5.
The results show that farms A and C continue to be the most
efficient farms to grow sugarcane, given the resource constraints
and the current energy cost scenario.
At the optimal activity
levels
farme A and C maximize their profit by utilizing aIl the
t
family and hired labor, 15% of A-4 fertilizer, 10% of Roundup, 6%
of gasoline, 42% of diesel, 0.02% of electricity and 0.01% of
residual fuel.
In addition, the dual values, with the exception of
famd1y and hired labor
have zero ehadow priee.
For instance, the
t
zero ahadow price of fertilizer i5 due to the large &mount of
fertl1izer that remalng unused.
Consequently. an attempt by the
farmer to use more fertilizer will merely leave the farmer'g net

---

104
revenue unchanged.
This 1s also true for aIl the direct and indirect
energy resources used ta produce sugarcane.
Only family and hired
labor have positive shadow priee which implies that the use ~f
additional unita of labor wi!! add $90 per acre ta the profit in
the case of faml1y labor and $191 per acre in the case of hired
labor.
Labar. the binding resource. can be substituted for the unused
resources and by reducing production cast increase the net revenue
of growers.
The second energy cast scenario represents a 50% increase in
energy cost (Appendix Table 6).
The results show that a 50% increase
in energy cost has not changed the number of acres of sugar and
consequently the amount of r8W sugar and molasses sold by farms A and
C.
However, the fo11owing changes were observed as a resu1t of a
50% increase in energy cost.
The first change observed in the primal solution resu1ts in a
65% decrease in net revenue.
This dec1ine in farm net revenues is
considerab1y 1ess than a 50% increase ln energy cost.
This auggests
that the farmer's net incame is not sensitive to the energy cost
increases under the high output price scenario.
An alternative
exp1anation is that the farmer's income is very ine1astic to the
changes in energy cost under the high price scenario.
The second change observed 15 a reductlon of the opportunity
cost of operator and hired Labor input.
Under this scenario, the
opportunity cost of fami1y and hlred 1abor ls estimated at $77 and
$187, respectlve1y.
This decrease ls in fact due to the higher cost

---

105
of energy which implies that a rational farmer can reduce his energy
consumption and use instead an addit10nal unit of family and hired
1abor, sinee the latter will increase his net revenues.
The third energy cost scenario (EC 100), in which the average
energy cost accounts for 26% of the total cast of production, has not
changed the optimal solutions (Appendix Table 7),
The earlier
conclusions reached for the first and second scenarios are aIs a
found valid in this case.
That is, although energy cost accounts for
26% of the total cast of production, the higher output priee tends
ta negate the effects of higher energy costs.
A mathematical
explanation 1s that, although the objective funetien coefficients
have changed. the resulting changes in the slope are not large enough
to shift the objective function to another feasible solution.
Although the solutions of the choice variables have not changed,
the net revenues have decreased by 13%.
This suggests that under
the given energy cost scenario, a 1007, increase in energy cost will
result in only a 13% decrease in net revenues which implies that net
revenues are very
inelastic to the changes in energy costs.
Similarly, a 100% increase in energy costs has resulted in 29% and
4% decrease in shadow prices of family and hired labor, respectively.
This imp1ies that an additiona1 unit of labor hired will add $64 to
net revenues in the case of family labor and $184 in the case of
hired labor.
Alternatively, given the EC 100 scenario. the farmers
can substitute energy inputs for hired and family labor and increase
net revenues by $184 and $64. respectively.

---

106
THE LOW OUTPUT PRIeE SCENARIO
Under the low output priee scenario in which the priee of raw
sugar and molasses 15 set at $264 and $40 per ton, respectively, a
sensitivity analysis 1s performed for the three energy cost scenarios
considered.
In aIl three cases, includ1ng the base period en erg y
cost scenario, none of the farms 15 found ta produce sugarcane.
AlI
the choice variables 1n the primal as weIl 36 in the dual are equal
ta zero which implies that, if the objective 1s ta maxim!ze net
revenues, sugar production 15 not profitable uoder this scenario.
The results of the base period scenario are presented in Appendix
Table 8.
The ab ove discussion leads ta the conclusion that the impacts of
higher energy costs appear ta be more critical ta farmers under low
output price scenario than under current high output price scenarios.
For instance, at identical increases in energy cost, say 50%, net
revenues have registered an 18% decrease under the current output
price scenario.
In addition, a stmi1ar increase in energy cost has
resulted in a 65% decrease in net revenue under the high output price
scenario, whereas net revenues vanish under the low output price
scenario.
This suggests that lov sugar priees tend to reinforce the
impacts of higher energy coste whereas higher suger priees tend to
negate them (Appendix Table 25).
Macadamia Nuts
The energy input accounted for about 16% of the cost of growing
macadamia nut in 1981.
Labor input, the most critical input J

---

107
conetituted about 60% of the total cast of production.
ln chis
section, the impacts of higher energy costs on the production of
macadamia nuts are examdned in terms of chree output priee scenarios
(current, high and law priees) and three energy cost scenarios
(EC 0, EC 50, EC 100).
THE cvaRENT OUTPUT PRIeE SCENARIO
In the current output priee scenario, the priee of macadamia
nuts 15 set at 90 cents per pound.
In the base period. in which the energy cost averages about 16%
of the production cost, farms B and E appear ta he the optimal sizes
ta grow macadamia nuts.
In facto an analysis of the production expenses shows a
consistency between the resules ohtained and the cost of production
on fares B and E as compared ta other farms.
Specially, these two
farme appe~r to be the least-cost producers of macadamia nuts or
under given out~ut priee, the revenue maximizers.
In production
theory, when all farms face the same output priee and the same or
different input priees, farms with the least cost of production are
selected as optimal farm sizes.
This is the only justification for
farms B and E sbowfng up in the programming solution.
The results of the base period scenario are shawn in Append1X
Table 9.
Although these results may appear unrealistic in light of
the performance of the macadamia nut sector today. they can be
justified for the follawing reasons.

---

108
First. these results are optimal solutions and reflect the
performance of rational farms, i.e., farms that have the sole goal
of maximizing net revenues.
In reality. we know that macadamia nut
farmers may have various economic and non-economic goals that cao
keep them ln production, even if they are not maximizing
net
revenUES.
Second, this study adopts a somewhat different approach ta
production cast accounting.
Family labor has been given imputed
mODetary values which mast farmers do not do in rea! life situations.
Both of these are important considerations that may explain the
divergence between the optimal solution and the actua! situation.
Viewed 1n this perspective, the results that are obtained here cao be
regarded as a framework within whleh the actual performance of eaeh
farm ean be examined.
There Is, therefore, a eonsistency between
these results and the real situation.
It ls important to point out here that the solution of the dual
problem shows a positive shadow priee for only diesel and eleetrlcity.
These shadow priees are $25 and $123. respeetively.
These indieate
that under the base energy cost scenario, farmers can inerease their
net revenues ta $25 for each additionsl gallon of diesel consumed
and $123 for eaehadditionalkilowatt of eleetricity used to grow
macadamia nuts.
Since aIl the other resources included in the study
have not been used, their shadow priees are zero
and are not
therefore effeetlve in Inereasing the net revenues of the growers.
It 1s implied, therefore, that diesel and electrieity, the binding

---

109
resources. cao readily he 6ubstituted for the unused resources and
the resu!ting decrease in cast may le ad to a greater profit.
The second energy cast scenario 16 EC 50, i.e., a 50% increase
1n energy casts.
The results show that a 50t increase in energy cast
has not changed the optimal acreage and production obtained
p revdous Ly , but has r-e su Lt ed in a 45% decr-ee.se in ne t
revenues.
This
lmplies that net revenues are not sensitive to the changes in energy
casts or are inelastic to the changes in energy casts.
However, some
changes were observed in the dual solutions.
While the shaclow priee
of diesel has decreased. that of electrlcity has slightly increased
(Appendix Table 10).
The thlrd energy cast scenario 15 EC laD, i.e •• a 100% increase
in energy costa.
The sudden inerease of energy costa has not changed
the optimal solutions but has œerely redueed the farmer1s net
revenues by 9%.
It has also deereased the shadow priee of diesel
by 64% and increased the ehadov price of elec.tricity by 2% (Appendix
Table Il).
It i5 important to note that in aIl three energy cost scenarios
considered, net reVenues are very insensitive to the changes in
energy ccs te ,
ln addition, only diesel and e.lectricity appear ta be
the binding constraints in aIl three scenarios.
A genera1 assess-
ment of this output priee scenario suggests that macadamia nut small
growers are not seriously affected by higher energy cast under the
current labor-intensive production technologies.
However, a shift or
a change iD the current technology to a more capital-intensive one

---

110
may affect the results. assuming that capital and energy are
complementary inputs.
THE HIGH QUTPUT PRICE
In the high output priee scenario, the priee of macadamia nut
(in-shell) 15 set at $1.26 a pound.
In the first series of I1near programming problems in which
energy casts reflect the base period, i.e., 1981 energy cast
situations, farms Band E continue ta be the mast efficient production
units.
A careful look reveals that farms Band E have the least cast
of production among the slzes cODsldered in this study.
The results
are presented in Appendix Table 12.
ln addition, the solutions of
the dual actlvltles reveal that diesel fuel and electricity are in
fact the bidning resources in growing macadam!a nut.
The shadow
priee of diesel is about $69 while that of electricity is about $187.
Diesel and electricity are therefore potential candidates to increase
net revenues.
Specifically, an additional amount of diesel and
electricity will add about $69 and $187. respectively. to the net
revenues.
Farmers may find it attractive to use more of the binding
resources (electricity and diesel) and less of the unused resources
which may result in a reduction of production cost and lead to
greater profits for the growers.
The second energy cast scenario (EC 50), i.e., a 50% increase in
energy costs, has not changed or affected the optimal solutions but
has reduced the net revenues from growing macadamia nut.
A 10%
increase in energy cast has reduced the net revenues by 3%.
This

