A bioeconomic systems approach to sustainability analysis at the farm level

Material Information

A bioeconomic systems approach to sustainability analysis at the farm level
Kelly, Terry C
Publication Date:
Physical Description:
viii, 210 leaves : ill. ; 29 cm.


Subjects / Keywords:
Farm layout ( lcsh )
Farms -- Economic aspects ( lcsh )
Food and Resource Economics thesis Ph. D
Dissertations, Academic -- Food and Resource Economics -- UF
bibliography ( marcgt )
non-fiction ( marcgt )


Thesis (Ph. D.)--University of Florida, 1995.
Includes bibliographical references (leaves 197-209)
General Note:
General Note:
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
Statement of Responsibility:
by Terry C. Kelly.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator ( with any additional information they can provide.
Resource Identifier:
002051822 ( ALEPH )
33436645 ( OCLC )
AKN9781 ( NOTIS )

Full Text

-----Original Message-----
From: Kelly, Terry []
Sent: Tuesday, May 23, 2006 10:20 PM
To: Hildebrand,Peter E
Subject: RE: Your dissertation

Hi Pete

I happily give my permission for the Library to digitize my
dissertation--hope this is all you need for me.

I hope all is well with you. When are you going to come to New Zealand?
There's some great trout fishing here!


-----Original Message-----
From: Hildebrand,Peter E []
Sent: Wednesday, 24 May 2006 9:17 a.m.
To: Kelly, Terry
Subject: Your dissertation

Hi Terry. Now the library wants to digitize your dissertation so if you
agree please send me permission. Thanks. Pete








First, I want to express my appreciation for the unwavering support and sound advice

provided by Peter Hildebrand, my committee chair, who supervised my work, and from whom I

learned a great deal about Farming Systems Research and understanding farmers. I also owe a

special thanks to my committee cochair, Chris Andrew, for his guidance through the early parts

of my graduate program at Florida, and for his openness. Both of them contributed in very special

ways to my growth, both personally and professionally.

I want to thank the other members of my graduate committee, all of whom contributed to

my being able to complete this work. Clyde Kiker challenged me with his ideas and always

encouraged me to explore creatively. David Zimet was extremely insightful with his knowledge

of North Florida farming systems, helping to keep me focused on the reality of farming there. Jim

Jones in agricultural engineering spent countless hours helping me understand systems modeling

and assisting me with the application of the crop models. And Ken Boote in agronomy helped

greatly in the early conceptualization of this work, and always provided sound feedback and advice

in response to my ideas and questions. In addition, I want to express my gratitude to the Suwannee

County growers who gave of their time and wisdom in providing me with insights into their

production systems and way of life; and to Jim Hansen, a colleague in agricultural engineering,

who collaborated with me in the farmer interviews, and who expanded my concepts of

sustainability through hours of discussion.

Most importantly, I want to thank my family, who put up with the long and irregular hours

thatI had to keep, and with the poverty that goes with being a graduate student. My wife, Judith

Kidd, made countless sacrifices, both personally and professionally, during the time that this work

was in progress, and still managed to keep our family on track. I hope that someday I can make

it up to her. To my children, Nathan and Lisa, I owe a special thanks for helping to keep my life

in proper perspective, and reminding me that there is more to life than dissertations and qualifying

exams. Through all this, they prospered beautifully. Finally, I would like to thank my parents, who

never interfered, but were always there when they were needed.


ACKNOWLEDGMENTS ....................................................ii

ABSTRACT ..................... ......... .................... .......... vii

1. INTRODUCTION .............. ............................... .......... 1

1.1. Farming System Sustainability ...... ............................... 2
1.2. Research Problem ............................................. 3
1.3. Hypotheses ........ .......... ... ................................ 5
1.4. Objectives ............... .......................... ............ 5
1.5. General Modeling Approach ...................................... 6
1.5.1. Brief Description of the Model ............................. 6
1.5.2. Representing a Farming System.............................. 8
1.6. A Systems Approach to Sustainability Analysis ................ ......... 9
1.7. How this Study Proceeds ........................................... 12


2.1. What is Sustainability? ............. ....... ....... ................. 14
2.1.1. Genesis of Sustainability ............ ................. ...... 16
2.1.2. Review of Sustainability Definitions ................ ......... 18
2.1.3. Characterizing Sustainability ............................... 21
2.2. Sustainability: A Relationship Between Agriculture, Society, and Ecosystem ... 25
2.2.1. Obligation to Future Generations ............................. 25
2.2.2. The Shape of a Sustainable Society ........................... 30

3. SUSTAINABILITY AND ECONOMIC THEORY .............................. 34

3.1. Sustainability in the Neoclassical Paradigm ............................. 34
3.2. Temporal Implications of Sustainable Agriculture ....................... 39
3.3. Some Microeconomics of Sustainability .............................. 46
3.3.1. Valuing Natural Resources and the Allocation Problem ........... 47
3.3.2. Resource Substitutability ...................................50
3.3.3. Accounting for Externalities ............................... 52
3.4. Measuring Sustainability ......... .................... ....... ....... 57


4.1. Indicators of Farming System Sustainability ............................ 58
4.1.1. Productivity Indicators ................................... 59
4.1.2. Indicators of Stability and Resilience .......................... 60
4.1.3. Indicators of Equity ................. ................... 63
4.1.4. Summary of Sustainability Indicators .......................... 64
4.1.5. Modeling Indicators of Sustainability .......................... 65
4.2. Review of Modeling Approaches which Include Sustainability-Related Issues .. 67
4.2.1. Recursive Programming ................................. 67
4.2.2. Modeling Biological Processes ............................ 69
4.2.3. Integrating Economic and Biological Models ................... 70
4.3. An Integrated Whole-Farm Sustainability Model ........................ 72
4.3.1. Simulation Component Details ............................ 73
4.3.2. Economic Model Specification .............................74

TO A NORTH FLORIDA FARMING SITUATION ........................ 81

5.1. A Suwannee County Peanut Farming System ........................... 81
5.1.1. Defining a Hypothetical Peanut Farming System ................. 82
5.1.2. Production Costs and Expected Returns ....................... 88
5.2. A Suwannee County Peanut Farm Model............................ 92
5.2.1. Crop Simulation Components .............................93
5.2.2. The Economic Model ................................... 96
5.3. Results ..................................................... 106
5.3.1. Preliminary Crop Simulation Results ...................... 106
5.3.2. The Simulated Sustainability of a Suwannee County Peanut Farming
System ............................................ 109
5.3.3. Potential Farm-Level Effects of a Discontinuance of the U.S. Peanut
Program ............................................ 111
5.3.4. Constraining Nitrogen Leachate ............................ 115

6. DISCUSSION AND CONCLUSIONS ....................................... 120

6.1. Implications of the Model's Results ................................ 120
6.2. More General Implications of the Whole-Farm Modeling Approach ........ 124
6.3. Conclusions and Recommendations .................................. 128

PRODUCTION COSTS ............................................ 132


OUTPUT PRICE DETERMINATION ................................ 142

GAMS FORMULATIONS OFMP1 ANDMP2 ......................... 144




BIBLIOGRAPHY ....................................................... 197

BIOGRAPHICAL SKETCH ....................... ......................210

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Terry C. Kelly

May, 1995

Chairperson: Peter E. Hildebrand
Major Department: Food and Resource Economics Department

An integrated, whole-farm simulation model that links biophysical process models with

a multi-period recursive programming model of farmer decision-making is developed to address

farm-level sustainability questions. This model provides a framework with which to analyze the

effects over time of a farmer's decisions regarding cropping patterns, cultural practices, and

farming technologies as they relate to the sustainability of the farming system. Sustainability

definitions and concepts are reviewed, particularly as they apply to farm-level analysis, which leads

to a definition of sustainability that is operational at the farm level. The moral philosophy

underpinning notions of sustainability is discussed and ethical issues relating to the obligation to

the future are explored. Serious questions arise regarding the validity of standard discounting

techniques and economic evaluation methods when applied to environmental and natural resource

problems. Ethical issues complicate sustainability analysis, suggesting that moral ordering

techniques should be employed together with economic valuation methods.

The farm sustainability model developed in this study consists of an economic decision-

making component and a simulation component that models biophysical processes and sets up the

succeeding economic component. The IBSNAT crop models are used to simulate crop growth

and the fate of soil nitrogen. The economic model is a multi-period nonlinear programming model

with a chance-constraint on nitrogen leachate. The farm model is developed with a recursive

structure and is locally optimizing, allowing for modification to parameters based on feedback and

results from previous periods. An application of the integrated farm sustainability model to a

hypothetical North Florida peanut farming system is illustrated. The results suggest that the

farming system is sustainable by a productivity criterion, but may not be sustainable according to

resilience criteria, depending on government policy for economic viability. The results also show

that for a peanut-based farming system, nitrogen leachate can be reduced in certain cases without

significant income loss.


Sustainability has been at the forefront of agricultural and economic development

issues for more than a decade. The issue of agricultural sustainability, however, has been

current for over a century, since von Liebig's 1840 work on the chemical basis for plant

nutrition (Rodale, 1990). The resulting controversy over care for the soil, as well as debate on

other sustainability issues such as farm structure, weed control, and erosion, have continued

since then, varying in intensity as the times and climatic cycles dictated. Many of the current

soil conservation programs have their roots in programs resulting from concern for the high soil

losses experienced during the dust bowl. Popular concern for the consequences of our actions

on others, present and future, and on the environment crystallized 30 years ago with the 1962

publication of Rachel Carson's Silent Spring. A decade later, publication of The Limits to

Growth (Meadows et al., 1972) further established the peril of failing to account for the

consequences of our actions. Since then, numerous environmental organizations have been

founded, governmental agencies supporting a host of environmental regulations have been

established, and countless environmental resolutions have been passed at the numerous national

and international conferences which have been held to address issues related to the


Today there is still a real need to enable the operationalization of sustainability concepts

(e.g., Batie, 1989; Harrington, 1991; Ashraf, 1991; Hildebrand and Ashraf, 1989). Farmers,

researchers, and extension workers need a definition of sustainability that is scientific, open to

hypothesis testing, and practicable (Conway, 1991). A general methodology for sustainability

analysis does not exist at any level, though more progress has been made at some levels (e.g.,

the field level) than at others.

The effects of a nonsustainable agriculture, such as soil erosion and water pollution,

transcend farm boundaries and are felt at the watershed and higher levels. Yet, remedies to

problems ofnon-sustainability in agriculture have tended to focus on the plant/animal,

field/paddock, or enterprise levels. It is only recently that analysts have begun to focus their

research efforts in sustainability at the farming system level. Since production decisions are

made at the farm level, the importance of farm-level sustainability research should not be


1.1. Farming System Sustainabilitv

In this study, a sustainable farming system is described as a farming system that will,

currently and in the long-run, meet the reasonable goals and objectives of the multi-

generational farm family; and will satisfy short- and long-term societal goals regarding

environmental quality, food safety and security, and general quality of life. A sustainable

farming system comprises three components:

1. financial viability in the short and long run, requiring non-decreasing net value-
of-production trends over time, which in turn depend on the quality and quantity
of the farm's natural, human, and capital resources;

2. non-degradation of the more general (off-farm) environment requiring a minimal
(or zero) level of negative externalities emanating from the farm; and

3. equitable distribution of farm-induced benefits and costs within the farm
household, across farms and across generations, and between farm and non-farm

The first component encompasses a narrow, production oriented definition of

sustainable agriculture in that environmental and resource quality are considered only to the

extent that they affect agricultural productivity. Expanding the definition of sustainable

agriculture to include the second component considers the off-farm effects of farming practices

on others. A large number of sustainable agriculture proponents would be satisfied with a

definition that includes only these first two components. Including the third component,

however, allows the sustainable agriculture concept to encompass the more normative issues

that contribute to a higher quality of life for the rural sector and society as a whole. The third

component also gives explicit consideration to the rights of future generations to the

opportunities afforded them from a healthy and resource-rich environment.

1.2. Research Problem

There is a perceived danger that if sustainability is not soon made more operational, the

public's concern for sustainable agriculture will diminish and existing momentum may be lost.

In the general rush to jump on the "sustainability bandwagon", sustainability is courted and

promoted from many different quarters, often in support of a particular ideology. While moral

convictions are laudable, they cannot substitute for rigorous analysis (Batie, 1989). Along with

the need to implement sustainable practices comes the need for an approach to analyze

practices which is based on sound science, and which quantifies, whenever possible, factors

influencing and indicating sustainability.

Economic theory identifies the socially optimal level of production as the point where

declining marginal social benefits equal rising marginal social costs. It also suggests that the

welfare gain to society from control of externalities is most cost-effective when decision makers

(i.e., farmers and researchers) have the most information and flexibility to make optimal

decisions (within the constraints of the "greater societal good") in response to their individual

production relationships and environmental conditions. The ability to evaluate the sustainability

of farming systems, practices, and technologies contributes vital information necessary to be

able to approach this social optimum in the most cost-effective manner.

The paucity of information and methodologies for evaluation of the sustainability

aspects of agricultural technologies and farming systems frequently results in suboptimal

decisions on the part of farmers, researchers, and policy makers. This leads to two interrelated

specific problems that are addressed in this study:

1. Researchers and analysts, and even farmers, from their differing perspectives, do
not consider the full range of sustainability issues and interactions in the
development and evaluation of agricultural technologies; and

2. Farmers do not have adequate information to make optimal management
decisions, especially for the long-run sustainable operation and development of
their farming systems.

This research addresses these problems using an integrated, whole-farm simulation approach

which links biophysical process models to an economic behavior model. Initially, the ideas,

concepts, and results derived in this study are likely to be most useful to those who are doing

research on the sustainability of agricultural systems and technologies. Ultimately it is hoped

that the concepts and methods will prove useful to those advising farmers and policy makers,

and that the approach eventually may be used by farmers themselves. Ruttan (1992) suggests

that at present, and for as long as three crucial sustainability issues remain unresolved (resource

substitutability, obligations to the future, and institutional design), sustainability is more

appropriately viewed as a guide for research rather than as a guide for practice.

In the development of a sustainability, whole-farm model, two hypotheses are

presented for investigation in this research:

Hypothesis 1: Inclusion of well-defined sustainability criteria in the farm-level decision
problem will significantly alter the optimal farm plan.

Hypothesis 2: The biophysical process models are sufficiently well-developed and
robust to be incorporated into a whole-farm decision model to facilitate the
investigation of long-term resource and sustainability problems.

Criteria are developed to measure or account for the productivity of the farming system over

time, its stability and resilience to both anticipated and unanticipated shocks, and for

externalities from farm production such as chemical leaching into groundwater. Parametric

programming and sensitivity analysis are employed to determine the effects of applying

sustainability criteria at varying levels.

1.4. Objectives

The overall goal of this research is to develop and demonstrate a means to assess

sustainability at the farm level. This necessitates defining sustainability criteria, which in turn

requires an understanding of sustainability concepts and issues, of sustainable processes, and of

farming systems. The analytical ability to measure and quantify sustainability depends on how

sustainability is conceptualized and defined (Harrington, 1991).

This research goal is accomplished through the following objectives:

Objective 1: Review and synthesize the numerous definitions of sustainability which
currently exist in the literature, and provide an operational concept of sustainability
which is applicable to farm-level analysis.

Objective 2: Develop a holistic, farm-level systems model for integrating societal and
farm-level goals and concerns regarding sustainability, which can be used to assess
sustainability issues and hypotheses at the farm level.

Objective 3: Apply this model to a hypothetical representative North Florida farming
system to investigate particular sustainability issues in the whole farm context.

1.5. General Modeling Approach

An integrated, whole-farm simulation model which links biophysical process models,

particularly crop growth models, with a multi-period recursive programming model of farmer

decision-making is developed to address farm level sustainability questions. This model

provides a framework with which to analyze the effects over time of a farmer's decisions

regarding cropping patterns, cultural practices, and farming technologies as they relate to the

sustainability of the farming system. Based on the definition of a sustainable farming system

provided earlier, consideration is given to effects on crop yields, farm income, the farm's

financial position, quality and quantity of water and soil resources, and on the farm's relation to

the broader community. Long-run and short-run considerations are included.

1.5.1. Brief Description of the Model

The dynamic farm sustainability model developed here is adapted from a stochastic

simulation model developed by Crawford (1982; also see Crawford and Milligan, 1982). A

diagrammatic representation of the sustainability model is presented in Fig. 1.1. The farming

year is divided into two decision periods. Period 1 is the period of spring and summer

production activities, including land preparation, planting, cultivating, and fertilizer application.

Period 2 includes harvest and marketing activities, winter- and cover-crop planting, and other

winter activities. Each decision period is further divided for the purpose of tracking cash flows,

and accounting for labor and equipment requirements.

Optimal production and marketing decisions for the year are simulated with two multi-

period programming (MP) models. In MPI, production decisions are made in period 1 based

Expected yields and
prices for year t

Simulate crop growth and resource use for period 1
and transfer inventories to period 2

Go to
next year
t = t+1

Figure 1.1.

Yes Stop simulation

Flow chart of farm sustainability model components and decisions, where t
represents year, t = 1,...,T, with T being the terminal year of the simulation.


Harvest, marketing, and winter decisions
Period 2t i Period 1t+1 Period 2t+1

Simulate winter crop growth and resource use for period 2,
compute year-end income and inventories,
make consumption and saving decisions, and
transfer inventories to year t+1

Simulated "actual"
yields & prices for year t,
expected yields & prices
for year t+1

on expected yields and prices in period 2. In period 2, harvest and marketing decisions are

made (MP2) according to "actual" yields and current and expected future prices. The two

simulation components, SIM1 and SIM2, generate crop yields and prices, and derive expected

prices and yields. SIMI calculates "actual" yields and prices and sets up MP2 by replacing

average-year coefficients with simulated values. SIM2 allocates surplus income to

consumption and savings, calculates final inventory balances to transfer to the next year, and

derives expected yields and prices for the next year.

