Embed Size (px)
Citation preview
Understanding Farmer Conservation Behavior: A Behavioral Economics Test of Tillage
Decisions in Nebraska and Kansas
by
Robert J. Sheeder
A THESIS
Presented to the Faculty of
The Graduate College at the University of Nebraska
In Partial Fulfillment of Requirements
For the Degree of Master of Science
Major: Agricultural Economics
Under the Supervision of Professor Gary D. Lynne
Lincoln, Nebraska
December, 2008
Understanding Farmer Conservation Behavior: A Behavioral Economic Test of Tillage
Decisions in Nebraska and Kansas
Robert Sheeder, M.S.
University of Nebraska-Lincoln, 2008
Adviser: Dr. Gary D. Lynne
The intent of this research is to discover what factors motivate farmers to adopt
conservation technologies that help reduce or eliminate non-point surface water pollution.
Particular attention is paid to the role that tillage decisions play in improving surface
water quality in the Blue River/Tuttle Creek Lake Watershed located in Nebraska and
Kansas.
Data for this research was collected via a survey instrument sent to farm operators
located in a four county target area situated upstream of Tuttle Creek Lake in
southeastern Nebraska and northeastern Kansas. These farmers were asked several
questions regarding personal beliefs and attitudes regarding the usage of conservation
tillage measures and other soil BMPs. Also, respondents were asked to provide more
technical information regarding farm processes, such as how many acres are farmed
under conservation tillage technologies.
Using the data obtained from the farmers in the watershed, logit and Heckman
(Heckit) models were constructed in order to empirically test an emerging theory in
Behavioral Economics that has been named “Metaeconomics.” This emerging theory
looks to transcend traditional microeconomic theory by incorporating not only financial
concerns, but also proxies for self-interest and shared other-interest psychological
tendencies, measures of autonomous and heteronomous control, and habitual tendencies.
Results from the logit models indicate that farmers who condition their pursuit of
self-interest with shared other-interest as represented in empathy and sympathy are more
likely to adopt conservation tillage strategies. We also find that farmers that believe they
cannot maintain autonomous control over farming practices when using conservation
tillage are less likely to use the technology. Also, habitual tendencies are found to be a
large driver in the conservation tillage adoption decision. Finally, results from Heckman
models show that preferences for control impact the conservation tillage intensity
decision. These findings, then, lend credence to the idea that an intricate mix of financial
incentives and moral suasion may be required in order to convince farmers to incorporate
conservation tillage strategies on working farms
i
TABLE OF CONTENTS
Table of Contents ________________________________________________________ i
List of Tables __________________________________________________________ iii
List of Figures __________________________________________________________ iv
Introduction
1.1.Questioning the Role of Profit-Maximization in Conservation Decisions _______ 1
1.2. View of Human Behavior from Neursoscience and Evolutionary Biology ______ 5
1.3. Parental Models and Overview of Metaeconomic Theory ___________________ 9
Review of Literature
2.1. Financial Motives for Adopting Conservation Practices ___________________ 17
2.2. Non-Financial Motives for Adopting Conservation Practices _______________ 21
2.3. Multiple-Motive/Multiple Utility Studies of Conservation Adoption _________ 24
Theoretical Model
3.1. Graphical Representation of Standard Production Economics _______________ 28
3.2. Mathematics of Standard Production Economics _________________________ 30
3.3 Graphical Representation of Metaeconomics ____________________________ 31
3.4 Mathematics of Metaeconomics ______________________________________ 36
Toward Empirical Testing
4.1. Physical Description of Study Area ___________________________________ 45
4.2. Description of Institutional Arrangements in Study Area __________________ 46
4.3. Empirical Models _________________________________________________ 50
4.4. Development of Survey Instrument and Data Collection ___________________ 52
4.5. Description of Variables ____________________________________________ 54
ii
Results and Discussion
5.1. Summary Descriptive Statistics ______________________________________ 68
5.2 Correlations _____________________________________________________ 75
5.3 Results of Logit Test of Microeconomic and Metaeconomic Theory _________ 79
5.4 Results of Heckman Test of Conservation Tillage Intensity ________________ 88
Conclusions, Implications, and Recommendations ____________________________ 99
Reference List ________________________________________________________ 110
Appendix A __________________________________________________________ 118
iii
LIST OF TABLES
Table Page
1. Mean Responses to Question 24 (Selfism Scale) 68
2. Mean Responses to Question 22 (Davis Empathy Scale) 69
3. Mean Responses to Question 23 (Sympathy Scale) 69
4. Mean Responses to Question 19 (Empathy/Others Water User) 69
5. Mean Responses to Question 19 (Empathy/Others Farm Entity) 70
6. Mean Responses to Question 19 (Empathy/Others Family) 70
7. Mean Responses to Question 15 (Farm Control Scale) 70
8. Mean Responses to Question 15 (Other Control Scale) 71
9. Mean Responses to Question 15 (Nature Control Scale) 71
10. Mean Results of Final Selfism Variable 71
11. Mean Results of Final Empathy/Sympathy Variables 72
12. Mean Results of Final Control Variables 72
13. Mean Results for Final Selfism*Emapthy/Sympathy Variables 72
14. Mean Results for Final Selfism*Control Variables 72
15. Mean Results of Final Income Variable 73
16. Mean Results of Final Soil Slope Variable 73
17. Mean Results of Final Habit Variable 73
18. Correlations between Various Behavioral Proxies 77
19. Logistic Estimation of No-Till Adoption Decision (Empathy Proxy) 80
20. Logistic Estimation of No-Till Adoption Decision (Sympathy Proxy) 81
21. Logistic Estimation of No-Till Adoption Decision (Empathy/Others Proxy) 82
22. Probit Estimation of No-Till Adoption Decision (Empathy Proxy) 89
23. Semi-Log Estimation of Individual No-Till Intensity (Empathy Proxy) 90
24. Probit Estimation of No-Till Adoption Decision (Sympathy Proxy) 91
25. Semi-Log Estimation of Individual No-Till Intensity (Sympathy Proxy) 92
26. Probit Estimation of No-Till Adoption Decision (Empathy/Others Proxy) 93
27. Semi-Log Estimation of Individual No-Till Intensity (Empathy/Others Proxy) 94
iv
LIST OF FIGURES
Figure Page
1.1 The Major Ranges/Modes of the CSN Model 16
3.1 Traditional Microeconomic Self-Interest Isoquant Curves 42
3.2 Metaeconomic Isoquant Curves 43
3.3 Metaeconomic Interests Frontier 44
4.1 Blue River/Tuttle Creek Watershed 67
5.1 Distribution of Mean Empathy Scale Responses 98
5.2 Distribution of Mean Sympathy Scale Respones 98
1
INTRODUCTION
1.1. Questioning the Role of Profit-Maximization in Conservation Decisions
Since the destruction and despair caused by the dust bowl of the 1930‟s,
Americans and their government have taken a keen interest in natural resource
conservation policy on agricultural land throughout the country. As a reflection of this,
the farm bill of 1936 entitled the “Soil Conservation and Domestic Allotment Act”
included for the first time provisions that provided payments and support to farmers
willing to employ soil conservation measures on their farms (Cain and Lovejoy, 2004).
