a farm-level approach to the methyl ... - getd.libs.uga.edu · a farm-level approach to the methyl...
TRANSCRIPT
A FARM-LEVEL APPROACH TO THE METHYL BROMIDE PHASE-OUT: IDENTIFYING
ALTERNATIVES AND MAXIMIZING NET-WORTH USING STOCHASTIC DOMINANCE
AND OPTIMIZATION PROCEDURES
by
MARK MCCULLOH BYRD
(Under the Direction of Cesar L. Escalante)
ABSTRACT
Methyl bromide (MeBr) is an effective yet toxic soil fumigant that is subject to an
accelerated phase-out under the Montreal Protocol on Substances That Deplete the Ozone Layer.
As a result of the phase-out, Georgia farmers now require a substitute fumigant and herbicide
combination that can control for weeds, pests, and diseases while delivering comparable yields to
those of MeBr. This study analyzes the economic viability of alternative production methods on
a representative Georgia pepper farm. The analysis is addressed through stochastic dominance
analysis, an enterprise budget, and the development of a linear programming model. Alternative
fumigants are ranked, analyzed within the budgetary framework, and subjected to production
constraints such as farm acreage, capital structure, and asset allocation within the programming
model. Compared with MeBr, the models prescribe results supporting the economic feasibility
of several alternative fumigants, and thus serve to educate Georgia vegetable farmers on
available input-substitution strategies.
INDEX WORDS: Methyl Bromide, Bell Pepper, Stochastic Dominance, Enterprise Budget, Linear Programming, Optimization-Simulation
iv
A FARM-LEVEL APPROACH TO THE METHYL BROMIDE PHASE-OUT: IDENTIFYING
ALTERNATIVES AND MAXIMIZING NET-WORTH USING STOCHASTIC DOMINANCE
AND OPTIMIZATION PROCEDURES
by
MARK MCCULLOH BYRD
B.S.A., University of Georgia, 1999
A Thesis Submitted to the Graduate Faculty of The University of Georgia in Partial Fulfillment
of the Requirements for the Degree
MASTER OF SCIENCE
ATHENS, GEORGIA
2005
© 2005
MARK MCCULLOH BYRD
All Rights Reserved
A FARM-LEVEL APPROACH TO THE METHYL BROMIDE PHASE-OUT: IDENTIFYING
ALTERNATIVES AND MAXIMIZING NET-WORTH USING STOCHASTIC DOMINANCE
AND OPTIMIZATION PROCEDURES
by
MARK MCCULLOH BYRD
Major Professor: Cesar L. Escalante
Committee: Michael E. Wetzstein Esendugue Greg Fonsah
Electronic Version Approved: Maureen Grasso Dean of the Graduate School The University of Georgia August 2005
DEDICATION
This thesis is dedicated to all of my friends within the College of Agriculture: Peter,
Katy, Hongsin, Feng (Frank), Rui (Carolyn), Xiaohui (Sarah), Yingzhuo, Marianna, Laxmi, Joe,
Jordan, Horacio, Katia, Tatiana, Dmitriy, R.J., Vahe, Doris, and Carrie. One of my greatest
future challenges will be to surround myself with such a diverse, kind, and intelligent group of
people. Thank you for all the good times.
iv
ACKNOWLEDGEMENTS
I cannot begin to thank Dr. Michael Wetzstein for his influence on my academic and
personal life. Your enthusiasm and commitment to the teaching profession, along with your
encouragement, respect, intellect, and laughter, is greatly appreciated and will never be
forgotten. To Dr. Cesar Escalante, it is a rarity to work with such a fine person. Your guidance
on this project has been invaluable. Thank you for your patience and support while juggling the
demands of a devoted father and beginning professor. I want to thank Dr. Esendugue Greg
Fonsah, for his valuable insights and contributions to this study.
Next, I want to express my deepest thanks to my parents, Paul McCulloh Byrd and Alice
Kathryn Gatlin, for their support and encouragement during my graduate education. You have
shown me more love than I could ever hope to repay. I love you, and I could not have done this
without you. To my brothers, Paul and Ben, I am proud of your accomplishments, and I love
you both. I wish you the greatest success the world has to offer.
Finally, no words can express my love for you Rhiannon. Your optimism and zest for
life has taught me to see the world in a different way. I will never forget our journey to the Lost
Coast, our plumerias and mangoes in Hawaii, thousands of miles of road, fireworks in Cuba,
backpacking in Canada, jazz clubs in New York City, Sunday reading with endless cups of
coffee, examinations, and the joy and laughter attached to each of them. Thank you for showing
me that unconditional love truly exists.
v
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS.............................................................................................................v
LIST OF TABLES....................................................................................................................... viii
LIST OF FIGURES .........................................................................................................................ix
CHAPTER
1 INTRODUCTION .........................................................................................................1
Background Information ...........................................................................................1
Objectives..................................................................................................................7
Organization ..............................................................................................................8
2 LITERATURE REVIEW ..............................................................................................9
Economic Impact Assessments .................................................................................9
Methyl Bromide Alternatives Outreach ..................................................................13
Additional Economic Impact and Viability Studies................................................20
3 COMMERCIAL PEPPER PRODUCTION ................................................................28
United States Pepper Industry .................................................................................28
Georgia Pepper Production .....................................................................................30
Marketing ................................................................................................................31
Farm Management...................................................................................................32
4 EMPIRICAL METHODOLOGY................................................................................41
Expected Utility, Risk Aversion, and Efficiency Criteria .......................................41
vi
Interpretation of Dominance Analysis, Enterprise Budget, and Programming
Model.................................................................................................................53
5 ANALYSIS OF RESULTS .........................................................................................69
Stochastic Efficiency-The Final Four Production Methods ....................................69
Results of the Enterprise Budget Analyses .............................................................71
Programming Solutions...........................................................................................77
6 SUMMARY AND CONCLUSION ............................................................................93
Study Summary .......................................................................................................93
Conclusions .............................................................................................................95
Future Research.......................................................................................................97
BIBLIOGRAPHY..........................................................................................................................98
APPENDICES .............................................................................................................................104
A Bell Pepper Fresh Market Variable Cost Budget for Methyl Bromide .....................105
B Bell Pepper Fresh Market Variable Cost Budget for C35 + KPAM..........................107
C Bell Pepper Fresh Market Variable Cost Budget for C35 + Chloropicrin.................109
D Bell Pepper Fresh Market Variable Cost Budget for Telone II + Chloropicrin.........111
E GAMS Program for the Base-Case Farm Model using Methyl Bromide..................113
vii
LIST OF TABLES
Page
Table 4.1: Pre-Operating Balance Sheet........................................................................................66
Table 4.2: Description of Constraints and Requirements ..............................................................67
Table 4.3: Description of Portfolio Activities................................................................................68
Table 5.1: Efficiency Ranking of Final 8 Production Methods Using Second Degree Stochastic
Dominance.....................................................................................................................83
Table 5.2: Cumulative Distribution Functions for Final 8 Production Methods ...........................84
Table 5.3: Financial Efficiency and Break-even Analyses Using Experimental Yield Results ....85
Table 5.4: Financial Efficiency and Break-even Analyses Using Constant Yield Results ...........86
Table 5.5: Comparison of Costs per Carton for Final Four Production Methods..........................87
Table 5.6: Programming Solutions and Financial Ratios of Five-Year Averages for all
Production Methods ......................................................................................................88
Table 5.7: Yearly Programming Solutions for Decision Variables and Key Financial Measures
Using Methyl Bromide..................................................................................................89
Table 5.8: Yearly Programming Solutions for Decision Variables and Key Financial Measures
Using C35 + KPAM as Methyl Bromide Substitute .....................................................90
Table 5.9: Yearly Programming Solutions for Decision Variables and Key Financial Measures
Using C35 + Chloropicrin as Methyl Bromide Substitute ............................................91
Table 5.10: Yearly Programming Solutions for Decision Variables and Key Financial Measures
Using Telone II + Chloropicrin as Methyl Bromide Substitute ....................................92
viii
LIST OF FIGURES
Page
Figure 3.1: Leading U.S. Bell Pepper Producing States (2002) ....................................................36
Figure 3.2: U.S. Fresh Bell Peppers, Production, Import, Export and Domestic Use (1979-2003)37
Figure 3.3: U.S. Bell Pepper Import Values from Selected Countries of the World (2000-2003)38
Figure 3.4: Georgia Farms, Farm Sizes, and Harvested Acreage (2002) ......................................39
Figure 3.5: U.S. Bell Peppers Season Average Prices and Per Capita Consumption (1979-2003)40
ix
1
CHAPTER 1
INTRODUCTION
1.1 Background Information
1.1.1 Historical Framework
On September 16, 1987, the signatures of 24 countries resulted in the Montreal Protocol
on Substances that Deplete the Ozone Layer. Former President Ronald Regan later committed
the United States to the international agreement on April 5th, 1988. Effective January 1, 1989,
the Protocol’s signatory nations began actively working toward fulfilling the guidelines
established under the Protocol. Between the meetings of the parties, scientists assessed the
toxicity of many chemicals and created an index which identified substances causing the most
damage to the ozone layer. As substances were listed, amendments were added to the original
treaty which recognized the need to accelerate the phase-out of these dangerous substances.
Methyl bromide (MeBr) was identified as one of the most toxic contributors to ozone depletion.
It was recommended during the Ninth Meeting of the Parties (1997) in Montreal, Canada that
MeBr face an accelerated phase-out schedule.
Methyl Bromide is an agricultural fumigant that is used to control weeds, nematodes,
soil-borne pests, and diseases. Methyl Bromide is widely accepted by agricultural producers
because they understand its chemical properties, it is inexpensive, and it is effective in most U.S.
climates. According to the United States Environmental Protection Agency, the U.S has "one of
2
the largest agricultural bases in the world and has historically used more MeBr than any other
country (EPA, 2005)."
A controversy now surrounds the use of MeBr and its phase-out schedule as developed
nations face a complete elimination of the substance by 2005. Within the U.S., California and
Florida consume the largest amounts of MeBr as their combined total domestic usage is greater
than 75% (Carpenter et al., 2000). Thus, Florida and California have conducted most of the
research on the costs and benefits of MeBr because they will absorb most of the total cost
associated with the phase-out. Florida and California use MeBr on fruits and vegetables to fight
microscopic parasitic roundworms known as root-knot nematodes, and major soil-borne diseases
such as bacterial wilt, southern blight, fusarium wilt (fungus), and fusarium crown and root rot.
Weeds are also effectively handled with repeated applications of MeBr (Carpenter et al., 2000).
Many of the problems faced by growers in Florida and California are the same as those faced by
Georgia producers.
This study addresses the potential economic impact to Georgia bell pepper producers at
the farm level under the MeBr phase-out. Georgia ranks third in the U.S. in acreage of fresh
market vegetables planted, and vegetables are the second most valuable crop in Georgia with an
approximate farm-gate value of $901.2 million (Boatright and McKissick, 2003, p.59).
Vegetable growers in Georgia state that eliminating MeBr will reduce yields and increase
production costs as a result of adopting more expensive and less effective alternatives (Seabrook,
2005). They requested that the United Nations Environment Programme (UNEP) consider their
nominations to continue using MeBr as defined under the “critical use exemptions” clause of the
Protocol.
3
1.1.2 Montreal Protocol and “Critical Use”
According to a USDA Economic Research Service (ERS) report (USDA, 2000), over 160
countries have agreed to reduce ozone depleting substances (ODS) by ratifying the Montreal
Protocol. Methyl bromide is individually classified within the Protocol as an Annex E controlled
substance with an ozone depleting potential (ODP) of 0.6 (UNEP, 1995). The ODP indicates the
amount of ozone destroyed by the emission of a particular gas relative to chlorofluorocarbon-11
(CFC-11), a major ozone depletor (Carpenter et al., 2000). Substances with an ODP over 0.2 are
considered Class I ozone depletors and are required to be phased out under the Protocol and the
United States' Clean Air Act. Additionally, Section 602(e) of the Clean Air Act states, “Where
the ozone-depletion potential of a substance is specified in the Montreal Protocol, the ozone-
depletion potential specified for that substance under this section shall be consistent with the
Montreal Protocol (USEPA,1990 )” However, MeBr's contribution to agriculture, and its
integral role in facilitating international trade have lead to a general agreement that there needs to
be allowances for “critical,” “quarantine,” and “pre-shipment” uses.
The term “critical use” was introduced to identify uses pertaining specifically to MeBr.
Decision IX/6 of Appendix 1 to the Protocol lays the foundation of “critical use” as decided upon
at the Ninth Meeting of the Parties (UNEP, 2000). Decision IX/6 states: (a) That a use of methyl
bromide should qualify as “critical” only if the nominating Party determines that:
(i) The specific use is critical because the lack of availability of methyl bromide for
that use would result in a significant market disruption; and
(ii) There are no technically and economically feasible alternatives or substitutes
available to the user that are acceptable from the standpoint of environment and
health and are suitable to the crops and circumstances of the nomination;
4
The Protocol delineates the phase-out schedules of MeBr according to parties considered
“developed” or “developing” nations. Article 2H of the Protocol defines the phase-out schedule
of MeBr to be administered by all Parties identified as “developed” nations, and stipulates that
these Parties “shall ensure…its calculated level of the controlled substance in Annex E does not
exceed, annually, its calculated level of consumption in 1991 (UNEP, 2000).” The 1991 baseline
figure also applies to those parties involved in the production of MeBr. The phase-out schedule
for Article 2H countries is:
25% reduction in 1999
50% reduction in 2001
70% reduction in 2003
100% reduction in 2005
Preshipment and quarantine uses exempt
Critical and emergency uses allocated after 2005.
1.1.3 Methyl Bromide Alternatives
Critical uses not withstanding, Georgia producers must identify alternative production
methods that incorporate chemical substitutes as a replacement to MeBr. As noted, California
and Florida have been the two states to predominantly research alternatives and conduct
economic impact analyses concerning the phase-out. These states rely on MeBr for the benefits
accruing to the production of strawberries and tomatoes. Florida uses 237% more MeBr in the
production of fresh market tomatoes than ten other states combined (Carpenter, 2000).
California uses 277% more MeBr for fresh market strawberry production than nine other leading
states combined (Carpenter, 2000). California alone accounts for half of the 35 million pounds
(active ingredient) of MeBr used annually for preplant fumigation in the U.S. (USDA ERS,
5
2000). Moreover, Florida consumes 30% of the total MeBr used for preplant fumigation in the
U.S. (USDA ERS, 2000). The total values to California and Florida for fresh market
strawberries and tomatoes are estimated at $1.4 billion and $920 million, respectively (USDA
NASS, 2005). These states combined account for 95% and 68%, respectively, of the total U.S.
fresh market value of strawberries and tomatoes (USDA NASS, 2005).
1.1.4 Technically Viable Alternatives
Before conclusions could be made regarding the economic impact of the MeBr phase-out,
substantial resources were invested in identifying technically viable alternatives. The USDA
Economic Research Service (ERS), the University of Florida, and the National Center for Food
and Agricultural Policy (NCFAP) worked together to develop studies that analyzed the
regulatory limitations and economic impacts of technically viable alternatives to MeBr (USDA
ERS, 2000). Scientists and growers are interested in chemicals that can produce yields
exceeding or comparable to MeBr. The majority of experiments that focused on yield were
conducted on California and Florida's strawberry and tomato crops. Thus, yield data concerning
other vegetables, vineyards, orchard, and nursery crops is limited (USDA ERS, 2000).
Early studies focused on the possibility of methyl iodide (MI) (iodomethane) as a
technically viable alternative to MeBr (Hueth et al., 2000). Studies conducted at the University
of California, Riverside indicate that MI is advantageous to MeBr because it is quickly broken
down and lasts in the atmosphere for a period of one to two days as opposed to MeBr's two years
(Zhang et al., 1998). Further, because MI is handled as a liquid and MeBr is handled as a gas,
worker safety is increased throughout the production process. Studies tested the performance of
MI under a broad range of environmental conditions such as soil moisture, temperature, texture,
and fumigation time. They found that MI was," consistently more effective than MeBr, on a
6
molar concentration basis, under a range of soil moistures, temperatures, soil textures and
fumigation times (Zhang et al., 1998, p.78)." An additional study by Becker et al. (1998) tested
the dose responses of MeBr and MI against three parasitic nematodes and one fungal plant
pathogen. California growers are susceptible to nematode infestations and fungi without the use
of an effective fumigant. In every case MI proved more effective at controlling for these pests
than MeBr. Finally, (Ohr et al., 1996) evaluated the effectiveness of MI for the control of fungi,
weeds, and nematodes in a series of fifteen field trials, and again found MI to be as efficacious or
more than MeBr.
Florida growers face different environmental conditions than those in California. In
addition to fungi, nematodes, and bacteria, the most problematic pests for growers in Florida and
the southeast are weeds. Yellow and purple nutsedge are common weeds that thrive in Florida
due to a long growing season and humid climate. Purple nutsedge is considered to be the most
noxious of all weeds found in tropical and sup-tropical regions of the world (Gilreath et al.,
2004), and few herbicides are registered for use throughout the entire growing season. Thus, it is
necessary to apply a combination of plastic mulch, fumigants, and herbicides to crops such as
tomato and pepper (Gilreath et al., 2004). A list of fumigants such as metam sodium (MNa),
chloropicrin (teargas) (Pic), anhydrous ammonia (AHN4), and 1,3-dichloropropene (1,3-D) were
analyzed in comparison to MeBr. These fumigants were tested in combination with the
herbicides napropamide, metolachlor, and pebulate. Gilreath et al. (2003) tested the effects of
these chemicals on bell peppers and cucumbers when applied under plastic polyethylene mulch
with drip irrigation. This work furthered studies performed on tomatoes by Locascio et al.
(1997). Both studies found the combination of 1,3-D + Pic to be effective at controlling for
nematode infestations, but the same combination failed to control for purple and yellow
7
nutsedges. With respect to tomatoes and peppers, combining the fumigants with pebulate
significantly reduced nutsedge levels up to 22% compared to those treatments not receiving
herbicides (Gilreath et al., 1994). The results of the 2003 study suggest that there are, "several
viable alternatives for MeBr in vegetable rotations (Gilreath et al., 1994, p.4)." However,
changes in rotations and planting seasons could affect the observed values. Unfortunately, the
erratic nature of fumigant efficacy on weed control does not allow for a single replacement for
MeBr (Gilreath et al., 2003).
Many field trials and indoor experiments conducted throughout the U.S. explored the
efficacy of alternative fumigants, fungicides, and herbicides with respect to MeBr. It is noted in
the scientific literature that local environmental conditions will dictate which combinations of
these chemicals and their respective application methods are most suitable for growers.
Agricultural economists use these results to conduct economic impact studies that account for the
regional demands and limitations of producers.
1.2 Objectives
The specific objectives of the research are as follows:
1) Identify and classify feasible fumigant alternatives to MeBr for Georgia growers through the
use of stochastic dominance analysis.
2) Combine available alternative fumigants with current production technologies to produce
economically viable production methods for Georgia growers.
3) Provide agricultural extension agents with a tool to educate Georgia vegetable farmers on the
resulting economic impact of varying input substitution strategies.
Stochastic dominance analysis, a tool that considers risk-return tradeoffs in identifying
more efficient methods among alternative production plans, will be used to address the first
8
objective. The second and third objectives are accomplished by employing an enterprise budget
developed for Georgia pepper producers in conjunction with an optimization/simulation
programming model. This combination recognizes the unique factors inherent to pepper
production and offers tailored solutions for Georgia producers. Enterprise budget analysis will
determine the comparative financial feasibility of a pepper production enterprise under
production plans involving MeBr and its substitutes. Optimization-simulation techniques
employed in a multi-period programming framework will determine the relative overall
economic viability and feasibility of optimal production and financial plans prescribed under
production systems that involve MeBr and its substitutes.
1.3 Organization
The remainder of this thesis is divided into five chapters. Chapter 2 examines the
economic literature with respect to stochastic dominance, and the optimization and simulation
procedures related to mathematical programming techniques. Chapter 3 identifies issues related
to commercial pepper production and management in the U.S. and Georgia. The development of
the theoretical model employed in this study is then discussed in Chapter 4. In addition to the
descriptive data and enterprise budgets used as support for the theoretical models, Chapter 5
presents the results of the stochastic dominance analysis and the simulation and optimization
procedures. The final chapter, Chapter 6, summarizes the study, presents the conclusions,
addresses limitations of the study, and offers suggestions for future research.
9
CHAPTER 2
LITERATURE REVIEW
2.1 Economic Impact Assessments
A chemical's technical feasibility is a necessary, but not independently sufficient,
condition for its economic feasibility. Published research on a crop's yield or survival, subject to
chemicals under various environmental or controlled conditions, does not satisfy requirements of
economic feasibility such as regulatory approval, sustainability, or a consistent positive return on
investment (Schaub, 2004). These requirements constitute potential barriers to economic
feasibility and dictate the use of economic assessment studies to evaluate the MeBr phase-out.
As noted in the previous chapter, regional differences in soil type and weather conditions may
produce large discrepancies among the efficacy of alternatives to control for a variety of pests
and diseases. Research conducted in California and Florida examines these differences and
constitutes a majority of the information concentrated on identifying MeBr alternatives
(VanSickle, 2000) (Carpenter, 2000).
In addition to these states, studies cited in this section investigate the economics of
pesticide use in the U.S., the economic values of alternative production methods and chemicals,
and the economic impact of the MeBr phase-out among regions, state, and countries. The
literature provides alternative empirical methods of assessing economic impacts through the use
of partial budgeting analysis, stochastic dominance analysis, and optimization/simulation
procedures. The use of integrated pest management (IPM) is of contemporary importance to
10
producers subject to the phase-out. The literature emphasizes that IPM will likely be a necessary
component of the production process for these growers. Literature surrounding the MeBr phase-
out acknowledges the variability and usefulness of IPM techniques among producers. However,
IPM is usually evaluated as a secondary consideration to the more important task of identifying
viable chemical substitutes (Lynch et al., 1997).
2.1.1 Economic Impact Assessments of Pesticide Use
Fernandez-Cornejo et al. (1998) investigated the economic value of pesticide use in
agriculture. The report summarized the empirical evidence related to the economics of pesticide
use. It emphasized the estimation of the value of pesticides in U.S. agriculture, the economic
effects of reducing or restricting pesticide use, and the promotion of IPM to reduce the potential
health and environmental effects associated with pesticide use. The study identified marginal
productivity calculations, the expected loss to pests, and the economic effect of banning
pesticides as three perspectives used to estimate the economic value of pesticide use.
It is generally recognized that producers' per acre expenditure for pesticides increases as
the value of the crop increases. For example, while wheat and corn producers spent
approximately $6 and $22 per acre in 1995, respectively, on pesticide use, cotton and strawberry
producers spent nearly $48 and $1600 per acre, respectively, during the same time period
(Fernandez-Cornejo et al., 1998). According to Fernandez-Cornejo et al. (1998), estimates of the
value of the marginal product (VMP) of pesticide use provide growers and policymakers with an
indirect measure of the cost "in terms of foregone agricultural output" of reducing pesticide use.
They note that under the usual assumptions, a farmer would maximize profits by increasing
pesticide use up to the point where the expected marginal return (VMP) equals the pesticide
marginal cost (Fernandez-Cornejo et al., 1998). They found that the VMP of pesticides appears
11
to be falling, but the methodology surrounding alternative empirical models is controversial.
Further, when farm risk is assessed in combination with pesticide use, they concede that the
conventional view recognized pesticide use to be risk reducing; however, there was no empirical
consensus that determined the effect of risk on the VMP. Studies found that uncertainty about
output price and yield leads to lower optimal levels of pesticide use by individual farmers,
whereas uncertainty about other variables such as pest density leads to a higher optimal pesticide
use under risk aversion (Fernandez-Cornejo et al., 1998). Thus, the authors concluded that the
VMP of most pesticides in the U.S. is higher than their corresponding price.