---

111
change ln net revenue 15 very ioslgnlf1cant compared to the 50%
increase in energy costs.
This shows that net revenues are very
insensitive ta the increases in energy costs.
Farms Band E there-
fore remain the optimal slzes ta grow macaclamla nut (Appendix
Table 13).
A careful look at the primal and dual activity solutions reveals
that electricity and diesel are the ooly binding reSDurces. indicating
a positive shadow priee ln the dual actlvity.
In this case, the
shadow priees of diesel and electricity are about $61 and $188.
repsectively.
This indicates that an additional amount of diesel and
electricity consumed will add about $61 and $188. respectively. ta
the farmers' profits.
If the objective ls ta lncrease net revenue.
more attention must be given to these ~o energy resources.
The same
argument advanced earlier is also valid in this case.
The third energy cost scenario (EC 100), i.e., a 100% increase in
energy costs, has resulted in only a 5% decrease in net revenue.
The optimal activity levels have remained unchanged (Appendix
Table 14).
As explained earlier, these results must be vieved with
care.
First, they reflect the situation of typical farms, i.e.,
efficient farms.
Second, it is assumed in this analysis that the
overall goal of the farmer is to maximize net revenues.
In real
life, we knaw that farmers can pursue various economic as weIl as
non-economic goals that may keep them in business, even though they
are not max1mizing net revenues.
In addition, this study adopts a
different accounting procedure, i.e., it imputes values to family

---

112
labor that farmers do not in fact accoUDt for in their cost studies.
These factors tend ta expIa in why on1y farms Band E appear in the
optimal solutions.
A close scrutiny of the high output priee scenario reveals that
energy cost increases have not caused a seriaus decrease in net
revenues_
This 15 due ta the fact that energy resources are not
essentiel inputs to graw macadamia nuts.
Macadamia nut growing 15 in
fact a labor-intensive operation in which labor cost constitutes
about 60% of the production costs.
These labor costs include unpaid
family labor as weIl as hired labor.
Since labor 15 a very expensive
input on the Big Island and ainee on1y diesel and electrlcity appear
to show positive shadow priee. a substitution of capital and energy
(with the exception of e1ectricity and diesel). for lahor may be a
possible way to reduce production costs and therefare increase the
profitability of macadamla nut production.
(The underlying
assumption here is that capital and energy are eomp1ementary inputs.)
THE LOW OUTPlfI' PRlCE SCENARIO
In the law output priee scenario, the output priee of macadamia
nuts (io shell) is aet at 54 cents a pound.
The first series of linear programming problems for the base
period revea1 that macadamia nut production is still profit~ble if
the farmers were ta face the current input prices.
The resu1ts of
the base period acenario reveal that ooly farm E appears ta he the
most efficient unit ta graw maeadamia nut under the low output price

---

113
scenario.
ln addition, only electricity has a positive shaclow priee.
The results of this scenario can be seen ln Appendix Table 15.
The second energy cast scenario (EC 50), i.e., a 50% in crea se
ln energy costs, has not changed the optimal levels of activity. but
has reduced the net revenues by 77..
The shadow priee of electricity
remalns positive.
The results of this scenario are summarized ln
Appendix Table 16.
The third energy cast scenario (EC 100), i.e., a doubling of
energy cast. has merely reduced the net revenues by 14% and the
shadow priee of electricity by 15%.
Farm E remains the optimal site
to grow macadamla nut uoder the low output priee scenario.
The
results are presented in Appendix Table 17.
The above analysis of the macadamia nut priee scenarios leads to
the conclusion that higher energy costs do not have a significant
impact on the production of macadamia nut in general.
For instance, a
50% increase in energy costs has resulted in a 4.5% decrease in net
revenue under the current or break-even priee scenario, only a 3%
reduction under the high output priee scenario and a 7% decline
under the low output priee scenario.
Similarly. a 100% increase in
energy costs has resulted in a 9% decline in net revenue under the
current output priee scenario, a 15% decrease under the low price
scenario, whereas only a 5% decline was observed under the"high
output price scenario.
The above findings warrant the conclusion that the higher the
output priee, the lower the impacts of higher energy costs on the

---

114
farmer's net revenues.
Law macadamia nut priees tend ta reinforce the
negative impacts of higher energy casts, whereas high priees tend ta
negate them.
It can also he readily inferred that. although higher
energy casts do have differential impacts depending on the ercp priee,
these impacts appear ta be more significant for sugae than for
macadamla nuts.
lt can further be pointed out here that although
energy cast represents about 10% of the total production cast of
sugar, as opposed to 16% for macadamia nut, the resulting decrease in
net revenue arising from hlgher energy costs 15 greater for sugar
than for macadamla nut.
Therefore. these results do not appear ta
warrant the conclusion that the more energy intensive the production
of an agricultural crop is, the more vulnerable it i5 ta energy costs.
Crop price seems to play a more important raIe in the reduction of
the farmer's income than do in creas es in energy cost as we compare
sugar with macadamia nut (Appendix Table 27).
Coffee and Macadamia Nut Interplanting
Coffee is becoming increasingly interplanted with macadamia nuts
on the Big Island.
With the exception of fertilizer, most farmers
interviewed are unable ta disassociate the inputs used to grow each
crop.
For this reason, coffee and macadamia nuts are treated here
aa joint products.
Bawever, in arder to asaess the impacts of higher energy costs
on the production of coffee alone, 1t is assumed here that the

---

11S
inputs used ta grow bath crops. vith the ~ception of fertilizer
which csn he easily disassociated. are identical in an Interplantlng
situation.
In this section. the impacts of hlgher energy costs are examined
vith reference ta three output priee scenarios (current or break-even.
high and lov priee) and three energy coat scenarios (EC O. Re 50.
EC 100).
THE CURRENT OUTPUT PRI CE
In the current priee scenario, the priee of coffee (parchment)
and coftee (cherry) la set st $2 and $1, respectively.
The first series of the linear pragramming are ta measure the
actlvity levels st the base perlod energy costs.
In this scenario.
in which the energy Cost averages about 18% of the total cast of
production. farms Band C appear to he the mast efficient units of
coffee growing.
The results are presented in Appendix Table 18.
It is important ta point out here that these results only show the
performance of the most efficient farma, given the available
resources and do not have to sum up ta the current acreage or
current production.
ln facto these results only reveal that if the
objective is te maxtmize net revenues given available resources.
farma B and C appear ta be the optimal sizes and they must produce
sa many pounds of cof f ee ,
In r ea Lf ty , we knov that farm A exists
and grows coffee.
The justification is that in the real world,
farmers do not only pursue the profit max1mization goal.
They may
pursue. ether eccncafc a:sW./or ncn-econcertc goals that ma.y keep them

---

116
in growing coiiee. even though they are Dot maximizing net revenues.
However, the profit maximization goal 18 assumed as the sole goal
here because ft reflects the general goal of the growers.
The
solutions of the primal problem CBn be seen in Appendix Table 18.
Similarly, the dual variables with the exception of the pro-
duction have zero shadow priee.
This implies that chose inputs for
which the cast 15 zero have no alternative uses, i.e., these inputs
vere Dot fully utilized in the primal problem.
Only the production
activitles show a positive shadow priee.
These reveal in the linear
programming framework that the net revenue can be Increased by
increasing the production of coffee.
These shadow priees are $1.85
and 76 cents for parchment and cherry, respectively (Appendix
Table 18).
The second energy cast scenario (EC 50), in which energy cost
has increased by 50% from the base period, has only resulted in an
insignificsnt decrease in net revenues and in the shadow price of
coffee (parchment) production.
No other changes were observed in
this scenario.
These findings show that the changes in energy costs have
insignificant impacts on the net revenues of coffee gravers.
This
decrease in net revenue, estimated et 0.08%, ia far 1ess than the
increase in energy costs.
Net
revenuea are therefore found ta be
inelestic to the changes in energy costa.
The reaults of this
scenario are presented in Appendix Table 19.