While the main purpose of this bill was to provide financial support to impoverished
farmers dealing with low commodity prices, the fact remains that natural resource
conservation was starting to become an important issue for the American public.
Over time, conservation titles in the farm bill have evolved into legislation that
not only protect soil from erosion, but they now include incentives for improving water
quality and water quantity problems, provisions that prohibit draining wetlands for
agricultural production, land retirement programs such as the Conservation Reserve
Program (CRP), and working land programs like the Environmental Quality Incentives
Program (EQIP). Expenditures for conservation measures have also significantly
increased over time. For example, the United States Department of Agriculture (USDA)
provided nearly 4.5 billion dollars for conservation programs in the farm bill for fiscal
year 2005, compared to 500 million dollars for the 1983 fiscal year (ERS, 2007). It also
appears this trend of increased conservation expenditures will continue, as the 2008 farm
bill will double the level of conservation funding under the previous farm bill if all
2
provisions are approved. Of this new money, nearly two thirds is scheduled to be
allocated to working land programs like EQIP (ERS, 2007).
While giving monetary payments to individual producers engaging in
conservation activities is ultimately a policy decision, the underlying assumption for
these payments is one borne out of traditional microeconomic theory. Specifically,
microeconomics theory assumes that all producers are rational agents engaging in
activities that will maximize profits. However, most conservation activities are not
inherently profitable to the individual farmer; so, conservation payments are provided
under the assumption that the only way to increase participation in conservation programs
is to increase profits received by the individual farmer. In effect, conservation payments
can be seen as incentives or “bribes” that should make conservation activity more
attractive to the individual producer.
If the profit-maximization theory of standard microeconomics is correct in
predicting individual farmer behavior, it would then be expected that the rapid expansion
in government expenditures for conservation payments to individual producers would
lead to great improvements in environmental quality throughout the country. Recent
empirical evidence, though, is showing that this is not the case. For example, modeling
of conservation behavior in the upper Mississippi River region indicated that increasing
conservation payments at the individual producer level would produce minimal change in
rates of soil erosion, nitrate leaching, and nitrate runoff in the area (Wu, Adams, Kling,
and Tanaka, 2004). These authors concluded that conservation payments, which were
modeled as an increase in profits to individual farmers, are not likely to be cost effective
on their own for addressing pollution problems in the Mississippi River and the Gulf of
3
Mexico. Evidence from Nebraska also indicates that increases in conservation program
expenditures do not necessarily lead to improved environmental quality. A recent report
released by the Nebraska Department of Environmental Quality noted that surface water
quality in the state did not improve from 2001 to 2005, and a higher percentage of lakes
in the state are not meeting environmental quality criteria established for their intended
uses (Link, 2005). While the report also indicates that surface water quality has not
deteriorated in Nebraska from 2001 to 2005, the evidence still suggests that increases in
conservation payments may have a minimal impact on improving water quality in the
state. Finally, a survey of eight Nebraska counties found increases in income, while
statistically significant, provided very little power in explaining what motivates farmers
to adopt conservation tillage technologies (Sheeder and Lynne, 2007). Therefore, as Wu
et al (2004) also conclude in their study of the upper Mississippi River region, the
acreage response to increases in conservation payments is, at best, highly inelastic, and in
absolute terms, the impact of conservation payments is extremely small.
Based on anecdotal evidence and the research cited above, it appears that
individual farmers are motivated to engage in (or not engage in) conservation strategies
by a multitude of factors. While it is undeniable that profits do play a role in
conservation decisions, the assumption that it plays the only role in economic decision
making is highly contentious. For instance, work by Nowak and Korsching (1998)
indicates that inadequacies in U.S. soil and water conservation policies can be attributed
to a misunderstanding of the human dimension (sociological and psychological factors)
of farmers, and not a lack of conservation expenditures. Work from Sen (1977) also
concludes that individuals may ultimately make choices based on sympathy and
4
commitment to others, even if the outcomes do not maximize a person‟s self-interest. He
even writes that a person pursuing only selfish interests, as is modeled in
microeconomics, is nothing more than a “rational fool” and a “social moron” (Sen, 1977).
Even the writings of Adam Smith indicate that there are fundamental elements of human
nature that transcend that of pursuing individual self-interest. Smith writes:
How selfish soever man may be supposed, there are evidently some
principles in his nature, which interest him in the fortune of others, and
render their happiness necessary to him, though he derives nothing from it,
except the pleasure of seeing it…That we often derive sorrow from the
sorrows of others, is a matter of fact too obvious to require any instances to
prove it; for this sentiment, like all the other original passions of human
nature, is by no means confined to the virtuous and humane, though they
perhaps may feel it with the most exquisite sensibility. The greatest ruffian,
the most hardened violator of the laws of society, is not altogether without it
(Smith, 1790, Part I, Section I, Chapter I, Paragraph 1; cited in Lynne, 2006a,
102).