1.5.2. Representing a Farming System

In this study, a "synthetic" farm approach is taken for the analysis, as opposed to a

typical farm, which strictly speaking, is a modal concept, or to a representative farm, which

denotes some weighted average (Feuz and Skold, 1991). A synthetic or hypothetical farm was

created from data and information gathered through study of individual similar farming systems

in a target area, and through discussions with extension personnel and suppliers of inputs and

farm credit. No attempt is made to claim that the synthetic farm is either typical or

representative, though it does fairly represent the farming situation in the study area. Given the

small number of farmers studied (seven in this instance) and the fact that they are all unique,

any one or none could be considered "typical." Further, using average data potentially biases

the results, particularly in farm-level analysis, since by averaging, both production and

production possibilities are generally overstated (Feuz and Skold, 1991). Averaging implicitly

assumes that excess resources on one farm are automatically available for use on other farms,

which is not usually the case.

1.6. A Systems Approach to Sustainabilitv Analysis

Since sustainable agriculture encompasses environmental, productive, economic, and

social concerns, its analysis is best served by a systems approach that accounts for component

interactions, and that can trace the consequences of an intervention through the entire system.

While in "high input" farming, interventions may disregard some important system interactions

and still obtain high yields (Edwards, 1989), such disregard is not possible in "sustainable"

agriculture where these interactions are key to decision making. A strong systems perspective

is a distinguishing feature of sustainable agriculture.

Spedding (1988) defines a system as

a group of interacting components, operating together for a common purpose, capable
of reacting as a whole to external stimuli: it is unaffected directly by its own outputs
and has a specified boundary based on the inclusion of all significant feedbacks. (p. 18)

While systems can be conceptualized at any of many scale levels, a systems approach defines a

specific scale of interest and an appropriate boundary of analysis. The systems approach to

analysis is marked by recognition of the whole system and the interactions within that system

rather than looking only at a system part. A systems approach employs specific techniques and

tools, such as rapid appraisal, pattern analysis, diagrams, and modeling, often in a

multidisciplinary fashion, to identify system boundaries and recognize component interactions.

Systems may be grouped together to form a larger system, and each system is itself

made up of component subsystems. This structure is referred to as the systems hierarchy.

Each level in such a hierarchy has a distinctive behavior which is related to the behavior of the

other levels, but not necessarily in a straightforward manner (Conway, 1985, 1991). A system

possesses properties inherited from its subsystems as well as emergent properties which do not

exist in any of its subsystems. Additionally, over time, a system is affected and shaped by

suprasystems within which it is encompassed. While a system inherits all properties of its

subsystems, many of these inherited properties lose some of their relevance as they are

embodied in properties of higher systems. As such, these properties are not readily understood

through the study of other levels. Consequently, each system level invites analysis in its own

right (Conway, 1985, 1991).

Much of the confusion in discussions of sustainability stems from the lack of

recognition of appropriate system levels (Lynam and Herdt, 1989). Choosing the system level

at which sustainability becomes a relevant characteristic creates the problem of boundary

specification. Systems analysts must be able to identify the system's boundaries, and how these

boundaries are positioned, depending on the objectives of the analysis. Although system

boundaries are artificial in the sense that they are really "bands of grey" within which factors

have diminishing effect on system behavior (Rountree, 1977), they are necessary for the utility

of the analysis.

Conway denotes an ecological system with a human management component as an

agroecosystem. In the agroecosystem hierarchy, the lowest level is the individual plant or

animal together with its immediate micro-environment and the people who tend it. The next

level is the crop or herd contained within a field or paddock, and these, alone or in various

combinations comprise a farming system. A farming system combined with non-farm activities

is considered to be a household system or livelihood system. A group of livelihood systems,

some with farming systems and some without, comprise a community or watershed system,

which is part of a regional agroecosystem, and so on.

A characteristic of system sustainability is that the system retain the integrity of its

organizational structure over time; that is, system parameters remain stable (Common and

Perrings, 1992). In this sense, sustainability at any level in the systems hierarchy is dependent

upon the sustainability of the levels above it This means that criteria used to evaluate

sustainability at lower levels must be compatible with criteria at higher systems levels. On the

other hand, while system sustainability is affected by the sustainability of lower systems levels,

it is not necessarily dependent upon their sustainability. In other words, a system will not likely

be sustained within a larger unsustainable system, though it can be sustainable even if one or

more of its components is not For example, a farm system is not sustainable within a

watershed system which is nonsustainable, but the converse is possible. A system which might

be unsustainable standing on its own may in fact be sustainable when combined with other

systems to form a larger sustainable system. For example, a crop production system with high

yield variability, like watermelons, might be unsustainable if produced alone, yet could be an

integral and necessary part of a larger sustainable farm.

It is not necessary (nor necessarily desirable) for the sustainability of a system to insist

on the sustainability of all system components, because of substitution possibilities among

resources (Harrington, 1991). This ignores, however, the normative question of whether it is

desirable to have unsustainable subsystems within a larger sustainable system. This is

particularly an issue in a situation where an agricultural system may be sustainable but

individual farms within it are not.

The implications of sustainability relationships across hierarchical levels are important

to the development of a farm-level sustainability model. A farm is not truly sustainable if the

greater systems in which it operates are not sustainable. Carrying this notion to the extreme, it

can be argued that the appropriate spatial scale for discussing sustainable agriculture is global

because of the possibilities for trade and migration, and the fact that most farmers are

connected to more distant markets through trade (Crosson, 1993). However, the implications

of this argument for research, intervention, and change preclude a thorough understanding of


the properties of the other system levels where the intervention and change will actually occur.

As noted earlier, each level warrants discussion and analysis, and the farm level is the level

where research seems most lacking, yet where intervention holds the most promise for effective

change. Of course, a whole farm model must account, to the extent possible, for the effects

from higher systems, as well as for the fact that a system inherits the characteristics of those

systems below it.

1.7. How this Study Proceeds

Analysis in this research is confined primarily to the farming system and, when

necessary, its immediate hierarchical neighbors. The study continues in chapter two with a

review of the sustainability literature, particularly as it applies to farm system analysis, and the

moral philosophy underpinning notions of sustainability. This review leads to a definition of

sustainability that is operational at the farm-level, and which contributes to (and is indeed

necessary for) the development of an integrated whole-farm model. This is followed in chapter

3 with a discussion of sustainability and economic theory, which includes a critique of the

neoclassical paradigm with respect to sustainability issues, particularly considering the moral

philosophical base of sustainability discussed in chapter 2. Chapter 3 concludes with a review

of the utility of environmental economics for the resolution of three crucial issues,

intertemporal equity, resource substitutability, and the externality problem.

Chapter 4 presents a conceptual model incorporating sustainability in the decision

criteria at the farm level. An integrated, whole-farm simulation model which links biophysical

process models to an economic, behavioral model is developed as a general tool for evaluation

of farming system sustainability. An attempt is made to make this model as general as possible

to be applicable to a wide range of farming systems, and the full range of sustainability issues


pertinent at the farm level. This is followed in chapter 5 by a specification and application of an

integrated whole-farm sustainability model to a North Florida farming situation. A hypothetical

peanut-based farming system is analyzed as to its sustainability, primarily as a test of the

conceptual model.

A discussion of the results and the implications of the whole-farm modeling approach

for sustainability analysis are presented in chapter 6. Strengths and limitations of the model are

examined in light of the sustainability issues explored in the earlier chapters, and the model is

evaluated for its ability and potential ability to measure indicators of sustainability. The study

concludes with a discussion of the potential of the modeling approach to analyze farm-level

sustainability and the need for further development of analytical capabilities in farming system



2.1. What is Sustainabilitv?

At its most basic level, sustainability is simply the ability to continue, to sustain. As

Robert Rodale (1990) put it, there is no middle ground on sustainability-something is either

sustainable or it is not sustainable. By definition, something cannot be half sustainable. While

the simplicity of this is attractive, and serves as a reminder from where sustainability comes, the

uncertainties as to what will happen, even tomorrow much less in years down the road,

preclude such a definitive concept of sustainability. In ex ante analysis, it is necessary to

consider the likelihood of a system being sustainable, and therefore more useful to think in

terms of degrees of sustainability.

Further, the dynamism of the world does not allow sustainability to be so simply

defined. Systems are constantly changing and adapting. What was sustainable yesterday often

is not sustainable today; an obvious example of this is the swidden (slash and bum) system of

agriculture. A system can only be considered sustainable within a particular context and in a

certain form. And the context and form are likely to change over time. Consideration must be

given to which system parameters are allowed to change and which are not.

In addition to the time dimension embodied in sustainability, the definition of

sustainability is itself temporal, changing according to "stages of development." In early stages,

the notion of sustainable growth is emphasized, as in the case of developing country


agriculture. As agriculture develops and higher production levels are achieved, the concept of

sustainability becomes agroecological, with more emphasis on preservation and less on growth.

Ultimately, in this development process, food and fiber production systems are achieved which

are healthy, efficient, and socially and environmentally sound. This way of looking at

sustainability generally mirrors individuals' attitudes toward the environment: as basic needs are

increasingly satisfied, concern for the environment increases. The importance of a temporal

concept of sustainability is that notions of sustainability early on in the developmental time-line

may be viewed not as ends, but as short-term steps toward an ultimate goal of a sustainable


These issues stress that, if the concept of sustainability is to be operationally useful, it

must be defined contextually, within certain bounds, and be sufficiently narrow in focus,

depending on the system's level of the analysis. A general global definition has very little

practical use at the farm level of analysis, for instance. On the other hand, narrow concepts of

sustainability, while useful, may easily miss more economical and environmental solutions to

farm problems (Keeney, 1990). A balance between broad and narrow needs to be struck when

characterizing sustainability, and the level of analysis must be defined clearly.

Further complicating any attempt to define sustainability or sustainable agriculture is

the fact that these terms have come to embody a host of normative issues not easily subsumed

under ecological or even productive notions of sustainability. Wendell Berry simply, yet

eloquently states that "a sustainable agriculture is one which depletes neither the people nor the

land"'. However, the ramifications of this definition are not simple at all. A further discussion

'From a flyer distributed by the International Alliance for Sustainable Agriculture. A similar definition
attributed to Berry is in the preface to Meeting the Expectations of the Land (Jackson et al., 1984).

of broad sustainability concepts and issues is useful for better understanding and appreciating

the context in which farm-level sustainability is characterized.

2.1.1. Genesis of Sustainabilitv

Part of the reason that it has been difficult to define sustainability satisfactorily for most

is that a useful definition of sustainability must be able to accommodate the several traditions

encompassed under its banner, namely the organic agriculture tradition, the land stewardship

movement, and the agroecology perspective, among others (Ruttan, 1988). The roots of

sustainability come from diverse and valid, yet often incompatible strains of thought (Kidd,


Concerns that human activity can degrade the entire planet go back at least two

centuries (Kidd, 1992). Benjamin Franklin stated that, "Whenever we attempt to amend the

scheme of Providence, and to interfere with the government of the world, we need to be very

circumspect less we do more harm than good." (quoted in Kidd, p. 7). In 1864, G. P. Marsh

wrote, "The scale of change initiated by man is no longer local, but global. The climatic and

hydrological effects of deforestation provide an example." (Kidd, p. 7).

The term sustainability first appeared as a major theme in a controversial book,

Blueprintfor Survival, a 1972 publication by the editors of the British journal, The Ecologist

(Kidd, 1992). The book's controversy stemmed from the fact that it was both utopian and

apocalyptic, and it was emotive and sensational in its presentation. Perhaps one reason that

"sustainability" is contentious, hard to define, and conceived of so differently by different

people is that it was first stated as a major goal of society in the polemical rather than the

academic literature (Kidd, 1992). It began controversially, and remains so today.


Early definitions ofsustainability came out of the limits-to-growth literature, and were

expressed in terms of a transition from growth to attaining a steady-state. Rather than being

narrow, ecologically bound concepts, the first uses of sustainability in the United States were

broad political and ethical statements justifying a no-growth society. The ecological component

was an integral part of this broader philosophy (Kidd, 1992). As the concept of sustainability

developed in the mid and late 1970s, no-growth was replaced by sustainable growth, especially

since many countries, particularly developing countries, could not accept a no-growth


Sustainable agriculture has its intellectual roots in organic agriculture (William

Lockeretz, communication via the Intemet, March 11, 1994). It was first used in today's

context in the late 1970s. In 1977, a conference in Switzerland of the International Federation

of Organic Agriculture was titled, "Towards a Sustainable Agriculture" (William Lockeretz,

communication via the Intemet, March 11, 1994). Wes Jackson probably fathered the

contemporary use of the term with his 1978 publication, "Toward a Sustainable Agriculture"

(Greg McIsaac, communication via the Intemet, March 9, 1994).

Sustainable agriculture was popularized in part by dissatisfaction with the term organic

agriculture and other terms common in the 1970s, such as biological farming and eco-

agriculture, terms which carried significant negative baggage among mainstream

agriculturalists (Youngberg et al., 1993). As circumstances changed dramatically in the 1980s

with the adverse side effects of conventional agriculture becoming widely recognized,

sustainability became a powerful symbol around which diverse interests were drawn.

Sustainability added legitimacy to and thus advanced low-chemical input approaches as it drew

nearly all agricultural interests within its embrace. Youngberg et al. note that "after several

decades of accelerating conflict over the future of American agriculture, sustainability has


provided a comforting, although largely unanalyzed, symbolic refuge for an incredibly disparate

array of agricultural interests." (p. 300). Sustainability has come to be the most accepted term

for acknowledging the need for basic changes in agriculture (Dahlberg, 1991). It implies that

the "...wisdom of current policies and of proposed actions should be assessed in terms of their

full long-range effects." (Kidd, 1992, p. 3).

One of the driving forces behind the sustainability movement in its infancy was the high

oil prices of the early and late 1970s, and the realization that an agriculture that was dependent

on petrochemicals and diesel fuel was not sustainable. In the time since then, however, farmers

have shown themselves very adept at using energy more efficiently, with the result that the

amount of energy used per unit of output has declined in American agriculture (Crosson,

1991). With increasing energy efficiency in agriculture coupled with the prevailing view that a

disruption of world oil flows is tantamount to an act of war, it is unlikely that energy costs will

pose a threat to agricultural sustainability in developed countries any time soon.

2.1.2. Review of Sustainability Definitions

Sustainable agriculture and the broader sustainable development have been

characterized in a number of ways; perhaps in as many ways as there are people who care to

conceptualize them. The term sustainable development was popularized by the World

Commission on Environment and Development (often referred to as the Brundtland

Commission after its chair, former Norwegian Prime Minister, Gro Harlem Brundtland), whose

definition has been widely cited and has become a standard: "Sustainable development seeks to

meet the needs and aspirations of the present without compromising the ability of future

generations to meet their own needs." (WCED, 1987, p. 43). Further, the Brundtland

commission contended that sustainable development "requires that the rate of depletion of non-

renewable resources should foreclose as few future options as possible." (WCED, p. 46).

Pearce et al. (1990) describe sustainable development as a situation in which a vector of

development characteristics does not decrease over time. Included in this vector are 1)

increases in real per capital income, 2) improvements in nutrition and health, 3) educational

achievement, 4) access to resources, 5) a more equitable distribution of income, and 6)

increases in basic rights and freedoms.

Sustainability in agriculture is typically defined as a set of characteristics contributing

toward a set of goals. For some people, the characteristics are important (i.e., minimum- or

no-till, leguminous crop rotations, integrated plant and animal systems, etc.), while others stress

what sustainable agriculture will do (i.e., reduced pollution and erosion, enhanced quality of

life, longevity of the system, etc.). For some people, sustainability is a philosophy, a moral

vision. For example, Arnold (1989, p. 21) considers sustainability as an "evolving vision

representing the coming together of a variety of different concerns, disciplines, and political

pressures." This vision encompasses among other things, 1) a long-term perspective, 2) a high

degree of uncertainty about both the value and functions of the natural environment for the

social system, and 3) a strong concern that the present market system seriously undervalues

many items essential to support sustainability. Francis and Youngberg describe sustainable

agriculture as

a philosophy based on human goals and on understanding the long-term impact of our
activities on the environment and on other species. Use of this philosophy guides our
application of prior experience and the latest scientific advances to create integrated,
resource-conserving, equitable farming systems. These systems reduce environmental
degradation, maintain agricultural productivity, promote economic viability in both the
short and long term, and maintain stable rural communities and quality of life. (quoted
in Flora, 1992, p. 38)

Stressing the stewardship concept, Strange (1984, p. 118) contends that, to be

sustainable, agriculture "must be organized economically and financially so that those who use

the land will benefit from using it well and so that society will hold them accountable for their

failure to do so." Others focus on productivity. Sustainability of agriculture means that it

remains the dominant land use over time, that its resource base be able to continually support

the necessary production levels for profitability and/or survival (Hamblin, 1992), and that the

net benefits that agriculture provides to society can be maintained for present and future

generations (Gray, 1991). Dobbs et al. (1992) refer to sustainable as "staying power," and they

go on to say that

unless decisions result in mankind surviving over the long term--able to live in an
environment with (1) non-degrading natural resources, (2) adequate food and incomes
to meet the basic needs of all people, and (3) human organizational/political institutions
that enable people to live in harmony--an agricultural production system will not be
"sustainable." (p. 1)

Official definitions of sustainable agriculture, because of their political ramifications,

tend to be all-inclusive, and, therefore, not particularly useful. The official U.S. Government

definition of a sustainable agriculture appeared in the Food, Agriculture, Conservation, and

Trade Act of 1990 as

an integrated system of plant and animal production practices having a site-specific
application that will, over the long-term, satisfy human food and fiber needs; enhance
environmental quality and the natural resource base upon which the agricultural
economy depends; make the most efficient use ofnonrenewable resources and on-
farm/ranch resources and integrate, where appropriate, natural biological cycles and
controls; sustain the economic viability of farm/ranch operations; and enhance the
quality of life for farmers/ranchers and society as a whole. (Title XVI, Subtitle A,
Section 1603)

The American Society of Agronomy defines sustainable agriculture as

one that, over the long-term, enhances environmental quality and the resource base on
which agriculture depends, provides for basic human food and fiber needs, is
economically viable, and enhances the quality of life for farmers and society as a
whole." (Neher, 1992, p. 53).