---

117
The third energy ~08t scenario (EC 100), in wh1ch energy cost
bas Buddenly increased by 100% from the ba~e per1od, has also
resulted ln a reductlon of net revenues.
This decrease, whlch 15
about 15%. 1s the only change observed ln this scenario.
The results
show that net revenues are still Inelastlc in the face of 100%
incresse in energy cast (Appendlx Table 20).
The major conclusion emerglng trom this analysis 15 chat under
the CUrTent priee scenario, the net revenues obtained irom the growing
of coffee are not vulnerable ta the increases in energy casts.
This
can he ~lalned by the nature of coiiee production technologies that
sre eS5entlally labar-intensive.
THE HICH OUTPUT PRICE SCENARIO
Varler the high output priee scenario ln which the priees oi
coffee (parchment) and coiiee (cherry) aTe set at $2.80 and $1.40,
respectively, energy cast averages about 18% of the total coat oi
production in the base period.
In the first series of the linear programming problem, only
faDnB B and C still remain the most efficient units ta grow cofiee.
The resu1ts can be aeen in Appendix 'Table 2l.
The justification advanced earlier under the current output priee
scenario ia also valid ln this case, i.e •• the optimal solutions
reflect the performance oi the most efiicient farms, given available
resources and with the singulat' goal to maximi:te profits.
The results are presented in Appendix Table 21.
Only coiiee
productioo activities show a positive shadow priee in the dual

---

118
problem.
These activities are, therefore. conducive to increasing the
net revenues under this scenario.
The second (Ee 50) and the thlrd (Ee 100) energy cost scenarios
have resulted 1n ooly 0.05% and 0.1IZ detrease in net revenues,
respectively.
The resulta cao he aeen in Appendix Tables 22 and 23.
THE LOloJ OUTPUT PRIeE SCENARIO
Under the lav output priee scenario. in vhich the priees of
coffee (parchment) and (cherry) are set at $1.20 and 60 cents,
respectively, energy costs represent about 18% of the production
costs.
Based on these priees. ou1y farma Band C continue ta be
the mast efficient units ta grow coffee, given avatiable resources.
The resulta are pre~eDted in Appendix Table 24.
The second energy cost scenario (Ee Sa) has resulted in only a
1.4% decrease in net revenues, although no other changee in the
optimal solutions vere observed (Appendix Table 25).
The third (Ee 100) energy cost scenario appears to have a
signif1cant impact on net revenues although the optimal solutions have
Qot changed.
In th1s scenar10 higher energy costa have reaulted 1n
a 2.8% deerease in net revenues.
The laver output price tends to
reinforce the impacts of hg!hQr euergy CO~t6 compared to those
observed under the high output price scenario.
Consequently, the
net revenues of the coffee gravera appear to be more vulnerable to
b1gher energy cosU under the lav output priee scenario than the
high output priee scenario.
The results of the third scenario are
summari:r.ed 1n Append1:J: Table 26.

---

119
The above analyais has led to the conclusion that higher energy
costs tend te have differential impacts depending aD crop priee.
For
instance, for coffee. a 50% increase ln energy cast has resulted in
a O.OSt decreaee ln net revenues under the current priee. whereas
a ~imilar Increase reduced net revenues ooly by 0.05% under the
hlgh crop priee.
However. similar Increases in energy costs have
resulted ln a 1.4% decrease in net revenue under the low output priee
scenario.
This confirms that low priees are 8seociated vith greater
impacts arteing trom higher energy costs.
Similarly. 100% incrseases
ln energy costs have resulted in a s!gnlflcant decline of revenues
by 23% under the law crop priee, whereas the observed decreases are
ouly 0.11% and 0.15% under the high and current output priee.
respectively.
This again confirms that law crop priees tend to
reinforce the impacts
arising fram higher energy costs. ~hereas high
crop priees tend to negate them (Appendix Table 27).
This analysis of the impacts of higher energy cost on the
production and D@t revenues of the three field crops (sugsr.
ma~adamia nut. and coftee) enables us to test the second hypothesis.
i.e., the more energy-intensive the production of a crop i5 in
relation to other crops, the more vulnerable it Is to energy cost
increases.
Ear1ier in our discussion. it ~aS shawn that coffee productiou
~aS more energy-intensive than that of macadamia nut and Bugar.
Specifica11y. energy costs constituted about 18% of the total coat.
vhereas they represented about 10% and 16% of the total cast of

---

120
growing sugar and macadamia nut. respectively.
Before proceedlng , it
18 worthwhl1e ta point out chat the seemlngly high energy Intenslty
of eoifee growing reflects the assumptlon made earller. i.e., sinee
eoffee and macadamla nut are interplanted and thst the energy and
non-energy inputs cannot be dissassociated, it was assumed that
input use 18 Ident1cal for bath crops.
Baaed on the above information. one can conclurle that at
Identlcal energy cost increases, the impacts must be greater on
eoifee revenues than on macadamls nut and sugar revenues.
The results
obtained do not appear ta corroborate the hypothesis that the more
energy-intenslve the production of an agricultural crop Is. the more
vulnerable it 18 to energy cost increases.
Such an 1nterpretat1on, hovever, must be made v1th care.
The
optim1zat1on problem that one 1a try1ng to solve here 1a a constra1ned
optimization problem, as opposed to a free opt1m1zation.
Although
the model that 1e ueed ta s1mulate the three crops is structurally the
eame, the constra1nte that face growers are not the same.
Con-
sequently, a direct compar1son may be m1sleading although lt 1e
usually doue.
However, it can be indirectly impl1ed that at equal
dollar received, sugar product1on or revenue Is more vulnerable to
hlgher energy cost than that of other crope.
It May therefore be
concluded that the above analysis does not seem to conflrm the
hypothee1e tested.

---

121
Policy Implications
The principal findings emerglng from the study resu!ts and thelr
poliey implications are examined ln this section.
A genersl scrutiny
reveals that hlgher energy costs have not greatly impacted on the
net revenues of 8mall growers, but do have differential impacts
depending on the resource endowments of each crop grower.
A generally
observed phenomenon ls that the lower the output priee, the greater
the impacts on the net revenue from crop growing under given energy
cost scenarios.
In sny case, net revenues appear ta be inelastic
ta the changes ln energy costs.
These resu!ts can he attributed to the followlng ressons.
Firat,
this atudy cavers only independent small growers of Buger. macadamla
nut and eoifee.
The production technologies of these three crops are
essentially labor-intensive.
Second, the production of these crops
consists largely of dry farming. i.e., most of the growers do not
irrigate.
Both of these factors tend to minimize the potential
impacts resulting from higher energy cast.
In fact, small scale growers do not usually use capital-intensive
technology.
Large farms, on the other hand, tend to take advantage
of the economies of scale and thus could afford a capital-intensive
tecbnology.
So, if the underlyiug assumption is that capital and
energy are complementary inputs, capital-intensive production is
associated with energy-intensive technology.
With given technology.
large growers can easily spread their cost over large acreage,
whereas small grovers have 11mited opportun!ties.
Cougequently, the

---

122
impacts of higher energy costs Are expected ta be greater on large
crop growers, sinee these farma are irrigated and are also capital-
intensive.
On the Big Island of Havaii, Laber 15 a very scarce resource.
Mbst of the independent growers, although they use a good desl of
family labor, are co=pelled ta hlre additional Labor ta carry out
tbeir harvesting operations.
In the case of sugar. Labor cast represents only 13% of the
total cost of production.
Ibis relatively Law cost of Labor may be
explsined by the fact that the harvesting operations of sugar growers
are mechaolcal and are carried out by Rila Coast Processing Company.
Consequently, the Labor input that la accounted for consista largely
of family Labor and a amell portion of hlred Labor.
Since Labor and energy contribute almost equally to the
production cost of growing sugar, i.e., the production of sugar does
not appear to be elther labor-intensive or energy-intensive, it is
difficult to recommend any input adjustment or substitution policy for
the independent growers.
However, since the results of the ltnear
programming model show a positive shadow price for labor in the dual
and since labor appears to be the only input that cao increase net
revenue, a larger budget must be allotted to the labor input.
We are
not recommending here that financial or lending institutions should
increase the total amount allotted to grawers.
Instead, we are
suggesting a reallocation of the current budget or input m1x in such
a vay that greater weight is given to the Labor input.
lt should be

---

123
noted here that the way the banks provide loans ta 8ugar growers does
not allov for input substitution st present.
Farmers are loaned B
fixed amount per acre ta carry out specified production activities.
The results of our study suggest that a more flexible bank
poliey vith respect ta the provision of loans la in arder.
This 15 Bn
important consideration that should be given greater emphasis in the
future in arder ta enhance the competitiveness of sugar.
Policy
makers seem to be overwhelmed by theproblems facieg suger al the macro
level~ without taking in ta account the coostraints that gravers face
at the micro level.
Theae problem5 are probably very crucial and
should not be ignored.
Celeris parlbus, 8 farm that has large input
substitution possibilities is certainly in a better competitive
position than one without such options as it faces increases in its
energy priees.
Perhaps. it provides at least a partial explanation
as to why sugar growers cannat afford even a slight decrease in their
crop priees.
It is important to note that this situation ie in no way fully
responsible for aIl the problems facing Rawaii'e sugar industry.
However, this fact should not be neglected.
It must be incorporated
into a comprehensive policy at the county or State level. sinee a
relaxation of these conatraints would permit sugar growers greater
input substitution posaibilities.
This could very weIl reduce
production cost and enhance the profitability of sugar in Hawaii and
its eompetitiveness in the vorld market.