Fernandez-Cornejo et al. (1998) examined studies that focused on the expected yield
losses relative to a current or potential yield and the respective impact these losses have on the
value of pesticide use. They recognized that, "estimates of crop yield losses that might result
without the availability of pesticides are difficult to obtain because these losses vary by crop,
soil, and weather condition. In addition, yields may vary by year because of technological
developments (i.e. new plant varieties), changes in cropping practices (destruction of crop
residues), appearance of new pests, and weather (Fernandez-Cornejo et al., 1998, p.470)." The
resulting empirical estimates are highly variable and tend to be based on judgments of experts in
different fields of natural science (Fernandez-Cornejo et al., 1998). The authors summarized
yield losses for several major U.S. crops and concluded that fruits and vegetables have high
losses when subject to the elimination of insecticides and fungicides. They estimated that the
elimination of herbicides results in yield losses ranging from 0% to 53%. Fernandez-Cornejo et
al. (1998) cite a study conducted by the USDA in 1997 that calculated the productivity of
pesticides to be approximately $3 to $4 of pesticide expenditure.
12
The article further assessed the economic effects of total or partial bans on pesticide use
in general or in individual case studies. The authors identified partial budgeting and large scale
econometric models as two methods that are generally used to estimate these effects. Partial
budgeting estimated the value of the production lost without pesticides assuming that output
prices remained constant, while the econometric models allowed for input and output
substitution during production (Fernandez-Cornejo et al., 1998). Fernandez-Cornejo et al.
(1998) concluded that a total ban of pesticides would increase annual consumer expenditures by
$228 per household (in 1989 dollars), amounting to approximately $30 billion per year. The
production of fruits and vegetables is examined as well as a corresponding total ban in these
industries. The result of a total ban in these sectors may require producers to increase production
acreage by 2.5 million acres (44%), and increase unit production costs by 75%. In turn,
wholesale prices of fruits and vegetables could increase 45%, returns to producers could decrease
by 30%, retail prices could increase by 27%, and domestic consumption could fall by 11%
(Fernandez-Cornejo et al., 1998). However, the studies received criticism for assuming a total
ban would be realistic, and for failing to acknowledge that future research may minimize
impacts. Fernandez-Cornejo et al. (1998) noted that authors of the original studies agree that
further research needs to identify intermediate points between current practices and a total ban.
The availability of alternative means of pest control influences the value of a pesticide's
use to agricultural producers and consumers (Fernandez-Cornejo et al., 1998). Thus, assessments
of partial bans of pesticides offer more realistic estimates of economic impacts. For example,
Fernandez-Cornejo et al. (1998) reviewed a study conducted in 1991 that assumed a 50%
reduction in pesticide use. This study concluded that the total costs of reduction would be
approximately $1 billion per year although the true impact would be difficult to determine
13
because there are different mechanisms in which to implement a ban (Fernandez-Cornejo et al.,
1998).
Finally, Fernandez-Cornejo et al. (1998) considered the value of integrated pest
management in the context of both governments' attempts to reduce the detrimental effects of
pesticides to human health, and as an option for producers to maintain viable business
operations. IPM uses multiple techniques to," maintain pest infestation at the most economically
sensible level rather than attempting to completely eradicate all pests (Fernandez-Cornejo et al.,
1998, p.478)." Some of the more important techniques adopted by producers are scouting, soil
testing, pheromone traps for pests, and cataloging data on weather patterns that identify the
development and activity of specific pests. Additionally, biological controls of pests such as
natural predators, parasites, and pathogens result in minimal environmental hazards, and may
result in an equivalent amount of pest control in some situations. Fernandez-Cornejo et al.
(1998) stated that the empirical evidence on the effect of IPM on the use of pesticide use is
mixed. The authors cited a study conducted by the National Foundation for IPM Education that
concluded IPM reduced pesticide use by 15% while delivering net returns to producers.
2.2 Methyl Bromide Alternatives Outreach
The Methyl Bromide Alternatives Outreach (MBAO) is an organization focused on
facilitating the exchange of information concerning viable alternatives for MeBr. The
organization began holding the Annual International Research Conference on Methyl Bromide
Alternatives and Emissions Reductions in 1994. Their goal is to establish a forum that promotes
the exchange of research information among agricultural and forestry specialists from
governmental, academic and private institutions. Several years after the beginning of the
14
conference, as scientists had had time to conduct experiments on alternatives and publish their
results, discussions began to incorporate the economic impact surrounding these developments.
The earliest economic impact assessments, analyzed and presented by Lynch et al.
(1997), provided researchers with a base-line from which to evaluate the impact of the MeBr
ban. Lynch et al. (1997) identified changes regarding the costs of technically feasible chemical
and non-chemical alternatives, the initial economic results of studies conducted in California and
Florida, the regulatory environment surrounding alternatives, and comparisons of the efficacy of
alternatives. Results presented that same year by Hueth et al. (1997) focused on the economic
impact of banning MeBr in California agriculture. Hueth's study measured the cost of
prohibiting MeBr use in California by directly measuring the change in consumer and producer
welfare. The methodology was based on data concerning the changes in per acre yields and
production costs resulting from the adoption of available alternatives.
Hueth et al. emphasize that their research was an estimation of the short-term impact of
removing MeBr completely, and that it adds to the literature by providing a more general
analysis of the impacts of banning MeBr. The authors further acknowledged that the impacts
reported in the study were uncertain because of the limited experimental opportunities and the
inability to forecast outcomes related to future innovations. The study concluded that the impact
on growers' profits may experience a wide range of values corresponding to the infiltration of
exotic pests. In good years growers may lose as little as $71 million while in years of high
infestation they may lose as much as $334 million in profits. The net impact, which accounts for
consumer benefits and losses as well as grower losses, exhibited the same dramatic range with
yearly impacts from $60 million to $242 million (Hueth et al., 1997).
15
Presentations made in 1998 and 1999 by Lynch and Carpenter focused on identifying the
most likely alternatives for producers, and investigating the impacts of phasing out MeBr for the
United States, respectively,. The 1998 study sought to identify the impact to producer yields and
costs by gathering information from research documents, expert opinions, and workshops. Their
research concluded that both Florida and California growers would increase usage of the
combination of 1,3-D and Chloropicrin. To maximize profits, the study found that Florida
growers would likely use an accompanying herbicide, and California growers would increase the
percentage of Chloropicrin in the mixture. Lynch and Carpenter (1999) then compare MeBr
with the next best alternative over a range of fruits and vegetables grown not only in California
and Florida, but also in Georgia and North Carolina.
Assuming regulatory restrictions, cost, and environmental factors dictate these
alternatives, the authors, "computed the increase in revenue from using one pound of MeBr
rather than the next best alternative assuming the market prices do not change for the crop
(Lynch and Carpenter, 1998, p.2)." These revenues serve as the value of MeBr to producers on a
per pound basis. After making assumptions for producer usage rates of MeBr, a grower's loss
was then calculated as a function of the amount of MeBr used per acre. The value of MeBr to
fruit and vegetable growers in California ranged from $3.85/lb. for peppers to $41.43/lb. for
southern coast strawberries. The value to Florida fruit and vegetable growers ranged from
$4.25/lb. for tomatoes to $33.48/lb. for strawberries.
Lynch and Carpenter's (1999) study identified these estimates as one approach to
estimating the economic impact of the MeBr phase-out. Their second approach examined the
annual crops that use MeBr extensively, and allowed for adjustments to acreage and prices.
Crops such as tomatoes, strawberries, peppers, watermelon, cucumber, squash, and eggplant are
16
crops primarily grown by horticulturalists in California and Florida with additional production
taking place in South Carolina, North Carolina, and Georgia (Lynch and Carpenter, 1999). The
authors used this presentation to introduce the phase-out as a spatial partial equilibrium problem.
Furthermore, Carpenter et al. (2000) fully developed the model used in the 1999 study
within a comprehensive study published the following year. The report by Carpenter et al.
(2000) was funded by the USDA's Economic Research Service and published through the
National Center for Food and Agricultural Policy. The authors developed a regionally
disaggregated model that emphasized California, Florida, Georgia, and South Carolina as the
primary users of MeBr, and identified Texas and Mexico as their competitors. The model used
calculations of baseline equilibrium production, monthly shipments between production areas
and markets, and monthly consumption given current technologies as a guide to compare with
changes resulting from the phase-out (Carpenter et al., 2000). Crop prices for each market were
introduced, and as MeBr is phased out, production technologies shift; resulting in changes in
production costs and expected monthly yields. The alternative technologies used in the study
were identified through previous research acknowledged during MBAO conferences. The study
defined the best alternative technology as the one with the lowest per-unit cost (Carpenter et al.,
2000). The authors use a simulation approach, rather than econometric, because data for
individual producers was unavailable. Costs were allowed to vary among regions and yields
were assumed to be non-stochastic. The goal of the model was to maximize producers' returns
and consumers' benefits while recognizing constraints on available land in each region, and the
amounts sold to consumers could not be greater than the amount supplied (Carpenter et al.,
2000).
17
California strawberry growers were expected to shift from using MeBr to a combination
of Chloropicrin and Vapam at an additional cost of $97.50 per acre. Florida strawberry and
tomato growers were expected to substitute Telone C-17 and various herbicides at an increased
cost of $227.50 per acre. Georgia and South Carolina growers would switch to Telone C-17 and
pebulate at an excess cost of $13.36 per acre (Carpenter et al., 2000). The study did not present
the additional costs per acre for pepper producers in Georgia, and acknowledged that little
research was conducted into alternatives for peppers in Florida. The authors assumed yield
losses among Florida pepper growers to be approximately 12.5%, and that they would use a
combination of Telone C-17 and herbicides costing $227.50 per acre.
The results of the study were presented in comparison to the pre-ban baseline. Carpenter
et al. (2000) stated that consumers incur the heaviest cost from the ban as consumer surplus is
expected to decrease by $158 million. The decrease in consumer surplus for strawberries and
tomatoes was estimated to be 10.3% and 1.7%, respectively. Consumer surplus for peppers was
expected to decrease $4.5 million, or 1.1% (Carpenter et al., 2000). With respect to producers,
the model predicted that California growers will increase strawberry production with a
corresponding decrease in tomatoes. Florida producers will increase the acreage planted of
strawberries, but will reduce the acreage planted of tomatoes and peppers (Carpenter et al.,
2000). Georgia producers were expected to increase their production of tomatoes, but the study
does not present any figures concerning increases or decreases in peppers. However, total U.S.
pepper production was estimated to decrease 14% (Carpenter et al., 2000). Alternatively, the
authors noted that changes in price may make it profitable for growers to continue operations.
They calculated the price impact to peppers resulting from the ban to be in the range of -$39.82
to $51.96 per ton (Carpenter et al., 2000).
18
Carpenter et al. (2000) analyzed the impact on producer revenues across regions and
compared them with the baseline. Revenues did not include the costs of production, and as a
result it was possible for a producer to experience post-ban revenues that were higher than pre-
ban revenues as a result of price increases (Carpenter et al., 2000). The model estimated that
Florida producers will experience decreases in revenues for tomatoes and peppers of $57.3
million and $12.6 million respectively (Carpenter et al., 2000). In California, strawberry and
tomato producers may experience a respective change of $38.8 million, and -$35 million in
receipts (Carpenter et al., 2000). The model prescribed that shifts would occur in production
among regions to compensate for dramatic losses in specific areas. However, the U.S. grower
that loses is expected to experience a decrease in revenues of $153.9 million (Carpenter et al.,
2000). Finally, the report indicated that if revenues were used to gauge the overall impact to
U.S. consumers and producers, the model estimated a total net decrease in welfare of $76.5
million (Carpenter et al., 2000).
More recent studies presented at the MBAO conferences evaluated alternative production
methods and used current data in an attempt to assess the options and impacts to growers facing
the phase-out. A study conducted by Jovicich et al. (2003) investigated the economic benefit
associated with the greenhouse production of peppers. The study found that growing colored
peppers for the specialty market can result in economically viable operations. The research was
conducted through the Protected Agriculture Project at the University of Florida and emphasized
the use of screened plastic greenhouses with passive venting as an economical alternative to field
production (Jovicich et al., 2003). The cost of these systems was estimated at $2 to $4 per
square foot, and it was demonstrated that fungal diseases decreased, fruit quality improved, and
yields were up to ten times that of field grown peppers (Jovicich et al., 2003). Although the
19
growing season could be extended for producers under this method, start up costs and knowledge
requirements may prohibit greenhouse adoption by established growers.
During the 2004 MBAO proceedings, Sydorovych et al. (2004) presented research
focusing on the economic impact of MeBr alternatives for strawberries. Their objective was to
evaluate the economic viability of chemical alternatives using a partial budget analysis. Several
characteristics of this study are relevant to the goals of our research. For example, researchers
first developed a cost model for a plasticulture production system on a five acre representative
strawberry farm in the Piedmont and Coastal Plain regions of North Carolina (Jovicich et al.,
2003). The production practices used in the study resulted from consultations with research
specialists and agricultural extension agents of North Carolina State University. An important
assumption of the Jovicich et al. (2003) model is that the machinery and equipment used in the
enterprise budget can be applied to other farming enterprises other than strawberries. This is
characteristic of the farming operations of many small to medium size growers found in the
southeastern United States. As a result, machinery expenses, not including fumigation and
irrigation costs, reflect the equipment costs for a total farm business (Jovicich et al., 2003).
Partial budgeting analyzes small production adjustments on farm profitability by
comparing the negative effects of applying a new treatment relative to a base or standard
treatment (Jovicich et al., 2003). The study found that production costs increased if an
alternative treatment resulted in either higher fumigation costs and/or higher yields because
higher yields result in both higher labor and material costs. The chemical alternatives tested in
the study were Telone-C35, Telone II, chloropicrin, Inline, and metam sodium. Methyl bromide
was recognized as the base treatment with respect to yields and revenues. Interestingly, MeBr
resulted in the highest cost to producers per acre at $1,267, while chloropicrin reduced costs
20
relative to the base by $92.00 per acre. Further, chloropicrin and Telone-C35, when used
independently, produced yields greater than MeBr. A comparison of revenues resulted in
chloropicrin, Telone-C35 and metam sodium applied by shank injection each providing
additional returns above MeBr of $1,767.60, $290.79, and $3.18 respectively. The study did not
list returns for Telone II and Inline.
2.3 Additional Economic Impact and Viability Studies
2.3.1 Economic Viability of Methyl Iodide
While most viability studies examining the efficacy and economic feasibility of
alternatives focus on a menu of choices available to producers, researchers may some times
focus on the economic impact of a single chemical alternative. Hueth et al. (2000) compared the
effectiveness of methyl iodide (MI) against MeBr for preplant fumigation, "through an analysis
that combines information on the current world market of iodine, the implications of limited
studies on MI effectiveness, and past analyses of the economic benefits of methyl bromide to
California producers (p.45)." An important assumption of the study was that the researchers
considered 1,3-D and chloropicrin to be the primary alternatives used in place of MeBr for
California growers. Thus, their study of MI was not an endorsement, and MI's demand schedule
was constructed by analyzing its cost and effectiveness against these alternatives (Hueth et al.,
2000). The efficacy of MI compared to MeBr was established through the studies noted in
Chapter one.
Hueth et al. (2000) assessed the potential market for MI by calculating the per-unit value
of MI, and developing demand curves for MI under six different scenarios. The study defined
the value of the marginal product (VMP) of MI as the value resulting from a discrete change
from one control material to another. The VMP represented the benefit of using MI compared to
21
the next best alternative (Hueth et al., 2000). The study restricted the demand of MI to
commodities where the VMP was positive. After calculating the VMP, demand schedules were
mapped out over a range of prices under optimistic, average, and pessimistic scenarios regarding
changes in application rates and non-chemical application costs (Hueth et al., 2000). The
expected price for MI was assumed to be the current price of $11.00/lb., and at least as high as
the $2.25/lb. price paid for MeBr.
The results indicated that as the price per pound of MI increases, the demand for MI
under each scenario decreases rapidly. The study evaluated demand schedules for those
producers using both tarping and non-tarping operations. As expected, the demand for MI
decreased at a greater rate for producers using a low density polyethylene tarp because tarping
added additional costs to operations and provided additional protection against pests. The
demand for MI was less across all prices and alternative scenarios when compared to non-tarping
operations.
The authors acknowledged that their analysis was not representative of the entire
California market, and therefore the potential demand for MI would likely be much higher
(Hueth et al., 2000). Their results indicated that at a price of $11.00/lb. demand could reach
four million pounds in California alone. As demand increased for MI due to the phase-out,
suppliers may be encouraged to innovate and enter this new market (Hueth et al., 2000).
However, limitations surrounding MI are its current costs of extraction and limited supply
compared to MeBr. Current production of MI could supply the demand of niche markets in the
U.S., but additional investments in extraction are needed in order to supply larger regional
markets (Hueth et al., 2000). Overall, the authors identify MI as a viable alternative to MeBr.
22
2.3.2 Economic Impacts within the U.S. Vegetable Industry
Analyses conducted by (Spreen et al., 1996) (VanSickle, 2000) investigate the impact of
the phase-out with respect to the U.S. winter fresh vegetable market and the U.S. vegetable
industry as a whole. Both studies make important contributions to the literature by recognizing
and quantifying the impacts to producers that rely heavily on MeBr.
Spreen et al. (1996) estimated the impact of the phase-out on the winter market for fresh
vegetables such as green peppers, squash, cucumbers, tomatoes, eggplant, and watermelons. The
primary impact of the phase-out to this market will be felt by Florida because they are the
leading domestic supplier of these products (Spreen et al., 1996). Previous research had failed to
assess the impact to Florida independently by region or season. Although California is a major
consumer of MeBr it does not compete with Florida's vegetable market due to the seasonal
differences between the two states. The study’s model focused on months in which Florida is a
supplier of vegetables to the U.S. market (Spreen et al., 1996). Spreen et al.'s model served as
the basis for the model used by (Carpenter et al., 2000). However, in this case researchers
characterize the North American winter fresh vegetable market as a spatial equilibrium problem
rather than the entire U.S. market (Spreen et al., 1996). The authors constructed the winter fresh
vegetable market by disaggregating Florida into four regions of production and acknowledging
Texas and Mexico as dominant suppliers during the winter season occurring from November
through May.
A quadratic programming model was developed to determine factors such as equilibrium
consumption of each commodity in every month and region, optimal levels of shipments
between areas of supply and demand, optimum acreage employed for each cropping system, and
optimum monthly production of commodities by region (Spreen et al., 1996). Solutions for total
23
supplies accounted for the double cropping systems common to Florida growers and the
production of commodities under each system. Total demand was estimated through an inverse
Rotterdam system of demand equations developed in previous studies, but was expanded to
include the additional commodities used in this study (Spreen et al., 1996). For simplicity, all
parameters of the model were assumed to be non-stochastic although the yields of fresh
vegetables were known to be highly variable. The authors noted that constraints imposed on the
model did not include land, labor, and machinery. They justified this approach because, "it could
not be established that effective resource requirements restricted the production of fresh
vegetables in Florida, Mexico, or other U.S. supply areas (Spreen et al., 1996, p.436)."
The model was initially solved for a base-case scenario without accounting for the
elimination or reduction of MeBr. This analysis generated results that conformed to past data
with respect to shipping among supply regions, quantity of the crop produced, and the acreage
employed for production of the six crops. However, the model did fail to produce realistic
results concerning pepper production in Florida. Under the base case, Southwest Florida was
expected to grow no peppers although it has historically been the largest pepper producing region
in the state (Spreen et al., 1996). After modifying the model to account for the loss of MeBr, the
solution indicated that the production of peppers, tomatoes, eggplant, and cucumbers would be
eliminated in one of Florida's four regions. All other regions throughout Florida were expected
to experience adverse effects from the ban due to a severe contraction in production of these
crops (Spreen et al., 1996).
The model predicted that Mexico and Texas would provide for most of the lost
production in Florida as neither region was affected by the ban (Spreen et al., 1996). Of the six
crops analyzed in the study, peppers and tomatoes were subject to the greatest reduction in
24
yields. In terms of revenue, Florida's free on board (FOB) revenues for all six crops combined
were expected to decline by approximately $548 million or 53%. Mexico's total (FOB) revenues
were expected to increase by $368 million while Texas gained revenues of nearly $24 million
solely through its increase in pepper production (Spreen et al., 1996). Spreen et al. (1996)
estimated that the loss of MeBr would have a $1 billion impact on the US winter vegetable
industry, with Florida accounting for nearly all of this impact. However, they emphasized that
viable substitutes may invalidate these conclusions and reduce future impacts to Florida growers.
A study conducted by VanSickle et al. (2000) expanded on the Spreen et al. model
developed in 1996. In the interim between studies, significant research was conducted on
alternatives to MeBr that provided more reliable data in terms of yields, costs, and regulatory
restrictions. The objective of VanSickle et al. (2000) was to evaluate the impacts of MeBr
alternatives on U.S. producers of fresh vegetables. Similar to Spreen et al. (1996), the study
characterized the North American vegetable market as a spatial equilibrium problem. However,
the new model introduced strawberries into the model in addition to the crops from the previous
study. Production areas identified by the model were expanded to include Florida, Mexico,
California, Texas, South Carolina, Virginia and Maryland combined, and Alabama and
Tennessee combined. Florida and California were then separated into regions (VanSickle et al.,
2000).
The model allocated production based on the cost of delivering the crops across regional
markets. The study used a similar system of inverse Rotterdam equations to estimate total U.S.
demand (VanSickle et al., 2000). Production costs were estimated according to the geographic
areas used in the model, and the chemical alternatives were selected by a team of scientists,
industry representatives, and environmental advocates (VanSickle et al., 2000). Strawberry
25
growers in Central Florida were expected to switch from MeBr to a combination of Telone C17
and Devrinol. Growers could expect a decrease in pre-harvest costs of $71 per acre and an
accompanying decline in yield of 15% (VanSickle et al., 2000). Florida tomato growers were
expected to use Telone C17 with the herbicide Tillam. If a double cropping system was
employed, costs ranged from a decrease of $61 per acre for tomatoes and cucumbers to an
increase of $255 per acre tomatoes and squash.
Florida pepper production was assumed to replace MeBr with a combination of Telone
C17 and Devrinol. The impact to costs varied depending on single or double cropping systems.
For single crops, costs ranged from a decrease of $41 per acre for spring peppers produced in
Central Florida to an increase of $397 when peppers were grown close to urban areas. For the
double cropping system, costs were expected to increase by $437 per acre. Additionally, the
associated yields for pepper were expected to decline 15% to 25% for most areas of Florida
(VanSickle et al., 2000).
The VanSickle model initially solved for a base-case that assumed the continued use of
MeBr. The base case allowed researchers to identify the accuracy of the model when compared
to the observed data of prior production seasons. Adjustments made to the parameters of the
base model reflected the changes in production costs and changes in the yield in switching to the
alternatives (VanSickle et al., 2000). Three scenarios were examined with respect to the base
case. Holding the costs constant throughout, Van Sickle et al. (2000) ran the model using the
expected yields identified by scientists, a pessimistic scenario using lower than expected yields,
and an optimistic scenario assuming larger yields than the expected case. The base solutions
approximated the production patterns recorded by the UDSA. VanSickle et al. (2000) found that
26
the total production of crops was not expected to dramatically change, but that changes would
occur in the allocation of production across regions.
The greatest losses occurred to tomato and pepper producers in Florida. Under the
expected scenario, Florida tomato growers lost $68.8 million in shipping point revenues and
pepper growers experienced a 65% decline in acres planted (VanSickle et al., 2000). In addition,
the optimistic scenario for pepper growers in Florida resulted in a $119.9 million decrease in
shipping point revenues (VanSickle et al., 2000). In terms of total losses, California strawberries
were greatest at approximately $192.8 million. In the aggregate, Florida and California could
lose $218.4 million and $218.1 million, respectively, in revenues for all crops under the expected
scenario (VanSickle et al., 2000).
Furthermore, continued research examined how much the impacts of the MeBr
alternatives must be reduced in order to minimize the changes in market share for existing
producers (VanSickle et al., 2000). The authors assumed that a "seamless transition" between
MeBr and its alternatives could occur if the largest impact to any producer's yield were less than
or equal to 10% of the baseline scenario. They maintained the cost impacts assumed in the
expected impacts scenario.