2.1.3. Characterizing Sustainability

Certain concepts emerge from these "definitions" of sustainability. One, there is a time

dimension--sustainability is concerned with the future as well as the present Two is the

concept of growth-as long as populations are increasing, agriculture cannot be sustainable

without increasing output. Three, there are ecological limits to growth: maximum sustainable

yields, carrying capacities, and assimilative capacities. Four, sustainability carries with it some

notion of equity, both intra- and inter-generational. Five, most characterizations of

sustainability contain an intrinsic value of nature. And finally, the systems approach inherent in

sustainable agriculture implies the involvement of multiple disciplines.

Batie (1989) contends that most interpretations of sustainability are encompassed

within two general definitions; one is the constrained growth definition, and the other is a

maintenance of the resource base definition. The first is a constrained maximization concept

while the other is a minimization concept that implies minimizing the use of the natural

environment. To better understand this distinction, O'Riordan and Turner distinguish four basic

world views concerning the environment (Turner, 1988, p. 1 [underlines added]):

(a) 'comucopian' technocentrism: an exploitative position supportive of a growth ethic
expressed in material value terms (e.g. Gross National Product); it is taken as
axiomatic that the market mechanism in conjunction with technological innovation
will ensure infinite substitution possibilities to mitigate long-run real resource

(b) 'accommodating' technocentrism: a conservationist position, which rejects the
axiom of infinite substitution and instead supports a 'sustainable growth' policy
guided by resource management rules;

(c) communalistt' ecocentrism: a preservationist position, which emphasizes the need
for prior macroenvironmental constraints on economic growth and favours a
decentralized socio-economic system;

(d) 'deep ecology' ecocentrism: an extreme preservationist position, dominated by the
intuitive acceptance of the notions of intrinsic (as opposed to instrumental) value in
nature and rights for non-human species.


While any such categorization is problematic (one can easily find his or her views expressed in

two or even three of the above groups), the distinctions are useful in understanding the different

interpretations ofsustainability. Turner also contends that a greater recognition and acceptance

of these different environmental rationalities will aid effective environmental decision-making.

Douglass (1984) suggests that sustainability can be conceptualized in three different

ways: 1) sustainability as long term food sufficiency, 2) sustainability as community, and 3)

sustainability as stewardship. These roughly coincide with the notions of productivity, equity,

and an environmental ethic. Again, as with Turner's distinctions, these are not mutually


A useful way of viewing the distinctions within the breadth of sustainability

characterizations is by the relative importance of people accorded in the definition; that is,

where the definition falls on the anthropocentric-ecocentric continuum. Among those who

subscribe to what is essentially an anthropocentric position, a further distinction can be made

between those who view people as important as individuals versus those who view people as

important in the aggregate. A common position in the sustainability debate is one that holds

that agricultural output must increase to accommodate growing populations if agriculture is to

be sustainable, yet rejects any notion of equity or distributive justice in the provision of that

output to populations. On the other hand, there are those who contend that equity

considerations are essential to sustainability. In the sustainability characterizations reviewed

above, most give some importance to the equity issue, and most contain some aspects of

ecocentrism, while still being primarily anthropocentric.

Another distinction, related to those above, is the relative importance one attaches to

social and economic issues in the sustainability discussion. Some argue that, although

economic and social factors and processes are important to sustainability, ecological


sustainability is the base upon which the others depend (Gliessman, 1990). Gliessman assumes

that sustainability "can be achieved in an agriculture that is ecologically sound, resource-

conserving, and not environmentally degrading." (p. 367). These sentiments are echoed by

Crews et al. (1991) who argue that the sustainability of agricultural systems is "constrained

solely by the ecological conditions of agriculture." (p. 146). They assert that "sustainability is a

measure of a system's potential to endure and is not the proper yardstick with which to measure

the desirability of a particular set of social relations." (p. 148). They contend that, rather than

constraining the sustainability of a system, social and economic structures must be examined

and, if necessary, adjusted so that ecologically sustainable agriculture is also socially and

economically sound. To the extent that social relations should not remain fixed, Allen (1993)

would probably concur. However, while ecological sustainability may be the base of

sustainable agriculture, it is not sufficient for sustainable agriculture. Guaranteeing the

ecological integrity of agriculture does not guarantee its sustainability. It must be remembered

that agriculture is a human endeavor, created by humans for humans, and its sustainability

cannot be isolated from the human element (Allen, 1993). It is likely that several different

forms of agriculture may be ecologically sustainable in any one context, but that fewer will be

acceptable socially and economically. It only makes sense to pursue those that are sustainable

from both an ecological and a socioeconomic perspective, rather than try to adjust social

systems to fit ecological sustainability. While long-term ecological sustainability is important, if

a system is not sustainable in the short run as well, it certainly cannot be sustainable in the long

run because it will not endure to the long term.

In light of this distinction, it is useful to distinguish sustainable agriculture from

agricultural sustainability, terms often used synonymously. Sustainable agriculture, as noted

above, embraces positive and normative concerns, relying to a significant degree on values and

social considerations. Its proponents do not argue that theirs is the only way that agriculture

can sustain, but rather that theirs is the best way for it to sustain. Few doubt that conventional

agriculture can survive--agriculture (in the large sense) has proven to be remarkably resilient

and able to change when it has had to. However, many do not like the scenarios of a

continuation of conventional agriculture. Agricultural sustainability, on the other hand, refers

simply to a system's ability to endure. In theory, agricultural sustainability need not depend on

values or social institutions. There are probably a number of ways which agriculture could be

organized to be sustainable or enduring. The debate is over which paths first and most

importantly avoid unsustainability, then which is the best path to take toward sustainability.

The first part of the question is more for science to decide, based on current knowledge and

best estimates of the future. The second part is normative. Sustainable agriculture

encompasses both.

As the meaning of sustainability is analyzed, certain contradictions and potential

conflicts appear. To some, growth and sustainability are incompatible. Growth is the

quantitative increase in the scale of physical dimensions of society, while sustainability implies

the qualitative improvement in the structure, design, and composition of society. Another

contradiction inherent in the concept of sustainability is that if the market cannot be relied on to

sustain the environment, then society is forced to rely on policies and international agreements,

without which individual and national dominance will dictate the terms of societies'

development (Redclift, 1987). Still other areas of contradiction include: 1) the differing

environmental objectives of individuals and across regions, and particularly between the

affluent and the impoverished; 2) the current productivity of food systems which is often

environmentally damaging, yet food is scarce in many areas; 3) the debate on technological

advancement vs. societal adjustment as the answer to sustainability questions; 4) the conflict

between growth and conservation; and 5) the familiar growth with equity dilemma.

Unfortunately, sustainable agriculture is too often perceived as describing the conflict between

agricultural productivity and the environment.

2.2. Sustainabilitv: A Relationship Between Agriculture. Society. and Ecosystem

Essentially, sustainable agriculture defines the relationship between farming and

society, and then with the greater ecosystem. The different views of sustainability have

different implications for these relationships. While justice to future generations is the main

moral issue in sustainability, distributive equity within present generations and justice to non-

human species are also sustainability concerns. Questions as to whether the present has an

obligation to the future, and whether humans have obligations to non-human species cannot

easily be answered directly because of the particular time-space scale humans have chosen to

describe the world in which they live (Giampietro and Bukkens, 1992). Yet these questions are

basic to a sustainability ethic.

2.2.1. Obligation to Future Generations

With respect to agriculture, intergenerational equity means that "...each generation must

provide subsequent generations with the opportunity to engage in agricultural production at

acceptable economic and environmental costs" (Crosson, 1993, pp. 38-39). In terms of social

capital, which consists of all the natural and human-made resources needed to produce the

goods and services valued by people, a sustainable agricultural requires that adequate supplies

of good quality energy, land, irrigation water, plant genetic material, climate, and knowledge

embedded in people, technology, and institutions, be passed from one generation to the next

(Crosson, 1993).

The essence of what is now called intergenerational equity was spelled out more than

80 years ago (and probably earlier) when Harvard Professor, Nathanial Shaler, wrote,

We may be sure that those who look back upon us and our deeds from the centuries to
come will remark upon the manner in which we use our heritage, and theirs, as we are
now doing, in the spend-thrift way, with no care for those to come. (quoted in Kidd,
1992, p. 7)

Intergenerational equity figured strongly in the early development of the sustainability concept.

The first time that the term sustainability appeared in a U.N. document was in a 1978 report

listing causes of environmental degradation, which noted that

a third type of threat to environmental quality arises from the satisfaction of immediate
requirements (often non-essential) at the expenses of long-term ones (often basic).
Sustainable development means that the needs of present and future generations must
be appropriately reconciled. (Kidd, 1992, p. 17)

Obligation to the future, particularly in western culture, has only recently become an

issue in the moral philosophy literature (Baier, 1984). Its concern has come about because of

our increasing power to perform acts with huge foreseeable consequences for future

generations, and our increasing ability to trace these consequences. The nature of the present's

obligation to the future defines the vision of sustainability, and may be cast in either a utilitarian

or a rights perspective.

Utilitarians see as crucial the effects of an action on peoples' happiness; their decision

rule is based on the notion of the greatest good. The classical utilitarian maximizes the sum of

present and future utilities, discounting each generation by the probability that it doesn't exist.

Neoclassical utilitarians maximize the sum of expected values of own utility, and, not knowing

to which generation one belongs, weighing each generation by the probability of being in that

generation. Generations are discounted for likelihood of extinction. A problem is that these

probabilities of extinction are treated as exogenous, unaffected by the present generation's

actions, which plainly is not true (Page, 1983).


In contrast to utilitarians, those who hold a theory of rights view the crucial question to

be the effect on people (or other species) as rights' holders. Their decision criteria are based on

the principle of non-violation of these rights. The intertemporal problem requires the

identification of the rights of future generations, which is not a straightforward task. Looking at

rights and responsibilities in the context of the future is complicated by the overlapping nature

of generations. It is difficult to delineate present from future. Young children may have a

greater interest in the future than would senior citizens, yet both are part of the present.

Delineating a "present" society that is distinct from a "future" society is problematic.

Given that future persons have rights that the present must respect in their actions, there

is a danger that the consequences of acting on those rights may place a tremendous burden on

the present, for example, by prohibiting the use of non-renewable resources (Burkhardt, 1989).

While resolution of situations of conflicting rights is messy (in this case, who has the right to

use the non-renewable resource, present persons or future persons?), one could appeal to the

"priority principle" (Shue, 1980) in which "basic" rights take precedence over "non-basic" rights

and individual preferences. The rights of future persons, in this situation, could require the

present to develop technologies which make the resource less critical to future persons'


While moral theory can describe present obligations to the future, and can assign rights

to future generations, practical treatment of intertemporal issues is not straightforward. In

conventional economic theory, this treatment is guided largely by the discipline's utilitarian

foundations. As Kathryn George (1991) argues, however, a rights-based theory may be more

consistent with economic theory, and is certainly more consistent with Western historical values

and political realities than is utilitarian theory. Nevertheless, both approaches to intertemporal

equity are plagued by the problem of a dictatorial first generation. Randall (1978, p. 291)

points out that "any rule for allocative decisions with multigenerational consequences must

necessarily place the present generation in a dictatorial role." He further suggests that a

discounting rule is a particularly selfish form of dictatorship: "How else could one explain the

attempt of the present generation to attribute normative significance for all time to the outcome

of its current capital markets?" (p. 292). Some neoclassicists suggest that the power of bequest

gives the present generation sufficient stake in the future. However, future generations are not

represented in an analysis of individual social preference orderings. By definition, the present

value criterion maximizes the welfare of the present generation (Page, 1977). Future

preferences are represented only insofar as what they are believed to be by the present.

On the grounds of democracy, Marglin (1963), in a frequently cited paper, rejects the

notion that preferences of future generations should be represented at all in a social welfare

function. He argues that democratic principles reflect only present preferences, and that

governments' only role in this should be to educate on the needs of the future, then let the

present include this in their preferences. Unfortunately, the democratic argument for societal

time preference does not address the ethical question of whether present peoples are justified in

being less concerned about the future.

Once obligation to the future has been established, and a sustainability ethic assumes

that it has, the nature of that obligation still must be determined. Ethically, the present might

ask, "would we praise our policies if we changed places with our descendants?" One might

apply the "Lockean" standard that the present generation has done enough morally if it leaves

"enough and as good" for future generations. However, Locke used this notion as a constraint

on the present generation's 'natural' right to appropriate natural resources by combining them

with their labor (Barry, 1983).

Contrary to Locke, Page (1983) takes the position of relative rather than absolute

ownership rights of resources. He contends that it is unjust to run down resources not justly

acquired by the present (i.e., those left by previous generations), and suggests "that if the

present generation provides a resource base essentially the same as it inherited (including

roughly the same lack of contamination), it has satisfied intergenerational justice." (p. 56). The

requirements of intergenerational justice have been fulfilled if"the present generation gives the

next an equal chance at what is jointly shared across time" (Page, 1983, p. 56).

Obligation to the future may be expressed in terms of productive capacity. Barry

(1983) contends that the present can compensate the future for resource use through

replacement of productive opportunities. Compensation is defined in terms of productive

potential, which is equal in two situations if the same effort would produce the same output.

Barry cautions that compensation does not imply a right to harm; however, unavoidable harm

requires compensation. (It is not clear just to what extent resource depletion is unavoidable.)

Intergenerational justice requires that those who consume the resources that are depleted

should bear the responsibility for compensation, which should go to all (Bany, 1983). A

caveat here is that a certain amount of substitutability between natural and man-made capital is

required in order for this compensation to be effective. At the very least, this substitutability is


A conservationist/preservationist view of obligation to the future would be one of

maintained or expanding productive potential constrained by a criterion of sustainability

(Pearce and Turner, 1990). This entails a two-fold obligation: saving to maintain or expand

productive potential and ensuring that it is sustainable indefinitely. The constraint must be

satisfied before applying economic evaluation procedures.

2.2.2. The Shape of a Sustainable Society

A sustainable society is not only concerned about its obligation to the future, but also

values the quality of the present. Of course this includes present consumption of goods and

services, including those provided by the natural environment, and tradeoffs with future

consumption. It also includes quality as represented by an environmental ethic and a social

ethic. The long-run survival of human society

depends on certain functional requirements that are met by a set of social norms (i.e.
principles of behavior that ought to be followed). Over time, such norms must be
consistent with the natural laws governing ecosystem maintenance if sustainability is the
accepted policy goal. (Pearce and Turner, 1990, p. 226)

Environmental values are a cornerstone ofsustainability. Three relationships underlie

the environmental ethics and policies adopted in industrialized societies: values expressed

through private preferences of individuals, public preference values, and physical ecosystem

values (Pearce and Turner, 1990). Social norms and legislation are based on public

preferences which involve opinions and beliefs about what ought to be the case, as opposed to

individual desires or wants. Many sustainability advocates also believe that nature has inherent

or intrinsic value which exists regardless of whether humans are around to experience it. This

distinction is the source of difference between two basic environmental ethical positions,

anthropocentrism and ecocentrism, that were mentioned earlier. The former is an expanded

rights view which sees nature as a storehouse of instrumental or use value, and is described and

defended in terms of human interests. Ecocentrists, on the other hand, take the position that

humans possess relative value determined by their relative contribution to biosphere integrity.

Animal-rights positions are founded on an ecocentric ethic. An extreme ecocentrist position

would have controversial consequences: even mildly intensive agricultural systems can be a

threat to the integrity and stability of ecosystems, implying that development aid to poor people

in marginal areas could be wrong. Another, less extreme vision of ecocentrism may be

expressed in the position that humans need not make sacrifices for the environment, but they

should not seek their own good at the expense of the biosphere (Cobb, 1984).

An environmental ethic implies moral concern for the environment, the nature of which

is determined by the underlying ethical position. If one holds that the value of the environment

is only as an instrument to the welfare of humans, then moral concern for the environment is

indirect, with humans the object of direct moral concern. This can be thought of as a weak

environmental ethic. On the other hand, nature having intrinsic value implies direct moral

concern for the environment (Booth, 1994), the position referred to here as a strong

environmental ethic. While there is extensive disagreement over the philosophical foundation

for this position, ample reasonable evidence exists that humans can emotionally identify with

environmental wholes, supporting the position of a holistic environmental ethic. Results of

contingent evaluation studies where willingness-to-accept is sometimes four times as great as

willingness-to-pay are consistent with a strong environmental ethic (Booth, 1994). Willingness-

to-accept could establish the lower bound on compensation which is attractive enough to give

up moral commitments, whereas willingness-to-pay establishes a minimum income needed to

support other moral ends, such as providing a decent standard of living. As discussed in the

next chapter, this distinction in ethical position has implications for the limitations on the

usefulness of economic valuation techniques.