While Smith obviously did not discount the role that individual self-interest plays in
motivating consumer and producer choice (see Smith‟s The Wealth of Nations, 1776), he
also recognized the duality of human nature and believed that a person could temper their
self-interest by empathizing with those affected by his or her choices. However, it is
important to note that Smith believed that the act of empathizing occurs within the self
and arises not because of concern for others, but rather concern with others for the self.
This is similar to the view held by Solomon (2007). He writes “We do not just have our
own interests. We share interests with others. Empathy is neither altruistic nor self-
interested. It rather exemplifies the implicit solidarity of human nature.” So, it appears
that both Smith and Solomon have the same understanding of human choice behavior:
individuals use empathy to temper their self-interest and then act in their own- interest
that accounts for the desires of the self and that shared with others.
5
1.2. View of Human Behavior from Neuroscience and Evolutionary Biology
Philosophers and other behavioral scientists (excluding traditional economists)
have been writing about the aforementioned alternative view of human behavior for
years. Yet, recent findings in neuroscience and evolutionary biology lend credence to the
idea that humans are, by nature, motivated by much more than maximizing individual
self-interest. For instance, Cory (2006b), who appropriately updated the work of
evolutionary neuroscientist MacLean (1990), developed the theory of the human triune
brain. In this theory, it is documented that the human brain has evolved into a three level
interconnected, modular structure. The three levels are named the reptilian complex, the
paleomammalian (or “old mammalian”) complex, and the neocortex.
According to Cory, the reptilian complex is the primal and innermost core of the
human brain. In ancestral fishes, amphibians, and reptiles, this portion made up the entire
brain. Today, this protoreptilian circuitry or self-preservation program, serves much the
same purpose as it did in our ancestral vertebrates. Namely, it governs the fundamental
life-support operations, including blood circulation, heartbeat, respiration, food
collection, reproduction, and defensive behavior (Cory, 2006b).
The next developmental stage of the human brain, referred to as the
paleomammalian brain or affectional program, can be identified with the structures
collectively designated as the human limbic system (Cory, 2006b). This portion of the
brain, which developed from gene-based continuities preexisting in the reptilian complex,
led to the development of distinctly mammalian features. Specifically, this complex led
humans to develop warm-bloodedness, nursing, infant attachment, and parental care. As
6
Cory (2006b, p. 26) notes, “these circuits became the basis of family life and our capacity
for extended social bonding.”
The most recent stage of human brain development is referred to as the neocortex,
or as MacLean (1990) names it, the neomammalian brain. The neocortex is a large mass
of brain tissue that dominates the skull case of all higher primates and humans. It has
evolved by “…elaborating the preexisting continuities present in the brains of early
vertebrates” (Cory, 2006b, 26). While the neocortex overgrew and encased the reptilian
and paleomammalian brain tissue, it did not replace them. It did, however, allow for the
evolution of greater complexity within the older parts of the brain, and it facilitated an
advanced interconnected circuitry between all three brain complexes. This produced
behavioral adaptations necessary for humans to deal with their increasingly sophisticated
circumstances (Cory, 2006b).
Once the triune brain had evolved, the unique features of the human brain were
refined over a period of several million years in kinship-based foraging societies. In
these societies, sharing and reciprocity were essential to human survival. This sharing in
society only served to strengthen the adaptive evolution of the now combined mammalian
characteristics of self-preservation and affection (Cory, 2006b). Ego and empathy, which
provide the basis of self-interest and shared other-interest, respectively, became key
features of our individual and social behavior.
While Cory and MacLean were able to determine the neurophysiological make-up
of the human triune brain, this strict biological view of the brain does very little to
explain actual human behavior in any given situation. For this reason, the Conflict
Systems Neurobehavioral (CSN) model of the human brain was developed (Cory,
7
2006b). In this model, the self-preservation and affective programs are interconnected,
and the core motivational (and emotional) circuits are cognitively represented in the
frontal regions of the neocortex, or executive program (Cory, 2006b). It is further shown
that empathetic and other-interested motivations and behaviors are born out of the
affectional program, while egoistic and self-interested acts are derived from the self-
preservation program. The executive program, then, is designed to cognitively represent
these egoistic and empathetic inputs, “…making what may be thought of as our moral as
well as rational choices among our conflicting, impulsive, and irrational or nonrational
motivations” (Cory, 2006b, 28).
The relationship between ego and empathy in the CSN model is very dynamic
(Figure 1.1). Both egoistic and empathetic motivations can trigger a range of behaviors
that can be brought out in the individual. In the egoistic range, behavior is dominated by
the self-preservational circuitry of the triune brain. The behaviors exhibited here are self-
centered in nature, and may tend to be dominating, power-seeking, or possibly attacking
(Cory, 2006b). However, it must be remembered that the behavioral programs of ego and
empathy are joint and non-separable, so empathy is in fact present in the egoistic range,
but to a much lesser degree. As empathy increases, though, behavior in the egoistic range
may become much less harsh, and could possibly be described as moderately
competitive, controlling, or assertive (Cory, 2006b). Still, the fact remains that in the
egoistic range of behavior the individual is putting his or her selfish interests ahead of the
interests of others.
In the empathetic range of the CSN model, behavior is weighted in favor of
shared other-interested, empathetic acts. Here, extreme behaviors may be characterized
8
as self-sacrificial and submissive (Cory, 2006b). Again, though, it must be remembered
that the egoistic and empathetic circuitry of the triune brain is inseparable. Therefore,
ego is present in the empathetic range, even if it exists in miniscule quantities.
Nevertheless, as ego increases in the empathetic range, other-interested behavior
moderates and can be described as “…supportive, responsive, or any of a variety of
„others first‟ behaviors” (Cory, 2006b, 29).