Next, adjustments to yield were made on each crop in the model until the largest impact
felt by any producing area for any single crop was a 10% loss in the baseline market share
(VanSickle et al., 2000). Results indicated that in order for tomato production to experience a
seamless transition, while holding costs equal to those under the expected scenario, Florida
growers would reduce yields by 55% to 60% (VanSickle et al., 2000). The model indicated that
pepper producers will need to reduce yield losses by approximately 93% to make the transition.
27
However, the authors noted that future alternatives may reduce the economic impact to producers
if they cause costs to decrease, yields to increase, or new market windows.
28
CHAPTER 3
COMMERCIAL PEPPER PRODUCTION
3.1 United States Pepper Industry
3.1.1 U.S. Production and Consumption
The bell pepper is a mild sweet pepper that is grown commercially on approximately
57,000 acres throughout the U.S. (USDA ERS, 2004). The primary destination for the wide
variety of bell peppers (yellow, purple, red, green) in the U.S. is the fresh market. The fresh
pepper market is the focus of this study because producers in this market face greater demands
by consumers to provide a consistent quantity of quality peppers for rapid distribution to market.
Further, the fresh market receives approximately 90% of all peppers produced domestically
(USDA ERS, 2001). Producers for the remaining 10%, the processed market, are mainly
concerned with generating consistent yields, and place less emphasis on quality.
Within the U.S., most domestic production is supplied by only eight states. The leading
five states for pepper production in terms of harvested acreage in 2002 were California, Florida,
North Carolina, New Jersey, and Georgia respectively (Figure 3.1) (USDA ERS, 2004).
California's production increased 9% from 2002 to 2003 and significantly influenced the 5%
growth in total production for the U.S. over the same period (Figure 3.2) (USDA ERS, 2004).
Peppers are grown throughout the entire year in the U.S., and shipments tend to increase during
the months of May and June. Conversely, import shipments usually rise during the winter
months as domestic producers' fall short of domestic demand (USDA ERS, 2001). The trend in
29
production for peppers has been moving upward over the past twenty-five years as total U.S.
production increased from 582 million pounds in 1979 to 1,648 million pounds as of 2003
(Figure 3.2). This represents an approximate total increase in production of 183% (USDA ERS,
2004). U.S. per capita consumption over the same period increased 141.4% and 2003 registered
as the second highest consumption year on record at 7.0 lbs. per person (USDA ERS, 2004).
3.1.2 Imports and Exports
As U.S. consumer demand for peppers trends upward (Figure 3.2), the contribution of
pepper imports to total U.S. supply exhibits the same trajectory. In 2003, the U.S. maintained its
position as a net importer of peppers with imports representing 25% of the total U.S. supply. Of
the total percentage of imports, 20% supply fresh market demand (USDA ERS, 2004). The
majority of peppers being imported for the fresh market are received from Mexico, Canada, and
the Netherlands with their respective contributions to the whole being 67%, 17%, and 9% in
2003 (USDA ERS, 2004).
Canada is the U.S.'s leading trading partner for fruits and vegetables. Pepper exports to
Canada, valued at $69.4 million, represent 94.5% of total U.S. pepper export value (Figure 3.3).
Mexico is our second most important trading partner for pepper exports. In 2003, Mexico
imported $1.3 million worth of peppers from the U.S. However, Mexico is also the primary
supplier of U.S. pepper imports. U.S. pepper imports from Mexico increased 8.75% from the
years 1999 to 2002 (USDA ERS, 2003). Exports as a percentage of total production increased
over the past twenty-five years, but, from 2002 to 2003, fell from 8% to 7% of total supply
(USDA ERS, 2004) (Figure 3.2).
30
3.2 Georgia Pepper Production
All commercial bell peppers grown in Georgia belong to the genus Capsicum annuum,
and approximately half of commercial brand peppers are openly pollinated or a hybrid variety
(UGA CES, 1990). Georgia's growing season usually begins in the spring during late February
and extends into early July. Fall crops are planted in July and may be harvested into November,
or until the first major frost (Kelly, 2003). Peppers grow well in Georgia due to the state's warm
and humid climate. However, producers must institute effective soil management practices and
pest control schemes in order to generate maximum yields.
Eighty-one percent of bell pepper production in Georgia takes place on approximately ten
farms. Each of these farms is greater than one-hundred acres in size. The remaining nineteen
percent of harvested acreage is spread among eighty-one farms, and almost half of these
producers farm approximately 0.1 to 0.9 of an acre (Figure 3.4) (USDA, 2002). Georgia's total
harvested acreage for bell peppers in 2003 was 5,230 with an accompanying farm gate value
estimated at $87.1 million (CAES, 2003). According to the University of Georgia's Cooperative
Extension Service, watermelons, sweet corn, and snap beans, respectively, lead all Georgia
vegetables in the total acreage employed.
In terms of farm gate value, tomatoes and onions lead the State at approximately $122
million and $104 million, respectively (CAES, 2003). These figures reveal that although bell
peppers rank thirteenth in terms of acres harvested among all Georgia vegetables, their ranking
as the third most valuable vegetable in the state reveals their importance to Georgia agriculture
(CAES, 2003).
31
3.3 Marketing
As with other commodities, the general price level for peppers is determined by supply
and demand. Measured in current dollars, average seasonal prices for U.S. bell peppers have
experienced fluctuations in the range of $19.90 to $30.70 per cwt. through years 1979 to 2003
(Figure 3.5) (USDA ERS, 2004).
Wholesale buyers of vegetables rely on industry standards in packaging to ensure
conformity with every purchase. Thus, bell peppers are commonly packaged in either 1 1/9
bushel wire-bound crates (1 1/9 buctns/crts.) or bushel cartons with each representing 25-30
pounds of peppers (Gast, 1991). An average price per carton of twenty-eight pounds is typically
assumed to be the weight of either packaging option. Thus, converting the U.S. prices measured
in ($/cwt.) into ($/crt.) results in a price range of $5.57 to $8.60 per carton for average U.S.
yields from 1979 to 2003. Georgia producers experienced prices above the national average in
2003. Prices for premium Georgia peppers (Jumbos) ranged from a low of $10.09/crt. for the
fall harvest to a high of $16.24/crt. for the spring harvest. Lower quality peppers (US 1&2)
commanded prices in the range of $7.51/crt. for the fall harvest to $9.93/crt. during the spring
harvest (CAES, 2003)
Variations in commercial pepper prices are primarily explained by supply-side factors.
For example, both seasonal changes in weather conditions and yearly changes in acres planted
influence total production. However, the demand for peppers is unlikely to be a motivating
factor for price variation because consumption patterns change little from year to year (CAES,
2003). Consumer preferences for peppers are represented in the quantity and quality of their
purchases. Although the specialty markets demand a variety of different colored peppers,
consumers tend to prefer green, and they are willing to pay a premium for Jumbo peppers.
32
Thus, marketing campaigns should be focused on these preferences (CAES, 2003). At
the same time, Georgia growers should also maintain focus on delivering a reliable, quality
product, to distributors. In a 2002 University of Georgia marketing survey, wholesalers and
distributors ranked the origin of the product last with respect to those factors that were most
important in their purchasing decision (Wolfe and Fonsah, 2002).
3.4 Farm Management
3.4.1 Technical Considerations
Bell peppers can be grown in a variety of soils and climates throughout the U.S.
However, in order to produce high quality peppers growers must consider climate, soil
conditions, and proper soil preparation before beginning operations. Peppers grow best in light,
fertile, well drained soils. Another requirement is that the soils are subject to adequate tillage.
Tillage is a cultural method of pest management that may be used in conjunction with chemical
and biological methods to reduce crop damage and enhance plant growth. Tillage directly
destroys weeds by disrupting or moving the soil within 10 to 12 inches of the soil surface
(USDA ERS, 2003). Because peppers have roots that reach depths of approximately 36 to 48
inches, tillage also supplies the plants with a greater soil volume, nutrients, and water than other
soil preparation methods such as disking (UGA CES, 1990).
In Georgia, growers usually choose to transplant 5 to 6 week old pepper seedlings rather
than directly seeding into the field. Transplanting allows for better weed control, less specialized
planting and bedding, and earlier harvest dates. Transplants in raised beds warm more quickly
and require greater irrigation and soil maintenance during drought conditions (UGA CES, 1990).
However, an advantage of this technique is that the heat accelerates plant growth. Many Georgia
farmers choose a production combination that uses both transplanting and plastic mulch to
33
increase yields. The combination restricts in-row weed growth while conserving water and
fertilizer (UGA CES, 1990). Drawbacks to this method are the additional costs for specialized
equipment and the inability to control for yellow or purple nutsedges, two particularly
problematic weeds for southeast growers (UGA CES, 1990).
Optimum pepper production in Georgia requires substantial irrigation in addition to
chemical methods of pest management. Producers may choose to use either sprinkler or drip
irrigation depending on the size of operations. While most Georgia growers use some type of
sprinkler system, drip irrigation is becoming more popular due to its effectiveness when used
with plastic mulch. Drip irrigation conserves water and can effectively deliver fertilizers directly
into the bed (UGA CES, 1990). In Georgia, fertilizer management programs are site specific,
but usually require liming the soil with calcitic or dolomitic limestone to lower soil acidity, and a
thorough application of nitrogen (UGA CES, 1990).
3.4.2 Diseases and Pest Management
Pepper producers in Georgia face a variety of diseases and pests that cause severe
economic damage to their crops. Fungi such as Cercospora capsici, Vermicularia capsici, and
Sclerotium rolfsii cause damage to the pepper plant through, respectively, defoliation, rotten
fruit, and rotten stems. In addition to fungi, there are viruses such as tobacco mosaic virus
(TMV), tobacco etch virus (TEV), cucumber mosaic virus (CMV), and potato virus Y (PVY)
(UGA CES, 1990). All of these viruses cause what is generally known as pepper mosaic virus.
As a result of this virus, plant growth may become stunted resulting in malformed and
underdeveloped fruits (Brunt et al., 1996). Pseudomonas spp., also known as bacterial leafspot,
is the most serious disease affecting peppers in Georgia because it results in defoliation and
brown spots on the fruit.
34
Moreover, producers must minimize the devastating effects of root-knot nematodes.
These parasites damage the host by disrupting vascular tissues at the root; thus, preventing or
reducing the uptake of water and nutrients necessary for plant growth. Infected plants display
galled roots, poor growth, wilted foliage, and poor yield. The damage can often be aggravated
by the parasite’s interaction with other micro-organisms such as fungi and bacteria (Brunt et al.,
1996). As a result of these problems, integrated pest management strategies should be employed
throughout the growing season. Scouting fields two to three times a week is the most cost
effective strategy for eliminating seedling pests such as cutworms or thrips, as well as foliage
feeders like aphids, flea beetles, potato beetles, and hornworms. On the other hand, weed
control is one of the more problematic issues facing pepper growers, and must be addressed
through seasonal applications of fumigants and herbicides. As noted, when used with black
plastic mulch, weed control can be attained, but nutsedges may persist as a problem.
3.4.3 Production Costs
The costs involved in maintaining or beginning many agricultural enterprises are
extensive. Producers able to effectively generate a consistent, high quality product for the
market must still consider the costs involved in production in order to maintain viable long-term
operations. Many agricultural commodities do not enjoy successful brand recognition, and as a
result, they do not enjoy the premium prices that such schemes deliver. Therefore, minimizing
production costs should be the main avenue for maximizing net income.
A general listing of costs facing pepper producers can be broken down into total variable
costs and total fixed costs. Land can be a variable or fixed cost. Most variable costs will differ
with respect to the acreage employed and the cultural practices of each producer. Some
examples of variable costs are seed, fertilizer, chemicals, fuel, and labor. These can then be
35
divided between pre-harvest and harvest costs. Pre-harvest costs are utilities, vehicles, and
associated user fees. Fixed costs include equipment owned, management, and general overhead
costs (UGA CES, 1990). Many of these costs can dramatically change across producers or
regions if owners enjoy economies of scale due to organization, specialization, or education.
Georgia producers have access to enterprise budgets prepared by agricultural extension
economists and specialists that allow them to estimate production and break-even costs (UGA
CES, 1990).
36
0
10,000
20,000
30,000
40,000
50,000
60,000
U.S. CA FL NC NJ GA
U.S. & STATES
Har
vest
ed A
cres
$0
$100,000
$200,000
$300,000
$400,000
$500,000
$600,000
Valu
e
Acres Harvested Value in 1000 Dollars
Figure 3.1: Leading U.S. Bell Pepper Producing States (2002) Source: USDA ERS 2004
37
0.0
500.0
1,000.0
1,500.0
2,000.0
2,500.0
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
YEARS
(Mill
ions
lbs.
)
Production Imports Exports Domestic
Figure 3.2: U.S. Fresh Bell Peppers, Production, Import, Export and Domestic Use (1979-2003) Source: USDA ERS 2004
38
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
1000 Dollars
2000: 2001: 2002: 2003:YEARS
Netherlands Canada Mexico World
Figure 3.3: U.S. Bell Pepper Import Values from Selected Countries of the World (2000-2003) Source: USDA ERS 2004
39
0
5
10
15
20
25
30
35
40
45
50
(0.1 - 0.9) (1.0 - 4.9) (5 - 14.9) (15 - 24.9) (25 - 49.9) (50 - 99.9) (≥ 100)
Farm Size (Acres)
Num
ber o
f Far
ms
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
Har
vest
ed A
crea
ge
Number of Farms Total Harvested Acres
Figure 3.4: Georgia Farms, Farm Sizes, and Harvested Acreage (2002) Source: USDA NASS 2002
40
0
5
10
15
20
25
30
35
40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Years
($/C
wt)
0
1
2
3
4
5
6
7
8
(Lbs
)
Season-average price Per Capita Use
Figure 3.5: U.S. Bell Peppers Season Average Prices and Per Capita Consumption (1979-2003)
Source: USDA ERS 2004.
41
CHAPTER 4
EMPIRICAL METHODOLOGY
4.1 Expected Utility, Risk Aversion, and Efficiency Criteria
4.1.1 Expected Utility Theory
The scientific and economic literature cited in the previous chapters provides this study
with a basis from which to extend the economic analysis of the MeBr phase-out. This chapter
describes the analytical methods used to answer the objectives noted in chapter one. This is
accomplished through the use of stochastic dominance analysis and a mathematical
programming model. These two methods are linked using an enterprise budget that represents
the financial operations of a southeastern vegetable grower.
Profit maximization is the goal of many agricultural producers. However, these
producers must successfully manage risk in order to maintain viable operations. Weather, yield,
pests, and price are types of uncertainty characteristic of the growers' decision-making
environment. To help producers minimize risk, agricultural economists integrate theoretical
tools such as expected utility, mean-variance, and stochastic dominance analyses into the actual
decision making process.
The expected utility hypothesis is the basis for much of the theory of decision making
under uncertainty, and provides a general decision rule for identifying agents' preferences (King
and Robinson, 1981). Risk can be defined as the variability in the outcomes associated with a
"state of nature" (Wetzstein, 2005). Each state of nature can be thought of as a lottery (N), and
42
defined as a set of probabilities, summing to 1, associated with all possible outcomes (Wetzstein,
2005). Similar to consumer theory, expected utility theory assumes that agents' preferences are
rational when choosing among states of nature. As such, the theorem relies on several axioms to
define an agent's behavior when faced with a choice that results in a probability distribution of
outcomes. When choices result in single-dimensioned consequences, the axioms of ordering and
transitivity, continuity, and independence are sufficient for deducing a decision maker's utility
function (Anderson et al., 1977).
The axiom of Ordering assumes that people prefer one of two risky choices a1 and a2 or is
indifferent between them (Anderson et al., 1977). Transitivity extends the concept of ordering
when the menu of available choices is greater than two. Transitivity implies "that an agent
preferring choice a1 to a2 (or is indifferent between them) and prefers a2 to a3 (or is indifferent
between them), she/he will prefer a1 to a3 (or be indifferent between them) (Anderson et al.,
1977, p.67)." The Continuity Axiom assumes an agent's ability to effectively determine the
subjective probabilities associated with outcomes involving risk. More formally, "if an agent
prefers a1 to a2 to a3, a subjective probability P(a1) exists other than zero or one such that she is
indifferent between a2 and a lottery yielding a1 with probability P(a1) and a3 with probability 1 –
P(a1) (Anderson et al., 1977, p.67)." This implies that an agent will risk receiving a bad
outcome, when given the choice between a good and a bad outcome, if the probability of
receiving this outcome is sufficiently low (Anderson et al., 1977). The Independence Axiom is a
fundamental assumption of many theories concerning choice under uncertainty (Wetzstein,
2005). The Independence Axiom states:
If N, N', and N'' are states of nature, and ( ρ ) is the probability of an outcome occurring,
43
then N f N' if and only if
ρ N + (1- ρ ) N'' f ρ N' + (1- ρ ) N'' (4.1) The state of nature N'' should not influence an agent's preference when choosing between state of
nature N and state of nature N' (Wetzstein, 2005). Finally, unlike commodities, states of nature
are mutually exclusive and can not be consumed jointly (Wetzstein, 2005).
In the mid-1940s, John von Neumann and Arthur Morgenstern developed the Expected
Utility Theory as a model to measure risk preferences. They expanded the more common utility
function into the expected or von Neumann-Morgenstern utility function. The expected utility
function is the weighted average of utility obtained from alternative states of nature, where the
weights are the probabilities of the states occurring (Wetzstein, 2005). In this model, each state
of nature is assigned its respective probability of occurring ( ρ i), and for all states of nature (Xi ),
the utility received in one state is added to the utility received in another state (Wetzstein, 2005).
For two possible states of nature the expected utility function can be represented as:
U(x1, x2, ρ 1, ρ 2) = ρ 1U(x1) + ρ 2U(x2) (4.2)
Agents make choices concerning the outcomes of each state of nature based on subjective and/or
objective probabilities. Thus, expected utility theory posits that a decision maker should act to
maximize her subjective expected utility if she is to be consistent with her expressed preferences
(Anderson et al., 1977).
4.1.2 Risk Aversion
Agents' preferences can be deduced according to his or her response to risk. Expected
utility employs the statistical concept of variance, a measure of the spread of a probability
distribution, to quantify the variability in the outcomes of alternative states of nature (Wetzstein,
2005). An increase in the variance of an outcome causes an agent's risk, associated with the
44
corresponding state of nature, to increase. The relationship between risk and variance is a
fundamental component of risk theory, and provides a basis for analyzing the concept of risk
aversion with respect to agents' preferences. Agents' preferences can be defined as risk neutral,
risk averse, or risk seeking. It is generally understood that most agents (i.e. firms, households, or
individual consumers) are risk averse. This idea, originally proposed by the mathematician
Daniel Bernoulli in the 1700s, posits that agents with risk-averse preferences will not play
actuarially fair games (Wetzstein, 2005). An actuarially fair game is defined as "a game with an
expected value of zero, or when the cost of playing the game is equal to its expected value
(Wetzstein, 2005, p.590)." Further, an experiment known as St. Petersburg's paradox reveals that
agents may not participate in games yielding a positive expected payoff (Wetzstein, 2005). The
paradox of the game occurs because the theoretical outcome results in both an expected payoff
and variance of infinity. However, no rational agent would be willing to pay an infinite amount
to play the game (Wetzstein, 2005). Moreover, because the variance of the game is infinite, the
paradox emphasizes that uncertain outcomes are worth less in utility terms than certain ones,
even when the expected payoffs may be large or equal.
With respect to an actuarially fair game, agents with risk neutral preferences are
indifferent between either playing or not playing the game, but those with risk seeking
preferences will play the game (Wetzstein, 2005). The measurement of these preferences
requires the derivation of an agent's utility function. These functions relate the outcomes of
choices to "single-valued indices of desirability" and in theory serve as exact representations of
preferences (King and Robinson, 1981, p.510). Although expected utility functions require
explicit information regarding agents' preferences, the inability to accurately measure these
45
preferences does not produce reliable evaluations of alternative choices (King and Robinson,
1981).
4.1.3 Efficiency Criteria
The shortcomings of expected utility theory motivated researchers to develop efficiency
criteria that more accurately determine the optimum choice among a set of risky alternatives.
The basis for this methodology are efficiency criterion, or decision rules, that provide a partial
ordering of choices for agents whose preferences conform to a specified set of conditions (King
and Robinson, 1981). When actual preferences are unknown, researchers determine the risk
efficient set of choices by making assumptions concerning an agent's preferences and/or various
approximations of the decision probability distributions (Wetzstein et al., 1988).
4.1.3.1 Expected Value Analysis
Wetzstein et al. (1988) and King and Robinson (1981) analyze popular efficiency criteria
cited in the literature. Among the most prominent are mean-variance (EV) and multiple versions
of stochastic dominance (SD) analysis. Expected value analysis is a necessary condition, for all
decision criteria under risk aversion, for one distribution to dominate another as it examines the
first moment of the decision density function (Wetzstein et al., 1988). The first through fourth
moments of a density function are defined as the mean, variance, skewness, and kurtosis,
respectively. The first moment measures the central value of the probability distribution. This is
accomplished by taking a weighted average of all the outcomes, where the weights are the
probabilities associated with each random value. Although the expected value, or mean, is the
most commonly used measure of the central value, both the median and mode of a distribution
may be employed when a distribution contains extreme values or researchers want to identify the
random value with the highest probability of occurrence. However, beyond the first moment,
46
necessary and sufficient conditions for identifying dominant alternatives depend on an efficiency
criterion's risk function (Wetzstein et al., 1998).
4.1.3.2 Mean Variance Analysis and Higher Moments
Mean-variance analysis is an efficiency criterion that ranks agents' choices according to
the first two moments of the probability distribution of the random variable (Bodie et al., 2005).
An assumption of mean-variance analysis is that agents are rational and act to maximize his or
her expected utility. In finance and economics, the random variable of interest is commonly the
payoff value of each outcome associated with a menu of risky choices. The probability
distribution is formed by creating a list for all the payoff values and their respective probabilities
of occurrence (Bodie et al., 2005). The expected value of the portfolio of all the uncertain
payoffs can be interpreted as the utility an agent receives as a function of the probability
distribution from his or her wealth:
E(r) =∑ (4.3) =
n
ssrs
1)()Pr(
Where s = 1,…..n represents all the possible scenarios, r(s) is the payoff from scenario s, and
Pr(s) is the probability of each payoff (Bodie et al., 2005). The risk of this portfolio is
determined by analyzing the distribution's higher moments. The variance, as the second central
moment, describes all the individual outcomes' deviations from the mean. The most frequently
used computation of the variance is the expected squared deviation from the mean:
= 22σ ∑=
−n
srEsrs
1)]()()[Pr( (4.4)
Mean-variance employs the first and second moment to determine an agent's utility
maximizing portfolio. More specifically, "for any two distributions G and H, G dominates H if
47
uG ≥ uH and holds with at least one strict inequality, where u and denote a
distribution's mean and variance, respectively (Wetzstein et al., 1988, p.171)." According to
(Samuelson, 1970), the first two moments are adequate to describe an agent's probability
distribution function if the distribution is "compact." Compactness assumes that an agent is able
to control most of the risk inherent to his or her portfolio. Thus, theoretically an agent can
construct a risk efficient set of alternative choices where no other set can provide a higher
expected value given the same level of risk for all sets.
22HG σσ ≤ 2σ
However, mean-variance may not be able to correctly describe utility functions of a
higher power than the quadratic or if the function violates the assumption of normality
(Wetzstein et al., 1988). Because the variance characterizes the risk of a portfolio on the basis of
the random variables' squared deviations from the mean, both good and bad deviations from the
mean are given the same weight (Bodie et al., 2005). If risk preferences are asymmetrical,
agents' do not perceive large rewards and losses to have the same effect on risk and this may lead
to an incorrect ranking of choices under uncertainty (Wetzstein et al., 2005).