---

124
Also. in chis analys1e, an sttempt 1a made to answer this
question:
What is the optimal si~e(a) at wh1ch suger cao be grown,
if the objective 13 ta max1m1ze net revenue?
The solving of the
firat 6eries of lineat ptogramming model reveals chat ooly farma A
and C~ i.e •• less chan 10 acres and 50-159 acres. are the optimal
farm slzes ta grow suger under the various resource constraints.
The
appearance of ooly these fsrms in the solution does not mean that
farms B and D should disappear.
A change in current resource
endowment may lead ta a change in the optimal scale as reflected in
the 1n1tial solution.
Perhsps, the existence of these farms appeats
ta he consistent with other economic and/or non-economic goals that
are not measured in chis seudy.
Macadamia nues and coffeé are found to be very profitable under
all output price scenarios.
The findings revesl that, although the
lobor cast of macsdamda nuts and coffee averages about 60% and 55% of
the total cast of production, energy casts are ooly 16% and 18%.
respectively.
The relatively large share of labor input may be
explained by the labor-intensive harvesting operation.
Durini the interview. moet of the farmers expressed their concern
about the lack of pickers dur1ng harvesting operations.
Sa labor
lnput poses a real problem.
In the face of tbis increasingly scarce
~esource, a reallocat1on of the farmer's budget 1n favor of energy
Lnputa and therefore capital inputs <assuming capital and energy
(re camplementary inputs) may increase the prof1tab1lity of
~cadamda nuts.

---

125
lt should be pointed out tbat in the linear programming model,
diesel and electricity are the binding reSQurces.
These exhibit
positive ehadow priees in the dual and therefore are the only inputs
that may contribute to an increase in net revenue.
50 a reallocatlon
of the farmerts budget in faver of these inputs, mostly diesel, may
grestly reduce production cost and augment the profitability of
macadamia nut on the Big Island and the State.
In the case of coifee, under the current and h1gh output priees.
aIl Chat 15 needed seems ta he an increase in the quantity of coffee
that 15 produced to enhance the profit of the growers.
It should
be noted that a!though coilee appears to have a larger percentage
of energy costs than macadamianuts. the impacts of higher energy
costs on the net revenues are greater on macadam!a nut chan on
coffee.
At least two reasons can be mentioned.
First, in recent
years, efforts vere made ta market Kona coffee as a gourmet item at a
price substant1ally above that of grocery-store grades.
Second,
the constraints that face coffee and macadamia nut grovers are not
the same. although the model used and its assumptions are almost
identical.
In Any case. the net revenues for both crops are in-
elastic to the changes in energy costs.
The algorithm of the linear programming model has also ensbled UB
to identify the optimal scales to growth bath macadamia nut and
coffee.
Only some farm sizes appear in the actual solutions.
These
do not Buggest that the faTm5 that do not appear should not ex!st.
Their existence may be found to he consistent vith other econoœic

---

126
and/or non-economic goals thet are not measured hers.
Hovever. if
the objective Is to maximize net revenues, these results provide some
indication of where the emphasis should he placed.
Alternatively.
these fanœs and their technology can readily serve as typical farms
and technology against which real farm performances can he tested.

---

127
CHAPTER V
SUHHARY AND CONCLUSIONS
In recent years, drsstlc changes have occurred in input priees,
output priees and in the institutional structure wlthln whlch
agriculturel producers opera te.
These changes are largely the upshot
of sharp increases in energy priees that are directly or Indlrectly
translated into higher output priees for the consumer.
Needless to
say. the energy situation has exerted and continues to exert pressure
on the U.S. economy by contributing to inflation.
The U.S. food system has developed into a system characterlzed
by intensive use of energy in ferm production, processing,
and
transportation of fsem products.
At eaeh stage, energy inputs are
used elther in the farm of gaso!lne. diesel, gas, electriclty or in
the form of pestl~ldes and fertlllzers as needed to grc~. manufacture
or transport the agricultural products.
As the priees of these
critical energy inputs in agriculturel production increase, output
priees have experienced cyclical changes that pose a serious threat
to the farm sector.
Summ.ary
The study examinea the interrelationships beeween the energy
Bector and tbe production of three crops (augar, macadamia nut,
and cofiee) by small gravera on the Big Island of Hawaii.
Specifi-
cally. lt attempts:
(a) to explore the patterna of energy use in
agriculture; (b) ta determine the relative efficiency of fuel uae by

---

128
s1ze amang the three field crops; and (c) ta investigate the impacts
of higher energy costs on the production and net revenues of the
t hr ee crops on the Big Island of Hawaii under three output priee
and
energy COBC scenarios.
Ta meet the objectives of the study, primary and secondary data
vere obtained.
Data collection procedure 15 stratified 't'andam
sampling with proportionsl allocation~ although in the case of Bugar,
secondary data vere obtalned and 6upplemented, where necessary, vith
primary data.
Eech crap la stratified by farm aize.
5ugar ferros are divided
into four s t ee categories, A. B. C. and D. co r respcndfng ta lees thsn
10. 10-49. 50-159 and 160 acres and over fatlD5. respec.tively.
Similarly, macadamia nut farms are Bubdivlded into five fa~ sizes.
A. B~ C. O. end E. corresponding to less than 5. 5-9. 10-19. 20-49.
and 50-499 acre farma. respective1y.
Coffee farms , on the other hand.
are represented by three farm categories. A. B. and C, that
correspond to the first three categories of macadamia nut.
This
makes stratified random samp1ing procedure more attractive than a
simple randOlll sampling.
In addition. 8 linear programming model is developed to
.tmulate the production of the three field crops under three output
price
and tnree energy cost scenarios.
The model includes only
production end selling activities for each farm size considered.
A s~ of the major findings based on an analysis of the
study results follows.

---

129
The patterns of indirect energy inputs use observed in the Bugar
sector appear to be an increasing function of farm size.
The patterns
of direct energy inputs use~ on the other hand~ suggest that the
energy inputs per acre used for harvesting and processing operations
are higher than those used for growing.
Based on the unit cast of the different types of energy inputs,
energy co9ts account for about 10% of the total cost of gro~ing
sugarcane.
The results of higher energy casts on the production of
field crapa reveal that sugsr, vith a low energy cost, appears to be
more vulnerable to higher energy costs than macadamia nut and cofiee
vith 16% and 18% of energy cost, respectlvely.
For exemple, at
Ideetlesl energy cost increases. say 50%, the net revenues of sugar
decrease by 18%, whereas the corresponding decresses sre only 4.5%
snd o.oa% for mseadamia nut and caffee under the eurrent output
priee scenario (Appendix Table 27).
Higher energy costs also tend to have differentisl impacts
depending on the output priee.
For instance. a 50% increase in
energy costs resulted in sn la% deerease in net revenues under the
current priee. whereas a similar increase reduced net revenues by
6.5% under the high output priee, and resulted in a 10s9 of net
revenues onder the low output priee.
In any case, a 50% and a 100%
!ncrease in energy eosts have not changed the optimal levels of
aetivity observed st base period energy costs.
In the case of macadamia nuts, the patterns of energy use
observed, vith the exception of gasoline and electricity, appear to be

---

130
a decreasing function of farm size.
For instance, farm E (50-499
acre) uees lees fertil1zer per acre than other types of farm.
S1milar conclusions are a1so reached for the patterns of herbicide
and direct energy use.
Specifieslly, the rate of gasoline and
electricity use takes the fo~ of a U curve, i.e., first starts
decreasing, reaches a minimum on the 10-19 acre and 20-49 acre farma
and then continues to increase.
The above description la based on
the 855umptfon that other variables are not held constant.
Energy costs constitute about 16% of the production cost of
macadamia nut.
An analyais of the study results indicates chat the
impact of increases in energy costa does not have a significant
impact on the farmer's revenues.
For instance, a 50% increase in
energy costs has resulted in only 4.5% decrease in net revenues under
the current output price, 3% reduction under the high output priee,
and 1% reduction of net revenues under the low output price
scenario.
It vas therefore found that lov output prices tend ta
reinforce the impacts of higher energy costs, whereas high output
prices tend ta negate them.
In addition, a1though energy.costs represent about 10% of the
production cast of sugar, as opposed to 16% and 18% for macadamia nut
and coffee, respective1y, the resulting decrease in net revenues is
greater for sugar than thoae for macadamia nut and coffee.
Output
priee appears ta pley a more important role in the reduction of
the farmerts incame than do increases in energy costs.

---

131
In the case of coffee~cadamia nut Interplantlng. the patterns
of input use. with the exception of fertl11zer, areassumed to be
Identl~al to those of macadam la nuts.
The patteTna of fertillzer use observed are not conclusive.
The
rate of use of cofiee cherry (10-5-20) ranges from a low 230 pounds
on the 10-19 acre farm to a high 454 paunds on the 5-9 acre farm.
The patterns of herbicide and direct energy input use, on the other
hand, show that larger farms use less energy per acre than sma1Ier
ones.
For instance, a lese than 5 acre farm uses about 4 and 2 times
more Psraquat per acre than the 5-9 acre and 10-19 acre farms,
respective1y.
The 9ame patterns are a1so observed for Roundup.
In
addition, the less than 5 acre farm uses about one and one-haIt and
two times more g8s011ne than the 10-19 acre and 5-9 acre farms,
respect1vely.
Sim118r re1at1coships are a1so observed for the rate of
e1ectricity used.
Based on the unit cost of the different types of energy inputs.
eoergy costa aCcoUDt for 18% of the cast of growing coffee.
The
results of hlgher energy costs ind1cate that coffee, with 8 high
energy cost, appears to be 1esB vulnerable to higher energy costs
than sugar and macadamia nuts.
For examp1e. at ident1cal energy cost
increaaes, say 100%, the net revenues of coffee decrease by 0.15%,
whereas the correspond1ng decrease is 9.1% for macadamla nut, and 33%
for sugar under the current output priee s~enaTio.
Output priee ia.
therefore, so important variable that influences the
magnitude of
impacts resu1tins from bisher energy costs.