In addition to holding certain environmental values, sustainable agriculture also

embraces a social ethic. Agriculture is a human construct within an ecological environment,

and as such, is governed as much by social rules as by ecological rules. The social rules of

sustainable agriculture are grounded in a communitarian ethic in which rights are defined in a

social context, implying that equity considerations are important. A sustainable society ensures

wise stewardship over resources, that poverty and hunger are minimized, and that wealth is

distributed in such a way that essential needs are met. "Even the narrow notion of physical

sustainability implies a concern for social equity between generations, a concern that must

logically be extended to equity within each generation." (WCED, 1987).

The communitarian stewardship position maintains that small family holdings are better

suited to land stewardship, and finds undesirable the long-term transition from broad-based

family farming to large-scale industrial farming (Strange, 1988). The planning horizon of

family farming systems is infinite, and squandering resources and damaging the environment

are checked by community embarrassment. Within the community, the rights of property carry

an obligation for good land use.

However, this vision that family farmers tend to husband agricultural resources more

sustainably is not necessarily supported by the evidence (Buttel, 1984). While some research

on farmers' attitudes toward pollution and soil conservation shows clearly that farm size is

inversely related to expression of environmental concern, other studies find that large, non-

family farms exhibit less soil erosion and greater implementation of best management practices.

An environmental ethic may be stronger in family farming systems, but the ability to act

according to ones values is dependent on many other factors, not the least of which are the

social and institutional characteristics of farming. In fact, the very characteristics of agriculture

which lead to environmental degradation--price instability, high risk, short planning horizons,

and cost-price squeeze--are the ones that have preserved its dispersed ownership structure

(Buttel, 1984).

This points to the need to address those institutional and social factors that constrain

farmers from practicing more environmentally sound farming. Sustainable agriculture strives

to revitalize once strong rural communities in order to build upon the social and environmental


ethical structure that is so important for fostering profitable and sustaining farming systems, and

to remove the policy barriers constraining this development


This chapter relates some of the sustainability issues discussed in the previous chapter

with conventional economic theory, and explores ways that the theory tries to deal with some of

them. It does not attempt a resolution of the contradictions and conflicts, although, hopefully, it

contributes to the discussion. It points to directions in which the development of an economic

theory of sustainability might proceed, and to implications as to its utility.

3.1. Sustainabilitv in the Neoclassical Paradigm

Critics claim that neoclassical economic theory for decision-making at micro-, meso-,

and macro-levels is inadequate for the examination of sustainability issues; it frequently leads to

analysis resulting in production and resource-use decisions which are not sustainable. Key

assumptions in the neoclassical paradigm ignore the fact that human economic activity is an

integral part of the materially-closed, evolutionary biosphere. As Common and Perrings (1992)


for over a hundred years the dominant metaphor of economic activity has been a
trophic one in which economic agents are conceived as 'feeding off the resources of a
quiescent and independently functioning natural environment in order to satisfy an
exogenously determined set of desires up to the limits permitted by an exogenously
determined set of endowments. But except over the very shortest of time scales, this is
not a useful characterization of economic activity. (p. 15)

Sustainability ideology challenges many of the concepts and traditional methods of

neoclassical economics. Sustainability thinking has its roots in ecology. Norgaard (1991)


argues that economic thinking and ecological thinking are rooted in irreconcilable differences;

neoclassical market economics is "an atomistic-mechanistic model which views systems as

consisting of parts and relations between parts which do not change," whereas ecology is based

on evolutionary thinking which "looks at systems as undergoing changes in their parts and

relations." (p. 641). Further, Norgaard contends, economic thinking gives value to things, the

more the better, while evolutionary thinking gives value to the diversity that sustains change. A

fundamental difference between an ecological approach and a neoclassical economics approach

is that the former acknowledges that it is not possible to understand the properties of an

ecosystem from the study of one of its populations, while the latter assumes that knowledge of

individual behavior is sufficient to understand the behavior of the system as a whole (Common

and Perrings, 1992).

Neoclassical economics, it is argued, is not a very satisfactory tool for analyzing

situations where externalities and non-market goods abound. Environmental problems are

considered as market failures, not as evidence of the applicable limits to the market model

(Norgaard, 1984; Batie, 1989). Neoclassical economic analysis is most useful when the

working rules of society are broadly agreed upon, and improving the efficiency of particular

economic systems is the primary objective. Sustainability issues do not conform well to this set

of guidelines (G. K Douglass, cited in Geng, Hess, and Auburn, 1990). Further, neoclassical

economics is firmly rooted in the concept of efficiency, and by making fairly restrictive

assumptions, equates economic efficiency with social optimum, and economic welfare with

social welfare. However, the closer society is to ecosystem limits, the less valid is the

assumption that economic welfare and total human welfare coincide (Batie, 1989).

A sustainable society requires certain sacrifices. In this sense, sustainability thinking is

more akin to classical economics, in which utility maximization means maximizing the sum of

utilities, sometimes directing people to act against their own self-interest. Pareto optimality

does not require people to act against their own self-interest, though, in theory, the

compensation principle provides the means to redistribute endowments (Russell and Wilkinson,


The principle of consumer sovereignty, most basic to Walrasian economics, often

conflicts with sustainability. Common and Perrings (1992) have strived to develop a model of

sustainable resource allocation by combining efficiency requirements of economic sustainability

with stability requirements of an ecological approach. They conclude that an economic system

can be ecologically sustainable only if preferences and technologies are themselves sustainable.

However, consumer sovereignty, which privileges existing preferences and technologies,

implies system non-sustainability if existing preferences and technologies are not ecologically

sustainable. They argue that a sustainability paradigm implies "an approach that privileges the

requirements of the system above those of the individual." (p. 32)

Sustainability within the utilitarian framework of neoclassical economics begins with

consumption (Common and Perrings, 1992). The nature of the economic problem of

consumption implies the satisfaction of a given set of wants which are exogenously determined

by a set of preferences, and that the satisfaction of wants is constrained by an exogenous

endowment of resources and accompanying property rights that are invariant with respect to the

production system. Relating consumption to income, Hicks defined income as the maximum

amount that may be spent on consumption in the present without compromising real

consumption expenditures in the future (Common and Perrings, 1992). Trying to find a

rationalization for the principle of constant consumption of exhaustible resources, economists,

such as Solow, Hartwick, Dasgupta, and others, came up with what is now known as the

Hartwick rule, which states that

consumption may be held constant in the face of exhaustible resources only if the rents
deriving from the intertemporally efficient use of those resources are reinvested in
reproducible capital. That is, a necessary condition for consumption to be maintained
over time is that the efficiency rents from exploiting exhaustible resources should be
reinvested in non-exhaustible assets. (Common and Perrings, 1992, p. 10)

Of course, this rule implies the substitutability of reproducible and exhaustible resources, a

matter which is discussed later in section 3.3.2.

Recently, economists have extended the neoclassical model to better address

sustainability issues. This explains the current popularity of natural resource economics,

environmental economics, and ecological economics. However, because of their neoclassical

welfare foundations, these subdisciplines are also criticized as inadequate for guiding society

towards sustainability. They still put the onus on individual behavior, implying that economic

goods and services have the correct economic valuation. Speaking of mainstream

environmental economics, Goodman (1993) argues that

this orthodoxy rarely ventures into the realm of social and political organization or
geopolitical structures and their relationship to sustainable development. The
operational challenge rather is to measure the contribution environmental assets or
"natural capital" make to social welfare, as revealed by individual preference systems,
and to express this valuation in monetary terms so that trade-offs between
environmental goods and "human-made" assets are explicit. In this view, market
failure and policy failure in appropriately valuing environment-economy linkage lead to
unsustainable development paths. Within mainstream environmental economics, the
practice of sustainable development frequently is reduced to a technical exercise in
policymaking to improve the workings of markets by incorporating the economic value
of environmental assets into the everyday calculus of individual production and
consumption decisions. It is significant that the initial or inherited distribution of
income and resource endowments is taken as datum. This is hardly the paradigm to
lead us down sustainable development paths to alleviate poverty among present
generations or to entertain other world views and forms of social organization. (p. 237)

Responding to the ethical and environmental challenges of sustainability, some

environmental economists have adopted the "steady-state" school of thought in their theoretical

reasoning. A steady state may be thought of as a stationary state, where GNP and population

are constant; a physical steady state, where the stock of renewable resources and the stock-


depletion ratio for non-renewable resources are constant; and a state of steady growth, with a

constant capital-output ratio (Heijman, 1992). Sustainability combines aspects of all three.

Steady-state reasoning is based on physical laws as opposed to the welfare or utility basis of the

neoclassical position, and is consistent with a strong environmental ethic. Neoclassical

economists traditionally focus on production and consumption processes, or the 'throughput',

whereas steady-state economists focus not only on throughput, but the complete input-

throughput-output process (Jongeneel, 1992).

The first and second laws of thermodynamics and the law of conservation of matter

provide the basic economic principles for the steady state school of thought. The first law of

thermodynamics and the law of conservation of matter state that energy-matter can neither be

created nor destroyed, but only altered in form. This means that the inputs or raw materials

used in the production process are equal in amount to the waste materials ultimately returned to

nature. However, by the second law of thermodynamics, or the 'entropy' law, the inputs and

outputs are not equal in quality. This law declares that whenever work is done, i.e. something

produced, energy is transformed from a stable, highly usable form to a dispersed, less usable

state. Low-entropy, highly usable raw materials from nature are transformed in the production

process to higher-entropy output, and ultimately to high-entropy, virtually useless waste. These

laws imply that the greater the reliance on material flows and the regenerative and assimilative

capacities of the biosphere to satisfy increasing human demands, the greater will be pollution

and its associated disamenities, and the lesser will be the future productive potential of the

biosphere. The principles of steady-state economics deal explicitly with the problems of

pollution and absolute resource scarcity.

Notwithstanding the limitations imposed by its neoclassical underpinnings, natural

resource economics has been and continues to be applied to sustainability issues. In particular,

there are three crucial issues regarding sustainability to which natural resource and

environmental economists can apply their analytical tools: 1) the temporal problem and the

obligations of the present to future generations; 2) the issue of substitution among natural

resources and between natural and reproducible resources; and 3) internalizing within

production and consumption systems the costs of actions that generate negative externalities.

Adequate technological and institutional responses to the issue of achieving sustainable growth

in agricultural production cannot be designed as long as these three issues remain unresolved

(Ruttan, 1992). The usefulness of environmental economics to resolving these issues should be

explored. These are broad issues which may only indirectly influence farming system

sustainability; however, they most likely affect the way that researchers and policy-makers view

sustainability at the farm level.

3.2. Temporal Implications of Sustainable Agriculture

Introducing the time dimension complicates economic analysis in several ways.

Technology is time-specific; its performance will vary with the time of its introduction

(Etherington and Mathews, 1983). Additionally, technical relationships change over time.

Annual crop yields often decline as soils degrade, and complementary relationships between

components change with changes in relative resource scarcity. Further, the intertemporal

assignment of rights to natural resources affects the socially optimal allocation of resources,

implying that efficient resource markets are not sufficient to ensure socially desirable

intertemporal resource allocation (Howarth and Norgaard, 1990). Finally, time preference

arguments suggest that goods and services are valued differently across time.

In section 2.2.1, it was noted that sustainability implies an obligation to the future, and

that the nature of that obligation varies depending upon the ethical position taken. In

economics, obligation to the future typically is defined by the discount rate. Decisions

frequently are made based on present value calculations. This requires that future benefits and

costs are discounted to the present. The problem arises as to what is the appropriate rate at

which to discount future values in light ofintergenerational equity issues.

Most economists and policy makers support the concept of a positive social discount

rate which reflects current preferences. One reason, they point out, is that people's behavior

reflects time preference. Almost without exception, individuals require a greater amount of

future consumption to forego current consumption. But is an individual's time preference the

same as society's? To the extent that individuals are mortal whereas society is not, it is argued

by most economists that society's time preference is less than the individual's. Also, even with a

given set of consistent preferences, individuals may collectively save more per person than each

would in isolation (the free rider concept), further strengthening the case for a social discount

rate lower than the individual private rate (Sen, 1982). In fact, if all individuals, present and

future, are weighted equally, it is easily argued that there should be a negative social discount

rate if population is expected to increase (Gray, 1991).

Pearce and Turner (1990) contend that impatience, which is frequently cited as a reason

for societal time preference, is a myopic argument which is incompatible with long-run welfare:

tomorrow's satisfaction matters, not today's assessment of tomorrow's satisfaction. Page (1982)

argues for an intertemporal discount rate, based on preferences of both generations taken

together, which is lower than the social discount rate. Since it is based on present and future

preferences, an intertemporal discount rate is Pareto superior to both private and social

discount rates.

In addition to appealing to pure time preference (the soundness of which is questioned

above and in section 2.2.1), those who advocate a positive social discount rate often argue on

the basis of the marginal productivity of the economy. This is the opportunity cost of capital

argument and, because of the compounding nature of the discount rate, is based on the

assumption that benefits are reinvested. Critics counter that many of the social benefits are not

and cannot be reinvested. They also argue that there is no meaningful way to measure growth

rate in productivity, which requires interpersonal and intertemporal comparisons of utility

(Pearce and Turner, 1990). They further note that the compensation principle in welfare

economics is only theoretical; that compensation rarely, if ever, takes place. In justice sense,

discounting of future costs only is valid insofar as compensation is actually paid.

In a related argument, diminishing marginal utility is used to justify discounting because

it is assumed that future generations will be better off than the present one. A notion of equity

also underlies this argument in the sense that it is "good" that less benefits go to those who are

better off, i.e., future generations. While "better off" is often indicated by material product, it

really implies that utility is an increasing function of time. There are several reasons to question

whether this is in fact the case today. When per capital income was increasing rapidly, unless

marginal utility was drastically diminishing, it was pretty safe to say that per capital utility was

increasing over time. However, today, there is an increasing importance of non-income factors

negatively affecting utility, such as environmental degradation, increasing crime, and high

health care costs, which have coincided with increased incomes. Growth rates of income are

not independent of degradation. Also, increases in income have not come to everybody

equally. The growing inequality in incomes over the past decade means that for the most part,

increases in income are going to those who have the highest diminishing marginal utilities of

income, suggesting that utility is not increasing as fast as income. There is growing suspicion

amongst the present generation that they are not better off than their parents, and it can no

longer be assumed that future generations are going to be increasingly better off.


While the opportunity cost argument and the diminishing marginal utility argument are

considered the most persuasive (Goodin, 1982), other arguments are commonly used as

justification for positive social discount rates. One is to account for the increasing risk and

uncertainty associated with the further future. However, a number of people, economists and

philosophers alike, question the validity of adjusting the discount rate to account for risk (e.g.,

Stiglitz, 1986; Marglin, 1963; Parfit, 1983; Pearce and Turner, 1990). They contend that time

preference and risk are two different issues and should be kept separate. In fact, Parfit (1983)

argues generally that using the discount rate to account for factors other than time preference

misrepresents our moral reasoning. Time is not a good proxy for risk, opportunity cost, etc.

The fact that present peoples are less concerned about remote effects is not because they are

remote, he argues, but because of other factors. Pearce et al. (1990) concur that the various

costs and benefits associated with the future should be appropriately accounted for and not

simply lumped under a discount rate.

There is also question as to whether a discounting rule should be applied uniformly

across time and goods. Economic efficiency arguments are presented in favor of a uniform

discount rate. Variable discount rates, it is claimed, distort resource allocation in the economy,

resulting in resource use which is intergenerationally inefficient (Pearce et al., 1990; Page,

1983; Lind, 1982; and Krutilla and Fisher, 1975). Recall, however, that unless future

generations' preferences are included in determining the social discount rate, the resulting

allocation is not Pareto optimal. The efficiency argument is further undermined in the case of

exhaustible resources and an indefinite time horizon: efficient use of such resources implies that

they be fully depleted at the end of the time horizon (Heal, 1986). Efficient use in an infinite

time horizon would mean that none of the resource is consumed.


Goodin (1982) contends that there is nothing "natural" about uniformity of discounting

across all goods and all periods. Psychologically, people attach more importance to some

periods than to others, and to flows of some goods over others. None of the arguments for

discounting discussed above compel one to use uniform discount rates. The opportunity cost

argument depends in part on the rate of technological progress, which is not constant; and with

the diminishing marginal utility argument, there is no reason to assume constant economic

growth rates. A constant discount rate applied across all goods implies that all goods can have

a monetary value attached to them so that they may be exchanged with one another. But some

goods are 'nontradable' in the sense that "their loss cannot be made good by gains (however

large) in stocks of other goods." (Goodin, p. 67).

Economists also argue for uniform discount rates from the contention that it is not the

good which is discounted, or even its value, but that it is the utility derived from the good

which should be discounted. Daly and Cobb (1989) find this concept of discounting future

utility absurd because

there is no real world operation by which satisfaction today can be stored in a fund and
even if there were, there is no reason to expect such a fund to grow to give greater
satisfaction tomorrow. Money grows in a bank account, fish populations grow in a
pond, trees grow in a forest, and so on, which may justify short-term discounting of
future amounts of money, fish, and wood to there equivalent present value. But to
extrapolate this procedure to discounting future satisfaction is problematic, and to
extend it into the long run leads to unrealistic infinitesimal numbers because of the
exponential nature of compound interest. (p. 153)

Looking at this debate in another way can serve to highlight the sustainability

implications of discounting. What are the consequences of using a positive discounting rule?