Due to the jointness of ego and empathy within the CNS model, behavioral
tension and stress can arise within the individual. Essentially, ego and empathy are
constantly engaging in a neurological “tug of war,” with each neurological circuit
subjectively evaluating a situation and seeking to express itself through objective
behaviors. If a single expression of ego or empathy is blocked, or if simultaneous but
mutually exclusive urgings of ego and empathy arise, the individual will experience
tension and stress, usually resulting in a subjective experience of frustration, anxiety, or
anger (Cory, 2006b). Therefore, it is up to the individual to strike a rough balance
between the expression of ego and empathy in order to reduce this behavioral tension.
When ego and empathy are balanced, the individual will engage in behavior
located within the dynamic balance range. Here, the individual‟s behavior is
characterized by equality, justice, sharing, and other acts that show respect for the self
and others. As Cory (2006b, 29) writes, “…respect for the self and others is the keynote
of the range of dynamic balance.”
1.3. Parental Models and an Overview Metaeconomic Theory
9
While writings in neuroscience and biology are increasingly recognizing that
behavior may be motivated by several tendencies or interests, mainstream
microeconomics continues to assume that the individual is motivated solely by profit or
utility maximization. Implicitly, then, traditional microeconomics also assumes a Strict
Father moral order (See Lakoff, 1996; cited in Lynne, 1999). In this moral order, the
world is modeled as inherently dangerous, with survival being the major concern.
Children in this model learn self-discipline through “tough love” by the Strict Father, and
a mature adult can only become self-reliant by applying this self-discipline while
pursuing their self-interest. Survival is assumed to be a matter of competition, and only
through self-discipline can a child compete successfully in life. Once the child of the
Strict Father reaches maturity, it is up to them whether they survive of perish. It is
assumed that the child knows what is good for themselves and their families, and that
they have the competancy to make their own decisions. Any meddling by the parents in
the lives of the child is highly resented (Lakoff, 1996).
It is relatively easy to see how this moral order fits into conventional
microeconomics. As noted by Lynne (1999):
It is but a small jump to microeconomics: A single, strict decision-maker in a
hosehold or firm, wherein the pursuit of self-interest is disciplined by the
market. Individuals maximize utility and profit in a constrained
maximization …Survival is through entry and exit in highly competitive
markets. The market rewards for good decisions, and it punishes for bad
mistakes…Also, reward and punishment, in itself, is moral: „Competition is
the crucial ingredient in such a moral system.‟ Without competition, the
motivation to be self-disciplined is removed. Restraints on competition…are
immoral.
10
By keeping the Strict Father moral order hidden in the invisible hand, traditional
economic theory carries an inherent bias within its framework. It seems as though any
kind of analysis or policy created within this traditional framework endorses only the
Strict Father moral order, “…even in cases where the invisible hand of the Strict Father is
directed toward a bad…” (Lynne, 1999, 271). By recognizing only self-interest and not
making the moral dimension explicit within its framework, microeconomics does not
allow for the expression of the multiple motives of ego and empathy within the
individual.
In contrast, the recent theory of Metaeconomics (Lynne, 1999, 2006ab) looks to
transcend, or “go beyond” the framework of conventional microeconomic theory.
Metaeconomics seeks to make explicit the moral and ethical order within its framework.
In doing so, the theory assumes that the individual is dually motivated by egoistic and
self-interested tendencies as well as empathetic and shared other-interested tendencies.
These egoistic and empathetic tendencies, though, may in fact be incommensurable.
Creating metaeconomics in this fashion allows the moral order of the Strict Father and
the moral order of the Nurturant Parent, which is rooted in being cared for and being
cared about (again, see Lakoff, 1996; cited in Lynne, 1999), to coexist in one model.
Intriguingly, metaeconomics now allows for more than one explanation as to what
drives human economic behavior. In standard microeconomics, every action, regardless
of how altruistic the act appears to be, is deemed to be caused by the pursuit of the
individual‟s self-interest. Meanwhile, metaeconomic theory, rooted in the
neurophysiological makeup described by Cory (2006b), can more realistically attribute
behavior to the pursuit of individual self-interest, the pursuit of an
11
empathetic/sympathetic shared other-interest, or both. In fact, the pursuit of self-
interested and other-interested tendencies is present in every situation that a choice must
be made. The degree to which each domain is pursued, though, is a subjective choice
within the individual in any given circumstance. For this reason, it is certainly possible
to see some actions based almost entirely on hedonistic tendencies, other actions rooted
almost entirely in other-interested tendencies, and still other decisions based not on
maximizing, but satisfying, both domains (Lynne, 1999).
It must be noted that metaeconomics does not entirely discount the value of
conventional microeconomics. In fact, the pursuit of individual utility/profit, which is a
core microeconomic principal, is still the foundation of metaeconomics. However, unlike
conventional microeconomics, metaeconomics realizes that utility or satisfaction may be
gained in the self-interest domain as well as in the shared other-interest domain. This
allows the theory to be analytically implemented under the guidance of methodological
individualism, yet still recognize the holism of the human experience, a component of
methodological holism (Lynne, 1999). This configuration of dual methodologies may
ultimately allow for the possibility of synergism, in that attempting to satisfy both
egoistic and empathetic domains may provide a solution or behavior that leads to a “sum
greater than the sum of the parts.” This allows the individual to move to a higher plane
that arises out of the symbiotic potential of ego and empathy (Lynne, 1999, 2006ab).
By integrating empathy and other-interested tendencies into the metaeconomic
framework, metaeconomics, by default, becomes a much more social theory than its
conventional counterpart. Nevertheless, the act of empathizing with others is still a
process that happens within the individual. This is to say, metaeconomics does not use
12
the construct of interdependent utility. The person making a decision may use empathy
to temper the self-interest, and possibly reach a conclusion that is different than when
self-interest is pursued for its own sake, but another person‟s utility (which is
unobservable) has no bearing on the decision to be made. All utility, whether in the
egoistic or empathetic domain, is pursued within the individual.