---

132
Conclusions
The principal conclusions trom the study are:
1.
Higher energy costs have not significantly affected the net
revenues of smal1 growers but do have a differential impact depending
on the re90UTce endowments of each crop grower.
2.
Law crop priees tend to relnforce the impacts of higher
energy costs on net revenues, whereas high priees tend to negate Them.
3.
Sharp increases in energy costs have not changed the
optimal levels of activlty for the crope studied under various energy
cost scenarios.
~.
Larger farms do not necessarl1y use lees energy per acre than
smaller ones.
5.
ln the case of sugar, 1abor appears to be the binding
resource whereas diesel fuel and electricity are the binding reg our ces
in the case of macadamia nut.
6.
Farmera are faced with many constraints that do not allow
factor substitution.
ConsequentlYI a reallocation of the total
budget in favor of the binding resources may reduce production cost
and lead to greater net revenues for the growers.
7.
An increase in the total budget allotted to the independent
gravera, snd for that matcer , an increase in the resource constrsints
has not affected the optimal levels of crop production. but has
merely increased the net revenues and the idle resources.

---

IJJ
Mode! Application for Further Research
This research has enabled us ta apply I1near programming in a
aomewh8t dlfferent vay.
The classlcal approach has been to app!y it
to study many creps slmultaneously at the micro or macro level.
In
thia atudy, a ne~ way to set up a I1near programming mode! has been
developed.
Each crop production has been dlsaggregated into different
farm slzes.
Bach farm sl~e Ig considered 8S a 8eparate activity or
alterna te way to grow and sell the crcp.
Consequent!y, for each
production activity. there 15 a corresponding selling activity.
There are many advantages to setting up the linear programmlng
in tbis manner.
First, it enables us to assess technologies!
differences and their attendant economles of Beale.
In real1ty,
technologies appear to be a function of farm sizes.
Large size farms
appear to be more capital intensive or less labor intensive than
small Bcale farms.
Consequently, although aIl of these farms are
engaged in crop growing, some appear ta he more efficient than others.
The setting up of the linear programming in this particular manner has
enabled us ta address these questions.
Second, it allows us ta scrutinize the whole system and
ldentify areas of deficiency.
Third, it enables us ta set up a framework in wbicb different
farme can test their performances.
Fourth. it permita us to inve9tig8te different output and input
'rice scenarios, i.e., to 6tudy the impacts of hlgher ecergy costs 00
:he production of each crop separately.

---

134
Fifch, although the mode! 15 used ta simulate the production of
~all Besle grawers. tts flexibil1ty allo~s lts application to
lar~e Beale 8ro~er5 or ta situations where there are lrrigated or non-
lrrigated farma.
Also, the model can easily Incorporate different
submodels in whlch output and input priees are endogenously
determined.
Finally, the model can he applied ta situations where output
priees or input priees are uncertain and determlned by uncertain
demand and supply situations.
AlI of thes€
arc possible research areas Chat can be investigated
vith this model.
Further reaearch efforts in the field of agri-
cultural economlC8 are therefore ~eeded ta enhance the usefulnesa,
the applicabl1ity and the validlty of thia model.

---

APPENDIX TABLES

---

136
Colvmns
Prcduc~1on ~ctlvlt1~s
PRMellEs
"_bar of acrea of IUliar flTO'm on A f,1Il''lIl
PRBACR,ES
n
b.r of ee r es
of IUI!,U
1'0"""' on ~ !.nll
PIlCACRLS
n
ber of lera. of sUfllr Iro,"", on C f.~
n.,.bll of Bcru of BUilJ Ir""'" on t· flm
P'"IAACR!.S
n_b.,T of "'cres of
I114CaclUIlI nur gloom I;In ;, fal1ll
P~",çR.E.5
nUIl\\~lI'r cl! aCrlOS of ... clel .....i. nut gTown on J!I t"'l1ll
P!"!CACIl..ES
nu.lllb"r of acres of mlc ...:\\amla nut, Ir""'" on C fal'1ll
PKll/l.CR.t:5
mllllber of UT"" of ~...c.damia nu: grcVIl on P tin
PI1EACRES
nUlllber of acr"a of IIICI<la.lllia Ou: gro"" on E hI1ll
PCM.:R1:S
n~b"r of aCrla of coft"" ITown on A f.rm
PCBACats
n
b.r o! .crls of CQ{fu. Iro"" On ~ f ....
l'CCACltS
e,
t..f of Icrn of
<othe 110wn on C hT'lll
Sel11"1 Actlvltlls
SJUS~G
'.0<\\& of n
lugar .old b~ ,l,. [nu
SlUlSl'G
ton
of
r
Ius"r .old bv 'il (al'lll
SRCSUG
ton. 01 raw Iys"r ,old by C larm
SIUlSue
tOD5 of Tav auglT .old by D f.rm
SAl'IOL
tona ol moll.I". lold by A l.~
"H'"
[onl ol mol •• luu .01d by Il tUlIl
SOlOL
,,,.,,
tilO. gl _ h . . es .ole by C flTIIl
tllDA
ol -01. . . . . lold by D fu...,
....cazs
pound. of MC.d.&Iai. !\\I1t (.kel11) 101d ky A fi",
SllRCUS
p~nd. of Mc.damil ~t ian.lle) Iole by ! faT'lll
SJlCCUS
pound. ol
c.d_l.. n'H
(!h .. lh)
.vld by C fATIIl
SrGlC1tES
pound~ ol
c.d...,,1 .. '"'" (."elb) .old Ily J t.na
SMEC1tES
pound. ol I114c.d...,,18 nllt (shella) lold Ily E la....
SCPr.CIl.ES
pound. of caltee (plrchll\\entl .old bv " f;l1'1ll
SC!lCRES
pound' of cart ....
(p.rchmer.t)
,old by !
flTIII
seCACR~
poonds .,f ".,llee
(p .. rdllRcnc)
lold by C l.rm
OC"EU
PO\\ltld.1 of coHl."- knerrv) "nid by A f ....
seKER!
?O\\U'lÙ of cofh.. (cl\\erry) lold by ~ f.~
sœesc
POllftG. of coftlll! (cil.ury) laid b", C bl."1ll
~
0..,
I)bj.lctive lunctlon to mu:1JIIlu nit r ....n.... (dc>l.1.arl)
"'ll!!>
oP"l'0rT;UDi~ CQIt 01 hU.d labor (dolan)
OI!5n
0ppOTt..nit"ll colt of dhl.l (dolhrsi
.=r
opportunity COlt of Ilectr1cit,.. (dollua)
'"o.
oppc>rtunit'V ecer al calf ... (plrcnmeot)
CPT<>,
opporc..nicy co.t al cotlee (cherry)

---

137
Appendix Table 2
Sugar:
Optimal Levels of Ac~ivity. Current OUtput Priee and EC 0
Acreage (Acres)
Producr:ion (Tons)
PRAACRES
152
SRASUG
1440
PRBACRES
0
SAHOL
446
PRCACRES
595
SRBSUG
0
PRDACRES
0
SEMOL
0
SRCSUG
7990
SCMOL
1440
SRDSUG
0
OBJ
959091
SDHOL
0
FAM
14
HIRED
92

---

138
Appendix Table 3
Sugar:
Optimal Levels of Activity. Current ou~put Priee and EC 50
Acreage (Acres)
Production (Tons)
PRAACRES
152
SRASUG
1440
PRBACRES
0
SAMOL
446
PRCACRES
595
SRBSUG
0
PRDACR,ES
0
SBMOL
0
SRCS];G
7990
SCllOL
}û4Q
SRDSUG
0
OBJ
786148
SDMOL
0
FAM
1.3
H1RED
89

---

139
Appendix Table 4
Sugar:
Optimal Levels of Activity. Current OUtput Priee and EC 100
Acrease (Acres)
Production (Tons)
PRAilCRES
0
SRASUG
0
PRBACRES
0
SAXOL
0
PRCACRES
630
SRBSUG
0
PRDACRES
0
SBMOL
0
SRCSUG
8461
SCMOL
1524
SRDSUG
0
ODJ
642551
SDHOL
0
FAX
0
HlRED
74

---

140
Appendix Table 5
Sugar:
Optimal Activ1ty Levels, High Output Priee and
3sse Period Energy Cost (Ee 0)
Acreage (Acres)
Production (Tons)
PRAACRES
152
SRASUG
1440
PRBACRES
0
SAMDL
446
PRCACRES
595
SRBSUG
0
PRDACRES
0
SBHOL
0
SRCSUG
7990
SCMOL
1440
SRDSUG
0
ODJ
2669660
SDMOL
0
FAM
90
BIRED
191

---

141
Appendix Table 6
Sugar:
Optimal Activity Leve Le , High Output Priee and EC 50
Acreage (Acres)
Produc.tion (Tons)
PRAACRES
152
SRASUG
1444
PRBACRES
0
SAMOL
446
PRCACRES
595
SR,BSUG
0
PRDACRES
0
SBl'fOL
0
SRCSUG
7990
S010L
1440
SRDSUG
0
OBJ
2496716
5RDMOL
0
FAM
79
HIRED
IB7