The implications of high discount rates are that the sustainability aspects of an activity or

project are undervalued. For example, Ehui et al. (1990) found that soil conserving methods

became less worthwhile as the discount rate increased. In fact, any commonly used discount


rate renders costs and benefits thirty years hence virtually valueless. For example, the present

value of $100 in thirty years discounted at ten percent is only $5.70! For the distant future, any

positive discount rate will ensure that their benefits and costs have no value to the present.

While discounting makes sense for capital goods which have a limited life and can be replaced

by other capital goods, the reasonableness of discounting for 'natural' capital is questionable.

For the same reasons it is obvious that any adjustment to the discount rate for environmental

considerations, as some advocate, is also insignificant to the distant future.

Various approaches have been offered to counter the 'present bias' in economic

evaluation techniques. One way to build temporal concerns into benefit-cost methodology was

developed by Krutilla and Fisher (1975). Their approach is based on the premises that 1)

benefits to the environment are likely to increase with time relative to other benefits in the

economy, and 2) technological change will tend to reduce the benefits from conventional

development over time through obsolescence. To account for these, their solution is to let

environmental benefits grow by some rate over time, and at the same time, let development

benefits depreciate by some rate over time. (Note that costs can be considered negative

benefits.) These "adjusted" benefits are then subjected to ordinary discounting.

Another method of integrating sustainability objectives into cost-benefit analysis is a

compensating project approach (Pearce etal., 1990). By including environmental damage as a

cost, the normal cost-benefit rule becomes


where B = benefits, C = ordinary costs, and E = environmental damage (E < 0 implies

environmental benefit). Bringing in the time dimension, the cost-benefit rule is

E2,{(B,- Ct,- Ed} >0


where t is the time period and d, is the discount factor. This says that the sum over time of the

discounted benefits minus costs minus environmental damage is positive. If this rule does not

hold, then the project will not be selected.

The compensating project approach sets a constraint on the depletion or degradation of

natural capital over a set of projects, such that

Zi Ei s 0 for i projects in a set.

This means that if the sum of the environmental damage for all projects is positive, then

projects with environmental benefits must be added until this constraint is met. These are the

compensating projects. The weak sustainability criterion specifies that over time, the sum of

the present values of the environmental damage must be non-positive, whereas the strong

sustainability criterion requires that the constraint be satisfied for each individual time period.

While the compensating projects are not subject to the ordinary cost-benefit rule, the costs of

these projects should be minimized while still satisfying the degradation constraint.

These approaches are subject to the same criticism noted earlier, namely that any

positive discount rate renders benefits and costs in the distant future virtually valueless, and that

merely adjusting the discount rate is insignificant in the distant future. Further, the

compensating project approach assumes that benefits from the compensating project can atone

for environmental damage elsewhere, an assumption with which many environmentalists would

not agree.

In conclusion, discounting, when justified, should be applied selectively, based on the

importance attached to the resource and to the time period, and based on the degree to which

other resources may be substituted for the resource in question. The important point is that the

rights of the distant future are protected, either by constraining resource use or through an

appropriate application of a discount rule. When questions of the future are reframed in terms

ofintergenerational distribution of rights to natural and other assets, the rate of discount

becomes moot; it is simply an equilibrating price (Norgaard, 1991). When rights to assets are

redistributed between generations, ways of consumption, savings, and investment change to

insure that real assets back the rights of future generations. This leads to the important

conclusion that optimal investment portfolios are determined according to the cost-

effectiveness of alternative combinations of ways of sustaining future rights over time

(Norgaard, 1991). This implies that benefits accruing to future generations from investments

made to assure their rights cannot be measured by current preferences and should not be

discounted; and that costs of assuring their rights must be subtracted from current benefits.

33. Some Microeconomics of Sustainability

Crosson (1991) contends that the greatest threat to the sustainability of developed

country agriculture comes from the environmental costs rather than the economic costs of

meeting future demands for food and fiber. His argument is essentially this. By 2030, the

demand for North American grains (including soybeans) is projected to increase 73 percent.

Over the same period, given a 0.5 percent annual growth in North American population and 1.5

percent annual per capital income growth, demand for environmental services (water quality,

wildlife habitat, etc.) would increase between 80 and 200 percent'. However, it will be easier

to increase the supply of grains than to increase the supply of environmental services because of

'Evidence suggests that demand for environmental services increases not only with population
growth, but also with per capital income. Estimates of this relationship vary widely, but most fall
between 0.5 and 1.5, indicating that for each one percent increase in per capital income there will be a
resulting 0.5 to 1.5 percent increase in demand for environmental services. With population growth in
North America projected to average 0.5 percent annually through 2030, and per capital income
conservatively expected to grow 1.5 percent annually, this means that total demand for environmental
services would increase between 80 and 200 percent by 2030 (Crosson, 1991).

the differences in their institutional conditions. Grains are supplied through markets which

register their prices while environmental services do not pass through markets, so they generally

are not priced. Under the induced innovation argument (Hayami and Ruttan, 1985), rising

grain prices under pressure of projected demand would induce technological development to

ease the pressure on supply. In the case of environmental services, there is no pricing

mechanism to induce innovation to ease their supply pressure. Further, although many

environmental services are essentially free, "they would, of course, be infinitely costly to

replace." (Ehrlich, 1989, p. 10)

Giampietro and Bukkens (1992) consider the dynamic equilibrium between biophysical

capital and human-technological capital as an indicator of sustainability. They note that

When human exploitation becomes so efficient that the energy flow left in the
biophysical compartment is too little to preserve the original complexity of biophysical
capital the process can undergo a stability crisis. In this case, the incalculable richness
of natural structures and functions, generated over millions of years of evolution, is
jeopardized by an excessive activity of human technology. (p. 30)

This interaction between natural capital and human-technological capital, and the effects of

human activity on natural capital are the subject of much discourse in the environmental

economics literature.

3.3.1. Valuing Natural Resources and the Allocation Problem

All the natural and human-made resources used to produce goods and services valued

by people collectively comprise what is known as social capital (Crosson, 1993). In the context

of sustainable agriculture, social capital includes supplies of land, water, energy, plant genetic

material, and knowledge embedded in people (human capital); technology; and institutions.

For convenience of analysis, social capital may be divided into natural and reproducible capital.

Of the natural capital, some is exhaustible and some is renewable; in the case of exhaustible

capital, some is replaceable with reproducible capital and some is not. Capital has value

according to the services it provides. All capital provides both environmental services and, for

lack of a better term, economic services; it would be expected that natural capital provides a

greater proportion of environmental services. While some exhaustible capital may be

replaceable for the economic services that it provides, it might not be replaceable in the

provision of environmental services. This fact makes the valuation of capital, especially natural

capital, very difficult.

In traditional neoclassical economics, the efficient intertemporal allocation of

exhaustible resources requires that the price of an exhaustible resource (neglecting extraction

costs) should increase over time at the rate of interest (Common and Perrings, 1992). This is

known as the "Hotelling Rule." The price of a natural resource depends on its user cost, which

is the net value of future foregone rents due to the current use of an additional unit of the

resource. User cost can be thought of as a dynamic shadow value from an optimal control or a

dynamic programming model, which mathematically is the partial derivative of the optimal net

present value function with respect to the state variable (Zilberman et al., 1993; Passmore and

Brown, 1991). The solution to the intertemporal resource allocation problem finds that the

price of the resource is equal to its marginal net benefit to society which is equal to the sum of

its marginal cost (extraction, processing, delivery, etc.) and its user cost. Presumably, user

costs include foregone value to future generations, though in practice, such value determination

is impossible, rendering the solution potentially suboptimal.

Pearce (1989) suggests that the way natural resources are priced can be corrected to

include not only extraction, processing, and delivery costs, as well as user costs, but also

environmental costs. At the margin, the price of a resource at time t should be

Pt= MC + MUCt + MECt


MC = marginal costs (of extraction, processing and delivery)
MUC = marginal user costs
MEC = marginal environmental costs

A price adjustment such as this is consistent with the 'polluter pays' principle. These charges

will be partly passed on to the consumer in the form of higher prices, but this is as it should be.

Consumers are, after all, the ultimate polluters: they signal to the producer what they want and

should pay the full costs of production.

The danger in relying on correcting prices of natural resources is that, unless the prices

are absolutely right, unsustainable practices of extraction and pollution may continue.

However, assigning values to non-market goods and natural resources has always posed a

problem for economists. Sustainability advocates view as absurd the notion that non-market

goods can be valued as services by revealed preference techniques (Batie, 1989). The issue in

accounting for natural resources is the optimal value of their stock. There is no reason to

believe that prices generated in surrogate markets are relevant to this measure. If the principle

of consumer sovereignty is accepted, Common and Perrings (1992) show that for economic

and ecological sustainability to converge, existing preferences and technologies must be

ecologically sustainable. If they are not, then the valuation of resources through price

corrections based on contingent valuation techniques deriving from unsustainable preferences

offer no advantage.

Further, non-market valuation techniques, when based on preferences of the present

generation, will not necessarily lead to socially optimal intertemporal resource allocation, even

if there are efficient resource markets (Norgaard, 1991; Howarth and Norgaard, 1990). Such

techniques merely show how the present generation values non-market goods and cannot

assure resources for future generations. An optimal intertemporal allocation of resources

requires that individual resource users either face a complete set of markets, including

contingent markets, from the present to infinity; or contract in current markets on the basis of

rational expectations about future prices, consistent with the clearing of all forward markets

(Common and Perrings, 1992). Practically, these two amount to the same thing.

3.3.2. Resource Substitutability

Scarcity in neoclassical economics is conveniently handled by prices: as a factor

becomes more scarce, its price increases resulting in a shift to other less scarce, lower priced

factors. Assumption of good factor substitutability is key to this conclusion, however.

Provided there is sufficient substitutability between reproducible and exhaustible resources,

economists have shown that it is possible to derive an optimal investment rule that will maintain

the productive capacity of the resource stock (Common and Perrings, 1992). Yet, the

substitutability assumption is increasingly being challenged. If factors are complements rather

than substitutes, production is constrained by the most limiting factor. More and more, the

productivity of manufactured capital is limited by the decreasing supply of complementary

natural capital.

Failing to recognize the increasing limits posed by decreasing supplies of

complementary natural capital is what Daly (1992) refers to as the problem of viewing an era of

full-world economics from the perspective of empty-world economics. The evolution of the

human economy has passed from the time when manufactured capital was limiting to an era

when the remaining natural capital has become the limiting factor. Economic logic suggests

that the productivity of the limiting factor should be maximized and its supply increased. Yet

empty-world economists continue to use up natural capital in the pursuit of increasing

productivity of manufactured and human capital, assuming the substitutability of technology for


natural capital. However, technology and other means of increasing the efficiency of resource

use should not be confused with capital. Defining capital as efficiency makes a "mockery of the

neoclassical theory of production, where efficiency is the ratio of output to input, and capital is

a quantity of input." (Daly, 1992, p. 29)

Giampietro and Bukkens (1992) illustrate the unreasonableness of the human dream of

substituting technological capital for environmental services in a global context using energy

analysis. For example, the biophysical compartment uses 44,000 teraWatt of solar energy for

the bio-geochemical cycle of water, while all the human activity on Earth uses only 10 teraWatt

of energy. Should water cycling be threatened because of human disregard for tropical forests,

"technology would be orders of magnitude short in being able to create a substitute."

(Giampietro and Bukkens, 1992, p. 40)

Recent applications of the neoclassical approach to sustainability have recognized the

limitations of these technological assumptions, in particular, the substitutability assumption.

Thus, the notion that some suitably defined stock of natural capital should be kept constant is a

crucial component of these approaches.

Pearce et al. (1990) consider the key necessary condition for sustainable development

as 'constancy of the natural capital stock'. The stock in question might be the existing stock at

the time a decision is made, or some optimal level that should exist. While the latter is clearly

the correct level according to neoclassical analysis, it can be argued that the existing level is the

one that should be maintained. In some cases, existing stocks are already below the optimum.

The uncertainty and irreversibility surrounding the use of natural stock provide another

rationale for conserving existing stocks. Still another is that optimality tends to be defined in

economic efficiency terms, while conserving the natural capital stock serves other social goals

as well.


A fourth reason for maintaining existing levels of stocks arises from recent research on

willingness-to-pay and willingness-to-be-compensated measures of benefit (Pearce et al.,

1990). Economic theory predicts that the difference between these two measures (the

equivalent and compensating variation measures of welfare gain) should be insignificant

Empirical work, however, suggests that there are very large discrepancies between willingness-

to-pay and willingness-to-accept. Essentially, people require much larger levels of

compensation to give up something than they are willing to pay to gain more of something.

(The philosophical underpinnings for this are discussed in section 2.2.2.) The theoretical

implications of this phenomenon are shown in Figure 3.1. Figure 3.1(a) shows the standard

neoclassical approach to determining the optimal level of natural capital stock, KM*, where

marginal benefits equal marginal costs. Figure 3.1(b), on the other hand, shows the optimal

stock level when the valuation function (B) is kinked due to the differential in the two valuation

measures. The kink is at the existing level of stock, making the existing and optimal stocks

probably coincident (Pearce et al., 1990).

3.3.3. Accounting for Externalities

Externalities arise when the activities of one economic agent affect the technology,

consumption set, or preferences of another agent other than through the pricing mechanism.

The potential market failure associated with externalities was first identified by Pigou in 1932

(Zilberman and Marra, 1993). He noted that the economically efficient level of production may

not be the socially optimal level because the producer of a good may not bear all the costs of

production. In other words, the marginal private cost of production is less than the marginal

social cost. This is illustrated in Figure 3.2 which shows the market for an agricultural good

before and after adoption of a cost-saving but polluting technology.

0 Ko KN 0 K=K*N KN
(a) (b)

Figure 3.1. Costs and benefits of conserving natural capital stock. B shows the benefits of
increasing natural capital stock; C is the cost of increasing it, or the forgone
benefits of using it for other purposes. (Adapted from Pearce et al., 1990)

The marginal cost of the good produced without the new technology is shown by MC

in Figure 3.2, and the marginal benefit, or demand for the good is indicated by MB, with the

resulting output and price given by qo and po respectively. Adoption of the polluting technology

reduces cost to MC', resulting in production increasing to qj. However, this does not account

for the pollution externality. The decrease in welfare can be represented by a downward shift

in the marginal benefit curve to MB', which reflects consumers' true willingness to pay for the

good, taking into account the pollution caused by its production. The socially optimal level of

production would be q2. Combining producer and consumer surpluses, total welfare in this

case is given by area FCH, and if this area is smaller than area EAG (total welfare without the

technology), then the technology should be banned. Otherwise, its use should be regulated to

achieve production at q2.

Alternatively, the pollution externality can be internalized through an imposition of a

Pigouvian tax on each unit of production such that the private marginal cost is increased to

q0q2 q1 Output

Figure 3.2. Effects of a negative externality on the production of an agricultural good. MC
and MB are the marginal cost and marginal benefit curves respectively of
producing a good without new technology. MC' and MB' are the marginal cost
and marginal benefit curves respectively of producing the same good with a new
technology which pollutes, and MSC is the marginal social cost curve of this

equal the marginal social cost, curve MSC in Figure 3.2. In this way, production is restricted to

q,, the socially optimal level, and the price will be P2, which now reflects the "true" cost to

society of production of the good. Similarly, a positive externality can be accounted for via

production subsidies.

In a world of clearly defined property rights, Ronald Coase in 1960 argued that

intervention through taxing or regulation was unnecessary in the case of externalities, and could

lead to an allocation which was not Pareto optimal (Binger and Hoffman, 1988). His premise

was that one agent (the one without the initial property right) could compensate the right holder

to an extent necessary to achieve the economically efficient allocation of exteralities. For

example, the polluting firm could bargain with the affected parties to pay them to allow it to

pollute. The essence of the Coase Theorem is that the initial allocation of property rights is

immaterial to the final outcome. Regardless of who owns the property right, the two parties

will bargain until the marginal benefit of pollution reduction is just equal to the marginal cost.

Given full information, low transactions costs, and strict enforcement of contracts, the two

marginal calculations would lead to the same Pareto optimal solution. Equity and intertemporal

concerns, however, are not addressed in a Coasian solution.

The assumptions underlying Coasian analysis infrequently hold in the real world. When

they do, externality problems often are solved by those involved and do not come to the

attention of analysts. It is when the assumptions do not hold, when information is limited and

transactions costs are high, that exteralities become a problem for economists and policy

makers. Solutions arrived at through Coasian bargaining or imposed as a socially optimal,

Pigouvian tax as discussed above are ideal, and referred to as first best solutions to an

externality problem. (A Pigouvian tax may be considered as a Coase optimal solution in the

sense that the tax is the result of bargaining between society, the initial right holder, and the

polluter who pays compensation in the form of a tax to society for the right to pollute.) Given

the difficulties associated with measuring marginal exterality cost so that a first best solution

can be imposed, some economists suggest instead a second best solution that involves setting

environmental targets and identifying the least cost policy to reach those targets.

The range of environmental issues to which economic valuation techniques may

appropriately be applied is limited by the environmental ethic inherent in sustainability

positions. The nature of the moral concern for the environment, whether it is based on nature's

instrumental or intrinsic value (see section 2.2.2), has significant implications for environmental

economics. The moral responsibility implied under an anthropocentric position requires that

compensation must be made to the "losers" in an externality problem (Booth, 1994). This is

contrary to the utilitarian neoclassical roots of environmental economics which only requires

that compensation be possible. In the case where compensation cannot be made, i.e.,

somethings are not "priceable," then economic valuation techniques such as cost-benefit

analysis cannot be applied. If the ethical position is that the environment has intrinsic value,

that the environment is the object of direct moral concern, then a price cannot be attached to

environmental damage, and some form of moral ordering rather than economic analysis is the

appropriate decision tool (Booth, 1994). Economists cannot assume that prices can be proxy

for moral value. Thus, economic valuation techniques are appropriate only in cases where

elements of the environment are instruments for which there are substitutes, and therefore can

be priced.