While it is true that metaeconomic theory puts the burden of choice on the
individual, we must be careful not to assume that economic choice occurs within a
vacuum. Economic choices impact citizens other than the individual making the
decision. These choices also take place in the context of “positive and negative”
freedoms (Sen, 1987), meaning that some people are allowed to do this or that, while
others have to do this or that. These freedoms are analogous in nature to the liberation
and restraint characteristics of economic institutions (Bromley, 2006). Therefore, it
seems that gaining insights into society‟s institutional makeup may also provide
information that can be useful in explaining economic choice behavior.
The primary goal of this research is to determine what tendencies, factors, and
internal characteristics or dispositions influence farmers to engage in conservation
strategies that enhance surface water quality, with particular attention being given to to
the role that empathy/sympathy plays in economic decision making. Antecdotal evidence
suggests that farmers are motivated to use conservation technologies by a heterogeneous
mix of factors that include both financial and non-finacial considerations. Specifically,
these factors include, but are not limited to, a need for profit, influences from friends and
family, information provided by equipment dealers, information from chemical and seed
13
suppliers, a desire for control over their farms, and a desire to “do the right thing” for the
environment, with the latter being borne out of empathy.
While it has long been realized that both financial and non-financial factors
motivate farmer behavior, the theoretical literature has been unable to evolve a settled
and unified account of egoistic-financial and non-financial social motives that ultimately
drive human behavior (Chouinard, Paterson, Wandschneider, and Ohler, 2008).
However, it seems that the theoretical framework of metaeconomics (Lynne, 1999,
2006ab) can provide a new and integrating theory that connects both financial and non-
financial motives into one coherent theory of human behavior. Therefore, this new
framework will be used to analyze farmer conservation behavior in the context of a water
quality conflict between upstream farmers and downstream water suppliers in the Blue
River/Tuttle Creek Watershed of Nebraska and Kansas.
The policy implications of this research extend nationwide. All regions of the
United States, especially those that rely heavily on traditional farm based economies,
experience problems with surface water quality. More accurately modeling farmer
conservation behavior could potentially lead to improved regional conservation programs
based on personality and behvioral characteristics of the region (this idea is also
suggested by Willock et al., 1999). A regional conservation strategy that accounts for
both behavioral and institutional characteristics of the area could ultimately yield greater
results in improving water quality than the “blanket” national conservation policy that is
currently in place.
Understanding the factors that motivate farmers to engage in conservation
behaviors may also lead to enhanced educational programs that target specific traits in
14
individual farmers and could lead to behavioral change. This, in conjunction with
improved conservation programs, may enhance surface water quality at a reduced cost.
These educational programs would also indicate that policy makers are willing to engage
and work with the farming community in order to solve water quality problems. It would
seem that this strategy may work much better than mandating change for the farming
community at large, as mandates are largely resisted and resented. Regardless of the final
outcome, though, it seems that a greater understanding of the factors that influence the
decision-making process of the individual farmer will help to yield better designed and
implemented conservation programs throughout the country.
While this research may have implications for overall conservation policy in the
United States, there may also be implications for both the farmers and water users located
in Blue River and Tuttle Creek Lake watershed. Specifically, by using metaeconomic
theory, we may find that farmers in the watershed have the ability to build shared other-
interest tendencies with those downstream that use the water in the Blue River and Tuttle
Creek Lake. This may suggest active involvement and engagement by conservationists,
farmers and downstream water users in groups, and organizations that work to build unity
with the causes of conservation, perhaps built upon a shared vision for better quality
rivers and lakes witin the region of interest, with each participant entering ever more in
sympathy with others then working to temper and condition the pursuit of self-interests.
The potential to build the described shared other-interest above can then be
determined by measuring the amount of empathy/sympathy present in farmers situated
above the Lake. If empathy/sympathy exists in ample quantities, it is possible that
farmers may condition their behavior and use conservation measures that can help
15
improve the water quality in the Blue River and Tuttle Creek Lake for the benefit of those
downstream. So, this research may in fact have the potential to help resolve an emerging
conflict between upstream farmers and downstream water users in the region.
16
Figure 1.1 The Major Ranges/Modes of Behavior of the CSN Model
(From Cory, 2006b)
17
REVIEW OF LITERATURE
The conservation literature is extensive and diverse, and would be impossible to
cover in its entirety. However, a sampling of the conservation literature is provided with
emphasis placed on studies that use financial, non-financial, and multiple-
motive/multiple-utility frameworks to explain conservation behavior.
2.1 Financial Motives for Adopting Conservation Practices:
Financial motives are the most widely cited account for conservation adoption on
farms. These motives most generally include a desire for greater profits, but may also
include other financial attributes including asset growth, risk reduction, and financial
liquidity (Chouinard et al, 2008). For instance, a model used by Cary and Wilkinson
(1997) hypothesized that five factors could explain the planting of trees and deep rooted
grasses on farms and pastures in south eastern Australia. Of these factors, the idea that
the conservation practice in question must be perceived as economically profitable before
adoption will occur was of paramount importance. In fact, the independent variable
measuring the degree to which a farmer believes tree and grass planting increases profits
in the long-run provided the largest positive coefficients in logistic regression models
constructed by the authors. In the end, the authors concluded that, in general, “…the best
way to increase the use of conservation practices to overcome land degradation…will be
to ensure the practices are economically profitable” (Cary and Wilkinson, 1997, p. 20).
Several other writings attempt to estimate the cost responsiveness of a farmer‟s
adoption of soil-conserving and/or runoff- reducing practices using data from surveys on
stated preferences. For example, Lohr and Park (1995) attempt to determine the cost
18
responsiveness of planting filter strips under the filter strip provision of the Conservation
Reserve Program (CRP) for farmers in Michigan and Illinois. Using a contingent
valuation (CV) framework, mail surveys were sent to farmers in Newyago County,
Michigan and Fayette County, Illinois. The surveys, which provided a hypothetical
payment offer to farmers for participation in the filter strip program, sought to evaluate
whether a respondent would participate in the program and determine the percentage of
eligible land that each willing participant would enroll in the program in response to the
proposed payment. Results of the study indicated that the “payment” variable, which was
defined as the per acre offer to farmers as inducement to join the filter strip program
(Lohr and Park, 1995, p. 485), had a positive and significant effect on the probability of a
farmer joining the filter strip program. Analysis also estimated that an $80 per acre
payment in Fayette County, Illinois and a $38 per acre payment in Newaygo County,
Michigan would be required to entice farmers to participate in the program.