Further, because mean-variance is limited to defining a distribution up to its second
moment, it may be unable to completely define distributions which are asymmetric, or skewed,
around the mean. Skewness is non-dimensional measurement describing the asymmetry of a
distribution. It strictly characterizes the shape of the distribution and can be mathematically
defined as:
M3= ∑ 3
=
−n
srEsrs
1)]()()[Pr( (4.5)
It is theoretically possible to have two distributions that have the same mean and variance, but
differing degrees of skewness. In this case, measurement of the third moment emphasizes
48
greater deviations about the mean, and thus gives more weight to the tails of the distributions
(Bodie et al., 2005). Risk-averse agents will act to minimize large negative deviations from the
mean. A positive skewness describes an asymmetric distribution characterized by frequent but
small losses and less frequent but extreme gains. Thus, a negatively skewed distribution can be
described as having a long tail extending outward toward the more negative values of the random
variable. Therefore, a risk-averse agent will choose a positively skewed distribution over a
negatively skewed distribution, assuming the distributions are mirror images around the mean, as
it is less likely to result in large unexpected losses (Bodie et al., 2005).
Similar to the variance, the fourth moment provides a measure of a distribution's extreme
values (Bodie et al., 2005). As with skewness, kurtosis is a pure, non-dimensional number that
describes the peakedness or flatness of a distribution. Measurements of kurtosis indicate the
likelihood of a distribution's extreme values, in addition to its shape, with larger values
indicating greater uncertainty (Bodie et al., 2005). Leptokurtic describes a distribution with
positive kurtosis, or peaked shape, and platykurtic describes a distribution that has a relatively
flat top when compared with the normal or Gaussian distribution (Bodie et al., 2005).
Mathematically the kurtosis is defined:
M4= 4 ∑=
−n
srEsrs
1)]()()[Pr( – 3 (4.6)
where the (− 3) provides for a kurtosis of (0) in the instance of a normal distribution. 4.1.3.3 Stochastic Dominance Analysis
Stochastic dominance analysis (SD) is an efficiency criteria used extensively by
agricultural economists to determine the risk efficient set of alternatives available to producers
when faced with uncertain outcomes. Different producers may experience similar production
49
problems across regions without regard to an operation's size or knowledge. Ranking
alternatives must therefore use a methodology that is not reliant on obtaining producers'
preferences of risk. Stochastic dominance allows for a ranking of alternatives based on a
minimal set of assumptions concerning producers' risk preferences. Researchers have developed
multiple variations of stochastic dominance but its two basic criterions are first-degree and
second-degree stochastic dominance.
4.1.3.3.1 First-Degree Stochastic Dominance
First-degree stochastic dominance (FSD) is the most general of all SD analyses. The
only economic axioms required for FSD are that agents maximize their utility of expected wealth
and are non-satiated in consumption, or agents' prefer more to less. The axiom of non-satiation
requires an agent's utility function to be monotonically increasing where the first derivative is
strictly positive, i.e. U1(y) > 0 (Anderson, 1977). A requirement of FSD is that population
probability distributions of the random variable be stated in terms of cumulative probability
distribution functions (CDFs) (Anderson, 1977). If (y) denotes a random variable with a
distribution function F(y), then the relationship between the distribution function F(y) and the
probability density function for the random variable (y), noted as F'(y) or f(y), can be expressed
as:
F(y) = (4.7) tdtfy∫ ∞−
)(
where t is used as the variable of integration (Wackerly et al., 2002). Comparing two continuous
CDF's F1 and G1 defined over an interval of [0,1], with each representing a set of risky
alternatives, " F is said to dominate G in the sense of first-degree stochastic dominance (FSD) if
F1(y) ≤ G1(y) for all possible (y) in the range [0,1] with at least one strong inequality (i.e., the <
holds for at least one value of (y) (Anderson, 1977, p.282)."
50
For example, if Y is a random variable representing an agent's given level of wealth, and
F1 and G1 represent the cumulative distribution functions of the rates of return on two mutually
exclusive assets, then for any Y, the probability that the rate of return on asset F is less than Y is
less than that for asset G (Huang and Litzenberger, 1988). Graphically, this scenario can be
observed when for a given level of wealth, the cumulative distribution of asset F lies to the right
of not only asset G but all other assets. This does not mean that the realized rate of return for
asset F is always greater than asset G (Huang and Litzenberger, 1988). It only infers that a
necessary condition for FSD is that the expected return of asset F must not be less than the
expected return of all other assets (Huang and Litzenberger, 1988). A necessary and sufficient
condition for FSD to hold is that the entire cumulative distribution of the dominating asset,
alternative, or state lie to the right of the dominated distributions.
The axiom of non-satiation proves insufficiently restrictive to produce a complete
ranking of risky alternatives in many cases of agricultural decision analysis. Although FSD is
transitive, producer preferences must be more strictly defined in order to reduce the number of
undominated alternatives that may result from this efficiency criterion (Huang and Litzenberger,
1988). FSD ranks alternatives by evaluating the first moment of the random variable's
probability distribution, but neglects the risk component when calculating the dominant states.
This may lead to conclusions where producers are not able to unambiguously state that he or she
prefers risky asset F to risky asset G (Huang and Litzenberger, 1988). In such situations the
cumulative distributions of the two assets may cross resulting in one asset having a higher mean
than the other, but also having a higher degree of risk associated with the payoff.
51
4.1.3.3.2 Second-Degree Stochastic Dominance
Second-degree stochastic dominance (SSD) eliminates dominated or inefficient
distributions from the FSD set (Anderson, 1977). This is accomplished by adding the
assumption of risk aversion to the decision making process with respect to agents' preferences.
Huang and Litzenberger note that," risk-averse agents may have utility functions that are not
monotonically increasing (Huang and Litzenberger, 1988, p.45)." However, they prove that
necessary and sufficient conditions required of second-degree stochastic monotonic dominance
are equivalent in theory to those of SSD. Therefore, the presumption that an agent's utility
function over the range [a, b] of possible payoffs is not only monotonically increasing but also
strictly concave can be understood as requirement of the measurements employed in this study
(Anderson, 1977).
Similar to first-degree stochastic dominance, SSD ranks alternative states by first
interpreting their respective probability density functions as CDFs. A distribution may be
considered dominant if it "lies more to the right in terms of differences in area between the CDF
curves cumulated from the lower values of the uncertain quantity (Anderson, 1977, p. 284)."
Huang and Litzenberger mathematically define the necessary and sufficient conditions for SSD
given two mutually exclusive distributions associated with assets A1 and B1. A1 is determined to
be dominant to B1 if and only if:
E [ r~ A] ≥ E [ r~ B] (4.8)
And
S(y) ≡ ∫ ≤−y
BA dzzFzF0
0))()(( ]1,0[∈∀y (4.9)
The inequality (4.8) describes the sufficient condition for SSD where the expectation associated
with the rate of return for asset A must be greater than or equal to the expected rate of return for
52
asset B. S(y), which describes the sum of the areas between the distributions, must be
continuous and negative. A necessary condition for A1 to be SSD over B1 is that S(y) must be
equivalent to the difference between the integrals of the two distributions A1 and B1 where (z)
represents a given state of wealth (Huang and Litzenberger, 1988). SSD works with a more
restrictive set of assumptions about risk preferences to determine the most efficient set of
alternatives, and this implies that any state which is FSD over another state is also guaranteed to
be SSD.
Risk-averse agents seeking to maximize utility will never prefer a dominated distribution
(Anderson, 1977). Therefore, a second-degree stochastically efficient set of alternatives will be
comprised of only non-dominated distributions, and any further reduction of this set will require
additional assumptions concerning risk preferences (Anderson, 1977).
4.1.3.3.3 Ranking Chemical Alternatives
This study used SSD in order to develop an efficient set of alternative chemical choices
for Georgia bell pepper producers faced with the MeBr phase-out. This procedure was chosen
instead of the mean-variance criterion (EV) for the following reasons: As previously noted, EV
may lead to an incorrect ranking if an agent's utility function is greater than a quadratic
functional form. Since SSD does not require a specific functional form of agents' utility
function, its use as a ranking mechanism for a broad section of producers is more justified. Next,
EV's assumption of a normal probability distribution of the random variable may not accurately
characterize agents' risk preferences. SSD overcomes this problem by making valid
generalizations concerning producers' behaviors. The additional assumptions concerning risk
preferences employed by SSD go beyond FSD to further reduce all available chemical choices
into a more efficient set.
53
4.2 Interpretation of Dominance Analysis, Enterprise Budget, and Programming Model
4.2.1 Dominance Analysis Our methodology began by using bell pepper price and yield data obtained from three
separate field trials during the spring and fall harvests of 2002 and 2003. The trials were
conducted at the Ponder Farm, University of Georgia's Experimental Plots Rural Development
Center located in Tifton, Georgia. The Center's agricultural economists, plant pathologists, and
horticulturalists determined the initial set of chemicals to be tested through research of the
scholarly literature: Thomson, 1991; Harris, 1991; Stall, 2000; Csinos et al., 1997 (Culpepper
and Langston, 2000). Each alternative fumigant, or combination of fumigant and herbicides, was
defined as a "production method" available to Georgia producers for the purposes of this study.
Before ranking each method, a database of total revenues corresponding to each method
was constructed using their respective wholesale prices and yield data for each harvest period.
Each trial generated peppers of varying quality ranging from the premium, or "Jumbo", pepper to
the lower grades of peppers classified as US 1&2's. Jumbo peppers are consistently preferred to
US 1&2 peppers and thus command higher prices. Jumbo peppers are categorized according to
size as extra-large, large, or medium with larger sizes commanding premium prices. An
additional factor affecting the market price of peppers and considered in the analysis was the
harvest period. Average prices for all grades of peppers, calculated using 2003 price data for
Georgia peppers under the assumption that this year represented a "normal" or representative
year for producers, were higher during the spring harvest than the fall harvest. The average total
revenues for each method were calculated by multiplying the yield of each grade of pepper by
their respective average prices during each harvest. The average prices served as weights to fully
characterize the distribution of peppers obtained from each method and harvest, and ensured a
54
continuity of prices across trials. The average total revenues of each method were then summed
across trials. The resulting figures served as the ranking criteria within the stochastic dominance
framework.
The initial yield data provided twenty-one different production methods that formed the
basis of this study. However, before ranking these methods assumptions were made in order to
increase the number of observations within each set, or method, and reduce the number sets to be
ranked. Those sets eliminated from the ranking procedure included any set of an individual
fumigant, or combination of fumigant and herbicides which was not tested across all three trial
periods. This resulted because these methods lacked a sufficient amount of observations to be
included in the ranking scheme. Additionally, methods that were composed of identical
fumigants or combinations of fumigant and herbicides across all trials, but differed in their
application methods were consolidated and labeled as an individual production method for the
purpose of the ranking procedure. Chemicals are usually applied in the field via broadcast or an
in-ground drip system. This study assumed that most producers will use the broadcast method
due to their familiarity with the system, its ease of use, and the cost savings accruing from its use
on multi-crop farms. After consolidation and elimination, the initial twenty-one methods were
reduced to eight methods.
The goal was to identify the two or three most efficient methods from this set of eight
methods through stochastic dominance analysis. The study employed the statistical software
package Simetar (Richardson et al., 2004) to perform the SSD analysis. The program
constructed the cumulative distribution functions using the average total revenues of each
method as the random variable. In this case, the average total revenues constituted observations
within a set, and could be interpreted as a measure of a given level of wealth corresponding to
55
each production method. The probabilities assigned to each observation were equally weighted
within each set in proportion to the observation's contribution to the whole set. For example,
given five observations within a set, each observation would be assigned a probability of (ρ=.20)
due to each observations equal likelihood of occurrence. After ranking the eight methods, the
three dominant methods that did not incorporate MeBr as the primary fumigant, as well as the
method incorporating MeBr, were chosen for further analysis in the enterprise budget. The set
employing MeBr was chosen to represent the base case from which to provide comparisons with
the dominant alternative methods.
4.2.2 Budget Analysis
Land grant universities develop enterprise budgets for a variety of agricultural
commodities. The budgets support academic research and provide agricultural producers with a
business decision tool tailored to their financial operations. This study developed four versions
of a pepper enterprise budget that represents production methods involving methyl bromide,
alternative fumigants, and herbicides. These versions use the format and base assumptions found
in the 2003 University of Georgia enterprise budget developed for Georgia's fresh bell pepper
and growers market (Fonsah and Rucker, 2003).
The enterprise budget attempts to control for incomplete knowledge concerning input and
output factors of production by systematically organizing data into a framework conducive to
more accurate decision making (Calkins and DiPietre, 1983; Fonsah and Chidebelu, 1995;
Fonsah and Rucker, 2003). For example, experimental data collected through test plots assign a
value to random variables such as yield per acre, while price forecasts attempt to accurately
predict future returns based on historical data. The budget can not reduce the randomness of
these variables, but it can assist producers by providing a more rigorous analysis concerning the
56
probability of future events. Obtaining the most accurate data available for all input factors is
imperative to implementing a successful budget. The enterprise budget thus provides the optimal
return per unit of enterprise using estimates of the combination of available inputs (Calkins and
DiPietre, 1983; Fonsah and Chidebelu, 1995; Fonsah and Rucker, 2003).
The enterprise budget represents a single point on the production function, and serves as
a piece of a whole farm budget that analyzes a producer's entire operations (Calkins and DiPietre,
1983). Using multiple budgets (i.e. enterprise, partial, and cash flow), producers' goals are to
anticipate all income and expenses of the organization in an effort to maximize net farm income
(Calkins and DiPietre, 1983; Fonsah and Chidebelu, 1995; Fonsah and Rucker, 2003).
Successful budgets allow producers to answer questions concerning farm operations such as the
optimal amounts of machinery and labor to employ in combination, whether to purchase or rent
additional capital inputs, and whether additions to or reductions in capital are needed in alternate
growing seasons (Calkins and DiPietre, 1983; Fonsah and Chidebelu, 1995; Fonsah and Rucker,
2003).
The enterprise budget used in this study is similar to those developed for other
agricultural crops. However, variables included in each component of an enterprise budget may
differ among producers depending upon whether production concerns a new investment, or
whether capital costs are distributed over a multi-crop or mono-crop system. The basic
components of an enterprise budget can be broken down into gross returns, total variable costs,
and total fixed costs per acre.
4.2.2.1 Gross Returns
The expected gross returns were calculated by multiplying estimates of the expected yield
times an expected price per unit (Calkins and DiPietre, 1983; Fonsah and Rucker, 2003). The
57
expected yields used in this study were calculated by summing the average yields for the Jumbo
and US 1&2 peppers under each production method and across trials. Moreover, test plots used
in field trials are not typically developed on an acreage basis. The calculations of a typical
enterprise budget, however, are based on the acre as the standard unit of measurement.
Therefore, original plot yields were extrapolated to conform to the acre unit using measurements
recorded by field scientists.
The expected prices used in this study were calculated by summing the average
wholesale prices obtained for Jumbo and US 1&2 peppers during the 2003 growing season.
Unlike yields, the expected price employed in each of the four final budgets analyzed remained
constant. Average prices and yields for bell peppers were recorded using a bushel-carton as the
standard unit of measurement. This unit is understood by producers to represent approximately
twenty-five to thirty pounds of peppers. This study assumed an average bushel-carton weight of
twenty-eight pounds for all calculations (Gast, 1991).
4.2.2.2 Total Variable Costs
The variable cost section is divided into two main parts: pre-harvest and
harvest/marketing costs. Total variable costs were calculated by summing these two
components. The fumigant and herbicide costs associated with each of the final four production
methods analyzed were addressed within the pre-harvest component of the variable cost section.
More specifically, four separate budgets were created that differed in terms of the yields and
costs associated with the fumigants and herbicides employed for each unique production method.
The marginal cost of fumigants and herbicides may vary according to the size of producers'
operations. However, this study assumes that a producer's cost schedule for fumigants and
herbicides applies to a forty acre representative farm for the purposes of the budget analysis.
58
This is because producers will not invest all of the necessary equipment if only one acre is
farmed. The prices for all chemicals used in this study were obtained from Hendrix and Dale
located in Tifton, Georgia. Chemical prices for 2003 were used in order to ensure continuity
with the wholesale prices received for 2003 peppers. Average chemical prices were calculated
according to a range that is inherent to the pricing structure of many chemical dealers. All prices
were based on typical application rates for Georgia pepper growers. Because growers purchase
chemicals in bulk quantities, those growers with smaller operations will experience a greater loss
of product through waste, thus resulting in a higher total cost, while large growers will benefit
from a input price discount. Additional costs within the pre-harvest section include plants,
fertilizer, insecticide, fungicide, plastic, drip tape, machinery, labor, land rent, irrigation, and
interest on operating capital. The harvesting costs component of the variable cost section
includes picking and hauling, grading and packing, containers, and marketing. The marketing
cost per unit was calculated as 8.50% of the expected price per carton. This value, when
multiplied by the expected yield per acre for each method, resulted in the total marketing cost per
acre.
4.2.2.3 Total Fixed Costs
The fixed cost section in this budget includes machinery, irrigation, land, and overhead
and management costs. Total overhead and management costs per acre were calculated by
multiplying the total pre-harvest variable cost by 15%. After establishing the total variable costs
and the total fixed costs, the total budgeted cost per acre was calculated by summing these
figures. Finally, the gross returns, variable costs, and fixed costs were used to calculate the gross
margins, and the base budgeted net revenue per unit of enterprise. Gross margins were calculated
by subtracting the total variable costs from the gross returns, and the base budgeted net revenue
59
was calculated by subtracting both the total variable and total fixed costs from the gross returns
(Calkins and DiPietre, 1983; Fonsah and Rucker, 2003).
4.2.3 Linking the Budget with the Linear Programming Model
The four risk-efficient production methods obtained from the stochastic dominance
analyses, which included both the MeBr method and the alternative methods, were analyzed
within the framework of their own enterprise budget. Calculations resulting from the budget
analyses provided variables to be transferred and used within a mathematical programming
model. These variables included the gross returns, total variable costs, total fixed costs, labor
costs, equipment costs, and net revenues. These figures and other financial parameters defined by
average farm financial conditions of a sample of crop farms participating under the Georgia
Farm Business Farm Management program served to construct the pre-operating balance sheet
representative of time (T0) within the context of the programming model as seen in (Table 4.1).
Forecasts of these variables, to be used in the programming model, were then developed over a
five year time horizon. Thus, the enterprise budget, representing the static position of producers'
financial operations, provided a platform from which the final analysis of the linear
programming model could be conducted.
4.2.4 The Linear Programming Model
4.2.4.1 Description of Optimization/Simulation Procedure
Mathematical programming models allow both researchers and producers to extend
analyses of agricultural operations beyond the enterprise budget. The goal of these models is to
construct a set of procedures that deals with the analysis of optimization problems (McCarl and
Spreen, 1994). More specifically, in this study the goal of the producer is to optimize his or her
expected utility of accumulated net worth over a specified planning period by solving for the
60
optimal values for a set of decision variables unique to their operations. The programming
model analytically investigates the decision maker's problem by developing a set of algebraic
expressions that attempt to fully characterize the relationships among the decision variables and
constraints inherent to the operation. This study used a linear programming (LP) model where,"
the decision variables were chosen such that a linear function of the decision variables was
optimized and a simultaneous set of linear constraints on the decision variables was satisfied
(McCarl and Spreen, 1994, p.21)."
The necessary components of a LP model are the decision variables, objective function,
and constraints. The resultant levels of a set of decision variables (xj), of which there are n (j=1,
2 …, n), quantify the amount employed of the respective unknowns (McCarl and Spreen, 1994).
The linear programming problem can be stated in matrix format as:
Max Z = CX
s.t. AX ≤ b
X ≥ 0
where Z is a matrix of the total objective value, C is a column vector representing the
contribution of each unit of X to the objective function, A is the use of the items in the ith
constraint by one unit of xj, and b is the upper limit imposed by each constraint (McCarl and
Spreen, 1994). The final inequality imposes a non-negativity constraint on the decision
variables.
The LP model assumes risk-neutrality on the part of the decision maker because the
model strictly accounts for the net changes in final wealth regardless of the agent's risk
considerations (Escalante, 2001). The model operates under a five-year planning horizon in its
determination of the producer's optimal net accumulation of wealth or net income over this
61
period. The final accumulation of net worth is calculated by accruing the values of farmland,
equipment, and cash balance at the end of the planning horizon less all financing charges
contracted over the same period (Escalante, 2001).
The final goal of optimizing producers’ net worth is predicated and works in conjunction
with a simulation procedure that begins with the analysis of a base-case representative farm
model. The base model accounts for the activity measures and financial conditions of a typical
producer’s operations. Adjustments to components of the base model reveal variations in the
profitability and cost structure of the farm’s operations. In this study modifications to the base
model are conducted by adjusting the production method as defined by the fumigants and
herbicides employed as chemical controls on the farm. More specifically, the production method
of MeBr and an accompanying menu of herbicides were chosen to serve as the base-case. The
three additional production methods as ranked by SSD were then tested against this base-case.
4.2.4.2 Characteristics of the Base Farm Model
In lieu of a singular analysis of the objective values resultant of all four production
methods as the sole determinant of efficiency, considerations were also given to measures of
liquidity and leverage when interpreting the viability of operations. The financial ratios used to
identify these measures are the tenure, current, debt-to-asset, and off-farm investment ratios.
The preferred production method will be characterized as the strategy that provides for the
largest accumulated net worth in proportion to the production’s total outstanding obligations.
The Base Farm Model as defined by Escalante (2001) is composed of sets of constraints
and requirements Table (4.2) that are in turn defined by the farm’s financial, marketing, and
production activities Table (4.3). Further, requirements and activities of a representative farm
must account for variability in cash flows over the planning period. Thus, the model assumes
62
two sub-periods for all cash transfers and defines these activities through the two cash transfer
equations (TRANSF1) and (TRANSF2).
4.2.4.3 Activities, Constraints, and Requirements
i) Initial Endowments
This model assumes the production requirements, constraints, and levels characteristic of
a Georgia pepper farm. Producers are subject to a set of resource constraints that define the
endowments of the operation at time T (0), one year prior to the five-year planning period.
These initial constraints are land (LAND0), equipment (EQPT0), existing long-term loans
(FINLAND), existing intermediate-term loans (MEDCRED), ending cash balance (CASH0), and
current asset position (CURAST0) (Escalante, 2001). The cash balances and other current assets
(OTHCUR0) at the end of year T (0) are transferred to the first sub-period of the first year to
insure coverage of operating expenses and investment requirements (Escalante, 2001).
Downpayment rates for both land (LNDDPMT) and equipment (EQDPMT) are assumed to be
20% and serve as scalars throughout the planning period.
ii) Production Constraints and Requirements
Total farm size in terms of production and marketing (ACPROD) throughout the
planning period can be identified as the sum of the number of acres owned at time (t) (ACBUY)
and the amount of acres that are cash rented (ACRENT) during the same period. This model
assumes an initial farm size of 362 acres to be divided between these two activities at time T (0)
as 300 acres purchased and 62 acres rented. Farm real estate values per acre (LANDVAL (t))
were obtained through the USDA, projected over the five-year horizon, and based on an initial
value of $2,150.00 per acre (USDA NASS, 2004). Figures were also obtained for the initial
value of a cash rented acre (RENT (t)), $105.00, and projected over the planning period. Both
63
the farm real estate values and cash rent figures were forecast using a growth rate of 5.6% under
the assumption that values would approximate the previous ten-year trajectory. Adjustments
were then allowed for these activities in each following period.
Other mechanisms to increase farm size other than purchasing or renting were not
included in this model. Further, a property tax (PROPTAX) of approximately $24.00 per acre
per year is assessed against all owned land throughout the planning period, and payments are
made in two installments over each year (t). All expenses accruing to operations other than cash
rents are entered into the model with positive coefficients with the assumption that payments are
made during the first sub-period (Escalante, 2001). Gross-crop revenues are realized during the
second sub-period, and identified as an inflow taking a negative coefficient for the purposes of
the model.
iii) Machinery and Labor Requirements
The model recognizes the machinery requirement imposed upon production activities
both through the owned equipment component (EQBUY) and additional purchases of new
machinery over the planning period. Adjustments to equipment values over the planning period
are made by acknowledging the cumulative effects of both depreciation and inflation (Escalante,
2001). Inflationary effects within the model were imposed at an approximate historical average
of 3%. Whereas machinery requirements are expressed in units of dollars per acre, labor
requirements are expressed in hours per acre. Total labor requirements are a combination of
direct and annual overhead labor (OVRLAB) requirements that can be satisfied through hired
labor (hirelab (t)). Hired labor is assumed to be the only form of labor available to producers in
this study, and the model disburses payments to labor over two sub-periods.