While conventional economic theory posits that the initial allocation of property rights

is immaterial to the resolution of an externality problem, the sustainability ethic suggests that

allocation of rights is very important. The implementation of sustainable agriculture, which has

focused on agricultural production rather than the entire agricultural sector, has resulted in,

among other things, a subtle shift in property rights (Dicks, 1992). Income protection for

farmers in the form of program benefits over time became institutionalized and translated into

property rights by farmers. Prior to 1985, exchange of property rights in agriculture was

conducted under rules which were voluntary and mutually agreeable by all parties.

Conservation programs were voluntary and provided cost-sharing and technical assistance. The

farmers' right to lose soil was implied. This was changed in the Food Security Act of 1985. To

minimize environmental disturbances, the Conservation Title used liability rules of property

rights exchange which did not require prior consent, and where prices were set by a third party.

Failure to comply meant the loss of program benefits. The rules of property rights exchange

were further shifted in the Food, Agriculture, Conservation and Trade Act of 1990. The new

wetlands provision revoked commodity program benefits whenever wetlands were altered, not

just when they were altered for agricultural production.

3.4. Measuring Sustainability

The limitations on marginal analyses imposed by a sustainability ethic discussed above

invite the search for alternative rules for making production decisions and measuring or

otherwise accounting for sustainability. When elements of the environment are objects of

moral concern and therefore cannot be assigned value which may be traded, then they must act

as constraints in the decision-making process. So-called second best solutions involving safe

minimum standards and similar threshold levels may in fact not be second best, but rather first

best according to an environmental ethic. Rather than strive for the Pareto efficient solution,

economists can use their constrained optimization and modeling techniques to find the "best"

way to produce within the environmental constraint. Norgaard (1991) argues that

economists need to explicitly recognize that sustainability is an equity question being
debated in various moral discourses utilizing ecological reasoning and that
sustainability will be chosen through politics. Economists in this framing can inform
the political process of the impacts of different equity decisions and the most cost
effective ways of reaching them. Economists can interact with moral discourse,
environmental lines of reasoning, and the political process but cannot "rationalize"
them. (p. 641)

Measuring sustainability at the farming system level is discussed in the next chapter.


To be sustainable, a farming system should produce at least as much in succeeding

years as it has in past years with the same level of factor input, accounting for changes in stocks

and flows of common property natural resources (Ehui and Spencer, 1992). In this study,

sustainability at the farming system level is characterized by a farming system that will,

currently and in the long-run, meet the reasonable goals and objectives of the multi-

generational farm family; and will satisfy short- and long-term societal goals regarding

environmental quality, food safety and security, and general quality of life. The three

components of a sustainable farming system are:

1. financial viability in the short and long run, requiring non-decreasing net value-
of-production trends over time, which in turn depend on the quality and quantity
of the farm's natural, human, and capital resources;

2. non-degradation of the more general (off-farm) environment requiring a minimal
(or zero) level of negative externalities emanating from the farm; and

3. equitable distribution of farm-induced benefits and costs within the farm
household, across farms and across generations, and between farm and non-farm

4.1. Indicators of Farming System Sustainability

Conway (1991) suggests that agroecosystems possess recognizable goals and strategies

to attain them, and that primary amongst these goals is increased social value, which has

measurable components expressed as agroecosystem properties. Conway's characterization of

agroecosystem properties provides a useful start toward an operational definition of farming

system sustainability. Since sustainability can be thought of as maintaining or increasing social

value, these properties provide useful indicators of sustainability.

Adapting from Conway (1991), a sustainable agroecosystem has four desirable

properties: productivity, stability, resilience', and equity. 1) Productivity may be defined as the

value of product output per unit of resource input, or, alternatively, as the net value of product

output. 2) Stability is the constancy of productivity in the face of small, usually cyclical,

disturbing forces. 3) Resilience is the ability of the system to withstand severe, unpredictable

shocks. And 4) equity refers to the evenness of the distribution, spatially and temporally, of the

benefits and costs from the productivity of the agroecosystem. These four properties constitute

the "social value" of the agroecosystem. Indicators of farming system sustainability should

reflect the well-being of people and the environment, and can be defined from these properties.

Sustainability indicators can be primary (ie., they directly measure one of the sustainability

properties), or secondary (ie., they measure a component of one of the properties).

4.1.1. Productivity Indicators

Productivity of a farming system may be indicated by the trend over time in the value of

output per unit of resource input, or the total factor productivity (TFP) of the system.

Sustainability requires that the time trend in TFP be constrained to be non-negative (Harrington

etal., 1994; Lynam and Herdt, 1989; Ehui and Spencer, 1992). Both Harrington et al. (1994)

and Ehui and Spencer (1992) note that TFP measures should include environmental costs and

changes in resource stocks. Economic sustainability also requires that the time trend in income,

'Conway refers to this third property as sustainability. Sustainability, as more broadly defined for this research,
encompasses all four of these properties. Thus, the third property is more accurately referred to here as resilience.


where productivity is defined as the net value of product output, be constrained by some lower

bound. In general, the values attached to outputs and inputs need not be monetary, though they

must be in some common unit that adequately measures productivity in situations where there

are significant changes in the farming system over time. In most situations in the United States

where output is produced almost entirely for sale, a monetary description of value is


A sustainable cropping system requires a balanced nutrient budget (Budelman and van

der Pol, 1992). In a situation where nutrients must come via a market (e.g., when nutrients are

not replaced through natural processes), sufficient value must be added to the inputs through

the production process to replace the lost nutrients. This assumes that, at the least, output price

equals the marginal cost of production plus the marginal user cost of the soil and nutrient

resources, and that marginal environmental costs are zero as a result of management of

outflows. The inability to replace lost nutrients is commonly the problem in many fanning

systems in developing countries where production is primarily for subsistence. Thus,

maintaining the nutrient balance of the soil is a possible secondary indicator of sustainable

productivity. The information for this indicator could be gained by modeling the soil nutrient

processes under particular soils, weather, and management, or through estimates from the

literature and crop statistics (Budelman and van der Pol, 1992).

4.1.2. Indicators of Stability and Resilience

Stability and resilience constitute a continuum, and may best be described by measures

of risk and uncertainty. Stability and resilience depend in part on the flexibility of the farming

system. Given the uncertainties involved in farming, both natural and economic, the farmer

must be able to respond and adjust to various outcomes. Production flexibility is very

important in sustainable agriculture (Gray, 1991).

Stability and resilience are distinguished by the level at which they apply. Following a

perturbation, stability refers to the propensity of a system to return to an equilibrium condition,

be that stationary or cyclical, while resilience is the propensity of a system to retain its

organizational structure. Stability pertains to system variables and resilience refers to the

stability of system parameters (Common and Perrings, 1992). Resilience permits multiple

equilibria for system variables but requires a unique equilibrium for system parameters.

Clearly, individual populations (or subsystems) within an ecosystem can only be stable if the

ecosystem is resilient (Common and Perrings, 1992). The distinction between stability and

resilience is of practical importance. Kidd (1992) notes that

Concentration on resilience implies less emphasis on trying to control systems by
reducing their variability. It implies more emphasis on taking steps to profit from the
natural capacity of systems to adapt by designing policies and institutions that can react
constructively to inevitable but unpredictable surprises. (p. 20)

The coefficient of variation in farm production or farm income, or some other measure

of variability in the productivity of the farming system, is a primary indicator of the stability of

the system. Generally, a farming system which is less variable in its output, that is, more stable,

is less vulnerable to the cyclical vagaries of nature and economics, and therefore more

sustainable. Also, it should be noted that high variability makes identification of time trends in

productivity more difficult, and hence less reliable.

Resilience may be measured in how well the farming system is protected from shock. It

is directly proportional to the magnitude of shock required to affect productivity, and to what

level and how quickly productivity will return if it is affected. While stability is represented by

variability, resilience is more akin to the "safety-first" notion of risk avoidance. Several issues

are important when assessing farming system resilience. First, how well is the system able to

resist a shock; how well is it protected from stress? The degree of resistance of the system may

be measured by the amount of reduction there is in production in response to a given shock. It

can also be measured by the probability of production or income dropping below some

minimum level, below which the system cannot recover, leading to system failure.

In a farming systems context, the minimum level in system productivity below which

the farming system cannot recover depends in part on the availability of external resources to

the farm family, and on the social and institutional environment in which the farm operates.

Two other issues important in assessing system resilience are the rapidity of recovery of the

system from a non-fatal shock, and the completeness of that recovery. It could be the case that

recovery is only partial, perhaps the farmer was forced to sell productive assets, for example,

and the system is in a more vulnerable, less resilient state than it was before the shock. In this

case, the farming system before the shock is less resilient than one that would recover

completely from the same shock.

The elasticity of supply may be an indicator of resilience of a regional or national

agriculture system which is primarily commercial. The ability to respond to external shocks is

largely a function of technologies and available inputs (Gray, 1991). An individual farmer who

is better able to respond to price changes resulting from a shift in supply is more resilient than a

farmer unable or less able to similarly respond.

Farming system resilience can be modeled by simulating the farm with a "failure

trigger" below which the system cannot recover, or by estimating the probability of farm

failure. Simulating farming systems over time, and subjecting them to hypothetical shocks can

give an idea as to the resilience of the system in terms of the ability of the system to recover and

the minimum level below which recovery is foreclosed. Secondary indicators of resilience

include measures of financial health, such as the degree of capitalization or the debt to equity

ratio. In assessing sustainability, one would also want to evaluate the likelihood of severe

shocks in a particular environment as it relates to the resilience of a particular farming system.

4.1.3. Indicators of Equity

Equity, as would be expected, represents the biggest challenge to measurement Well-

known macro indicators of distribution such as the Gini Coefficient and the Lorenze Curve are

not well-suited for firm level analysis. Equity covers a broad range of issues, and at the

farming system level may refer to the distribution of wealth among farms, between the farm

and non-farm sectors, and within the farming system itself; and it refers to the distribution of

the costs and benefits of farm production between the farm and the community. It is with this

last aspect of equity that this research is primarily concerned. Specifically, this includes the

environmental and social off-farm effects of production, both good and bad effects. These are

commonly referred to as production externalities.

An assessment of equity, more than with any of the other properties of sustainable

farming systems, can only be made within the context of the prevailing sociopolitical

environment. Since this environment is particularly volatile and laden with ethical concerns,

equity assessment can be precarious, and should be undertaken with as much care and

objectivity as possible, limiting analyses to investigation of the possible scenarios without

making a prescription. Based on an appraisal of the political and institutional environment,

which includes the goals and objectives of society as well as those of farmers, equity may be

indicated by defining acceptable levels of wealth or property accumulation, of chemicals

leaching through the soil profile or running off into surface waters, of minimum resource stocks

to be maintained, or of other externalities to constrain the farm's activities. Welfare analysis

can help reveal the distribution between farming and non-farming sectors of the costs and

benefits of adopting or not adopting sustainable farming practices.

4.1.4. Summary of Sustainabilitv Indicators

While the socio-political-economic environment is generally considered as given in

farm level analysis, it must be recognized that this external environment often constrains

progress toward a more sustainable agriculture. For example, common agreement exists that

commodity programs and base requirements inhibit farmers from adopting beneficial rotations

and other sustainable practices. For this reason, some sustainability analysts prefer to look at

ecological sustainability independent of the social environment, preferring to describe a

sustainable system and prescribe the type of socioeconomic system needed to accommodate it.

However, these types of theoretical constructs rarely gain passage in the real world, leaving

farmers with more limited options for sustainable farming. Therefore, it makes sense to

consider sustainability indicators which reflect constraints imposed by the socio-political-

economic environment as well as by the natural environment.

A summary of possible sustainability indicators, both primary and secondary, is

presented in Table 4.1. Some of these may be used to constrain a system to be sustainable,

while others may indicate a level of sustainability. It is important to recognize the

complementarities and trade-offs among these properties when assessing the sustainability of a

farm system. This necessarily involves value judgement and qualitative assessment. In most

situations, it is preferred to have agroecosystems in which all properties are high and the trade-

offs are minimized (Conway, 1991).

In practical application, quantifying indicators to measure sustainability is beset with

problems (Harrington, 1991). The range of possible indicators discussed above mirror the


multiple goals associated with sustainable agriculture. The impacts of the various goals occur

in different time frames, some well into the future. The degree to which sustainability can be

measured depends on the ability of analysts to reliably forecast relevant future events.

Satisfactory measurement of sustainability requires the simultaneous measurement

(quantification) of control as well as state variables in order to link problems with their causes.

Simulation modeling of farming systems can help to overcome certain quantification problems

in sustainability analysis by portraying possible future scenarios more clearly.

Table 4.1. Indicators of Farming System Sustainability


Productivity Trend in value of output per unit of resource input as measured by TFP
Trend in net value of output as measured by gross margin or income
Trends in the nutrient balance of the soil
Energy efficiency ratios, & energy flows over time
Stability Coefficient of variation in productivity
Probability of farm failure
Resilience Probability of farm failure
Minimum productivity/consumption constraint
System diversity
Capacity for recovery
Availability of external resources, both private and public
Financial health
Estimated probability of severe environmental or economic stress
Equity Restrictions on levels of pollution
Maintenance of natural resource base
Restriction on rate of real property accumulation

4.1.5. Modeling Indicators of Sustainability

Sustainability can be considered in a farm sustainability model in two ways. One,

sustainability concerns can appear as constraints to the model's performance. Production can

be constrained by limits on allowable nitrogen leachate, for example, or by restricting soil

organic carbon and organic nitrogen levels to be non-decreasing. A minimum level of

consumption or income can be imposed on the model, and borrowing can be restrained to a

certain percentage of assets, or of a level of a particular asset Environmental constraints can

appear either as deterministic or stochastic.

A second way that sustainability may be considered is by analyzing system performance

over time as simulated by the model. The ultimate test of sustainability is whether the farming

system survives--whether it continues to operate productively. Considering the second law of

thermodynamics, survivabilty requires that the flow of fundamental resources into the system

also be maintained. In addition, the trend in productivity as measured by yields, TFP, or

returns, and the variability within these trends can be observed and evaluated. Model

parameters and constraint levels can be varied to observe system behavior under different

conditions and different management practices to evaluate the resilience of the system. In this

way, the limiting constraints to adopting more sustainable practices, as well as the associated

trade-offs, can be examined.

Productivity, as indicated by gross margin or some other income measure, appears in

the model as the objective to be maximized. Trends in the income measure are observed and

analyzed over time as simulated in the model. The income objective may be constrained in the

model by non-decreasing TFP, for example, or by resource stock levels, or by levels of

resource flows. Sustainability advocates often argue that sustainability is a goal and that

sustainability issues should appear as objectives of the decision-maker, resulting in a multiple

objective type of a problem (Flora, 1992). However, because of the difficulties encountered in

assigning value to resources and externalities, and in accounting for future preferences,

sustainability concerns are modeled as constraints to the income maximizing problem. This is

consistent with a sustainability ethic in which the environment is an object of moral


consideration, and also avoids the problem of assigning weights to various goals or objectives.

Although, ideally, sustainability should be a goal of farmers, in reality, farmers are more likely

to feel constrained by sustainability issues as defined through public policy.

4.2. Review of Modeling Approaches which Include Sustainability-Related Issues

Natural resource use and degradation issues have concerned agricultural economists for

several decades, and numerous modeling techniques have been employed to address them.

Though much of the discussion has involved qualitative analysis, attempts to quantify the

parameters of the debate have increased recently as scientific research and analytical

capabilities have increased. Econometric methods have been used to estimate impacts of

current production practices on future soil productivity (Orazem and Miranowski, 1994), to

ascribe the effects of certain conservation methods on resource prices (Palmquist and

Danielson, 1989), and to identify various factors explaining the use or overuse of natural

resources (Lockwood et al., 1994). Dynamic optimization techniques are frequently employed

to model the dynamic aspects of resource degradation or rehabilitation and other sustainability

concerns. Passmore and Brown (1991) used a stochastic dynamic programming model to

analyze the degradation of rangeland in an arid grazing area of Australia. A multi-period linear

programming model was employed by Miranowski (1984) to find the optimal choice of tillage

method and crop rotation for farmers who consider the yield-decreasing effects of soil erosion.

4.2.1. Recursive Proaramming

Recursive programming (RP) techniques have been used in dynamic farm analysis to

model feedback in the decision process to study farm growth and agricultural change. A

recursive programming approach, as described by Day (1963, 1983) and Day and Cigno


(1978), is selected for this research over dynamic programming and optimal control procedures

because it more accurately represents the way decision-makers act in situations with limited

information. Recursive programming "is the theory of partial economizing or suboptimization

with feedback that describes economic behavior in disequilibrium or temporary equilibrium."

(Day and Cigno, 1978, p. 8). This approach assumes that decision-makers optimize, not over

some extended time period, but over relatively short periods of time, distinguishing it from

intertemporally optimizing models. The feedback structure, which is not completely accounted

for in the specification of the optimizing operator, is based on the premise that real world

economic behavior proceeds by "decomposing large decision problems into smaller, simpler,

approximate or local decision problems which are modified by feedback on the basis of

behavioral rules, convenient computational formulae and observed changes in the decision-

making environment." (Day and Cigno, 1978, p. 8).

Recursive programming was first described in the seminal work by Day (1963) in

which he developed a dynamic systems modeling approach based on a synthesis of linear

programming and difference equations. He applied this approach to successfully explain and

then predict changes in production in the Mississippi Delta region during the volatile post

World War II period.