Like the study conducted by Lohr and Park (1995), Cooper and Keim (1996) also
use a CV framework in order to determine farmers‟ cost responsiveness to five different
practices that protect water quality. Mail surveys were sent to farm operators in four
critical watershed regions: the Eastern Iowa and Illinois Basin areas, the Albermarle-
Pamlico Drainage Area in Virginia and North Carolina, the Georgia-Florida Coastal
Plain, and the Upper Snake River Basin Area. Again, survey participants were asked to
divulge their current usage of preferred practices (if at all) as well as their willingness to
adopt these practices in response to a stated hypothetical bid price if they do not currently
use the practice in question. The authors then used the data obtained from the survey to
develop a model of conservation practice participation as a function of bid price. The
19
authors modeled only the data from those respondents that did not use the preferred
conservation practices, and results indicate that incentive payment offers ranging between
35 and 65 dollars would be required to entice 50 percent of the sample to participate in
the surveyed conservation practices. Intriguingly, though, about 12 to 20 percent of the
sample indicated that they would participate in conservation programs in the absence of
an incentive payment (Cooper and Keim, 1996, p. 62). The authors did not elaborate or
attempt to substantively explain this result, but it does perhaps suggest that motives other
than considerations of cost and profit influence the conservation adoption decision in
farmers.
Cooper (1997) extended the work of Cooper and Keim by adding data on the
actual users of conservation practices (i.e. farmers who use the BMPs with no incentive
payments) to the contingent behavioral analysis produced by Cooper and Keim (1996).
The results of this work show that adoption rates of the selected conservation practices
are significantly higher over a wide range of incentive payment offers than the results
predicted form the hypothetical CVM data only. Nonetheless, the results still indicate
that government expenditures would need to increase substantially over current levels in
order to obtain much higher levels of conservation practice adoption in the survey areas
(Cooper, 1997).
All three of the aforementioned studies find a significant degree of cost
responsiveness and downward sloping demand for conservation practices. This suggests
that subsidies for conservation technologies applied on working farmland are likely to
yield substantial increases in the use of such practices. However, models of stated
preferences do not always provide good predictors of actual behavior, making it desirable
20
to validate the results with studies using revealed preference data. Lichtenberg (2004)
uses such a study to analyze conservation decisions by farmers in Maryland. The
analysis conducted by Lichtenberg (2004) uses survey data combined with information
on standard unit costs of installing the following seven soil-conserving and/or runoff-
reducing conservation practices as identified by a Maryland state cost-sharing program:
critical area seeding, contour farming, strip-cropping, cover crops, waterways, terraces,
and diversions. Latent demand models for each of the seven practices were developed
and all exhibit a downward slope, suggesting that cost sharing could have a strong impact
on the adoption of these conservation practices in Maryland. The results also show
strong complementarity among critical area seeding, cover crops, waterways, and
terraces, indicating that increases in incentive payments for one of these practices could
ultimately yield greater adoption of the other three practices without subsequent increases
in incentive payments. In the end, though, the conclusions provided by Lichtenberg
(2004) do not differ substantially from the conclusions drawn by Lohr and Park (1995),
Cooper and Keim (1996) and Cooper (1997). Specifically, all of these studies proffer
that increases in conservation incentive payments should lead to substantial increases in
the adoption of conservation practices by farmers.
2.2 Non-Financial Motives for Adopting Conservation Practices:
While the conservation literature appears to be dominated by work citing financial
motives as the primary driver of the adoption of conservation practices, a considerable
amount of work shows that other, non-financial factors can play a role in the conservation
21
decision made by farmers. For example, a study of Missouri farmers conducted by Ervin
and Ervin (1982) indicates that personal factors attributed to the individual farmer may
have a substantial impact on the number of adopted conservation practices. In fact, the
authors note that the two most important variables in explaining the number of
conservation practices employed on an individual farm were “…either personal
characteristics or related to personal characteristics: education and perception of the
degree of erosion problem” (Ervin and Ervin, 1982, p. 286). The results of this study
show that higher education levels yield a greater likelihood of perceiving an erosion
problem, and thus younger and more educated farmers will be more willing to engage in
conservation technologies that prevent erosion. This finding also suggests that
governmental assistance to farm operators in erosion prone areas should possibly vary
depending upon operator characteristics. The authors note that younger farmers, on
average, tend to be more receptive to a wider range of conservation practices than older
operators, but probably require cost sharing in order to accomplish high levels of farm
conservation effort. On the other hand, in lieu of financial incentives, older and/or less
educated farmers may require technical information programs to explain the
consequences and benefits of unfamiliar conservation practices.
Similarly, work from Gould, Saupe, and Klemme (1989) applies the model used
by Ervin and Ervin (1982) to explain the use of conservation tillage schemes in
southwestern Wisconsin. Like Ervin and Ervin, the authors find that farmer
characteristics are critically important in explaining the adoption of alternative tillage
technologies. Specifically, age and education of the farmer were found to be significant
variables, with younger and more educated farmers being more likely to adopt
22
conservation tillage measures. However, they also indicate that because younger
operators have less farm experience, they may also be less likely to perceive the existence
of an erosion problem in their fields. Therefore, it is suggested that educational efforts
regarding conservation tillage should be directed at these younger, less experienced
farmers (Gould, Saupe, and Klemme, 1989).