64
iv) Credit, Cash, and Financing Components
Credit is available to the producer for three main types of financing requirements. Short-
term loans (SHTCRED) are available for the immediate financing of operations. The financing
terms for these loans are for six months with the principle and interest payments (1 + i) accruing
over the first period and coming due on the second sub-period of the year. Intermediate-term
loans (MEDCRED) are typically used to finance equipment purchases and are assumed to carry a
fixed interest rate over the five-year planning period. Long-term loans (FINLAND) carrying
fixed interest rates over a 20 year period are used to finance additional purchases of land. Both
the (FINLAND) and (MEDCRED) components are amortized annually (P & I) with payments
due at the second sub-period of the year (Escalante, 2001). Finally, the model assumes a 7%
interest rate for the duration of all outstanding loans.
The model further allows for purchases of equipment and land through cash components
(casheq (t)) and (cashland (t)), respectively. However, financing involving cash is subject to a
minimum down payment requirement and upper limits on distributions according to the
producer’s cash position (Escalante, 2001). Thus, solutions of the model may allow for
financing of up to 90% of purchases, or self-financing according to the producer’s current cash
position.
v) Consumption and Non-farm Investments
The expenditures (famcon (t)) and off-farm investments (offinv (t)) of the representative
farm family modeled in this study are based on the historical levels of family living expenses and
the historical yields of United States Treasury bills. Family consumption is accounted for in
each sub-period of each year, constrained by a fixed upper bound in each year, and assumed to
increase over each period (t) (Escalante, 2001).
65
The total off-farm income generated from off-farm investments (i.e. government
securities) is not a significant activity measure in this model. Greater emphasis is given to on-
farm revenue generating activities (Escalante, 2001). However, capital gains (1 + i) are
recognized by the cash transfer equations as inflows in the second sub-period resulting from the
outflow of investment made during the first sub-period.
vi) Taxable Net Income
The producer’s gross taxable income is calculated by summing net operating margins and
off-farm income generated through investments. The taxable net income (txincm (t)) considers
allowable deductions for such items as the interest payments on intermediate and long-term
loans, property taxes, cash rents, depreciation of business equipment, and wages (Escalante,
2001). A tax rate of 40% accounts for all applicable local, state, and federal taxes imposed on
the farm enterprise. The value of the taxable net income for time T (0) was $20,171.
66
TABLE 4.1
Pre-Operating Balance Sheet
Current Assets Current Liabilities $241,688 Cash $124,255 Non-Farm Investments $100,000 Other Current Assets $379,861 Intermediate Liabilities $422,010 Total Current Assets $604,116
Long-Term Liabilities
$350,099Intermediate Assets Machinery & Equip. $468,900 Total Liabilities $1,013,797
Fixed Assets Land $645,000
Buildings $125,218$770,218 Net Worth $829,437
Total Assets $1,843,234 Total $1,843,234
67
TABLE 4.2
Description of Constraints and Requirements
Sign of
Constraint Unit Right
Hand Side Value
INITIAL RESOURCE AND DEBT LEVELS
=
Acres
B
Initial Land Owned = Dollars B Initial Equipment Owned = Dollars B Initial Long-Term (Real-Estate) Loan = Dollars B Initial Intermediate-Term (Equipment) Loan = Dollars B Initial Cash Available = Dollars B Initial Amount of Current Assets = Dollars B LAND REQUIREMENT Number of acres produced and marketed ≤ Acres 0 Long-term financing for real estate purchases = Dollars 0 Land downpayment requirement in year t ≥ Dollars 0 EQUIPMENT REQUIREMENT Machinery requirement to prod. & mkt. crops in year t
≤ Dollars 0
Intermediate financing for equipment purchases in year t
= Dollars 0
Machinery downpayment requirement in year t ≥ Dollars 0 LABOR REQUIREMENT ≥ Hours 0 FAMILY LIVING EXPENSES = Dollars B OFF-FARM INCOME = Dollars B CASH FLOW PER YEAR (TRANSFERS) = Dollars 0 TAXABLE INCOME = Dollars 0
68
TABLE 4.3
Description of Portfolio Activities
Code Unit Description
ACPROD (t) Acres Number of acres of peppers produced and marketed in year t ACBUY (t) Acres Number of acres purchased in year t ACRENT (t) Acres Number of acres rented in year t FINLAND (t) Dollars Long-term loan at fixed rate over 20 years used in year t EQBUY (t) Dollars Amount of equipment purchased in year t MEDCRED (t) Dollars Intermediate-term loan at fixed rate used in year t SHTCRED (t) Dollars Short-term (operating) loan used in year t cashland (t) Dollars Cash used to purchase land in year t casheq (t) Dollars Cash used to purchase equipment endcash (t) Dollars Ending cash balance at period p of year t famcon (t) Dollars Family consumption in year t offarm (t) Dollars Off-farm income in year t txincm (t) Dollars Taxable net income at year t offinv (t) Dollars Investments in treasury bills (off-farm investments) at year t hirelab (t) Hours Amount of hired labor in year t
69
CHAPTER 5
ANALYSIS OF RESULTS
5.1 Stochastic Efficiency-The Final Four Production Methods
As discussed in Chapter four, the initial set of twenty-one production methods was
consolidated into a more efficient set of eight production methods Table (5.1). Two of these
eight methods consisted of 1) MeBr as a stand alone fumigant and 2) MeBr used in combination
with a menu of herbicides. The remaining six methods were composed of three classes of
fumigants without the use of herbicides, corresponding to three mutually exclusive production
methods, and the same three classes in combination with an identical menu of herbicides as
found in the base-case using MeBr. Therefore, the four classes of fumigants identified in the set
of eight production methods are as follows:
i) Methyl Bromide
ii) KPAM in combination with C35
iii) Telone II in combination with Chloropicrin
iv) C35 in combination with Chloropicrin
Employing SSD as an efficiency criteria resulted in the ranking scheme identified in Table
(5.1). The resultant ranking scheme is consistent with the expectation that those fumigants
employed in combination with a menu of herbicides will lead to greater yields as a result of their
ability to more efficiently control for a host of weeds and pests. None of the four production
methods not employing herbicides ranked as one of the final four efficient methods.
70
An expected result of the analysis was that the method employing MeBr with herbicides
would rank as the most efficient set out of the final four sets. However, the final four efficient
methods resulted in the ranking scheme listed below as identified by their coefficients of
variation (C.V.). The following listed numbers in bold italics correspond to the methods’
identification numbers in Tables (5.1) and (5.2), and hereafter will uniquely identify the methods
within the text.
3. C35 + KPAM (.058)
1. Methyl Bromide (.268)
5. Telone II + Chloropicrin (.308)
7. C35 + Chloropicrin (.323)
The coefficient of variation is a traditional economic measurement that allows for comparing an
investments standard deviation to its expected return. In this case, the expected return or mean is
the average total revenue per test plot per production method. The (C35+KPAM) method ranked
as the most dominant method with its average revenues exceeding those of MeBr by $1.77,
(Telone II and PIC) by $3.17, and (C35 + PIC) by $3.25.
The cumulative distribution chart shown in Table (5.2) gives a graphical representation of
SSD. As can be seen, methods (3, 1, 5, & 7) lie significantly to the right of methods (2, 4, 6, &
8) in terms of the total areas under the cumulative functions. In this graph, the (X) axis
represents the average total revenues per plot and the (Y) axis represents the probability of
receiving this amount of wealth or less. It is obvious that the conditions for FSD do not hold due
to the crossing of the cumulative functions. However, for method (3) one can observe that for
most levels of wealth (x), the probability of obtaining that amount or less is greater than almost
71
all other methods. After performing the SSD calculations, methods (3, 1, 5, & 7) were analyzed
within an enterprise budget.
5.2 Results of Enterprise Budget Analyses
The results of the enterprise budget allow for a comparison of the costs, revenues, and
financial efficiency measures among the final four production methods. Similar to the
programming model, analyses of the budgets considered the production operation employing
MeBr as the primary fumigant as the base-case model from which the other three methods were
compared. Tables (5.3) and (5.4), respectively, identify the activity measures, efficiency ratios,
and breakeven statistics when the budgets were analyzed using experimental yield results and
when expected yields were held constant. The purpose of separate analyses was to isolate key
activity measures within the production process and assess their individual contribution to the
entire budget.
Activity measures analyzed for each budget were the representative farm’s yield, gross
returns, total variable costs, total fixed costs, and net returns. Total variable costs were broken
down into two components, fumigant costs and other variable costs, in order to examine each
fumigant’s contribution to the farm’s economic viability. These contributions were appraised
through three farm efficiency ratios; 1) Net returns / Gross Returns, 2) Variable Cost / Gross
Returns, and 3) Fumigant Cost / Gross Returns.
Finally, break-even statistics were calculated for each method. Break-even analysis is
traditionally a type of simulation or “what if” analysis with the objective of determining what
volume of production, value of sales, or value of production inputs will result in the firm
experiencing zero net revenue or “breaking-even.” This study calculated the farm’s break-even
revenues, yields, and prices according to each production method under both the experimental
72
and constant yield scenarios. All activity measures, efficiency ratios, and break-even statistics
were calculated on a per acre basis.
5.2.1 Examination of Costs
5.2.1.1 Other Variable Costs
The final four production methods analyzed in the budget were subject to identical
variable cost measures. Most pre-harvest variable cost values other than fumigants remained
through each budget. Variations in the pre-harvest variable costs were primarily attributed to the
differences in the interest charged on operating capital. Appendices (A-D) give a visual picture
of the basic components required for the representative pepper farm. Each acre required 174
pepper plants, costing $5.50 a plant, or $957 an acre. Lime was applied at a rate of one ton per
acre with an associated cost of $26/acre. The combination of base and sidedress fertilizers cost a
total of approximately $214/acre. This study assumed that the prices attributed to insecticides,
fungicides, nematicides, and herbicides remained constant for all four methods and cost the
grower an approximate total value of $996/acre. However, producers may in actuality
experience discounts in these areas due to business negotiations, the ability to store extra
product, or bulk price reductions.
Low-density polyethylene plastic (LDPE), which is used to control for weeds and pests,
is charged to growers by the roll, and when combined with the cost of its removal cost nearly
$267/acre. Operations required 8700 feet of drip tape per acre at a cost of $.02 / foot or
$174/acre. Machinery costs are calculated at five hours per acre, $21/acre, for a total of
$105/acre. The labor component of the budget is broken into two parts; transplant labor and
non-transplant labor. For this study the cost of both were calculated at $8.00/hour in accordance
with prevailing regional prices. A sum total of 53 man-hours were required per acre resulting in
73
a total cost of $424/acre. Land-rents were expected to cost the grower $105 per acre while
irrigation, consisting of machinery and labor costs, held a total cost per acre of $65. Finally,
differences in the interest owed on producers’ operating capital were significant with MeBr
paying ($168), C35 + KPAM ($185), C35 + PIC ($213.48), and Telone II + PIC paying ($174).
Thus, pre-harvest variable costs for all methods ranged from a low of $3,905 for MeBr to a high
of $4,534 for the C35+PIC method.
Next, the total harvest and marketing costs vary according to yield as greater yields
require more picking and hauling (P & H), grading and packing, containers, and marketing. The
cost per unit for all components remained the same for all methods. Measured in cartons, the
price of (P &H), $1.25, accounts for approximately 20% of the $6.29 total harvesting and
marketing costs. Grading and packing, at a price of $2.75/ctn. represents 44% of the total,
containers ($1.30/ctn.) equal 21%, and marketing at $0.99/ctn makes up 16%.
5.2.1.2 Fixed Costs
The fixed costs for the producer consist of machinery, irrigation, land, and overhead and
management (O & M). Overhead and management, costing $645/acre, is the largest fixed cost
for producers and represents approximately 70% of all total fixed costs ($919/acre). The second
largest cost component is irrigation at $221 per acre, or 24% of the total. Fixed machinery costs
of $54/acre represent 6% of total fixed costs. The total budgeted cost per acre is calculated by
summing both the total variable costs and the total fixed costs per acre. The range of this cost
over all methods was a low of $12,107 for MeBr to a high of $13,388 for C35+KPAM.
5.2.1.3 Costs per Carton
The enterprise budget then calculates for several of the previously listed costs on a per
carton basis (c.p.c.). Table (5.5) illustrates the differences in costs per carton among the four
74
methods. Differences in the pre-harvest variable costs, due to the discrepancies in the prices of
the fumigants, and the fixed costs, caused by the fluctuating charges associated with overhead
and management, produce a range for the total budgeted cost per carton of $10.23 for C35 +
KPAM to $11.05 for C35 + PIC.
5.2.2 Net Revenues
In this study, calculations of gross and net revenues were dependent on a constant
expected price of $11.68 received by producers. Gross revenues were calculated by multiplying
expected yield times expected price. Base budgeted net revenues for each method were
calculated after deducting all variable and fixed costs from the gross figures. Gross and net
revenues differed for both the experimental and constant yield scenarios due variations in the
overall cost structures of each method.
5.2.3 Experimental Yield Results
Experimental yields were obtained through field trials conducted in Tifton, Georgia.
Yields indicated in Table (5.3) indicate the number of cartons per acre. C35 + KPAM produced
the largest yields with 1,298 cartons per acre. Methyl bromide and C35 + PIC varied slightly in
yield at 1,167 ctn./acre and 1,115 ctn./acre respectively. Telone II + PIC produced the smallest
yield of 1,119 ctn./acre. Although the C35 + KPAM method incurred total variable costs that
were approximately $1,200 more than MeBr, its greater yield resulted in net returns that were
$248 more than MeBr and $1,036 greater than the C35 + PIC combination, which experienced
the lowest net returns at $732 per acre.
5.2.3.1 Efficiency Ratios-Experimental Yields
The efficiency ratios indicated in Table (5.3) measure a method's ability to generate net
revenue, or profits, from its gross returns, as well as inefficient cost allocations among the four
75
methods. For example, MeBr was able to generate a net return of $0.13 for every $1.00 of gross
returns while C35 + PIC only generated $0.02 per $1.00 from approximately the same base of
gross returns. This results predominantly from the discrepancies in fumigant costs between the
two methods. MeBr incurs a fumigant cost of $405/acre while C35 + PIC cost 2.5 times more
per acre at $1,007. Although C35 +KPAM only earned $0.12 profit for every $1.00 of gross
revenues, and experienced total variable costs nearly 2 times that of MeBr at $786/acre, its
ability to generate 131 more cartons of peppers compensated this method for its higher cost
structure. Telone II + PIC generated $0.09 per $1.00 of gross revenues but its variable costs as a
percentage of gross revenues exceeded MeBr and C35 + KPAM by 2% and 3% respectively.
5.2.3.2 Break-even Statistics-Experimental Yields
Break-even statistics for these four methods give an indication of the units of production
that are needed at a given price in order to maintain economic viability. Consequently, the
reverse question also allows producers to calculate the price that is necessary, for a given level of
production, in order to ensure the farm's profitability. The goal of break-even analyses is to
determine what value of a given variable (x), holding other variables constant, assures a net
return to the firm of zero. From this value, producers can determine what values of (x) must be
attained in order to maintain profitability. Break-even revenues can be found by calculating the
total budgeted cost per acre. Revenues in excess of this figure will lead the firm into profitability
while revenues below this figure will eventually cause the firm to fold. In this study, break-even
yields were calculated holding the expected price of $11.68 per carton constant. In order for
producers to maintain viable operations (MeBr) must yield a minimum of 1,037 ctns. /acre, (C35
+ KPAM) 1,146 ctns. /acre, (C35 + PIC) 1,092 ctns. /acre, and (TII + PIC) must yield 1,024 ctns.
/acre. Alternatively, holding each method's unique experimental yield constant, (MeBr) must
76
receive a minimum of $10.38 per carton, (C35 + KPAM) $10.32, (C35 + PIC) $11.05, and (TII +
PIC) must receive $10.68 per carton. Due to its comparatively high cost of production, (C35 +
KPAM) must generate revenues of $13,388 to maintain profitability. This figure is $1,281
greater than (MeBr), $633 greater than C35 + PIC, and $1,433 more than TII + PIC.
5.2.4 Constant Yield Results
The study assumed constant yields of 1200 cartons per acre as a result of consultations
with field scientists and producers in Georgia concerning realistic assumptions involving the
expected yields generated by MeBr. Because MeBr was taken as the base-case for this study,
holding these expected yields constant allows producers to completely isolate costs within the
budget and determine the effects that these costs have on net revenues (Table 5.4). As with the
analysis on experimental yields, pepper prices were again held constant at $11.68 / ctn. for these
calculations. A result of holding both the price and yield constant was that the gross revenues for
each method were identical in this analysis. However, in this experiment MeBr replaced C35 +
PIC as the production method generating the largest net returns. More specifically, MeBr
generated net returns of $1,700/acre, (C35 + KPAM) produced $1,242/acre, (C35 + PIC)
produced $976/acre and (TII + PIC) generated $1,552 per acre. The total fixed costs remained
the same for each method due to the constant yields.
5.2.4.1 Efficiency Ratios-Constant Yields
Under a constant yield scheme, net returns as a percent of gross returns increased 5% for
(C35 + PIC), 2% for (TII + PIC), and 3% for (C35 + KPAM), but decreased 1% for (MeBr).
Fumigant costs as a percent of gross returns increased 1% for (C35 + KPAM), decreased 1% for
(C35 + PIC) and remained unchanged for (TII + PIC) against the figures from the experimental
analysis.
77
5.2.4.2 Break-even Statistics-Constant Yields
The break-even revenues, as expected, did not change dramatically from the experimental
case. All the variation among these figures was due to the interest charged on operating
expenses contained in the variable cost component of the budget. As expected, break-even
yields for all methods except (C35 + KPAM) increased compared to the experimental yields.
Under the new scenario, (MeBr) needed to yield 1,054 ctns. /acre, (C35 + PIC) 1,116 ctns. /acre,
and (TII + PIC) needed 1,067 ctns. /acre. Assuming yields were held at 1200 ctns./acre, (MeBr)
had to receive $10.26 per carton, (C35 +KPAM) needed $10.64 per carton, (C35 + PIC) required
$10.87 per carton, and (TII + PIC) needed to receive $10.39 per carton to maintain profitability
(Table 5.4). The break-even figures for price reveal that if production were able to generate at
least 1,200 cartons of peppers per acre, growers should have little problem maintaining viability
as the highest price required for profitability out of all four methods was $10.87, which was
$0.81 below the average price received by growers across spring and fall harvests of year 2003.
5.3 Programming Solutions
The LP model delivers solutions to the optimization problem for each period throughout
the five year time horizon. Additions and/or reductions to both the farm’s assets and liabilities
dictate adjustments to the final value of accumulated net worth. These adjustments can be made
through increases or decreases in owned land, cash-rented acreage, equipment, investments, debt,
or any combination of these various decision variables bound within the constraints of the model.
Each simulation of the four production methods began with an identical set of
assumptions concerning certain attributes of the representative farm. For example, beginning
land values and cash rent levels, equipment costs, family consumption, off-farm income and
yields on off-farm investments, labor costs, depreciation schedules, and interest on credit
78
facilities were constant values at time T(0) for all methods. Variables such as the projected
variable costs per acre, gross returns per acre, overhead costs per acre, and net margins per acre
were forecast over the planning period based on calculations made in each method’s enterprise
budget. The enterprise budget analyses employed for each method indicated the economic
viability of operations under all four scenarios. Therefore, the expectation for the solutions to
the programming model was that the farms would increase production over the planning period
until maximum production could be achieved according to the constraints.
5.3.1 Solutions-Activity Measures
5.3.1.1 Production Solutions
Conveying the LP solutions for the production variables as five-year averages, Table
(5.6) indicates that the two dominant methods (MeBr) and (C35 + KPAM) yield similar solutions
to the production problem over the planning period. The LP set an upper limit constraint on total
production of 1,000 acres per method. Although both (MeBr) and (C35 + KPAM) produced the
limit of 1,000 acres each year, (MeBr) purchased seven acres more and rented seven acres less
than (C35 + KPAM) in year T(1).
Note that the solutions for activity measure (ACBUY) indicated in Tables (5.7, 5.8, 5.9,
5.10) are incremental measures that express the additional acres purchased in excess of the initial
endowment of 300 acres. The values for (ACRENT) do not accumulate during the planning
period. Producers rent land for a period of one year, and at the beginning of the next period must
again decide how much acreage to devote to renting. This figure is then carried over and added
to purchases made during the next period, with this process being repeated over the life of the
planning period. Thus, the total owned acres for a given period is the prescribed value for
(ACPROD) less (ACRENT).
79
During year T (2) the model predicted that (MeBr) purchased an additional 93 acres, nine
less than that purchased by (C35 + KPAM). However, (MeBr) rented two more acres than the
latter over the same period. For the remaining three years T (3) through T (5), the amount of
land purchased and rented by each method remained nearly identical. Finally, it is expected that
relaxing the constraint on farm size would result in increased production under (MeBr) and (C35
+ KPAM). However, the results of the model suggest growers would plant more under the (C35
+ KPAM) method when compared to (MeBr) due to a more favorable financial position over the
planning period.
The model solutions for (C35 + PIC) and (T2 + PIC) suggested increases in production
through cash-renting although neither reached the 1,000 acre limit. Over the planning period
(C35 + PIC) cash-rented an average of 586 acres per year and (T2 + PIC) rented an average of
587 acres per year. The farm size solutions for these methods were nearly identical. Thus, the
model did not suggest any additional purchases of land for either method over the planning
period.
Of the four financial ratios provided in Table (5.6), the tenure ratio illustrates the
relationship between the total land owned and total land produced. Tenure ratios of (0.4530) and
(0.4350) for (MeBr) and (C35 + KPAM) indicate that the returns and cash flow available to
producers using these methods encourage land ownership. (C35 + PIC) and (T2 + PIC) realize
tenure ratios of (0.3386) and (0.3382) suggesting that the higher costs associated with ownership,
acquisition costs and taxes, force these methods to rely more heavily on cash-renting.
Further, the model did not prescribe any additional allocations to equipment for methods
(C35 + KPAM) and (MeBr) over five years. For (C35 + PIC), solutions indicated that the
80
producer averaged additional purchases of equipment of $22,416 over five years, and (T2 + PIC)
averaged $32,276 over the same period.
5.3.1.2 Financial Solutions
Discrepancies among solutions for the financial decision variables were pronounced
between the methods (MeBr)/(C35 + KPAM) and (C35 + PIC)/(T2 + PIC). Each of the four
farms began with an initial allocation of off-farm investments totaling $100,000. The former two
methods realized off-farm investments of approximately $1.5 million per year over five years
Table (5.6). Methods (C35 + PIC) and (T2 + PIC) were prescribed equal solutions of $393,840
invested in off-farm activities per year when averaged over the planning period.
All farms were prescribed total current assets, assets that can be easily converted into
cash in less than one year, at time T (0) of $604,116 from a total asset pre-operating base of
$1.85 million. The model revealed that growers using methods (MeBr) and (C35 + KPAM) will
realize cash, non-farm investments, and other current assets positions over the five-year period
up to an approximate average of $45.5 million for (MeBr) and $48.5 million for (C35 + KPAM).
The solutions for the farms’ total assets, or amount of current and fixed assets, for the same two
methods increased to approximately $47 million and $50 million, respectively, as averaged over
the planning period.
However, the current assets position increased less on average for (C35 + PIC) and (T2 +
PIC) to approximately $9.9 million and $16.3 million over five years. Increases in total assets to
$18 million and $17.6 million, respectively, were generated by the model for these two methods.
During the planning horizon decisions are made by producers to either increase investment in
machinery and land, or to maintain his or her current capital position. For (C35 + PIC), heavy
investments were made during years one and five in machinery and equipment. Alternatively,
81
although the five-year averages for the total assets are similar, producers using method (T2 +
PIC) distributed investments in fixed assets more evenly over the planning horizon.