De Haen and Heidhues (1978) employed a recursive programming approach to model

the interdependent firm-household decision process in their study of agricultural change in

Germany. Kingma (1978) utilized a RP model in which parameters for any one time period are

dependent upon decisions taken in the previous time period to examine farm growth in an

Australian case. In a regional planning study of the Peruvian Sierra, Schaefer (1974) extended

a recursive linear programming model to include the ability to determine optimal investment

strategies in agriculture. Kolajo and Martin (1987) found that a recursive linear programming

model more accurately projected the growth of an Alabama row-crop operation than did a

multi-period linear programming model.

4.2.2. Modeling Biological Processes

An important aspect of this research is the use of crop growth models to investigate,

under varying management practices, not only what happens above the soil surface (i.e., crop

growth and yields), but what takes place in the soil. Assessing the environmental impacts of

various practices demands knowledge of the fate of nitrogen and other chemicals applied as

part of these practices.

Crop growth simulation models substitute for traditional production functions, which

express output as a function of economically scarce resources, assuming given technological

and agro-climatic conditions (Burrell, 1991). By simulating the plant's response to its

environment on a daily or hourly basis, crop models essentially break down the production

function into a large number of small sequential links, the parameters for which can be obtained

from a variety of sources. Furthermore, the user can vary the assumptions regarding

technology and agro-climatic conditions.

Biophysical process models, especially crop growth models and hydrologic models, are

increasingly being used in the analysis of management and resource use questions in

agriculture. The marriage of biophysical process models to economic behavioral models is a

relatively new procedure primarily because the process models have only recently achieved the

level of development and confidence necessary for their widespread use. Numerous recent

applications of this procedure, particularly in resource management research, attest to its

increasing applicability and acceptability.

Biophysical process models are not without limitations, however. For example, crop

growth models cannot presently account for certain important factors such as phosphorous,

minor soil nutrients, and weed competition, although improvements in these areas are

anticipated. However, biophysical simulation models do permit the investigation, in a relatively

short period of time, of field-level resource effects of farming practices, and they can be made

readily available to many potential users. Moreover, they make it possible to extend the

analyses into the hypothetical future. They have been extensively validated over a wide range

of environments, and are increasingly used to answer questions related to sustainability (Bowen

etal., 1992).

4.2.3. Integrating Economic and Biological Models

Regional agricultural models which incorporate process models have been used to

investigate regional agricultural sustainability. Ereshko et al. (1988) found optimal sustainable

production strategies for a region using a regional simulation model in which outputs of a crop

production simulation model were inputs for production and resource accounting models.

Wind and water erosion were estimated using the universal soil loss equation and the Kansas

Manhattan wind erosion model; soil transformations were updated annually. Kitamura et al.

(1988) employed a recursive linear programming model with a runoff simulation model to

maximize regional gross receipts constrained by water resources and water pollution.

At the sectoral level, Kaiser et al. (1993) used a "unifying" model made up of a

weather generator, a dynamic simulation crop model, and an optimizing farm-level economic

model to examine climate change impacts on the farm sector. In this effort, since the crop

models simulate potential yield rather than actual farm yields, all simulated yields were

normalized to county averages for the 1980s by multiplying each simulated yield by the ratio of

the 1980s county average yield to the simulated potential 1980s average yield. They then

applied their methodology to examine the effects of several climate change scenarios on a

hypothetical southern Minnesota grain farm.

Faeth et al. (1991) developed a framework which combined the EPIC

(Erosion/Productivity Impact Calculator) simulation model with an agricultural sector model

and an accounting model in order to quantify economic, fiscal, and environmental costs and

benefits of agricultural policy options. The EPIC model was linked with a farm-level

programming model and regional estimates of off-site damage per ton of eroded soil to

calculate soil erosion, off-site damages, and a soil depreciation allowance, as well as crop yields

and sales, production expenses, government deficiency payments, and net farm incomes for

each cropping pattern.

In a farm-level model, Parsch et al. (1991) combined biophysical simulation and

stochastic dominance to analyze wheat followed by soybean double-crop. They interfaced

CERES-Wheat (Ritchie etal., 1990) and SOYGRO (Jones etal., 1989) models to account for

wheat drydown after maturity, and soil moisture from wheat maturity to soybean planting.

Stochastic dominance techniques were used to analyze the effects of wheat harvest timing

(grain moisture content) and the resultant soybean planting date on overall net returns.

Farm models incorporating crop growth simulation have been applied to resource

management problems. Johnson et al. (1991) used the nitrogen loss output of the CERES

family of crop simulation models integrated with dynamic economic models to evaluate on-

farm costs of strategies to reduce nitrate groundwater pollution. Kim and Mapp (1993)

employed the EPIC-PST model to complete the coefficient matrix of a linear programming

model in order to evaluate alternative scenarios for achieving a certain level of water pollution

control. A crop and erosion simulation model was integrated into a recursive programming


model to look at wind erosion susceptibility (Lee et al., 1989) and irrigated cropping decisions

under exhaustible groundwater supply (Lee and Lacewell, 1990) for typical crop farms in the

Texas high plains. EPIC was used in a study of rice-wheat rotations in the Indian Punjab in

order to estimate soil erosion and current and future crop yields for each rotation under

different tillage and fertilization alternatives (Faeth, 1993). Hombaker and Mapp (1988)

combined a crop growth model with recursive programming to analyze the potential irrigation

water savings from adopting scheduling and low pressure technologies for three different

irrigations systems in Oklahoma. Crop production decisions under differing risk attitudes were

examined by Dillon et al. (1989) for the Texas Blacklands using a quadratic programming

model into which simulated yield data were integrated.

Boggess and Ritchie (1988) used crop simulation models [CERES-Maize (Jones and

Kiniry, 1986) and SOYGRO] to link information about the response of corn and soybean to

irrigation with economic analysis and risk assessment. Boggess and Amerling (1983) used a

net present value model, into which crop simulation models are embedded, to analyze the risks

and returns of irrigation investments in humid regions, with particular attention paid to the

impact of variations of weather patterns. In both cases, stochastic dominance techniques were

used to identify risk efficient outcomes.

4.3. An Integrated Whole-Farm Sustainability Model

The whole-farm sustainability model for this study consists of two basic components,

an economic decision-making component and a simulation component that models biophysical

processes and sets up the succeeding economic component (see Figure 1.1). The integrated

model is run over a sequence of years, the time frame of which depends on the specific

situation and objectives of the analysis. A twelve-year time frame is used for this research since

two-, three-, and four-year rotations cycle completely in twelve years, and technology can be

assumed constant over this time frame. For longer time periods, technological change would

have to be considered. To avoid this modeling problem, longer-term and multi-generational

sustainability concerns are captured in the constraints or other sustainability criteria, by

providing for limits on resource degradation and use.

4.3.1. Simulation Component Details

The simulation component models crop growth, soil water, and soil nutrient processes,

and calculates "actual" prices (see Determination of prices and yields in section 4.3.2) to be

used in harvest and marketing decisions. It also transfers inventories to the next period, and

computes year-end income, consumption, and savings. The core of this component is the

simulation of biophysical processes, especially crop growth and soil nitrogen changes.

The crop growth models which are used in this research are those included in the

Decision Support System for Agrotechnology Transfer, or DSSAT (IBSNAT, 1989).

Currently, DSSAT models exist for maize (CERES-Maize, Jones and Kiniry, 1986), soybean

(SOYGRO, Jones et al., 1989), peanut (PNUTGRO, Boote et al., 1989), wheat (CERES-

Wheat, Ritchie, 1985), drybean, and other crops. They are mechanistic models which integrate

component processes in the soil-plant-atmosphere system. They respond to weather, soil

moisture, and soil nitrogen to simulate crop growth from emergence to physiological maturity.

Using user-supplied soils and management variables and stochastic weather inputs, the crop

models generate empirical distributions of crop yields, organic carbon and nitrogen levels,

nitrogen leaching, and other variables.

Stochastic weather for each year is generated by a weather generator internal to DSSAT

v3, with coefficients estimated from actual weather data for the area. This multi-year


simulation then is replicated a number of times using different stochastic weather sequences. In

this way, a distribution of system behavior is generated and probabilities can be attached to

different model outcomes.

The nitrogen model used in the DSSAT suite of crop models has been tested under a

variety of environments from diverse locations spanning the world's cropping regions (Godwin

and Jones, 1991). The CERES-N model is the basic nitrogen model adapted and used for the

DSSAT. It describes mineralization and/or immobilization ofN associated with the decay of

crop residues, nitrification, denitrification, urea hydrolysis, leaching of nitrate, N-fixation (for

legume models), and the uptake and use of N by the crop. In this model, only nitrate and urea

are leached; ammonium is assumed not to be transported across soil layers (Godwin and Jones,

1991). Bowen et al. (1992) tested and adapted the CERES-N model to realistically simulate

legume N availability to succeeding maize from green manures.

The DSSAT v3 recently has been adapted to allow sequencing of the crop models,

including fallow periods, which permits more realistic simulation of crop sequences and

rotations. The particular rotation and number of years to be simulated are specified. The soil

model then runs continuously over this period, allowing assessment of the soil processes and

the resulting effects on crop growth over time.

The DSSAT v3 grain legume models also have been adapted to account for pest

damage. The type and amount of damage must be specified by the user, either by inputing pest

populations and pest feeding rates, or by inputing measurements of actual pest damage.

4.3.2. Economic Model Specification

A multi-period programming model with a recursive structure is used to model the

farmer decision processes over time. The objective function of the decision model is to

maximize annual before-tax family income, constrained by resource limitations as well as

sustainability concerns. The farm model actually consists of two separate decision models,

MP1 and MP2 (see Figure 1.1). MP1 models planting decisions based on expected yields and

prices, and covers both decision periods for the year. MP2 models harvesting decisions based

on simulated "actual" yields and prices, and covers decision period 2 of the current year. If fall

planting is an option, decision periods 1 and possibly 2 of the following year are included in

MP2, in which case expected yields and prices for the winter crop are appropriate for decisions

made in MP2.

In model MP1, two types of activities are required, production activities and harvest

activities. Production activities are by rotation and harvest activities are by crop. Consistency

constraints are required to ensure that hectares of crop harvested do not exceed hectares

planted in rotation, and crop balance equations convert hectares harvested to kilograms sold or

stored. Production and subsequent harvest are constrained by available cropping land, labor

and equipment resources, cash availability, and environmental considerations. The model is

also constrained by a consumption or minimum income requirement and a borrowing

limitation, both of which impact cash availability. For each year model MP1 and MP2 are

structured as follows:

Max Z = PgS + U xJ- EE Yh-F (1)
m g m j m h

subject to (2)
A4x + E_Ay b (2)
f h

EDj s ek (3)

cE + E, + R. EmS W o (4)
J h g

W, L,

Yh Vxj 0

ES EYhqh
m h

x., yh 0

Where j =
g =
h =
k =
m =
P =
x =
y =

P =
S =
F =
A =
b =
D =
e =
R =
L =
v =
q =

index of production activities (rotations),
index of crops sold,
index of harvest activities,
index of resource or constraint type,
index of accounting periods,
index of production periods,
vector of production activities,
vector of harvest activities,
vector of costs,
vector of prices,
crops sold,
other income, including off-farm income,
total fixed costs,
resource input coefficients for unit levels of the activities,
levels of constraining resources,
environmental input coefficients for unit levels of the activities,
constraining levels of environmental resources,
consumption requirements,
amount borrowed,
maximum loan limit,
ratio of number of years of harvested crop h to number of years of total rotation,
yield of crop g in harvest activity h.

The objective function (1) maximizes total farm family income before taxes. Farm resources,

such as land, labor, and equipment are constrained in (2), and environmental resources are

constrained in (3), where e might represent a maximum allowable level of pollution. A cash

flow constraint in each accounting period is represented in (4), with borrowing limited in (5).

The consistency constraints (6) requires that the hectares of crop harvested not exceed the

hectares planted, while the crop balance equations (7) convert hectares of each crop harvested

to kilograms of crop available for sale. MP1 and MP2 differ in that MP2 will generally use

actual rather than expected prices and yields, time periods indexed by m andp will be different

in each model, and certain activity levels are known in MP2.

The recursive structure of the economic component of the farm model allows

parameters and decision rules to change over time in response to past results and to changes in

economic and environmental conditions. Provisions are made to change constraints over time

to reflect changing goals, objectives, and circumstances; and to account for period-to-period

carry-over effects. Examples of these include changes in soil characteristics, farm resources,

financial conditions, off-farm income sources, and farm and environmental policies; and in

structural changes in the farm household and farming enterprises.

This farm model provides, through simulation, estimates of the biological and physical

effects of different farming activities and practices, analyses of critical economic factors under

stochastic yields, and investigation of optimal decisions under various environmental and

socioeconomic constraints. Moreover, it allows these analyses to be performed for both short-

run and long-run considerations. Information, including results and knowledge from past

periods and expectations about the future, is used to the extent it is available.

Determination of prices and yields. Expected yields and prices are determined from

weighted averages of recent years. An exponential smoothing function with a weight of 0.33

on the most recent observation, which results in an estimate similar to a 5-yr moving average

(Dobbins and Mapp, 1983), is used to estimate expected yields and prices for a particular year.

For example, expected yields for year t are estimated by

EY, = 0.33Y,. + 0.67EY,. (9)

where E is the expectation operator and Yis yield.

Simulated "actual" yields are taken from crop simulation models in the case of crops

for which models are available. For other crops, yields under each management regime should

be calculated as a function of weather if this relationship is known or such data are available.

Current "actual" prices may be specified and varied parametrically to determine the sensitivity

of system sustainability to price fluctuations; they may be determined stochastically through a

random draw from empirical distributions for program crops and for other crops whose local

production does not affect prices; or in the case of crops whose prices depend on local

production, the functional relationships may be estimated econometrically where data are


Modeling environmental uncertainties. Certain sustainability constraints will require

limiting pollution originating from the farming system. Improved monitoring capabilities and

recent advances in computer simulation techniques provide both farmers and regulators with

much more information regarding contamination effects of management practices. Yet,

because of stochastic rainfall patterns, a great deal of uncertainty surrounds this information. In

mathematical programming, uncertainty in the technical constraints, either in the technical

coefficients or in the constraint levels (RHS) can be dealt with using stochastic programming

(SP) techniques, such as discrete stochastic programming (DSP) and chance-constrained

programming (CCP).

Ideally, all the risks regarding resource requirements and supplies (ie., those in the

constraint set) should be transferred to the objective function and a single risk decision rule

applied. However, this requires that these resources be freely tradable so that stochastic

discrepancies between resource supplies and requirements can be captured in the objective

function through buying and selling activities. DSP techniques (Cocks, 1968), often referred

to as stochastic programming with recourse (SPR), may be used in these cases where

decisions are sequential and the probabilities of events are known or can be estimated. Such

models may often be represented with decision trees.

When resources are not freely tradable, any risk in resource use or supply cannot be

transferred to income risk through buying and selling activities. This describes the situation

of an imposition of a constraint on the production of a social or environmental "bad," such as

pollution. Such problems are more easily handled with chance-constrained programming

(Charnes and Cooper, 1959). In this approach, the objective function is optimized subject to

a set of constraints in the form

Pr{ aJ, x bk 1-a, for all k (10)

where at is the acceptable probability of exceedence of the kth constraint, Pr is the

probability operator, a, is the stochastic requirement of the kth resource under thejth

activity, and bk is the imposed constraint level of the kth resource.

In the case of uncertain technical coefficients, if their variability is ignored and only

the mean is used in the model, assuming normality, the resulting optimal farm plan will fail

to meet the constraint requirements 50 percent of the time. Probabilistic or chance

constraints can be used to raise the probability of meeting specified constraints. In the case

considered here, this is important if regulations require that the upper limit on contamination

be observed with a prescribed degree of certainty or if penalties are imposed for violation of

this upper limit.

The simplest form of the chance-constrained problem is when all but one of the

constraints are deterministic. If more than one constraint is stochastic, it becomes necessary


to assume that the stochastic coefficients in all the constraints are statistically independent of

each other across constraints (Hazell and Norton, 1986). (Coefficients within a constraint do

not need to be statistically independent.) Tractable methods for handling jointly dependent

constraints have not yet been developed.

Another problem with chance-constrained programming is its weak theoretic

underpinnings. CCP only implies satisficing rather than utility maximizing. Further, it

ignores recourse decisions which generally characterize real decision problems involving risk

(Hogan et al., 1981). CCP does not indicate what to do if the recommended solution is not

feasible. It is, in this sense, an incompletely specified stochastic programming model that

does not account for the expected value of perfect information. Finally, CCP suffers from

the arbitrary choice of probability levels, which may be inconsistent with the type and degree

of risk averse behavior assumed of the decision maker.

Despite these shortcomings, chance-constrained programming still remains attractive

in certain situations for certain types of problems. Examples of CCP applications include

accounting for uncertain protein content of ingredients when looking for least-cost feed

formulations (Chen, 1973), and considering stochastic pollution discharges when trying to

determine optimal regulatory control of nonpoint water pollution sources (Milon, 1987).

Chance constraints on environmental resources may be included in (3) in model MP1.

In the next chapter, an application of the whole-farm sustainability model is discussed.

The simple example demonstrates the richness and flexibility of the approach to analyzing the

sustainability of a farming system.