Finally, a study of the Central Platte River Valley of Nebraska conducted by
Supalla (2003) suggests that the best way to convince producers of irrigated agriculture to
use conservation practices that protect water quality is through expanded educational
programs. In this study, Supalla notes that producers in the area have developed a
stewardship ethic and are willing to forego profits in order to use best management
practices (BMPs) that enhance water quality in the region. In fact, it was discovered
through an attitudinal survey that 85 percent of producers in the region were willing to
voluntarily accept lower net returns in exchange for reduced groundwater pollution
(Supalla et al., 1995, cited in Supalla, 2003, p. 96). This seems to indicate that producers
care about the environment and are willing to sacrifice profits in order to be stewards of
the land. In other words, farmers in the Platte Valley region of Nebraska have been able
to evolve shared values (i.e. a shared other-interest) regarding the use of BMPs in the
watershed. Therefore, it appears that a lack of knowledge concerning BMPs, not income,
may be the deciding factor in the decision to adopt conservation technologies along the
Central Platte River Valley. Thus, policy instruments that stress education may be more
likely to increase conservation technology adoption rates than policies that stress
financial incentives or direct regulation.
23
In addition to human capital studies that stress general farmer characteristics like
age and education levels, several studies have been completed that analyze the
importance of farmer values and attitudes when making a decision regarding the adoption
of conservation practices. Wallace and Clearfield (1997) examine the reasons why
producers adopt stewardship practices and determine that many voluntarily install
conservation practices on their land because it is the “right thing to do.” They also
indicate that many farmers and ranchers see private ownership not as a right to do what
they please with the land, but as a right to be stewards of the land. Therefore, it is their
responsibility to protect the land and pass it on in a condition that will benefit future
generations. Finally, the researchers show that, even when facing difficulties, many
agricultural producers have maintained an attitude and ethic that treats farming and
ranching as “a way of life,” and not a venture to maximize profits (Wallace and
Clearfield, 1997, p. 4).
Maybery, Crase, and Gullifer (2005) also show the importance that values and
attitudes can have in shaping conservation behavior. Survey responses from farmers in
New South Wales, Australia showed that producers in this region had three distinct
values in regard to their farming operations: economic, conservation, and lifestyle. Also
of importance was the fact that a clear separation existed between economic and
conservation values, as well as economic and lifestyle values. This suggests that
“…ideologically different policy approaches may have separate pathways of influence
within landholder decision making” (Maybery, Crase, and Gullifer, 2005, p 68).
Therefore, it can be reasonably inferred that those with strong conservation goals and
weak economic goals are unlikely to respond to financial incentives as motives to engage
24
in conservation practices. Conversely, those with weak conservation values are not likely
to buy into volunteer conservation practices that will sacrifice profit. Yet, this study still
shows that financial motives alone cannot entirely predict participation in conservation
programs in this region.
2.3 Multiple-Motive/Multiple-Utility Studies for Adopting Conservation Practices:
As evidenced from the studies cited above, it is clear that farmers can be
motivated to adopt conservation practices by both financial and personal/attitudinal
considerations that are not directly related to profit or financial capacity considerations.
However, as shown by Chouinard, Paterson, Wandschneider, and Ohler (2008), the
literature has largely stepped around using a systematic integration of these two types of
goals to describe conservation behavior, either by assuming that only maximum profits
and/or minimum costs matter, or by adding social and stewardship factors in an ad hoc
way. However, a recent subset of the conservation literature has started to use such an
integrated approach in an attempt to substantively explain a wide variety of conservation
behaviors.
Lynne, Shonkwiler, and Rola (1988) were among the first researchers to use a
multiple-motive framework and apply it to an analysis of conservation decision-making.
They collected attitudinal data, as well as context variable data including income and
farm terrain, from farmers in the panhandle of Florida. Results from the study showed
that attitudes toward conservation, perception of environmental problems, farm
ownership, current profitability, income per effort, and risk were all important in
predicting the effort of conservation adoption in the region.
25
In a study similar to that of Lynne, Shonkwiler, and Rola (1988), Cutforth,
Francis, Lynne, Mortensen, and Eskridge (2001) use a multiple motive framework to
model crop diversity decisions on farms in Saunders County, Nebraska. The authors
used an ordinary least squares (OLS) regression model to measure crop diversity in the
county for the 1998 growing season. The model included independent variables that
measured both attitudes toward using crop rotations and social norms in the region, as
well as economic variables like net household income. In the end, the full regression
model was determined to be significant and explained 24 percent of the total variance of
crop diversity. The economic variable “net household income” proved to be significant
and indicated that those farmers with higher incomes were more specialized in their
cropping decisions. However, results also show that positive attitudes toward the use of
rotations were significantly associated with crop diversity in 1998. Obviously,
respondents in Saunders County are motivated (and conflicted) by both economic and
diversity concerns on their farms.
As an evolution to multiple-motive studies like those cited above, Lynne (1995)
developed a new behavioral economic model in which a farmer (or person in general) is
proposed to pursue multiple-utilities or multiple interests. Taking cues from Sen (1977)
and Etzioni (1986), Lynne proposed that individuals pursue not only a self-interest utility
as modeled in microeconomics, but also a shared other-interest utility rooted in social
norms and the ideas of sympathy, meta-ranking, and commitment (Sen, 1977). He labels
each of these utilities as “I” and “We” utilities, respecitvely. It was hypothesized that the
addition of the “We” utility to conventional economic models could greatly improve the
explanatory power of studies intended to describe farmer conservation behavior.
26
Over time, the multiple-utility model proposed by Lynne has been refined and
given the name “metaeconomics” (Lynne, 1999, 2006a). The metaeconomic model,
which will be described in greater detail in the next chapter, has been tested in several
different settings. Lynne (1995) used this model in an attempt to detail the adoption of
irrigation technologies that improve water use efficiency by strawberry growers in
Florida; Lynne and Casey (1998) and Casey and Lynne (1999) test the model in order to
understand the adoption of drip irrigation technology by Florida tomato farmers; and
Sautter, Ovchinnikova, Kruse, and Lynne (2008) use the meateconomic framework to
explain the adoption of conservation tillage in portions of Nebraska. The theory has been
applied in areas oustide of the conservation literature as well, suggesting that it may be
applicable in describing many different types of behavior. For instance, Artikov and
Lynne (2005) use metaeconomics to elucidate farmers‟ use of weather information, and
Kalinowski, Lynne, and Johnson (2006) use the theory to explain recycling behavior by
citizens in Nebraska. In all of the examples cited above, empirical results indicate the
presence of joint self-interest (“I”) and shared other-interest (“We”) utilities or interests,
together forming an internalized own-interest. Also, as hypothesized, the predictive
capacity of the behaviroal models significantly improved with the inclusion of variables
that served as proxies for a shared other-interested utility.