The current liabilities facing producers represent all short-term loans, interest, unpaid
expenses, accounts payable, and other debts coming due within one year. Current liabilities
under each method at time T (0) were $241,688. Note that the figures for (CULIAB) indicated
in the tables are marginal values representing the short-term liabilities of the operation. These
liabilities are not carried over from one period to the next, but are incurred at the beginning of
the period and repaid at the end of each year. However, intermediate-term (MEDCRED) and
long-term (FINLAND) liabilities accounts follow an amortization schedule, and their respective
figures at time T (5) indicate the final balance on these loan accounts. The average five-year
balance on short-term loans decreased to $146,386 for (MeBr) and (C35 + KPAM), while
decreasing to $156,714 for (C35 + PIC) and $162,110 for (T2 + PIC). Total liabilities also
decreased for all methods from a pre-operating base of approximately $1 million. (MeBr) and
(C35 + KPAM) averaged total liabilities of $420,840 over five years. (C35 + PIC) averaged
$453,946, and (T2 + PIC) averaged $469,268 during each period.
The current ratio is a measure of a firm’s liquidity, or an indication of a firm’s ability to
cover its short-term obligations. It is calculated by dividing the firm’s current assets by its
current liabilities. The measure provides a method for comparison among the alternative
production methods. (C35 + KPAM) with a current ratio of (331.59) had the highest liquidity.
(MeBr) was also highly liquid with a ratio of (310.72). (C35 + PIC) measured (221.17), and (T2
+ PIC) registered the lowest liquidity of all methods at (100.60). The ratios for each method
were well above the critical value of two times liabilities and indicated that each had the ability
to quickly pay off short-term debts.
82
Measurements of each method’s total debt to total assets (debt-to-asset ratio) provide a
value from which to compare firms’ financial risk by indicating how much of a firm’s assets
have been financed by debt (Bodie et al., 2005). Debt-to-asset ratios exceeding 65% indicate
excessive debt. Due to the current assets position as prescribed by the model, no method
incurred a significant amount of debt in proportion to its overall financial portfolio. Debt-to-
asset ratios ranged from less than one percent (.008) for (C35 + KPAM) to a high of
approximately (.027) for (T2 + PIC). Finally, the off-farm investment ratio reveals that more
money is able to be allocated outside of the primary revenue generating activities of the farm for
(MeBr) and (C35 + KPAM).
Finally, the accumulated net worth of each simulated farm increased over the planning
horizon. The pre-operating net worth attributed to all methods was $829,437. The net worth
solution prescribed by the model for (C35 + KPAM) ranked first among all methods and
indicated that a producer employing this method would realize an average of $46.6 million over
the planning period. The net worth solution for (MeBr) determined in the first year of simulation
was approximately $11,688,000, and after five years increased to a maximum value of $80.8
million in T (5), indicating an average five-year value of $46.7 million. The third highest five-
year average net worth was prescribed to (C35 + PIC) at $17.5 million. The model designated
the net worth of (T2 + PIC) to be $17.1 million over the planning period. The dominant net
worth positions prescribed by the model for (MeBr) and (C35 + KPAM) can be attributed to the
gains resulting from greater production. A final comparison of the solutions specified by the
model establishes (C35 + KPAM) as not only a viable alternative to MeBr in terms of liquidity,
leverage, and tenure, but also the preferred method in terms of the accumulated value of net
worth.
83
TABLE 5.1
Efficiency Ranking of Final 8 Production Methods Using Second Degree Stochastic Dominance
Production Methods*
(1) (2) (3) (4) (5) (6) (7) (8)
Methyl
Bromide with
Herbicides
Methyl
Bromide without
Herbicides
C35 + KPAM
with Herbicides
C35 + KPAM without
Herbicides
Telone II +
Chloropicrin with
Herbicide
Telone II +
Chloropicrin without
Herbicide
C35 +
Chloropicrin with
Herbicide
C35 +
Chloropicrin without
Herbicide Mean (u)
14.84
12.96
16.61
10.65
13.44
9.98
13.36
9.30
Standard Deviation (S.D.)
3.97
4.70
0.96
5.53
4.14
5.81
4.31
4.79
Coefficient of Variation (C.V.)
0.268
0.363
0.058
0.519
0.308
0.432
0.323
0.515
Ranking**
( 2 )
( 5 )
( 1 )
( 8 )
( 3 )
( 6 )
( 4 )
( 7 )
* Production methods ranked according to the total revenues generated per plot across all trials ** Increasing rank number corresponds to decreasing level of efficiency
84
TABLE 5.2Cumulative Distribution Functions for Final 8 Production Methods
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
0 2 4 6 8 10 12 14 16 18 20
Cum
ulat
ive
Prob
abili
ty
1 2 3 4 5 6 7 8
PRODUCTION METHODS
85
TABLE 5.3 Financial Efficiency and Break-even Analyses Using Experimental Yield Results
Production Methods Methyl
BromideC35 &KPAM
C35 &
Chloropicrin Telone II &
ChloropicrinActivity Measures Yield (cartons) 1,167
1,298
1,155 1,119
Gross Return $13,627 $15,156
$13,487 $13,070
Fumigant Cost $405 $786
$1,007 $528
Other Variable Cost $10,869 $11,682
$10,794 $10,548
Total Variable Cost $11,247 $12,468
$11,801 $11,076
Total Fixed Cost $860 $920
$954 $879
Net Returns $1,520 $1,768
$732 $1,115
Efficiency Ratios (Rounded to the nearest % )
Net Returns / Gross Returns 13.0% 12.0%
2.00% 9.00% Variable Cost / Gross Returns 83.0% 82.0%
88.0% 85.0% Fumigant Cost / Gross Returns 3.0% 5.0%
8.0% 4.0% Break-even Statistics
Break-even Revenues $12,107 $13,388
$12,755 $11,955
Break-even Yield 1,037 1,146
1,092 1,024
Break-even Price $10.38 $10.32
$11.05 $10.68
86
TABLE 5.4 Financial Efficiency and Break-even Analyses Using Constant Yield Results
Production Methods
Methyl Bromide
C35 &KPAM
C35 &
Chloropicrin Telone II &
Chloropicrin
Activity Measures
Yield (cartons)* 1,200
1,200
1,200 1,200
Gross Return $14,016 $14,016
$14,016 $14,016
Fumigant Cost $405 $786 $1,007 $528
Other Variable Cost $11,051 $11,068 $11,078 $11,057
Total Variable Cost $11,456 $11,854
$12,085 $11,585
Total Fixed Cost $860 $920
$954 $879
Net Returns $1,700 $1,242
$976 $1,552
Efficiency Ratios (Rounded to the nearest % )
Net Returns / Gross Returns 12.0% 9.0%
7.0% 11.0% Variable Cost / Gross Returns 82.0% 85.0%
86.0% 83.0% Fumigant Cost / Gross Returns 3.0% 6.0%
7.0% 4.0% Break-even Statistics
Break-even Revenues $12,316 $12,774
$13,040 $12,464
Break-even Yield 1,054 1,094
1,116 1,067
Break-even Price $10.26 $10.64
$10.87 $10.39
*Yields of 1200 ctns./acre have an associated probability of occurrence of (p=.50)
87
TABLE 5.5
Comparison of Costs per Carton for Final Four Production Methods
Production Methods
Cost per Carton MeBr C35 + KPAM C35 + PIC Telone II + PIC
Pre-harvest Variable Cost
$3.35
$3.32
$3.93
$3.60
Harvest and Marketing
$6.29
$6.29
$6.29
$6.29
Fixed Cost
$0.74
$0.71
$0.83
$0.79
Total Budgeted Cost
$10.38
$10.23
$11.05
$10.68
88
TABLE 5.6
Programming Solutions and Financial Ratios of Five-Year Averages for All Production Methods
Five Year Averages Production Methods
Activity Measures MEBR C35 + KPAM C35 + PIC T2 + PIC ACPROD 1,000 1,000 886 887ACBUY (cumulative) 435 435 300 300ACRENT 565 565 586 587EQBUY $0 $0 $22,416 $32,276OFFINV $1,528,000 $1,520,800 $393,840 $393,840CASSETS $45,485,200 $48,541,400 $9,857,504 $16,309,720TOTASSTS $47,081,400 $50,137,600 $17,961,600 $17,622,800CULIAB $146,386 $146,386 $156,714 $162,110TOTLIAB $420,840 $420,840 $453,946 $469,268NETWRTH $46,660,800 $49,717,000 $17,507,660 $17,153,500 Financial Ratios
Tenure 0.4530 0.4350 0.3386 0.3382
Current 310.7210 331.5987 62.9012 100.6090
Debt-to-Asset 0.0089 0.0084 0.0253 0.0266
Off-farm Investment 0.0325 0.0303 0.0219 0.0223
89
TABLE 5.7
Yearly Programming Solutions for Decision Variables and Key Financial Measures Using Methyl Bromide
Time Periods
T(1)
T(2)
T(3)
T(4)
T(5) Activity Measures
ACPROD 1,000 1,000 1,000 1,000 1,000 ACBUY* 55 93 0 11 7 ACRENT 645 552 552 541 534 EQBUY $0 $0 $0 $0 $0 OFFINV $0 $1,873,500 $1,876,100 $1,921,200 $1,969,200 CASSETS $10,910,000 $29,618,000 $45,459,000 $62,057,000 $79,382,000 TOTASSTS $12,310,000 $31,240,000 $47,065,000 $63,707,000 $81,085,000 CULIAB* $144,210 $155,090 $144,210 $144,210 $144,210 TOTLIAB $622,200 $530,100 $406,890 $284,530 $260,480 NETWRTH $11,688,000 $30,710,000 $46,658,000 $63,423,000 $80,825,000 * Marginal values
90
TABLE 5.8
Yearly Programming Solutions for Decision Variables and Key Financial Measures Using C35 + KPAM as Methyl Bromide Substitute
Time Periods
T(1)
T(2)
T(3)
T(4)
T(5)
Activity Measures
ACPROD 1,000 1,000 1,000 1,000 1,000 ACBUY* 48 102 0 10 7 ACRENT 652 550 550 540 533 EQBUY $0 $0 $0 $0 $0 OFFINV $0 $1,837,500 $1,876,100 $1,921,200 $1,969,200 CASSETS $11,597,000 $31,436,000 $48,465,000 $66,300,000 $84,909,000 TOTASSTS $12,981,000 $33,062,000 $50,076,000 $67,954,000 $86,615,000 CULIAB* $144,210 $155,090 $144,210 $144,210 $144,210 TOTLIAB $622,200 $530,100 $406,890 $284,530 $260,480 NETWRTH $12,359,000 $32,532,000 $49,669,000 $67,670,000 $86,355,000
* Marginal values
91
TABLE 5.9
Yearly Programming Solutions for Decision Variables and Key Financial Measures Using C35 + Chloropicrin as Methyl Bromide Substitute
Time Periods
T(1)
T(2)
T(3)
T(4)
T(5)
Activity Measures ACPROD 949 905 892 863 822 ACBUY* 0 0 0 0 0 ACRENT 649 605 592 563 522 EQBUY $0 $0 $77,461 $34,620 $0 OFFINV $0 $0 $0 $0 $1,969,200 CASSETS $604,120 $9,134,400 $13,038,000 $13,283,000 $13,228,000 TOTASSTS $10,410,000 $14,304,000 $14,583,000 $14,541,000 $35,970,000 CULIAB* $144,210 $144,210 $160,270 $167,440 $167,440 TOTLIAB $622,200 $519,220 $463,370 $353,970 $310,970 NETWRTH $9,787,300 $13,785,000 $14,120,000 $14,187,000 $35,659,000 * Marginal Values
92
TABLE 5.10
Yearly Programming Solutions for Decision Variables and Key Financial Measures Using Telone II + Chloropicrin as Methyl Bromide Substitute
Time Periods
T(1)
T(2)
T(3)
T(4)
T(5)
Activity Measures
ACPROD 949 838 912 890 844 ACBUY* 0 0 0 0 0 ACRENT 649 538 612 590 544 EQBUY $0 $0 $109,040 $52,341 $0 OFFINV $0 $0 $0 $0 $1,969,200 CASSETS $8,972,600 $11,752,000 $13,152,000 $13,205,000 $34,467,000 TOTASSTS $10,248,000 $13,019,000 $14,481,000 $14,558,000 $35,808,000 CULIAB* $144,210 $144,210 $166,810 $177,660 $177,660 TOTLIAB $622,200 $519,220 $486,400 $384,900 $333,620 NETWRTH $9,625,500 $12,500,000 $13,994,000 $14,173,000 $35,475,000 * Marginal Values
93
CHAPTER 6
SUMMARY AND CONCLUSION
6.1 Study Summary
This study began by addressing the phase-out of methyl bromide within the historical
framework of the Montreal Protocol on Substances that Deplete the Ozone Layer. It was
recognized that United States’ agricultural production has come to rely on this fumigant for the
benefits accruing to both pre and post-harvest activities. Many states have now proposed the
continued use of this chemical through the applications of “critical use exemptions.” These
exemptions have provided a short-term solution for a problem whose conclusion will likely be a
complete phase-out by member countries within the immediate future. As developed countries
begin to eliminate MeBr completely, U.S. producers will be forced to identify alternative
methods of production that result in economically viable operations.
Further analysis of the problem reveals that Georgia pepper producers face significant
losses from the phase-out as growers struggle to identify alternative combinations of agricultural
chemicals that provide a comparative yield structure to that of MeBr. Therefore, the objective of
this study was three-fold. First, yield and price data obtained by plant scientists and agricultural
economists was analyzed using stochastic dominance analysis in order to classify an efficient set
of feasible alternative fumigants for Georgia growers. Next, the results of this analysis were
combined with current production technologies within the context of enterprise pepper budgets
with the expectation of identifying a set of economically viable production methods. Finally,
94
results obtained from the enterprise budget were subjected to the simulation/optimization
procedures of a linear programming model. The results provided for by the General Algebraic
Modeling System (GAMS) software will be used to provide Georgia growers with a comparative
analysis of additional input-substitution strategies.
The related literature surrounding the problem was then reviewed in an attempt to both
identify and eliminate sources of error within the study. It was recognized that studies conducted
in Florida and California composed the bulk of the research with regard to both technically and
economically feasible alternatives to MeBr (Deepak et al., 1996; Carpenter et al., 2000;
VanSickle et al., 2000). However, these analyses predominantly focused on optimizing a
“spatial equilibrium” problem that dealt with cross-boundary impacts on a national or
international scale. Additional studies provided a biological assessment of the technical
feasibility of MeBr substitutes (Gilreath et al., 1994; Becker et al., 1998).
Although these studies identified substitutes for MeBr, regional growing conditions and
environmental policies dictate that plausible solutions can not always be implemented that satisfy
the production requirements of every U.S. producer. As a result, a general assessment of U.S.
pepper production was coupled with a more thorough investigation of Georgia pepper
production, marketing, and farm management.
The study then considered the analytical tools needed to address the problem, and
described the efficiency criteria employed by agricultural economists to determine the optimal
choice among a set of risky alternatives. Expected value, mean-variance, and stochastic
dominance analyses were compared and contrasted according to their various assumptive
restrictions. Second-degree stochastic dominance was chosen for use in this study due to the
imposition of more restrictive assumptions that served to reduce an initially broad set of
95
alternative chemicals into a more efficient set. Each method composing this final set was
analyzed within an enterprise pepper budget for variations in gross returns, variable costs, total
costs, and base budgeted net revenues. A determination of the economic viability of each
production method allowed for an initial comparison that would be more fully scrutinized within
the linear programming model.
The programming model was developed according to the characteristics of a
representative Georgia pepper farm. The simulation/optimization procedures of the model
prescribed solutions that were conditional on an initial set of production and financial
constraints. These constraints, although tailored for pepper production, addressed a full menu of
components inherent to many farm operations such as acres produced, equipment and machinery
purchases, cash and credit allowances, and labor. Expectations were that the model would
prescribe a feasible set of solutions for each alternative method from which to make a
comparative analysis based on the maximization of producers’ accumulated net worth.
6.2 Conclusions
The comparative analysis of alternative production technologies was based on methyl
bromide representing the base-case due to its current position as the dominant fumigant choice
for pepper growers throughout Georgia. Our initial expectations were that production methods
employing MeBr either as a stand-alone fumigant or in combination with a set of herbicides
would dominate within the final efficient set.
After consolidating an initial set of twenty-one production methods into a set of eight, the
latter set was ranked using Simetar, a software package that provides for second-degree
stochastic dominance analysis, into a final set of four methods. The ranking of these eight
methods according to their average total revenues per experimental plot did not conform to one
96
of our initial expectations. Of the average total revenues for the final four fumigant
combinations, C35 + KPAM dominated all other methods by a minimum of $1.77 per test plot.
Further, comparisons according to the methods’ coefficients of variation revealed that C35 +
KPAM remained the dominant method followed by MeBr, T2 + PIC, and finally C35 +PIC.
However, the ranking did substantiate our expectation that those methods incorporating a menu
of herbicides into their production technologies would dominate non-herbicide technologies.
Each of the final four dominant methods employed a corresponding menu of herbicides.
These rankings followed through to the enterprise budget analysis where C35 + KPAM
resulted in the largest base budgeted net revenue ($1,768/acre) of all methods. MeBr resulted in
net revenues per acre of ($1,520) while the final two methods T2 + PIC and C35 + PIC resulted
in net revenues of ($1,115) and ($732) per acre, respectively. After constructing a pre-operating
balance sheet from the budget data, projections were made of specific components of the budget
to be incorporated as variables within the five-year planning horizon of the programming model.
Finally, the linear programming model prescribed solutions against initial expectations.
Net worth solutions averaged over the five-year planning horizon resulted again in C35 + KPAM
being prescribed as the dominant production strategy with a value of approximately $49.7
million. MeBr was prescribed a solution of nearly $46.6 million while T2 + PIC and C35 + PIC
resulted in average net worth’s of approximately $17.1 million and $17.5 million respectively.
Although both SSD and the enterprise budget accorded T2 + PIC the dominant position when
compared with C35 + PIC, the linear model reversed this trend and produced solutions that
recognized the latter as the preferred method for producers.
Financial ratios were created from these solutions that measure the firm’s liquidity and
solvency, as well as its asset allocations. The current ratios suggested that each firm was highly
97
liquid as assets were well in excess of liabilities for each method. The two leading methods,
MeBr and C35 + KPAM, had ratios above the critical value of (2X) liabilities at approximately
311 and 332 respectively. Debt-to-asset ratios for all four methods ranged from a low of (0.008)
for C35 + KPAM to a high of (0.025) for C35 + PIC. These figures indicate that none of the
firms relied heavily on the use of debt to finance operations as the critical value for this ratio is
typically 65%. As a result of the highly liquid positions associated with methods C35 + KPAM
and MeBr, the tenure ratio suggests that growers using these options will choose to add
additional acres to production through purchases of land rather than cash-renting. On the other
hand, growers employing C35 + PIC and T2 + PIC will choose to cash-rent as the overall
financial characteristics of these farms do not suggest either additional land purchases or
maximizing the bound production constraints of prescribed by the model. Finally, the dominant
current assets position of C35 + KPAM and MeBr as compared with C35 + PIC and T2 + PIC
over the five-year planning period allows growers to dedicate a larger proportion of their overall
financial portfolio to off-farm investments. The model prescribes solutions for off-farm
investments that constitute approximately 3.0% of the growers total net-worth for C35 + KPAM
and MeBr, and approximately 2.0% for C35 + PIC and T2 + PIC.
Our results suggest that economically viable alternatives exist to replace MeBr.
However, growers have not moved to adopt solutions recognized in the literature due to several
reasons. First, growers may not have access to information concerning the alternatives cited in
scholarly journals. Next, irrigation levels, soil conditions, diseases, or pests may negate the
efficacy of the alternatives cited both in the literature and this study. This forces producers to
uniquely identify the fumigant or combination of fumigants and herbicides that results in the
same yield and quality levels as MeBr. An advantage of MeBr is its ability to eradicate diseases,
98
weeds, and pests over a wide range of environmental conditions and growing conditions.
Finally, MeBr has maintained consistent yields for producers. Although alternatives have been
found to be equally effective in experimental trials, there is not enough information that they can
deliver consistent yields over the long-term.
6.3 Future Research
Additional research concerning methyl bromide's role in Georgia pepper production can
expand on the findings of this study through several outlets. First, continued research into
economically viable fumigant alternatives must begin with the generation of additional field trial
data from which to conduct economic impact assessments. A limiting factor in this study was
the availability of historical yields and prices for Georgia bell peppers. In addition to yield and
price data, future studies must focus on the residual effects of alternatives over several harvests
and the effects crop rotations have on these substitutes.
A benefit of MeBr is that growers may apply the fumigant once during the growing
season. Subsequent crops then benefit from its ability to remain active in the soil. Tests on
alternatives have not confirmed the same results and may indicate higher costs to producers over
the long-term. Moreover, comprehensive data would allow researchers to more accurately
characterize the operations of a representative Georgia pepper farm. For example, the linear
programming model used in this study was limited by its inability to account for producers'
varying degrees of risk-aversion.
Next, the development of a quadratic programming model as employed by (Escalante,
2001) would provide for the effects of risk on such components of the farm model as debt-levels,
share-leasing, and credit reserves. Moreover, initial rankings can be subjected to a more rigorous
analysis within the framework of stochastic dominance. Stochastic dominance with respect to a
99
function accounts for an Arrow-Pratt risk aversion coefficient when evaluating the ranking
scheme. Finally, as advancements are made into the research and development of alternatives,
enterprise budgets and programming models can be modified to account for the changing cost
structure of these new substitutes. Georgia growers will then be able to adjust to these changes
in an attempt to remain profitable and competitive on a regional and international scale.
100
BIBLIOGRAPHY
Anderson, J.R., J.L. Dillon, and B. Hardaker. Agricultural Decision Analysis, 1st ed. Ames, Iowa: Iowa State University Press, 1977. pp. 65-108. Bertelson, D. “The U.S. Strawberry Industry” USDA, ERS. Statistical Bulletin Number 914, January 1995. Bodie, Z., A., Kane, and A. Marcus. Investments, 6th ed. New York: McGraw Hill Book Co., 2005. Brooke, A., D. Kendrick, and A. Meeraus. GAMS: A User's Guide, Redwood City, California: The Scientific Press, 1988. Brunt, A.A., K. Crabtree, M. Dallwitz, A.J. Gibbs, and L., Watson. Viruses of Plants: Descriptions and Lists from the VIDE Database. United Kingdom, C.A.B. International, 1996. p.1484. Calkins, P. H., and D.D. DiPietre. Farm Business Management-Successful Decisions in a Changing Environment, New York: Macmillan Co., 1983. Carpenter J., L. Gianessi, and L. Lynch. "The Economic Impact of the Scheduled U.S. Phaseout of Methyl Bromide." National Center for Food and Agricultural Policy, 2000. Chyen, D., M.E. Wetzstein, R.M. McPherson, and W.D. Givan. "An Economic Evaluation of Soybean Stink Bug Control Alternatives for the Southeastern United States." Southern Journal of Agricultural Economics 24, no. 2 (1992): 83-94. Csinos, A.S., W.C. Johnson, A.W. Johnson, D.R. Sumner, R.M. McPherson, and R.D. Gitaitis. "Alternative fumigants for methyl bromide in tobacco and pepper transplant production." Crop Protection 16 (1997): 585-594. Culpepper, S. and D.B. Langston. "Replacement of Methyl Bromide by Integrating the Use of Alternative Soil Fumigants, Cultural Practices, and Herbicides for Tomato, Pepper, and Watermelon." Proposal for CSREES/IFAFS for Program Area Entitled Critical and Emerging Pest Management Challenges. Pro. No.14. (2000). Deepak M.S., T.H. Spreen, and J.J. Van Sickle. "An Analysis of the Impact of a Ban of Methyl Bromide on the U.S. Winter Fresh Vegetable Market." Journal of Agricultural and Applied Economics 28, no. 2 (1996): 433-443.