---

142
Appendix Table 7
Sugar:
Optimal Activity Levels. High Priee and EC 100
Acreage (Acres)
Production (Tons)
PRAACRES
152
SRASUG
1440
PRBACRES
0
SAHOL
446
PRCACRES
595
SRBSUG
0
PRDACRES
0
SBMOL
0
SRCSUG
7990
SCMOL
1440
SRDSUG
0
OBJ
2326002
SDMOL
0
FAH
64
HlRED
184

---

14)
Append1x 8
Sug8Y:
Opttmal Levels of ActivitY7 Low OUtput Priee and
EC 0, or EC 10 or EC 100
Acreage (Acres)
Production (Tons)
PRAACRES
0
SRASUG
0
PRBACRES
0
SAMOL
0
PRCACRES
0
SRBSUG
0
PRDACRES
0
SBMOL
0
SCRSOG
0
SCMüL
0
SRDSUG
0
OBJ
0
SOHOL
0
FAM
0
RI RED
0

---

144
Append Ut Tab le 9
Mac4dam1a Nut:
Optimal Levels of Activity, Curreu~ OUtput
Priee and te 0
ëcxeege (Acres)
Production (Pounds)
PMAACRES
0
SMACREs
a
PHBACREs
337
5MBCRES
1353251
PMCACRES
0
SHCCRES
0
PMDACRES
0
SMDCRES
0
PMEACRES
945
5MBCRES
5442693
aDJ
3347520
DIESEL
25
ELEn
123

---

145
Appendix Table 10
Kacadamla Nut:
Optimal t~vels of Activity, Current Output
Priee and Ee 50
àc r eege (Acres)
Production (Pounds)
Pl!AACRES
0
SMACRES
0
PIlBACRES
337
5MBCRES
1353251
PMCAlŒS
0
SMCCRES
0
P.KDACRES
0
SMIlCRES
0
PIlEACRES
945
SHEC&ES
5442693
OBJ
3196898
DIESEL
17
ELETY
124

---

146
Appendix Table Il
Mscadamia Nut:
Optimal Levels of Activity, Current Output
Priee and EC 100
Acree,ge (Acres)
Production (Pounds)
PMAACRES
0
SHACRES
0
PMEACRES
337
5MBCRES
1353250
PMCACRES
0
SMCCRES
0
l'!lOACRES
0
S'MDCRES
0
PMEACRES
945
SMECRES
5442693
OBJ
3043713
DIESEL
9
ELETY
125

---

147
Appendix Table 12
Macadamia Nut:
Optimal Levels of Activity, 81gh Output
Priee and E.C 0
Acreage (Acres)
!roduction (Paunds)
PMAACRES
0
SMACRES
0
PMBACRES
337
5MBCRES
1353251
PMCACRES
0
SHCCRES
0
PHDACRES
0
SHDCRES
0
PMEACRES
945
SHECRES
54442693
oBJ
5794059
DIESEL
69
ELEn
187

---

148
Appendix Table 13
Kacadam1a Nut:
Optimal Activity Levels. High Output Priee and EC 50
Acreage (Acres)
Production (Pounds)
PMAACRES
0
SMACRES
0
PMEACRES
337
5MBCRES
1353251
PHCACRES
0
SHCCRES
0
PMDACRES
0
SMOCRES
0
PMEACRES
945
SMECRES
5442693
OBJ
5643438
DIESEL
61
ELEn'
lBB

---

149
Appendu Table 14
Mscadam1a Nut:
Opt1ma.l Levels of Act1vity. High Output
Priee and EC 100
Acreage (AcTes)
Production (Pounds}
PMAACRES
0
SMACRES
0
PMBACRES
337
5MBCRES
1353250
PMCACRES
0
SKCCRES
0
PKDACRES
0
Sr-IDCRES
0
PKEACRES
945
SKECRES
5442693
OBJ
5490253
DIESEL
54
ELEn
1B9

---

150
Appendix Table 15
Macadamia Rut:
Optimal Level. of Activity, Law Output
Priee and EC 0
Acreage (Acres)
Production (Pounds)
PMM.CRES
0
5MACRE5
0
PMBACRES
0
5MBCRES
0
PMCACRES
0
5MCCRES
0
PMDACRES
0
SMDCRES
0
PMEACRES
982
5IŒCRE5
5653836
OBJ
1074867
DIESEL
0
ELETY
48

---

151
Appendix Table 16
Hac:adamia Nut:
Optimal Acttvity Leve.Ie , Lev Output
Priee and le 0
Acreage (Acre§)
Production (Pounde)
PIlAACRES
0
SMACRES
0
P!!BACRES
0
5MBCRES
0
PMCACRES
0
5MCCRES
0
PIlDACRES
0
51lDCRES
0
PMEACRES
982
5MECRES
5653836
OBl
998291
OIESEL
0
ELETY
44

---

152
Appendix 1able 17
Macadamia Nue:
Optimal Levels of Activity. Law Output
Priee and Be 100
Acreage (Acres)
Production (Pounds)
PHAACRES
0
SMACRES
0
PMBACRES
0
5MBCRES
0
PMCACRES
0
SMCCRES
0
PMOACRES
0
SMIlCRES
0
PMEACRES
982
SMECRES
5653835
OBJ
919752
DIESEL
0
ELETY
41

---

153
Appendix Table 18
Coffee:
Optimal Levels of Activity, Current Output
Pr-t ce and sc 0
Acreage (Acres)
Produetion (Pound)
PCACRES
0
SCACRBS
0
PCBACRES
97
SCBACRES
1408416
PCCACRES
24
SCCACRES
41884
SCHERA
0
SCHE'RB
286510
SenERC
196490
OBj
3057043
PPTON
1.86
CPTON
0.76

---

154
AppeDdl.x Table 19
Coffse:
Optimal Levels of Ac.tivity, Cur rent; Output
Priee and EC 50
Ac:reage (Acres)
Production (Pounds)
PCACRES
0
SCACRES
0
PCBACRES
97
SCBACRES
1408416
PCCACRES
24
SCCACRES
41584
SCHERA
0
SCHERB
286510
SCHERC
196490
OBJ
JOJJJ54
PPTON
l.B5
CPTON
0.74

---

155
Appendix Table 20
Coffee:
Optimal Levels of Activity, Current Output
Priee and Be 100
-eage (êcr es)
Production (Pounds)
PCACRES
0
SCACiŒS
PCBACRES
97
SBACRES
1408416
PCCACRES
24
SCCACRES
41584
SCHERA
a
SCHERB
286510
SCHERC
196490
OSJ
3009762
PPTON
1.84
cnON
0.71

---

156
Appendix Table 21
Coffee:
Optimal Levels of Act1v1ty. H1gh Output
Priee and EC 0
Acreage (Acres)
Production (Pcunds )
PCACRES
0
SCACRES
PCBACRES
97
SBACRES
1408416
PCCACRES
14
SCCACRES
41584
SCHERA
0
ScHEPJl.
186510
SCHERC
196490
OB]
4410243
PPTON
2.66
CPTON
1.16

---

157
Appendix Table 22
Caffee:
Optimal Levels of Activity, High OUtput
Priee and EC 50
êcr eage (Acres)
Production (Pounds)
PCACRES
0
SCACRES
PCBACRES
97
SCBACRES
1408416
PCCACRES
24
SCCACRES
41584
SCHERA
0
SCHERB
286510
SCHERC
196490
OBJ
4386554
PPTQN
2.64
CPTQN
1.14

---

156
Appendix Table 23
Caffee:
Optimal Levels of Activity. High Output
Priee and EC 100
Acreage (Acres)
Production (PouI!.~
PCACRES
0
SCACRES
PCBAeRES
96
SCBACRES
1408415
PCCACRES
24
SCCACRES
41584
SCHERA
0
SCHERB
286570
seMERC
196490
06J
4362962
PPTON
2.64
CPTON
1.12

---

159
Appenrlix Table 24
Coffee:
Optimal Levels of Activity. Law Output Priee anà EC 0
Acreage (Acres)
Production (Pounds)
PCACRES
0
SCACRES
0
PCBACRES
96
SCBAGRES
1_08_15
.
PCCACRES
24
SCCACRES
41584
SCHERA
0
SCHERB
286510
SCHERC
196490
OBJ
1)03843
PPTON
l.05
CPTON
.35

---

160
Appendix Table 25
Colfee.
Optimal levels of Activicy, Law Output Priee and EC 50
acreage (Acres)
Production (Pounds)
PCACRES
0
SCACRfS
PCBACRES
96
SCBACRES
1408415
PCCACRES
24
SCCACRE5
41.584
SOlERA
0
SCHERB
286510
SCHERC
196490
OBJ
1680155
PPTON
1.04
CPTON
.34

---

161
Append1X Table 26
eaffee:
Optimal Levels of Act1vity, Law Output Priee and EC 100
Acreage (Acres)
Production
(Pounds)
PCACRES
0
SCACRES
PCBACRES
16
SCBACRES
1408415
PCCACRES
l4
SCCACRES
41584
SCHERA
0
SCHERB
286510
SCHERC
196490
OBJ
1656562
•
PPTQN
1.03
CPTON
.32

---

Appendix Table 27
!Dergy C08~ Scenarios. Crop Priee Scenario and Their Impacts on Net Revenues
Percentage of Energy ln To~al
Decreaae in He~ Revenue
Decre88e in Net Revenue
Production Coat eX)
Under EC 50 (%)
Under EC 100 (~)
S.!!8.8r
cueeene
10
lB
33
H1gh
10
6.5
13
Law
10
VaniBh
Vanish
MacadamtA Mut
Corrent
16
~.5
9.1
81gh
16
3
3
Law
16
1
15
Coffee
Current
lB
O.Og
0.15
H1gh
16
0.05
0.11
1011
16
1.4
2.6
...~
'"

---

163
Appe.ndu Table 28
Loan Prog!'8Dl:
Amaunt Allotted pel' Acre and. Total 'Budget'" (Sugu)
(Dollars)
Fert.ilizer
340
1.160,420
Herbicide
65
221,845
seede aee
45
133,585
Labor
35
119,455
Planting Cast
90
J07,170
Kililcellaneou8
(Energy. Rent , tu)
157
535,841
Total
732
2,496,316
SOurce:
TelephoDe interview vith ,irac Hawaiian Bank, Rilo Branch,
1982.
'Baaed on the 6825 acres gtown by t~ independeat gtowers. 1981.