5.1. A Suwannee County Peanut Farming System

Suwannee County is a traditional agricultural county characterized by small- to

medium-scale mixed crop and animal farming systems. Its soils are predominantly deep rolling

sands, characteristic of much of North and Northcentral Florida. It is the largest tobacco

producing county in Florida, annually generating $8.2 million in sales (Robert S. Tervola,

Suwannee County Extension Director, personal communication). However, difficulties in

getting labor in a timely manner has led many farmers to give up tobacco. Peanuts are another

important agricultural commodity in Suwannee County, accounting for over $5 million in sales

in 1991 (Tervola, personal communication). Finally, as traditional row crops such as corn and

soybeans have become less profitable, there has been recent growth in specialty and vegetable

farming, and in pine plantations (Tervola, personal communication).

Nitrate contamination of groundwater is the environmental problem that currently is of

greatest concern in Suwannee county. Farmers have traditionally applied high doses of

nitrogen fertilizer to their crops due to the inherent infertility of the sandy soils and frequent

hard rain events. Paraphrasing a Suwannee County tobacco grower, "we put on the fertilizer as

recommended, then it rains for two days, and a couple of days later the leaves start turning

yellow, so we have to put the nitrogen on again." In addition, dairy farming is on the rise in the

county with milk production increasing from 26.9 million pounds in 1984 to 81 million pounds


in 1993, and poultry production is increasing with the recent expansion of the Gold Kist plant

near Live Oak, the county seat. The result is increasing amounts of animal waste which

potentially can be used as fertilizer, and which create potential groundwater contamination

problems. Finally, non-farming people have been increasingly moving into the rural areas of

the county due at least in part to the natural beauty of the Suwannee River which borders the

county on three sides. This is causing some not unexpected land use conflicts, as well as

increasing concern over water quality for both drinking and recreation.

5.1.1. Defining a Hypothetical Peanut Fanning System

A peanut-based farming system was selected for analysis because of the crop's relative

importance in Suwannee County, because of the peanut farmers' stability and viability relative

to growers of other traditional commodities, and because peanuts can be modeled using the

DSSAT crop models. Peanut production increased significantly in the county in the early

1980s, from about 590 ha harvested in 1980 (and a 9th ranking among Florida counties) to

over 1,800 ha harvested in 1984 (a ranking of 4th in the state). Production has remained

relatively steady at this level since then.

A synthetic, hypothetical peanut farming system is defined for this study based on

interviews with seven Suwannee County peanut growers, as well as with extension personnel

and suppliers of farm inputs and farm credit. (See section 1.5.2 for a discussion of representing

a farming system.) Additional information was obtained from bulletins and publications

provided by the Florida Cooperative Extension Service and from surveys conducted by

University of Florida Agricultural Economists, Steve Ford and Tim Hewitt, in the late 1980s.

To protect the confidentiality of the growers, no names are revealed here, nor are any individual

farms described. All of the farms were unique in certain aspects, yet there were many common

features as well. The information gathered through the interviews was synthesized into one

farm for the purpose of defining the whole-farm model described here. While the synthetic

farm is not claimed to be either typical or representative, it does fairly represent the peanut

farming situation in the region, and could describe a viable peanut producing operation.

The survey process. The names of peanut growers were provided by Robert S.

Tervola, Suwannee County Extension Director. Seven of the growers contacted agreed to

participate in the interview process, and visits were made during the first two weeks in March,

1993. Informal survey techniques were employed along the lines of farmer visits in rapid rural

appraisal or Sondeo methodology (Hildebrand, 1982). The objective of these visits was to get

a general feel of the farming operations, and to gain an understanding of the circumstances and

constraints under which the growers operate. Detailed descriptions of the cropping systems

were solicited as time and circumstances permitted. However, since a synthetic fannrming system

was sought, details that could be obtained from secondary sources were not explicitly requested

in order to avoid taking more of the farmers' time than necessary. In general, the growers

interviewed were very open and willing to share information about their farms. Sustainability

per se was not discussed with them because of the negative connotations surrounding

sustainable agriculture among conventional farmers in this region. However, all the farmers

were very interested in the stability and future viability of their farms, which are crucial

sustainability issues. All but one of the growers were visited a second time for follow-up

interviews in late March.

Qualitative information. One of the strengths of the informal survey approach is the

ability to obtain qualitative information, which guides model building by helping to define

constraints and to determine appropriate questions to "ask" the model. All growers expressed

concern for the future of the peanut program with the pending (at that time) NAFTA and

GATT agreements. Only one farmer indicated that he could continue to grow peanuts if the

quota were removed, assuming that the price would be that currently received for additional

peanuts (peanuts grown in addition to quota peanuts). The farmers generally felt that they were

entitled to the high and stable prices provided through the commodity programs, contending

that their inputs were provided through markets that were not free, dominated by large

companies that controlled prices, yet they were expected to compete in open, free markets for

their output. They argued that farming was different than other businesses in that if there was

an oversupply, they did not have the option to hold back their products until the price went up.

Another concern commonly expressed by the growers was over the difficulties of

keeping the farm intact in the family in the face of rising property and inheritance taxes. Those

who had children who were interested in maintaining the farm expressed a keen desire to pass

the farm on intact, more for the quality of life and the tradition that went along with fanning

than for the business potential. In fact, one grower's son planned on doing whatever it took to

keep the farm intact for his children, even if it meant diversifying into more nontraditional types

of enterprises, such as vegetable production, though such prospects scared him. All indicated

that, in the face of financial difficulties, they would liquidate their land holdings only as a last


Concerns for the natural environment were not expressed by any of the growers, and

there was no evidence of any farmer-caused degradation. Some identified government

regulations of chemicals as a problem, but none expressed serious concern over this issue.

Economics, not regulations seemed to dictate their choices not to use chemicals. None of the

growers used nematicides, yet all rotated out of peanuts at least three years to control

nematodes. Availability of reliable labor was also cited as a problem, and the reason most have

gotten out of tobacco, which seasonally is very labor intensive.

A Suwannee County peanut farming system. Peanut farming is most common in the

southeastern part of the county. A common soil in this region, Blanton fine sands, is a member

of the loamy, siliceous, thermic family of Grossarenic Paleudults. It consists of moderately

well drained, moderately permeable soils that formed in sandy and loamy marine materials.

They are nearly level to moderately sloping soils with no surface runoff; water drains rapidly

through the soils into the porous substrata. The underlying limestone is typically 100 cm to 400

cm below the soil surface. Farmers indicated that their soils were deep and well-drained, and

did not vary significantly from field to field. On individual farms, management of peanuts was

the same on all fields.

The hypothetical farming system defined for this study consists of 162 hectares (400

acres) of available cropping land, all on Blanton fine sands, and the farmer has a peanut quota

totalling 68,000 kg (150,000 lbs.). The farmer rotates peanuts such that a field is in peanuts

one year out of every four in order to control disease and insects, primarily nematodes. He has

the choice of two basic rotations: one year in peanuts followed by three years in bahia pasture,

or one year in peanuts followed by two years in corn and then one year in weed fallow. The

corn in the second rotation may be produced under varying fertilizer regimes, and is typically

grown without herbicides or insecticides. All crops are grown without irrigation, the common

practice in the region, though some farmers do have irrigation available for a portion of their

fields and will use it to save their crop. The farming system also includes a 100 cow-unit cow-

calf operation that utilizes the bahia pasture. It is assumed that pasture is available for rent

should inadequate pasture be available on the farm, and conversely, that surplus pasture can

either be rented or hayed for returns equivalent to those realized by grazing.

Peanuts is not a labor intensive crop, so labor typically only needs to be hired at harvest

at which time it is important to get the peanuts out of the ground when they are ready. None of

the farmers interviewed had equipment constraints, and some even had surplus equipment.

Since peanuts are so profitable, it pays for the grower to invest in the necessary equipment to

get the job done. Apparently, good used equipment is relatively easy to find as long as needs

are identified far enough in advance so that the equipment may be purchased when it becomes

available and not under duress. Peanut equipment, however, is very specialized and cannot be

used for other crops. This applies a certain rigidity to the farming system which could be

problematic if the peanut quota system were eliminated and farmers were forced to consider

producing an alternative crop.

The farm family consists of the grower and his wife (all farmers interviewed were

male) and their two children. The woman works full time off the farm at a salary comparable

to that earned in clerical/secretarial work, bringing in $4,930 every four months. Minimum

annual living expenses are $24,000 for the family. The farmer may borrow up to $40,000 for

operating and/or living expenses, based on roughly half the value of the projected peanut crop.

Typically, however, peanut farmers have enough savings that they rarely need to borrow for

operating expenses. Most debt incurred by peanut farmers is intermediate term for equipment

or additional peanut allotments (Russell Pope, personal communications). As was the case for

all the farmers interviewed, the land was inherited and is owned outright. (Most farmers

commented that they could not operate if they had to pay a mortgage on their land. They all

concurred that it would be virtually impossible for someone to start out farming and make it if

they had to purchase the land at today's prices.)

Peanuts are generally planted in the first or second week of April, depending on

temperature and rainfall. Land preparation for peanuts typically involves three tillage

'Personal interview with Russell Pope, Branch Manager of North Florida Farm Credit Service in Live
Oak, FL, on March 4, 1993.

operations during late winter and early spring: incorporating residue with a cutting disc,

plowing, and then a second discing. Prior to planting, a pre-emergent herbicide, usually a tank

mix of Dual and Prowl, is sprayed and incorporated, and fertilizer is broadcast and

incorporated as well. A commonly used fertilizer is a 3-9-18 mix with micronutrients, applied

at a rate of 336-784 kg per ha (300-700 lb/acre). Lime is typically applied to North Florida

soils at the rate of 2.25 metric tons per hectare (about one ton per acre) every three to four

years. Some farmers perform a drag and sweep operation about five days after planting.

Frequently, a post-emergent herbicide like Gramoxone is applied about two weeks after

planting. Peanuts are typically cultivated twice during the season, once in mid-May and again

in early to mid-June. Growers spread gypsum in mid- to late May at the rate of about 2.25

metric tons per hectare (1 T/acre). Biweekly fungicide applications (usually Bravo 720) begin

in early June and continue through August. The peanuts are usually ready to harvest in early

September, and when the time is right, it is imperative that the peanuts be gotten out of the

ground to dry, then combined in as timely a fashion as possible. Harvest is generally the only

time that peanut growers often have to hire some help.

Since peanuts are so profitable, they drive the cropping system. The other crops in the

rotation are of secondary importance, and hence, are not managed intensively. bahiagrass is

commonly rotated with peanuts as improved pasture, producing good grazing from April to

November. It is established in the spring the year following peanuts and is usually kept for

three years. Land preparation is similar to that for peanuts, with two to three tillage operations

required to produce a good seedbed. Planting is in late March to early April in order to avoid

the weed competition that results from a later planting. In the establishment year, fertilizer is

applied at planting to provide 34 kg/ha of nitrogen, 15 kg/ha of phosphate, and 30 kg/ha of


potash. Approximately 40 days after planting, 78 kg/ha of additional nitrogen are applied. In

the second and third years, 240 kg/ha of 23-5-10 are applied in the spring.

Another suitable rotation crop for peanuts is corn, though corn has become less

common in recent years as its price has dropped, making it virtually impossible to cover

production costs. The corn that is still grown is primarily by those farmers who still raise hogs.

Normally, three tillage operations are performed in late winter to prepare the ground for corn.

Planting is in mid to late March, with 336 to 560 kg/ha of 3-9-18 or 5-10-15 applied during

planting. Dryland corn grown in rotation with peanuts is typically produced without herbicides

or insecticides. Two cultivations are performed, one in late April and another three to four

weeks later. The corn is sidedressed or injected with anhydrous ammonia in early May with 70

to 90 kg/ha of N. Corn harvest generally takes place in late August or September, but gives

way to the higher priority peanut harvest if the time conflicts.

5.1.2. Production Costs and Expected Returns

Farmers reported that it costs from $300 to $500 per acre to grow peanuts, or about

$740 to $1200 per hectare. At an estimated yield of 3360 kg per hectare, a price of $0.22 to

$0.36 per kg, or about $200 to $325 per ton, is required to cover costs. In many years this is

more than the price received for additional peanuts. However, the support price for quota

peanuts is consistently much higher than this, assuring growers who either own or rent quota a

healthy profit in all but the worst years.

Costs of production in this study are estimated from information provided by growers,

supplemented with data from "Cost of Production Tables for North Florida" prepared by

Timothy D. Hewitt, Extension Economist, Food and Resource Economics Department,

University of Florida, NFREC, Marianna (Hewitt, 1990-93), "Costs of Owning and Operating


Farm Tractors and Equipment, Florida, 1990" prepared by Timothy D. Hewitt, and "Machinery

Operating Costs, 1992" (Doane's Agricultural Service, 1992). Production costs for each of the

crops are detailed in Appendix A. The non-labor variable costs of producing one hectare of

peanuts are estimated to be about $890, of which just over $700 are pre-harvest production

costs with the remainder attributable to harvest and preparing for sale. The most expensive

production inputs for peanuts are crop protection chemicals, which total almost $250 per

hectare, and seed that is about $190 per hectare. Drying and cleaning account for nearly $150

per hectare at expected yields. Although extension publications suggest that management, and

hence production costs, are greater for quota than for additional peanuts, all of the growers

who were interviewed reported that they treated their quota and additional acreage the same.

Variable non-labor costs for producing corn and bahiagrass are considerably less than

for peanuts on a per hectare basis. Three different fertilizer options are considered for corn

production in this study: low nitrogen that is 10 kg/ha at planting and 70 kg/ha topdressed,

medium nitrogen that is 25 kg N/ha at planting and 70 kg N/ha topdressed, and high nitrogen

that is 25 kg N/ha at planting and 90 kg N/ha topdressed. Other options were simulated, but

under North Florida environmental conditions, dryland corn did not respond significantly to

higher nitrogen levels. The variable costs of growing a hectare of corn at the low nitrogen level

is about $185 (of which $15 are harvest costs), at the medium nitrogen level is about $195, and

at the high level is about $205. In the establishment year for bahiagrass, non-labor variable

costs are about $165 per hectare, and it costs about $50 per hectare to maintain bahiagrass in

each of the succeeding two years.

Labor usage is estimated based on farmer interviews, farmer surveys conducted by

Ford and Hewitt, and "Cost of Production Tables for North Florida" prepared by Timothy D.

Hewitt. Details of the labor estimates for each rotation are presented in Appendix B.


Additional labor is available and can be hired for five dollars an hour. The only period of the

year in which hired labor must be used is during the peanut harvest. It is not uncommon for

growers to have two peanut digger/inverters and two, three or even more peanut combines in

order to keep the harvest going smoothly. Farmers may share harvest equipment and help each

other with the peanut harvest, but they typically hire outside labor as well. Lack of harvest

equipment does not appear to be a constraint.

Initially, expected yields for peanuts are estimated at 3000 kg per hectare, about 12

percent less than the 3360 kg/ha (3000 lb/acre) used by the farmers interviewed, but higher

than the average yields for non-irrigated peanuts in the county. Because this figure is used for

making planting decisions and growers want to be sure to make their quota, it is a conservative

forecast of yield. Some peanut growers contracted their crop before planting to lock in a price.

In 1993, the contract prices for quota and additional peanuts varied some, but typically was

$0.716 per kg ($650/ton) for quota peanuts and $0.386 per kg ($350/ton) for additional. At

an expected yield of 3000 kg per hectare, quota peanuts can provide a gross margin of over

$1200 per hectare (excluding labor costs), while additional can return nearly a $270 gross

margin exclusive of labor costs.

For corn, initial expected yields range between 3100 and 3300 kilograms per hectare,

or about 50 bushels per acre, which is in the expected range for non-irrigated corn in the

county. At a corn price of $0.10 per kg, or $2.55 a bushel, and expected yields of 3125 kg/ha,

3275 kg/ha, and 3325 kg/ha for low, medium, and high nitrogen applications respectively, corn

returns are between $127.50 and $132.50 gross margin per hectare exclusive of labor. No

value (or cost) is assigned to the weed fallow that follows two years of corn, although cattle are

probably allowed to forage on this land.

The value of bahiagrass pasture on a per kilogram basis was calculated from the

average value of calves sold per hectare for 181 days of grazing, plus half the value of culled

animals, less a portion of other direct costs, divided by the average bahiagrass yield. The value

of calves sold per hectare for the grazing period is derived from 60 calves at 200 kg each on 81

hectares at the prevailing price per kilogram of weanlings. Other direct costs contributing to

calf production during grazing include soybean meal and grain, salt and minerals, veterinary

costs, one replacement bull, and operation and maintenance, totalling $8,000, of which $3,250

can be assigned to the grazing period, giving a total of $40.12 in other direct costs per hectare

of grazing.2 The salvage value of culled animals totals $10,230, which contributes $63.15 per

hectare of grazing. At a price for calves of $1.98 per kg ($90/cwt.), the average value of one

hectare of bahiagrass pasture is $171, returning an average gross margin of $83 per hectare

(see Appendix A for details). The value of a hectare of bahiagrass pasture varies depending on

bahia yields and cattle prices.

Cattle income is from the cow-calf enterprise. A 100-cow unit operation typically

yields 60 calves for sale at a weaning weight of 200 kg. At $1.98 per kg, gross income from

sale of calves is $23,760. An additional $10,230 can be expected from the sale of culled

animals (21 cows, 3 replacement heifers, and 1 bull) of which half has been credited to the

bahiagrass pasture, yielding a total gross income from this operation of $28,875. Costs of

production include hay at $7280, other direct costs of $4750, and the value of the bahiagrass

pasture which is $13,851 at average yields and expected beef prices, leaving a gross margin

from the cow-calf operation at $2,994. This figure will change with changing bahia yields and

beef prices.

'Calculated from enterprise budgets for cow-calf operations, assuming 100-cow units on 81 hectares
(200 acres) and selling 60 calves (Simpson, in press).