101
Escalante, C.L. "Farm-Level Evidence on Risk Balancing Strategies of Illinois Grain Farmers Against Increasing Income Risk." PhD Dissertation, University of Illinois at Urbana-Champaign, 2001. Fernandez-Cornejo, J., S. Jans, and M. Smith. "Issues in the Economics of Pesticide Use in Agriculture: A Review of the Empirical Evidence" Review of Agricultural Economics 20, no. 2 (1998): 462-488. Fonsah, E.G., and Angus SND Chidebelu. “Economics of Banana Production and Marketing in the Tropics,” Minerva Press Publishers, London, pp.1-230. Fonsah, E.G., and K. Rucker. “Bell Pepper on Plastic Enterprise Budget,” Department of Agricultural and Applied Economics, University of Georgia, 2003. Gast, K.L.B. "Containers and Packaging: Fruits & Vegetables," Kansas State University, Agricultural Experiment Station, MF-979, March 1991. Gilreath, J.P., J.P. Jones, and A. J. Overman. "Soilborne pest control in mulched tomato with alternatives to methyl bromide." Proc. Fla. State Hort. Soc. 107: pp.156-159. 1994. Goodhue, R.E., S.A. Fennimore, and H.A. Ajwa, University of California, Davis. "Economics of VIF Use with Inline and PIC EC in California Strawberries." 2004. http://mbao.org/2004/Proceedings04/099%20GoodhueR%20MBAOVIF.pdf(Last Accessed July 2005) Harris, D.C.. "A Comparison of Dazomet, Chloropicrin and Methyl Bromide as Soil Disinfectants for Strawberries." Journal of Horticultural Science 66, no.1 (1991): 51-58. Huang, C., and R. Litzenberger. Foundations for Financial Economics. Upper Saddle River, New Jersey: Prentice-Hall, 1988. Hueth, B. M., L. Lynch, B. McWilliams, M. Roberts, J. Siebert, D. Sunding, and D. Zilberman. "Economic Impact of Banning Methyl Bromide Use in California Agriculture." Annual International Research Conference on Methyl Bromide Alternatives and Emissions Reductions. 1997 http://www.epa.gov/ozone/mbr/airc/1997/030hueth.pdf (Last Accessed July 2005) Hueth, B., B. McWilliams, D. Sunding, and D. Zilberman. "Analysis of an Emerging Market: Can Methyl Iodide Substitute for Methyl Bromide?" Review of Agricultural Economics 22, no. 1 (2000): 43-54. Jorn Ole Becker, Howard D. Ohr, Nigel M. Grech, Milton E. McGriffen, Jr & James J. Sims. "Evaluation of Methyl Iodide as a Soil Fumigant in Container and Small Field Plot Studies." Pesticide Science. 53 (1998): 71-79.
102
Jovicich, E., D.J. Cantliffe, N. Shaw, M. Smither-Kopperl, and P. A. Stansley. University of Florida, IFAS, Protected Agriculture Project, Gainesville, FL 32611. SWFREC, Immokalee, FL 34142. Economic Benefit of Greenhouse Production of Peppers in Soilless Culture. 2003. http://mbao.org/2003/027%20Smither-KopplerM2MBA%20gh%20pepper.pdf(Last Accessed July 2005) Kelly, W. "Land and Resources-Vegetable Production." New Georgia Encyclopedia. University of Georgia Press, 2003. http://www.georgiaencyclopedia.org/nge/Article.jsp?id=h-1005 (Last Accessed July 2005) Kelly, W. "Quiet, deliberate growth: Georgia vegetable production advances." Southeast Farm Press, December 2002, p.1. King, R.P., and L.J. Robinson. "An Interval Approach to Measuring Decision Maker Preferences." American Journal of Agricultural Economics 63, no. 3 (1981): 510-520. Locascio, S.J., D.W. Dickson, and T.A. Kucharek. "Evaluation of multi-purpose soil fumigants for root-knot nematode and soilborne disease management on tomato in Florida." Proc. of the International Resource Conference on Methyl Bromide Alternatives Emissions Reduction, Orlando, FL, pp.3-4. 1994. Lucier G., C. Plummer, and A. Jerardo. USDA, ERS. Vegetables and Melons Situation and Outlook Yearbook. July 2004. http://ers.usda.gov/publications/vgs/Jul04/VGS2004.pdf(Last Accessed July 2005) Lynch, L., B. McWilliams, and D. Zilberman. "Economic Implications of Banning Methyl Bromide: How Have They Changed With Recent Developments?" 1997. http://www.epa.gov/ozone/mbr/airc/1997/029lynch.pdf (Last Accessed July 2005) Lynch, L., C. Greene, and C. Kramer-LeBlanc. USDA, ERS. Proceedings of the Third National IPM Symposium/Workshop. May 1997. http://www.ers.usda.gov/Publications/mp1542/(Last Accessed July 2005) Lynch, L. and J., Carpenter. The Economic Impacts of Banning Methyl Bromide: Where Do We Need More Research? 1999 http://www.epa.gov/ozone/mbr/airc/1999/55lynchl.pdf(Last Accessed July 2005) Lynch, L. and J., Carpenter. Economic Impacts of Switching From Pre-Plant Fumigation with Methyl Bromide. 1998 http://www.epa.gov/ozone/mbr/airc/1998/039lynch.pdf (Last Accessed July 2005) McCarl, B.A., and T.H. Spreen. Applied Mathematical Programming Using Algebraic Systems. College Station, Texas: Texas A&M University, 1994.
103
Ohr, H.D., J. J. Sims, N. M. Grech, J. O. Becker, and M. E. McGiffen, Jr. "Methyl Iodide, an Ozone-Safe Alternative for Methyl Bromide as a Soil Fumigant." Plant Disease 80(1996):731-35. Osteen, C. "Agricultural Resources and Environmental Indicators: Pest Management Practices." USDA, ERS. No. AH722. September 2000. http://ers.usda.gov/publications/arei/ah722/ arei4_3/DBGen.htm (Last Accessed July 2005) Richardson, J.W., K. Schumann, and P. Feldman. Simetar “Simulation for Excel to Analyze Risk.” Version 2004.01.20 Beta. Agricultural and Food Policy Center at Texas A&M University. January 2004. Samuelson, P.A. "The Fundamental Approximation Theorem of Portfolio Analysis in Terms of Means, Variances, and Higher Moments." Review of Economic Studies 37, no. 4 (1970): 537-542. Schaub, J. D. "Economic Forces and Progress with Post-harvest Methyl Bromide Alternatives" Paper presented at the International Research Conference on Methyl Bromide Alternatives and Emissions Reductions, Orlando FL, 1-18 November 2004. Seabrook, C. "Vegetable growers rely on ozone-depleting gas." The Atlanta Journal and Constitution. February, 2005. Stall, W. Methyl Bromide Alternatives Field Day. Live Oak, FL: North Florida Research and Education Center-Suwannee Valley, 2000. Sydorovych, O., C. D. Safley, E. B. Poling, L. M. Ferguson, G. E. Fernandez, P. M. Brannen, and F. J. Louws. Departments of Agricultural and Resource Economics, Horticultural Science, and Plant Pathology, North Carolina State University, Raleigh NC 27695; University of Georgia, Plant Pathology, Athens GA 30602. Economic Evaluation of Methyl Bromide Alternatives for Strawberry Production. 2004. http://mbao.org/2004/Proceedings04/028%20SafleyC%20Louws %20economicsl%202004.pdf (Last Accessed July 2005) Thompson, W.T. Agricultural Chemicals: Book IV-Fungicides. Fresno, CA: Thompson Publications, 1991. United Nations Environmental Program, Report of the Seventh Meeting of the Parties to the Montreal Protocol on Substances that Deplete the Ozone Layer. New York, 1995. United Nations, Environment Programme. The Montreal Protocol on Substances that Deplete the Ozone Layer. New York, 2000. University of Georgia, College of Agricultural and Environmental Sciences. 2003 Georgia Farm Gate Vegetable Survey Report. Compiled by Susan R. Boatright and John C. McKissick. SR-04-01 (pp.1-10)
104
University of Georgia, Cooperative Extension Service. Commercial Pepper Production. Bulletin No. 1027, February 1990. http://pubs.caes.uga.edu/caespubs/pubcd/b1027-w.html(Last Accessed July 2005) USDA. Vegetables Final Estimates 1998-2003. Statistical Bulletin # 987. May 2004. http://usda.mannlib.cornell.edu/reports/general/sb/sb987.txt (Last Accessed July 2005) USDA, ERS Publications and Outlook Reports 2004. http://www.ers.usda.gov/publications/ so/view.asp?f=/specialty/vgs-bb/ (Last Accessed July 2005) USDA, ERS. Vegetables and Melons Situation and Outlook Yearbook. July 2004. http://www.ers.usda.gov/publications/so/view.asp?f=/specialty/vgs-bb/ (Last Accessed July 2005) USDA, ERS. Vegetables and Melons Outlook. December 2001. http://www.ers.usda.gov/publications/so/view.asp?f=/specialty/vgs-bb/(Last Accessed July 2005) USDA, ERS. Vegetables and Melon. June 2003. http://www.ers.usda.gov/publications/so/view.asp?f=/specialty/vgs-bb/(Last Accessed July 2005) USDA, NASS. Land Values and Cash Rents, 2004 Summary. August 2004. http://usda.mannlib.cornell.edu/reports/nassr/other/plr-bb/land0804.pdf(Last Accessed July 2005) USDA, NASS. Vegetables Annual Summary. January 2005. http://usda.mannlib.cornell.edu/reports/nassr/fruit/pvg-bb/2005/(Last Accessed July 2005) USDA, NASS. 2002 Census of Agriculture, State Data. (http://www.nass.usda.gov/census/census02/volume1/ga/st13_1_035_035.pdf)] (Last Accessed July 2005) USDA, NASS. Statistics of Vegetables and Melons. 2004. http://www.usda.gov/nass/pubs/agr04/04_ch4.pdf (Last Accessed July 2005) USDA, NASS. Vegetables, Final Estimates 1998-2003. Statistical Bulletin Number 987. May 2004. http://usda.mannlib.cornell.edu/usda/reports/general/sb/sb987.pdf (Last Accessed July 2005) U.S. Environmental Protection Agency. Methyl Bromide Alternative Case Studies. Volumes 1, 2, and 3. http://www.epa.gov/spdpublc/mbr/casestudies/ (Last Accessed July 2005)
105
U.S. Environmental Protection Agency. Methyl Bromide Fact Sheet. May 2005. http://www.epa.gov/ozone/mbr/CueFRfactsheet.html (Last Accessed July 2005) U.S. Environmental Protection Agency, “Clean Air Act, Section 602. Listing of Class I and Class II Substances,” 1990. http://www.epa.gov/ozone/ods.html(Last Accessed July 2005) VanSickle, J.J., C. Brewster, and T.H. Spreen. "Impact of a Methyl Bromide Ban on the U.S. Vegetable Industry." Bulletin No. 333, University of Florida, 2000. Wackerly, D., and W. Mendenhall III, and R.L. Scheaffer. Mathematical Statistics with Applications, 6th ed. Pacific Grove, California: Duxbury Resource Center, 2002. Wetzstein, M.E. Microeconomic Theory-Concepts and Connections, 1st ed. Mason, Ohio: Thomson Southwestern, 2005. pp.584-618. Wetzstein, M.E., P.I. Szmedra, R.W. McClendon, and D.M. Edwards. "Efficiency Criteria and Risk Aversion: An Empirical Evaluation." Southern Journal of Agricultural Economics 20, no. 1 (1988): 171-178. Wolfe, K. and E.G. Fonsah. "Wholesale and Distributors Outlook for Fruit and Vegetables Produced in Georgia" GFVGA News Vo. 7, No. 4, Fall 2002. Zhang, W., M.E. McGriffen, J.O. Becker, H.D. Ohr, J.J. Sims, and S. Campbell. "Effects of Soil Physical Factors on Methyl Iodide and Methyl Bromide." Pesticide Science. 53 (1998): 71-79
106
APPENDICES
107
APPENDIX A
Bell Pepper Fresh Market
Variable Cost Budget for Methyl Bromide Number of acres (1)
Expected Yield (cartons) 1167 Expected Price per carton $11.68 Item Unit Quantity Price Amount per acreVariable Costs Plants Hundred 174.00 $5.50 $957.00Lime, applied Ton 1.00 $26.00 $26.00Base Fertilizer Lbs. 12.00 $9.50 $114.00Sidedress Fertilizer Gal. 0.54 $184.00 $99.36Insecticide Acre 1.00 $186.22 $186.22Fungicide Acre 1.00 $231.36 $231.36Nematicide Acre 1.00 $510.00 $510.00Herbicide Acre 1.00 $68.08 $68.08Plastic Roll 2.80 $68.50 $191.80Plastic removal Acre 1.00 $75.00 $75.00Drip Tape Ft. 8700.00 $0.02 $174.00Fumigation Acre 1.00 $405.00 $405.00Machinery Hour 5.00 $21.00 $105.00Transplant Labor Hour 20.00 $8.00 $160.00Labor Hour 33.00 $8.00 $264.00Land Rent Acre 1.00 $105.00 $105.00Irrigation (Machinery and Labor)
Acre
1.00 $65.08 $65.08Interest on Operating Capital $ 3716.89 $0.09 $168.16Pre-Harvest Variable Costs $3,905.05 Harvest and Marketing Costs Unit Quantity Price Amount per acre Picking and Hauling Ctn. 1167 $1.25 $1,458.33Grading and Packing Ctn. 1167 $2.75 $3,208.33Container Ctn. 1167 $1.30 $1,516.67Marketing Ctn. 1167 $0.99 $1,158.27Total Harvest and Marketing $6.29 $7,341.60Total Variable Costs $11,246.65
108
APPENDIX A (cont'd)
Bell Pepper Fresh Market
Variable Cost Budget for Methyl Bromide Number of acres (1)
Item Unit Quantity Price Amount per acre Fixed Cost
Machinery
Acre
1.00 $53.59 $53.59
Irrigation
Acre
1.00 $220.65 $220.65
Land
Acre
1.00 $0.00 $0.00
Overhead and Management
$ 3,905.05 $0.15 $585.76
Total Fixed Costs
$860.00
Total Budgeted Cost / Acre
$12,106.64
Costs per Carton Pre-Harvest Variable Cost per Carton $ 3.35 Harvesting and Marketing Cost per Carton $ 6.29 Fixed Cost per Carton $ 0.74 Total Budgeted Cost per Carton $ 10.38 Base Budgeted Net Revenue $1,520
109
APPENDIX B
Bell Pepper Fresh Market
Variable Cost Budget for C35 + KPAM Number of acres (1)
Expected Yield (cartons) 1298 Expected Price per carton $11.68 Item Unit Quantity Price Amount per acreVariable Costs Plants Hundred 174.00 $5.50 $957.00Lime, applied Ton 1.00 $26.00 $26.00Base Fertilizer Lbs. 12.00 $9.50 $114.00Sidedress Fertilizer Gal. 0.54 $184.00 $99.36Insecticide Acre 1.00 $186.22 $186.22Fungicide Acre 1.00 $231.36 $231.36Nematicide Acre 1.00 $510.00 $510.00Herbicide Acre 1.00 $68.08 $68.08Plastic Roll 2.80 $68.50 $191.80Plastic removal Acre 1.00 $75.00 $75.00Drip Tape Ft. 8700.00 $0.02 $174.00Fumigation Acre 1.00 $785.75 $785.75Machinery Hour 5.00 $21.00 $105.00Transplant Labor Hour 20.00 $8.00 $160.00Labor Hour 33.00 $8.00 $264.00Land Rent Acre 1.00 $105.00 $105.00Irrigation (Machinery and Labor)
Acre
1.00 $65.08 $65.08Interest on Operating Capital $ 3716.89 $0.09 $185.29Pre-Harvest Variable Costs $4,302.93 Harvest and Marketing Costs Unit Quantity Price Amount per acrePicking and Hauling Ctn. 1298 $1.25 $1,622.02Grading and Packing Ctn. 1298 $2.75 $3,568.45Container Ctn. 1298 $1.30 $1,686.90Marketing Ctn. 1298 $0.99 $1,288.25Total Harvest and Marketing $6.29 $8,165.63Total Variable Costs $12,468.57
110
APPENDIX B (cont'd)
Bell Pepper Fresh Market
Variable Cost Budget for C35 + KPAM Number of acres (1)
Item Unit Quantity Price Amount per acre Fixed Cost
Machinery
Acre 1.00 $53.59 $53.59
Irrigation
Acre 1.00 $220.65 $220.65
Land
Acre 1.00 $0.00 $0.00
Overhead and Management
$ 4302.93 $0.15 $645.44
Total Fixed Costs
$919.68
Total Budgeted Cost / Acre
$13,388.25
Costs per Carton Pre-Harvest Variable Cost per Carton
$3.32
Harvesting and Marketing Cost per Carton
$6.29
Fixed Cost per Carton
$0.71
Total Budgeted Cost per Carton
$10.23
Base Budgeted Net Revenue $1,768
111
APPENDIX C
Bell Pepper Fresh Market
Variable Cost Budget for C35 + Chloropicrin Number of acres (1)
Expected Yield (cartons) 1155 Expected Price per carton $11.68 Item Unit Quantity Price Amount per acreVariable Costs Plants Hundred 174.00 $5.50 $957.00Lime, applied Ton 1.00 $26.00 $26.00Base Fertilizer Lbs. 12.00 $9.50 $114.00Sidedress Fertilizer Gal. 0.54 $184.00 $99.36Insecticide Acre 1.00 $186.22 $186.22Fungicide Acre 1.00 $231.36 $231.36Nematicide Acre 1.00 $510.00 $510.00Herbicide Acre 1.00 $68.08 $68.08Plastic Roll 2.80 $68.50 $191.80Plastic removal Acre 1.00 $75.00 $75.00Drip Tape Ft. 8700.00 $0.02 $174.00Fumigation Acre 1.00 $1,007.05 $1,007.05Machinery Hour 5.00 $21.00 $105.00Transplant Labor Hour 20.00 $8.00 $160.00Labor Hour 33.00 $8.00 $264.00Land Rent Acre 1.00 $105.00 $105.00Irrigation (Machinery and Labor)
Acre 1.00 $65.08 $65.08
Interest on Operating Capital $ 4743.94 $0.09 $213.48Pre-Harvest Variable Costs $4,534.19 Harvest and Marketing Costs Unit Quantity Price Amount per acrePicking and Hauling Ctn. 1155 $1.25 $1,443.45Grading and Packing Ctn. 1155 $2.75 $3,175.60Container Ctn. 1155 $1.30 $1,501.19Marketing Ctn. 1155 $0.99 $1,146.42Total Harvest and Marketing $6.29 $7,266.66Total Variable Costs $11,800.86
112
APPENDIX C (cont'd)
Bell Pepper Fresh Market
Variable Cost Budget for C35 + Chloropicrin Number of acres (1)
Item Unit Quantity Price Amount per acre Fixed Cost
Machinery
Acre
1.00 $53.59 $53.59
Irrigation
Acre
1.00 $220.65 $220.65
Land
Acre
1.00 $0.00 $0.00
Overhead and Management
$ 4534.19 $0.15 $680.13
Total Fixed Costs
$954.37
Total Budgeted Cost / Acre
$12,755.23
Costs per Carton Pre-Harvest Variable Cost per Carton $ 3.93 Harvesting and Marketing Cost per Carton $ 6.29 Fixed Cost per Carton $ 0.83 Total Budgeted Cost per Carton $ 11.05 Base Budgeted Net Revenue $732
113
APPENDIX D
Bell Pepper Fresh Market
Variable Cost Budget for Telone II + Chloropicrin Number of acres (1)
Expected Yield (cartons) 1119 Expected Price per carton $11.68 Item Unit Quantity Price Amount per acreVariable Costs Plants Hundred 174.00 $5.50 $957.00Lime, applied Ton 1.00 $26.00 $26.00Base Fertilizer Lbs. 12.00 $9.50 $114.00Sidedress Fertilizer Gal. 0.54 $184.00 $99.36Insecticide Acre 1.00 $186.22 $186.22Fungicide Acre 1.00 $231.36 $231.36Nematicide Acre 1.00 $510.00 $510.00Herbicide Acre 1.00 $68.08 $68.08Plastic Roll 2.80 $68.50 $191.80Plastic removal Acre 1.00 $75.00 $75.00Drip Tape Ft. 8700.00 $0.02 $174.00Fumigation Acre 1.00 $528.00 $528.00Machinery Hour 5.00 $21.00 $105.00Transplant Labor Hour 20.00 $8.00 $160.00Labor Hour 33.00 $8.00 $264.00Land Rent Acre 1.00 $105.00 $105.00Irrigation (Machinery and Labor)
Acre 1.00
$65.08 $65.08
Interest on Operating Capital $ 3859.89 $0.09 $173.70Pre-Harvest Variable Costs $4,033.59 Harvest and Marketing Costs Unit Quantity Price Amount per acrePicking and Hauling Ctn. 1119 $1.25 $1,398.81Grading and Packing Ctn. 1119 $2.75 $3,077.38Container Ctn. 1119 $1.30 $1,454.76Marketing Ctn. 1119 $0.99 $1,110.97Total Harvest and Marketing $6.29 $7,041.92Total Variable Costs $11,075.51