---

164
Appendix 7able 29
ReBouree Constraints <Suger)
Labor
Direct Energy
Ope:rator
11243 (hours)
Gasoline
241545 (gal)
Famil,
8151 (hours)
Diesel
296296 (gal)
Hind
8713 (hours)
Fert.il1zer
~
6825 (acres)
A-l
4486960 (lbs)
11-104
1547227 (lbs)
Seedcene
7314 (tons)
1I-28
309447 (lbs)
A-5
511893 (lbs)
Capital
11392631 (dollsl's)
Berb1cide
Production
Dalapen
10487 (gal)
Raw Bugar
367000 (tons)
larmes
37617 (lbs)
Molasses
107000 (toDS)
!ltraziae
37646 (lbs)
ROundup
410 (gal)
Surfactant
1268 (gal)
Ametrytle
3262 (gal)
St.icker
1585 (gal)
Paraquat
148 (gal)

---

165
Appendix Table 30
Lean Progrsm:
Amount Allotted per Acre. 1981
(Dollars per Acre)
Hacadamia Nut
Coltee
PrunJ..ng aod Fert1l1z1ng
280
280
SprBylng
250
250
Processlng Fee
265
265
RarvesUna
661
500
Puel Expellses
31
31
Depreciation CORt
122
122
Iasurance
50
50
LeuiD<! Coete and Taxes
B6
86
Repaire and Maintenance
lB
lB
'U8cellaneous
537
720
:Ootal
1900
2200

---

166
Appendix Table 31
Resource Constra1nts (Kacadamia Nut)
Labor
637362 (hours)
Flilirtilizer
1-16
9922327 (lbs)
1-20
7222871 (lbs)
Herbicide
Paraquat
32198 (gal)
Roundup
14361 (g&1)
Warfa.r1n
12127 (g&1)
2. 4-D
4888 (gal)
Atrazine
32483810 (lba)
Diuron
6393 (lb.)
Land
12210 (acres)
Capital
2100120 (dollars)
Direct Energy
Gasol1ne
40797 (gal)
Diesel
22797 (gal)
!lectrietty
22580 (laohl
Production Constra1nts
33270000 (lbs)

---

167
Appeadtz Table 32
Re.ource Constra1nts (Coffee)
Laber
910517 (hours)
Fertil1r:er
Hac-8
12439388 (lbs)
10-5-20
8100650 (Ibe)
Herbicide
Paraquat
31652 (gal)
Roundup
16156 (gal)
2, 4-D
4488 (gal)
Atta&1ne
19490 (lbs)
D1uron
4265 (lbs)
Direct Enerp
Gasol1ne
45333 (g.l)
Diesel
18926 (gal)
Electr1city
22283 (kwh)
Land
1800 (acres)
Capital
2100120 (dollars)
Production
1450000 (lbs)

---

168
LlTERAIURE CITED
L
Adams, R., G. lC.1Di. and W. Johnson.
"Effects of Energy Cost
InereaseB an Regional Allocation Polle1es and Agricultural
Production."
Amerlcan Journal of Agriculture! Econam1cs,
59 (1977),445-454.
2.
Baker, B.e.rold L., and J. T. Ke.eler.
"An Econondc 'Report on the
Production. of Kons Caffee,"
Agriculture! t:xperhent Station,
University of Ravaii, 1974.
3.
Barlowe, Raleigh.
Land ReSOUtee Economies: The Economies of
Real Estate, 3rd Edition.
Englewood Cliffs, N.J.: prentlce-Hall,
IDC., 1978.
4.
Baumol, W. J.
Economie Theory and Operations Analysis, 4th
Edition.
Englevood Cliffa. N.J.: Prantiee-Hall, Ine., 1977.
5.
Be11ock, Richard Ph111ip.
"A Structural Model and Suulstlons
of the U.S. Potato Industry vith Special Emphas1s on Examinina
the Interregloual Effects of Changlng Energy Costs."
Ph.O.
DiaeertatioD, The Penuaylvania St_te University. 1981.
6.
Casey, James 1., Ronald Lacewell, and Lonnie L. Janea.
"Impact of L1lD.1ted Fuel Supplies on Agricultural Output aDd Net
Returtls: Sout'nern R1gh Plains of TUas."
Texas Agricultural
Exper1ment Station, MP 1175, January 1975.
7.
Chiang. Alpha C.
Fvndamental Methode 01 Mathematical Economica,
2nd Edition.
New Yorl: KcGraw-Rill Company. 1974.
8.
Christensen, lAurit R., aad D. W. JorgeIlSon.
"The Measurement
of O.S .. bal Capital Input, 1929-1967."
Review of Incarne
~T-"h"e...:::==-",-,=="
and Wea1th, 15 (1969), 293-320.
9.
Cochran, W. G.
S:Bll!pling TeChniques. 3rd Edition.
New 'York:
John WiJey & Sous, 1977.
10.
COlmIlOl1er, Barry, H. G.., R. napper, and W. Lockeretz.
"The
Vulnerabil1ty of Crap Production ta Energy Problems. 11
Center
for t~e B101ogy of MaturaI Systems, Washington University, Sc.
Louis, M1Bsouri, JanuaYy 1975.
11.
Depertment of Havaitan TelephoDe Company.
Bonolvlu ConsumQr
Priee Index.
Hoaolulu: April 1981.
.2.
Departalant of Plann:Lng and Economie Development •
Hawaii
Integrated Eners: A.aessaent (Execvtive SummAry).
Honolulu.
Savaii. 1980.

---

o _
... ' . . . . . UO,,,"'",,,,,,",
169
13.
Deparaaeat of Planning and Economic Developqent.
Hawaii
Integrated Energy Assessment (Executive Summary).
Honol.ulu,
HawaH, 19B1.
14.
Dorfman. Robert. Paul A. Samuelson, and R. M. SolOW'.
L1near
Programm1ng and Economie Ana1ysie.
New York: McGralot-Bl11
Book Campany. 1958.
15.
Dvoskin. D•• and E. O. Heady.
U.S. Asrieu1tura1 Production
Under Limited !nergy Supplies, 81gh Energy Priees and Expand1ng
Agrieultural Exports.
The Center for Agricultura1 and Rural
Deve!opment
Gard Report.
Iowa State University, 1976.
J
16.
Duncan. M., and K. Webb.
Energy Use in Amer1can Agricu!ture.
Kansas City: Federa! Reserve Bank, February 1980.
17.
Eidman, Vernon, Craig Dobblns, and R. Schvartz.
"The Impact of
Chang1:a.g bergy Priees on Net Returus, Production Methods and
Kilocalories of Output for Representative Irrigated Farms. ft
Profeaaloaal Paper (PP-230) of the Oklahoma AgTicu1tura1 Experl-
=ent Station. 1915.
18.
Field, Barry C•• and C. Grebenetein.
"capital Enersy
Substitution in U.S. MAnufacturing."
The Revlew of Economies and
StatLstlco, 63 (1980): 207-212.
19.
Firat Hawaiian Bank.
''Monthly Report of the Research Division."
Honolulu. Bawaii. Sepcember 1981.
20.
PUkunaga, Edward t.
Coffee.
Honolulu, Raw~i, 1954
(unpub1ished) •
21.
Copal.4k.rishnan, Chennat.
Natural Resoureea and Energy: Theory
and Polley.
AnD Arbor: Ann Arbor sctence , 1980.
22.
Gopalakr1shnan., Cheanat. and P. Lasturl.
"The Economics of
Biomass Bougy."
Agricu1tural Vastes, 2 (1980): 83-'91.
23.
Gopalakrbhnan, Chennat, and M. Nahan.
'IEconom.1c Potentia1 of
Baga.se as an Alternate Energy Source: The Bawaiian Experienee. Il
Agriculture and Inergy.
H., York: Academic Press. 1977.
24.
Gopalakr1.shnan~ CheDDat, E. N. [ofU et al.
"Optiaal Use of
Hawaii'. Biuaaas Sources for Enersy: A Bovel Proposal."
Pr0ceedfasa of the Internatlonsl Symposia on Altero.te Energy
Source. and technolosy (IASTED) , San Francisco, 1981. pp. 68-72.