114
APPENDIX D (cont'd)
Bell Pepper Fresh Market
Variable Cost Budget for Telone II + Chloropicrin Number of acres (1)
Item Unit Quantity Price Amount per acre Fixed Cost
Machinery
Acre
1.00 $53.59 $53.59
Irrigation
Acre
1.00 $220.65 $220.65
Land
Acre
1.00 $0.00 $0.00
Overhead and Management
$
4033.59 $0.15 $605.04
Total Fixed Costs
$879.28
Total Budgeted Cost / Acre
$11954.79
Costs per Carton Pre-Harvest Variable Cost per Carton
$3.60
Harvesting and Marketing Cost per Carton
$6.29
Fixed Cost per Carton
$0.79
Total Budgeted Cost per Carton
$10.68
Base Budgeted Net Revenue $1,115
115
APPENDIX E
GAMS Program for the Base-Case Farm Model using Methyl Bromide
*GAMS PROGRAM FOR PEPPER OPTIMIZATION $OFFSYMXREF OFFSYMLIST OPTION LIMCOL=0; OPTION LIMROW=0; *SIMULATION BASE MODEL FOR RISK BALANCING STUDY SETS T YEARS IN PLANNING HORIZON /T0*T5/ X DECISION VARIABLES /ACPROD, ACBUY, EQBUY, FINLAND, ACRENT, MEDCRED, SHTCRED/ ALIAS (T,J) ALIAS (X,Y); SCALARS LAND0 INITIAL LAND OWNED IN ACRES /300/ EQPT0 INITIAL EQUIPMENT OWNED /468900/ CASH0 INITIAL ENDING CASH BALANCE /124255/ CURAST0 INITIAL TOTAL CURRENT ASSETS /604116/ INV0 INITIAL AMOUNT OF NON-FARM INVESTMENTS /100000/ OTHCUR0 OTHER CURRENT ASSETS /379861/ BUILD0 INITIAL VALUE OF BUILDINGS /125218/ OVRLAB ANNUAL OVERHEAD LABOR /900/ LABREQ LABOR REQUIREMENT MANHOURS PER ACRE /53.0/ LNDDPMT LAND DOWNPAYMENT RATE /0.20/ EQDPMT EQUIPMENT DOWNPAYMENT RATE /0.20/ STRATE STOCK PURCHASE RATE /0.02/ PROPTAX PROPERTY TAXES PER ACRE /24.19/ XPRATE PERCENT USE OF EXPENSES PER SUBPERIOD /0.5/ INCTAX INCOME TAX RATE /0.4/ LIMSTOCK LIMIT ON FCS STOCK /1000/ LIMSP LIMIT ON OFF FARM INVESTMENTS /200000/ UPSIZE UPPER LIMIT ON FARM SIZE /1000/; PARAMETER LANDVAL(T) PROJECTED LAND VALUE PER ACRE /T0 2150 T1 2270 T2 2397.54 T3 2531.80 T4 2673.59 T5 2823.31/;
116
PARAMETER RENT(T) PROJECTED LAND CASH RENT LEVELS /T0 105 T1 110.88 T2 117.09 T3 123.65 T4 130.57 T5 137.88/; PARAMETER EQCOST(T) PROJECTED EQUIPMENT COST PER ACRE /T0 1304.25 T1 1343.38 T2 1383.68 T3 1425.19 T4 1467.94 T5 1511.98/; PARAMETER FACONS(T) FAMILY CONSUMPTION /T1 114417 T2 116477 T3 118923 T4 121777 T5 124822/; PARAMETER MOFFINC(T) MAXIMUM OFF-FARM INCOME /T1 63177 T2 64314 T3 65665 T4 67241 T5 68922/; PARAMETER VARCOST(T) VARIABLE COSTS PER ACRE OF OWNED LAND /T0 10674.17 T1 10994.40 T2 11324.23 T3 11663.95 T4 12013.87 T5 12374.29/; PARAMETER LABCOST(T) PROJECTED LABOR COST PER MANHOUR /T0 8.00 T1 8.24 T2 8.49 T3 8.74 T4 9.00 T5 9.27/; PARAMETER RETURN(T) PROJECTED GROSS RETURNS PER ACRE /T0 14016.00 T1 14436.48 T2 14869.57 T3 15315.66
117
T4 15775.13 T5 16248.39/; PARAMETER INVGAIN(T) PROCEEDS FROM NONFARM INVESTMENTS /T1 1.035 T2 1.035 T3 1.035 T4 1.035 T5 1.035/; PARAMETER YIELD(T) YIELD ON NONFARM INVESTMENTS /T1 0.035 T2 0.035 T3 0.035 T4 0.035 T5 0.035/; PARAMETER OVRHD(T) PROJECTED OVERHEAD COSTS /T0 856.86 T1 882.57 T2 909.04 T3 936.31 T4 964.40 T5 993.34/; PARAMETER MARGIN(T) PROJECTED NET MARGINS PER ACRE /T0 1723.63 T1 1775.34 T2 1828.60 T3 1883.46 T4 1939.96 T5 1998.16/; PARAMETER PRINTS(T) PRINCIPAL AND INTEREST PER DOLLAR SHORT-TERM CREDIT /T1 1.070 T2 1.070 T3 1.070 T4 1.070 T5 1.070/; PARAMETER INPMTS(T) INTEREST PAYMENT PER DOLLAR SHORT TERM CREDIT /T1 0.0725 T2 0.0750 T3 0.0775 T4 0.080 T5 0.080/;
118
TABLE EQDEP(T,J) EQUIPMENT DEPRECIATION RATE (ACRS10) T0 T1 T2 T3 T4 T5 T1 0.00 0.10 0.00 0.00 0.00 0.00 T2 0.10 0.18 0.10 0.00 0.00 0.00 T3 0.18 0.14 0.18 0.10 0.00 0.00 T4 0.14 0.12 0.14 0.18 0.10 0.00 T5 0.12 0.10 0.12 0.14 0.18 0.10; TABLE DEPVALEQ(T,J) DEPRECIATED VALUE OF EQUIPMENT T0 T1 T2 T3 T4 T5 T1 1.0000 0.9000 0.0000 0.0000 0.0000 0.0000 T2 0.9000 0.7380 0.9000 0.0000 0.0000 0.0000 T3 0.7380 0.6350 0.7380 0.9000 0.0000 0.0000 T4 0.6350 0.5580 0.6350 0.7380 0.9000 0.0000 T5 0.5580 0.5030 0.5580 0.6350 0.7380 0.9000; TABLE FPRIN20(T,J) PRNCPL AND INTRST 20 YR FIXED LOAN T0 T1 T2 T3 T4 T5 T1 0.0996 0.0989 0.0000 0.0000 0.0000 0.0000 T2 0.0996 0.0989 0.1019 0.0000 0.0000 0.0000 T3 0.0996 0.0989 0.1019 0.1027 0.0000 0.0000 T4 0.0996 0.0989 0.1019 0.1027 0.1027 0.0000 T5 0.0996 0.0989 0.1019 0.1027 0.1027 0.1027; TABLE INTF20(T,J) INTEREST PAYMENTS 20 YR FIXED TERM LOAN T0 T1 T2 T3 T4 T5 T1 0.0522 0.0761 0.0000 0.0000 0.0000 0.0000 T2 0.0485 0.0744 0.0801 0.0000 0.0000 0.0000 T3 0.0446 0.0725 0.0784 0.0811 0.0000 0.0000 T4 0.0403 0.0705 0.0765 0.0793 0.0811 0.0000 T5 0.0358 0.0683 0.0744 0.0775 0.0793 0.0811; TABLE PRF20(T,J) PRINCIPAL PAYMENTS 20 YR FIXED TERM LOAN T0 T1 T2 T3 T4 T5 T1 0.0474 0.0228 0.0000 0.0000 0.0000 0.0000 T2 0.0511 0.0246 0.0218 0.0000 0.0000 0.0000 T3 0.0550 0.0264 0.0236 0.0216 0.0000 0.0000 T4 0.0593 0.0284 0.0255 0.0233 0.0216 0.0000 T5 0.0638 0.0306 0.0275 0.0252 0.0233 0.0216;
119
TABLE PROUTF20(T,J) PRINCIPAL OUTSTDG 20 YR FIXED TERM LOAN T0 T1 T2 T3 T4 T5 T1 0.6300 0.9772 0.0000 0.0000 0.0000 0.0000 T2 0.5789 0.9526 0.9782 0.0000 0.0000 0.0000 T3 0.5239 0.9262 0.9546 0.9784 0.0000 0.0000 T4 0.4646 0.8978 0.9291 0.9551 0.9784 0.0000 T5 0.4008 0.8672 0.9016 0.9298 0.9551 0.9784; TABLE PRINTMED(T,J) PRNCPL AND INTRST PMTS MEDIUM CREDIT T0 T1 T2 T3 T4 T5 T1 0.2591 0.2558 0.0000 0.0000 0.0000 0.0000 T2 0.2591 0.2558 0.2584 0.0000 0.0000 0.0000 T3 0.2591 0.2558 0.2584 0.2591 0.0000 0.0000 T4 0.2591 0.2558 0.2584 0.2591 0.2591 0.0000 T5 0.2591 0.2558 0.2584 0.2591 0.2591 0.2591; TABLE INTMED(T,J) INTRST PMTS ON MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T1 0.0930 0.0880 0.0000 0.0000 0.0000 0.0000 T2 0.0776 0.0732 0.0920 0.0000 0.0000 0.0000 T3 0.0607 0.0572 0.0767 0.0930 0.0000 0.0000 T4 0.0422 0.0397 0.0600 0.0776 0.0930 0.0000 T5 0.0220 0.0207 0.0417 0.0607 0.0776 0.0930; TABLE PRMED(T,J) PRNCPL PMTS MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T1 0.1661 0.1678 0.0000 0.0000 0.0000 0.0000 T2 0.1815 0.1825 0.1664 0.0000 0.0000 0.0000 T3 0.1984 0.1986 0.1817 0.1661 0.0000 0.0000 T4 0.2169 0.2161 0.1985 0.1815 0.1661 0.0000 T5 0.2371 0.2351 0.2167 0.1984 0.1815 0.1661; TABLE PROUTMED(T,J) PRNCPL OUTSTNDG MEDIUM CREDIT T0 T1 T2 T3 T4 T5 T1 0.8339 0.8322 0.0000 0.0000 0.0000 0.0000 T2 0.6524 0.6497 0.8336 0.0000 0.0000 0.0000 T3 0.4539 0.4511 0.6518 0.8339 0.0000 0.0000 T4 0.2371 0.2351 0.4534 0.6524 0.8339 0.0000 T5 0.0000 0.0000 0.2367 0.4539 0.6524 0.8339;
120
TABLE PRINTM3D(T,J) PRNCPL AND INTRST PMTS MEDIUM CREDIT T0 T1 T2 T3 T4 T5 T1 0.1627 0.1627 0.0000 0.0000 0.0000 0.0000 T2 0.1627 0.1627 0.1627 0.0000 0.0000 0.0000 T3 0.1627 0.1627 0.1627 0.1627 0.0000 0.0000 T4 0.1627 0.1627 0.1627 0.1627 0.1627 0.0000 T5 0.1627 0.1627 0.1627 0.1627 0.1627 0.1627; TABLE INTMED3(T,J) INTRST PMTS ON MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T1 0.0617 0.1000 0.0000 0.0000 0.0000 0.0000 T2 0.0516 0.0937 0.1000 0.0000 0.0000 0.0000 T3 0.0404 0.0868 0.0937 0.1000 0.0000 0.0000 T4 0.0282 0.0792 0.0868 0.0937 0.1000 0.0000 T5 0.0148 0.0708 0.0792 0.0868 0.0937 0.0100; TABLE PRMED3(T,J) PRNCPL PMTS MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T1 0.0101 0.0627 0.0000 0.0000 0.0000 0.0000 T2 0.0111 0.0690 0.0627 0.0000 0.0000 0.0000 T3 0.0122 0.0759 0.0690 0.0627 0.0000 0.0000 T4 0.0134 0.0835 0.0759 0.0690 0.0627 0.0000 T5 0.0148 0.0918 0.0835 0.0759 0.0690 0.0627; TABLE PROUTM3D(T,J) PRNCPL OUTSTNDG MEDIUM CREDIT T0 T1 T2 T3 T4 T5 T1 0.5159 0.9372 0.0000 0.0000 0.0000 0.0000 T2 0.4047 0.8682 0.9372 0.0000 0.0000 0.0000 T3 0.2825 0.7923 0.8682 0.9372 0.0000 0.0000 T4 0.1480 0.7088 0.7923 0.8682 0.9372 0.0000 T5 0.0000 0.6169 0.7088 0.7923 0.8682 0.9372; TABLE PRINTM2D(T,J) PRNCPL AND INTRST PMTS MEDIUM CREDIT T0 T1 T2 T3 T4 T5 T1 0.2538 0.2538 0.0000 0.0000 0.0000 0.0000 T2 0.2538 0.2538 0.2538 0.0000 0.0000 0.0000 T3 0.2538 0.2538 0.2538 0.2538 0.0000 0.0000 T4 0.2538 0.2538 0.2538 0.2538 0.2538 0.0000 T5 0.2538 0.2538 0.2538 0.2538 0.2538 0.2538;
121
TABLE INTMED2(T,J) INTRST PMTS ON MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T1 0.0850 0.0850 0.0000 0.0000 0.0000 0.0000 T2 0.0707 0.0707 0.0850 0.0000 0.0000 0.0000 T3 0.0551 0.0551 0.0707 0.0850 0.0000 0.0000 T4 0.0382 0.0382 0.0551 0.0707 0.0850 0.0000 T5 0.0119 0.0119 0.0382 0.0551 0.0707 0.0850; TABLE PRMED2(T,J) PRNCPL PMTS MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T1 0.1688 0.1688 0.0000 0.0000 0.0000 0.0000 T2 0.1831 0.1831 0.1688 0.0000 0.0000 0.0000 T3 0.1987 0.1987 0.1831 0.1688 0.0000 0.0000 T4 0.2156 0.2156 0.1987 0.1831 0.1688 0.0000 T5 0.2339 0.2339 0.2156 0.1987 0.1831 0.1688; TABLE PROUTM2D(T,J) PRNCPL OUTSTNDG MEDIUM CREDIT T0 T1 T2 T3 T4 T5 T1 0.8312 0.8312 0.0000 0.0000 0.0000 0.0000 T2 0.6481 0.6481 0.8312 0.0000 0.0000 0.0000 T3 0.4494 0.4494 0.6481 0.8312 0.0000 0.0000 T4 0.2339 0.2339 0.4494 0.6481 0.8312 0.0000 T5 0.0000 0.0000 0.2339 0.4494 0.6481 0.8312; TABLE PRMEDD(T,J) ADVANCED PRNCPL PMTS MEDIUM TERM CREDIT T0 T1 T2 T3 T4 T5 T0 0.1661 0.1678 0.0000 0.0000 0.0000 0.0000 T1 0.1815 0.1825 0.1664 0.0000 0.0000 0.0000 T2 0.1984 0.1986 0.1817 0.1661 0.0000 0.0000 T3 0.2169 0.2161 0.1985 0.1815 0.1661 0.0000 T4 0.2371 0.2351 0.2167 0.1984 0.1815 0.1661 T5 0.0000 0.0000 0.2367 0.2169 0.1984 0.1815; TABLE PRF20D(T,J) ADVANCED PRINCIP PYMNTS 20 YR FIXED TERM LOAN T0 T1 T2 T3 T4 T5 T0 0.0474 0.0228 0.0000 0.0000 0.0000 0.0000 T1 0.0511 0.0246 0.0218 0.0000 0.0000 0.0000 T2 0.0550 0.0264 0.0236 0.0216 0.0000 0.0000 T3 0.0593 0.0284 0.0255 0.0233 0.0216 0.0000 T4 0.0638 0.0306 0.0275 0.0252 0.0233 0.0216 T5 0.0687 0.0329 0.0297 0.0273 0.0252 0.0233
122
VARIABLES WLTH CHANGE IN NET WORTH OVER PLANNING PD; POSITIVE VARIABLES decvar(X,T) Decision Variables begcash(T) Amount of cash existing in first subperiod endcash(T) Amount of cash existing in second subperiod cashland(T) Cash used to purchase land casheq(T) Cash used to purchase equipment famcon(T) Family consumption txincm(T) Amount of taxable income invest(T) Amount invested in Treasury bills offarm(T) Amount of off farm income offinv(T) Amount of off farm investments hirelab(T) Amount of hired labor taxes(T) Amount paid in taxes in year t casset(T) Amount of Current Assets iasset(T) Amount of Intermediate Assets *ownland(T) Acreage owned lasset(T) Amount of Long Term Assets tasset(T) Amount of Total Assets culiab(T) Amount of Current Liabilities intliab(T) Amount of Intermediate Liabilities lgliab(T) Amount of Long Term Debt totliab(T) Total Debt nworth(T) Net Worth cureq(T) Current Equity inteq(T) Intermediate Equity lgteq(T) Long Term Equity mscredit(T) Maximum Short Term Credit Reserves micredit(T) Maximum Intermediate Term Credit Reserves mlcredit(T) Maximum Long Term Credit Reserves interest(T) Total interest charges paid depre(T) Depreciation; EQUATIONS OBJ Objective Function PRODMARK Acreage Produced and Marketed FINALAND Long-Term Financing for Land LABORCON Labor Constraint LANDDWP Land Downpayment Requirement MACHREQT Machinery Use Constraint FINAMACH Machinery Financing MACHDWP Machinery Downpayment Requirement FAMICONS Family Consumption OFFARINV Off-Farm Investment OFFARINC Proceeds from Off-farm Investments OFFIDENT Off-Farm Income Identity TRANSF1 Cash Transfer in First Subperiod of the year TRANSF2 Cash Transfer in Second Subperiod of the year
123
TAXINCOM Taxable Income Constraints CURRAS `Current Assets INTAS Intermediate Assets LNGTAS Long Term Assets TOTAS Total Assets CUDEBT Short Term Debt INTDEB Intermediate Term Debt LNGDEBT Long Term Debt TOTDEB Total Debt EQUITY Owner's Equity TAXPAYMT Total Taxes Paid in a year DEPN Depreciation Charges FSIZE Upper Limit on Farm Size CEQUITY Current Equity IEQUITY Intermediate Term Equity LEQUITY Long Term Equity MAXSDEBT Maximum Short-Term Debt MAXIDEBT Maximum Intermediate-Term Debt MAXLDEBT Maximum Long-Term Debt INTREST Interest Charges paid; OBJ.. WLTH =E= SUM(T,(nworth(T)-nworth(T-1))); PRODMARK(T)$(ORD(T) GT 1).. decvar("ACPROD",T)-decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J))=E=0; FINALAND(T)$(ORD(T) GT 1).. LANDVAL(T)*decvar("ACBUY",T) -cashland(T)-decvar("FINLAND",T)=E=0; LABORCON(T)$(ORD(T) GT 1).. hirelab(T)- (LABREQ*(decvar("ACRENT",T)-SUM(J$(ORD(J) LE ORD(T)), decvar("ACBUY",J))))=G=OVRLAB; LANDDWP(T)$(ORD(T) GT 1).. cashland(T)- (LNDDPMT*LANDVAL(T)*decvar("ACBUY",T))=G=0; MACHREQT(T)$(ORD(T) GT 1).. EQCOST(T)*(decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J)))- SUM(J$(ORD(J) LE ORD(T)),DEPVALEQ(T,J)*decvar("EQBUY",J))=L=0; FINAMACH(T)$(ORD(T) GT 1).. (1-EQDPMT)*decvar("EQBUY",T) -decvar("MEDCRED",T)=E=0; MACHDWP(T)$(ORD(T) GT 1).. casheq(T)- EQDPMT*decvar("EQBUY",T)=G=0; FAMICONS(T)$(ORD(T) GT 1).. famcon(T)=E= FACONS(T); OFFARINV(T)$(ORD(T) GT 1).. offinv(T)-invest(T)=E=0; OFFARINC(T)$(ORD(T) GT 1).. offinv(T)*YIELD(T)-MOFFINC(T)=L=0;
124
OFFIDENT(T)$(ORD(T) GT 1).. offinv(T)*YIELD(T)- offarm(T)=E=0; TRANSF1(T)$(ORD(T) GT 1).. VARCOST(T)*(decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J))) +0.5*RENT(T)*decvar("ACRENT",T)+OVRHD(T) +0.5*PROPTAX*SUM(J$(ORD(J) LE ORD(T)), decvar("ACBUY",J)) +cashland(T) +invest(T) -endcash(T-1) +begcash(T)-decvar("SHTCRED",T) +XPRATE*famcon(T)+casheq(T) +(LABCOST(T)*(hirelab(T)/2))=E=0; TRANSF2(T)$(ORD(T) GT 1).. -RETURN(T)*(SUM(J$(ORD(J)LE ORD(T)), decvar("ACBUY",J))+decvar("ACRENT",T)) +0.5*RENT(T)*decvar("ACRENT",T) +0.5*PROPTAX*SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J)) +SUM(J$(ORD(J) LE ORD(T)),FPRIN20(T,J)*decvar("FINLAND",J) +PRINTMED(T,J)*decvar("MEDCRED",J)) +PRINTS(T)*decvar("SHTCRED",T) -invest(T)*INVGAIN(T) -SUM(J$(ORD(J)LE ORD(T)), EQDEP(T,J)*decvar("EQBUY",J)) +endcash(T) -begcash(T)+XPRATE*famcon(T) +INCTAX*txincm(T-1) +(LABCOST(T)*(hirelab(T)/2))=E=0; TAXINCOM(T)$(ORD(T) GT 1).. -MARGIN(T)* (decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J))) +PROPTAX*SUM(J$(ORD(J) LE ORD(T)), decvar("ACBUY",J)) +RENT(T)*decvar("ACRENT",T) +SUM(J$(ORD(J) LE ORD(T)),INTF20(T,J)*decvar("FINLAND",J) +INTMED(T,J)*decvar("MEDCRED",J))+INPMTS(T)*decvar("SHTCRED",T) +SUM(J$(ORD(J)LE ORD(T)), EQDEP(T,J)*decvar("EQBUY",J)) +LABCOST(T)*hirelab(T) -offarm(T) +txincm(T)=E=0; CURRAS(T)$(ORD(T) GT 1).. casset(T)-endcash(T)-OTHCUR0 -SUM(J$(ORD(J) LE ORD(T)), invest(T)) =E=0; INTAS(T)$(ORD(T) GT 1).. iasset(T)- SUM(J$(ORD(J)LE ORD(T)), DEPVALEQ(T,J)*decvar("EQBUY",J))=E=0; LNGTAS(T)$(ORD(T) GT 1).. lasset(T)- LANDVAL(T)*SUM(J$(ORD(J) LE ORD(T)), decvar("ACBUY",J)) -BUILD0 =E=0;
125
TOTAS(T).. tasset(T)-casset(T)-iasset(T)-lasset(T)=E=0; CUDEBT(T)$(ORD(T) GT 1).. culiab(T) -SUM(J$(ORD(J) LE ORD(T)), PRINTMED(T,J)*decvar("MEDCRED",J) +FPRIN20(T,J)*decvar("FINLAND",J)) -INCTAX*txincm(T)=E=0; INTDEB(T)$(ORD(T) GT 1).. intliab(T)- SUM(J$(ORD(J) LE ORD(T)),PROUTMED(T,J)*decvar("MEDCRED",J)) +SUM(J$(ORD(J) LE ORD(T)), PRMEDD(T,J)*decvar("MEDCRED",J)) =E=0; LNGDEBT(T)$(ORD(T) GT 1).. lgliab(T)- SUM(J$(ORD(J)LE ORD(T)), PROUTF20(T,J)*decvar("FINLAND",J)) +SUM(J$(ORD(J) LE ORD(T)), PRF20D(T,J)*decvar("FINLAND",J)) =E=0; TOTDEB(T)$(ORD(T) GT 1).. totliab(T)-culiab(T)-intliab(T)-lgliab(T)=E=0; CEQUITY(T)$(ORD(T) GT 1).. cureq(T)-casset(T)+culiab(T)=E=0; IEQUITY(T)$(ORD(T) GT 1).. inteq(T)-iasset(T)+intliab(T)=E=0; LEQUITY(T)$(ORD(T) GT 1).. lgteq(T)-lasset(T)+lgliab(T)=E=0; EQUITY(T)$(ORD(T) GT 1).. nworth(T)-tasset(T)+totliab(T)=E=0; MAXSDEBT(T)$(ORD(T) GT 1).. mscredit(T)-5*cureq(T) -MARGIN(T)*(decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J))) +RENT(T)*decvar("ACRENT",T)+LABCOST(T)*hirelab(T) =E=0; MAXIDEBT(T)$(ORD(T) GT 1).. micredit(T)-2*inteq(T) -2*SUM(J$(ORD(J) LE ORD(T)), PRMED(T,J)*decvar("MEDCRED",J)) -0.30*MARGIN(T)* (decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J))) +0.30*RENT(T)*decvar("ACRENT",T)+0.30*LABCOST(T)*hirelab(T) -2*SUM(J$(ORD(J) LE ORD(T)), DEPVALEQ(T,J)*decvar("EQBUY",J)) =E=0; MAXLDEBT(T)$(ORD(T) GT 1).. mlcredit(T)-2*lgteq(T) -0.12*MARGIN(T)* (decvar("ACRENT",T) -SUM(J$(ORD(J) LE ORD(T)),decvar("ACBUY",J))) +0.12*RENT(T)*decvar("ACRENT",T) +0.12*LABCOST(T)*hirelab(T) -2*SUM(J$(ORD(J) LE ORD(T)), PRF20(T,J)*decvar("FINLAND",J)) -.05*LANDVAL(T)*SUM(J$(ORD(J) LE ORD(T)), decvar("ACBUY",J)) =E=0;
126
TAXPAYMT(T)$(ORD(T) GT 1).. taxes(T)- PROPTAX*SUM(J$(ORD(J) LE ORD(T)), decvar("ACBUY",J)) -INCTAX*txincm(T)=E=0; INTREST(T)$(ORD(T) GT 1).. interest(T)-INPMTS(T)*decvar("SHTCRED",T) -SUM(J$(ORD(J) LE ORD(T)), INTF20(T,J)*decvar("FINLAND",J) +INTMED(T,J)*decvar("MEDCRED",J))=E=0; DEPN(T)$(ORD(T) GT 1).. SUM(J$(ORD(J) LE ORD(T)), EQDEP(T,J)*decvar("EQBUY",J)) -depre(T)=E=0; FSIZE(T)$(ORD(T) GT 1).. decvar("ACPROD",T)-UPSIZE=L=0; decvar.FX("ACBUY","T0")=LAND0; decvar.FX("ACPROD","T0")=362; decvar.FX("EQBUY","T0")=EQPT0; decvar.FX("FINLAND","T0")=350099; decvar.FX("MEDCRED","T0")=422010; decvar.FX("SHTCRED","T0")=241688; decvar.FX("ACRENT","T0")=362-LAND0; endcash.FX("T0")=cash0; casset.FX("T0")=CURAST0; iasset.FX("T0")=EQPT0; lasset.FX("T0")=BUILD0+(LAND0*2150); nworth.FX("T0")=829437; txincm.FX("T0")=20171; invest.FX("T0")=100000; cureq.FX("T0")=362428; inteq.FX("T0")=46890; lgteq.FX("T0")=420119; MODEL MEBR METHYL BROMIDE STUDY /ALL/; SOLVE MEBR MAXIMIZING WLTH USING LP;