---

170
25.
Bamilton, R. A•• and E. T. Fukunaga..
Crawlas Mac:ed&llia Nuts
in Bavai!.
Bulletin 21, University of Hawaii Agrlculcural
Experiment SCation, January 1959, 51 pp.
26.
Rs.val1 Agrlculcural Reporting Set'Vlce.
Sc.ciseles of Hawal1an
Agriculture.
Rawaii Deparonene of Agriculture and V.S. Deparc-
ment of Agriculture, 1981.
27.
Haval1an Electrlc Company. lue.
Annual Report.
HODolulu~
Baw.U, 1980.
28.
Bawaii 'Beural Energy Institute.
Plagrees Report on Renewable
Energy in Havaii.
Honolulu, Rawaii, April 1982.
29.
Heady, Earl O., and ~. Chandler.
Linear Programmlng Methods,
6th Edition.
Ames: Iowa Scace University Press. 1969.
30.
't&.!CO Raises Rate Because of Lag ln Alcernate Energy Sources."
The Honolulu Advertlser, 8 Sepcember 1981, A-6.
31.
Helchel. G. H., and C. R. Frink.
IIAnticipaUIl8 the Energy Needs
of American Agriculture."
Journal of 50il and Water Conservation,
30 (Jan-reb 1975).
32.
Benderaon. J. M.• and R. K. Quandt.
Microeconomie Theory, 2nd
Edition.
New York: HCGrav-Bill Book Co., Inc., 1911.
33.
Birst, Eric.
"Food-Related Eoel"gy Requirements."
!.ç.ience, 184
(April 1974). pp. 134-138.
34.
Bolderness, J. S., G. R. Vieth, F. S. Scott, Jr., and H. Briones.
"Cost of ProduciIl3 and Proee8sing 5ugarcane Amons: Hawaii' e
lndependent Cane Gravera vith Caœpar1soos to the Plantation
Growers. "
College of Tropical Agriculture and. Buman Resources
and Agr1cultural Reporting Service, June 1981 (unpublished).
35.
RoldernesB, J. S., G. J.. Vieth. and F. S. Scott, Jr.
"Viabllity
of Independent Sugareaoe Fame on the Rilo Coast," 1979
(unpubl1shed) •
36.
Hud8on,!. A•• and. D. 'W. Jorgm8on.
''U.S. Policy and Economie
Growth, 1975-2000. "
Tbe Bell Journal of Economies (1978):
461-514.
37.
I.B.M. Corporation.
An Introduction to Linear Programming. Ist
Edition.
White PlaiDS, Mew York (1964).
38.
Johnson, Bruce B., and P. Â. Rend_rson.
Energy Priee Lev.ls and
the Economies of Irrigation.
Deparment of Agricultural Eeonam
ies, Report (79), University of Nebraska, September 1977.

---

171
39.
Kan1a, John Joseph.
"An I1IIpact Assument Mode! of Umiting
Pest1cides in AgrIculture."
Ph.D. DhaertaUon. The
University of Nebraska, 1980.
40.
Reeler, Joseph T., and Wen-Yuan Huang.
Cost and Returns of
Hacadam1a Production.
CoIlQge of Tropical Agriculture.
Oepartment of Agrlcultural Econam1ce, 1976 (unpubllshed).
41.
I.eeler, Joseph T'
and E. T. Fukuoaga.
The ECOn01llic. and
t
Horticulcural Aspects of Gravina Macadamia Nuts Commercially in
Ha~81i.
Agricultural Economies Bulletin 27, University of
Hawaii. Agricultural Expertmeot Station, JUDe 1968, 47 pp.
42.
Reeler. Joseph Tot J. 1. Iwaoe. and D. K. MStsuœoto.
An
Economie Report on the Production of KODa Coffee.
Agricultural
Economl,s Bulletin (12). Hawaii Agrlcultural Exper~ent Station.
December. 1958. 36 pp.
43.
Kelvin, Lancaster.
Mathematic.a1 EeoDOlIIics.
New York:
The
McMillsn Company~ 1968.
44.
kioc.h, D. x., J. IL Wana. and R.. E. Strobman.
Eguipment for
Husking Macadam1a Nuta.
!ullêtio (1261) Hawaii AgricultuTsl
Exper.1lllent Station, June 1961.
45.
Lacewell, Ronald D.
"SOIIle Effects of Alternstivs Energy
I85ue an St.abllity ln the Great Plalu. 1I
Proc.eedings of Seminar
on
ricultural Polie
Great PlaiDS
ricultural CouDci1
Publications (7~ , 1975.
46.
Lacewell. Ronald D., G. D. Condra. and B. Fish.
"Impact of
Energy Cost on Food and 'iber Production."
CoUege Station,
Texas:
Texas A & H University, 1976.
47.
Layard, P. R. G•• and A. A. Walters.
Microeconomie The0!Y'
New York:
McGr.Y-Hill look COIlIpaDY, 1978.
'18.
Litterm&n, Mary Atm.
"Energy Subst1tution in U.S. Kanufaeturing
and Agriculture."
Ph.D. Dissertation, University of K1D.nesota.
1980. 460 pp.
9.
MapP, Harry P., Jr •• and Craig L. Dobbins.
"Implications of
lising Energy Costs on Irrlgated. rarms in tbe Ok:lahoma P8n-
baDdlê."
Amer1can Journal of A&r1cultural Econl.llllics, 59 (1976):
971-977 •
1.
l'Ie.rl1n. EricitsoD.
"The Economie h1pact of !oergy Priee tnerealel
on erop Production in Sovthveat Nebraska."
Pb .. D. Dissertation.
Department of Agrieultur&1 ECODOIIic8. Ou1vereicy of Nebraska,
L1ncoln, October 1979.

---

172
51.
Misra~ Arun lCumar.
"an Economie Evaluation of Prob Leas and
Production Potentials of Small GrO'ilers in Ori668, 'Ind La , Il
University of Missouri, Columbia, 1980.
52.
National Economies Division, Economies, and Statistic9 Service.
Cast of Producing and Processing Sugarcane and Sugsrbeets in
the O.S.
E.R.S. Staff Report.
U.S. Department of Agriculture,
April 1981, 36 pp.
53.
Nes8, Wu, and Coppina, Richard.
Linesr Programming and
Extensions.
New York: McGraw-Hill Book Company, 1981.
54.
NoIlll8.n, Rask.
"Crtticsl Cholces in Energy: Discussion."
Amerlcan Journal of Agricultural Economies, 62 (1980): 976-977.
55.
Office of Science and Technology, Executive Office of the
President, Stanford Research lnstitute.
Patterns of Energy
Consumption in the United States.
Stanford, Ca1ifornia,
Appendix C•• January 1972.
56.
"011 Priee Upturn Seen Within Year."
Honolulu Star Bulletin,
21 April 1982, C-10.
57.
Perelman, M.
"Farming vt tb Petroleum."
Environment 14 (1972):
134-138.
58.
Pimentel, David, et al.
"Food Produe tion and the Bnergy Crfs Ls . Il
Science, 182 (1973), 443-449.
59.
Pindyek, Robert S.
The Structure of Wor1d Energy Demand.
London: The MIT Press, 1979.
60.
Pope, W. T.
The Maeadamia Nut in Hawaii.
University of Hawaii
Agrieultura1 Experiment Station, Bulletin 59, Nov. 1929, 23 pp.
- 61.
Reno, Robert.
"üeapi t e Reagan, Energy is a Problem."
Honolulu Advertiser, April 1981, A-8.
62.
Scott, Frank S., Jr., and H. K. Marutani.
Economie Viabi1ity of
Small Macadamia Nut Fards in Kona.
Hawaii Institute of Tropical
Agriculture and Human Resourees, Research Series 009, 1982.
40 pp.
63.
Sc'hef f LeLd, Les Lde F.
"Economie Ana1ysis of the Costs and
Returns for tbe Production of Corn Using Center-Pivot
Irrigation System in Southwest Nebraska."
Ph.D. Dissertation,
University of Nebraska, Lincoln, 1970.

---

173
64.
Short. Charles Cameron.
"ürounëvater Mining in the Og1a18
Aqulfer in Relation to Rising Energy PrieES and Agr1cultural
Production"
Ph.D. Dissertation, lo...a State University, 1980,
178 pp.
65.
Skold, Melvin D.
"Farmer Adjustments to Higher Energy Pr âces ;"
Natural Resouree Economies Division, Economies Reaearch
Service, V.S. Department of Agriculture (ERS-66), Nov. 1977.
66.
SRI International.
Encrgy Self-Sufficiency for the Big Island of
Hawaii.
Final Report.
January 1980.
67.
State of Hawaii, Depactment of Planning and Economie Development
Bnd Lawrence Berkeley Laboratory.
Hawaii Integrated Energy
ASSEssment: Executive Summary.
University of California,
Berkeley, January 1981.
6B.
Steinhart, S •• and C. Steinhart.
"Energy Use in the U.S. Food
System."
Science, 183 (1974): 301-316.
69.
Thompson, Gerald E.
Linear Programming.
N~ York: McMillan
Company, 1911.
10.
Young, Kenneth B.
"The Impact of Iucreaae in Power Source
Pr Lcea on Irrigation."
Western Irrigation Forum on Irrigation
Energy and Conservation.
Tri State Generation and Transmission
Association, IDC., Denver, Colorado, 1919.

---