In sum, it appears reasonable to hypothesize that egoistic-financial and social-
moral factors can influence conservation decisions made by farmers. However, it is the
contention herein that models that integrate selfish and social motivations into one
coherent behavioral theory may ultimately provide a better model for explaining
conservation adoption on farms than models proposed by standard microeconomic
27
theory. Therefore, the next chapter of this paper presents the theoretical models of both
standard microeconomic theory and metaeconomic theory. Then, after examining both
theoretical engines, the theory that best explains conservation behavior in the Blue River
watershed can be pragmatically chosen.
28
THEORETICAL MODEL
Metaeconomics, as first proposed by Lynne (1999), is an emerging theory in
behavioral economics that looks to transcend the traditional framework of standard
microeconomics (Lynne, 2006ab). Traditional microeconomic theory assumes that an
individual pursues only one interest, which is ultimately derived from pursuing selfish
and hedonistic tendencies. Conversely, metaeconomics, which uses a framework similar
to that introduced by Etzioni (1986), proposes a dual-motive, dual-utility model of the
individual. Specifically, metaeconomics proposes that an individual jointly pursues both
an egoistic-hedonistic based self-interest utility/tendency and an empathy-sympathy
based other-interest utility/tendency, with the latter being shared with others.
Importantly, these dual interests are viewed as non-separable and are jointly internalized
within the own-interest of an individual (Sautter, Ovchinnikova, Kruse, and Lynne,
2008). So, given the similarities and differences exhibited between metaeconomic and
traditional economic theory, it appears that an overview of both theoretical models is
necessary in order to be pragmatic in choosing which theoretical engine may be best in
explaining farmer conservation behavior. Therefore, graphical and mathematical
representations of both economic theories are provided.
3.1 Graphical Representation of Standard Production Economics
The essence of the traditional microeconomic production model is illustrated in
Figure 3.1. There is only one set of isoquants, as the individual producer is assumed to
have only one interest (IG) that is rooted in egoistic-hedonistic tendencies. It is further
assumed that the individual producer intends to maximize these selfish interests, and
29
therefore will move along the expansion path 0G in response to price ratios. It must also
be noted that this model cannot address behaviors that require self-sacrifice or altruism
depicted at points B and C because a producer cannot maximize profit at these points
(Lynne, 2006b); this is to say, one cannot simultaneously sacrifice and maximize profits
A producer pursuing the self-interest path 0G can be characterized as Homo
economicus as described in traditional microeconomics. A person portrayed in this
fashion is presumed to be motivated only by egoistic-hedonistic tendencies and to pursue
maximum profits while minimizing costs. This quest for profit is assumed to provide the
decision-maker with maximum satisfaction. Therefore, the pursuit of the shared other-
interest is ignored entirely, even though as argued earlier, there is generally a strict father
ethical system operating in the background. In the context of farming, this presumes that
producers will only modify their behavior in response to monetary payments along path
0G, with the possibility for changing the underlying ethical system (e.g. introducing parts
of the nurturant parent family system) not considered. Therefore, farmers settle upon
point A, the point of tangency of the highest self-interest isoquant IG and their budget
constraint RoR
o. This leads to growing a substantial amount of crops while using
relatively few conservation techniques e (conservation tillage, buffer strips, etc.) and
relatively high amounts of industrial inputs d (intensive tillage, chemicals, etc.) on
cropland. Notice, too, that in the case of extreme free riding, no conservation techniques
would be used and the expansion path 0G would be the vertical axis.
30
3.2 Mathematics of Standard Production Economics
While the above graphical representation is useful in understanding conventional
microeconomic producer theory, it is also important to add mathematical precision to the
model. Consider the following production and objective functions:
(1) 1 2( , )Y Y X X
(2) 1 2 1 1 2 2( , ) ( )L pY X X R r X r X
where rj refers to the input prices paid for attributes Xj by the individual farmer; and p is
the market generated price for providing the product or crop; and R is the capital/budget
constraint.
Next, we obtain the first order conditions (FOC):
(3) 1
1 1
L Yp set r
X X
(4) 2
2 2
L Yp set r
X X
(5) 1 1 2 2 0L
R r X r X set
The least cost condition is found by dividing Equation (3) by Equation (4):
(6) 1 1
2 2
( )( )
( )( )
p Y X r
p Y X r
Finding the expansion path 0G in Figure 3.1 from equation (6):
(7) 2 2 1 2( , , , )X X r r p R
The derived demand function for an input becomes:
(8) 1 1 1 2( , , , )X X r r p R
31
Unlike other types of firms or producers, though, the individual farmer must also
account for the general physical characteristics of the ground being farmed when making
production input decisions. These characteristics include, but are not limited to, soil
slope, soil slope length, and organic matter. Therefore, the final derived demand function
for the input decision becomes:
(9) 1 1 1 2( , , , , )X X r r p R N
where the input demand of factor X1 is a function of both input prices (r1 and r2), the
market generated price for the produced crop (p), the capital/budget constraint (R), and
the physical characteristics of the individual farmer‟s land (N).
3.3 Graphical Representation of Metaeconomics
The essence of the metaeconomic framework is presented in Figure 3.2. Careful
inspection of the figure reveals several major differences from the traditional
microeconomic production model (Figure 3.1). First, note that there are two sets of
isoquants that represent both the egoistic-hedonistic self-interest (IG) and the empathetic-
sympathetic shared other-interest (IM). Also note the absolute overlap of the dual self-
interest and shared other-interest isoquants. The intersection of both interests at every
point in the space is represented at points A, B, and C. This overlap of the dual interests
represents the foundation of the metaeconomic model. In the context of the farming
community, traditional profit-bearing outputs like corn and soybeans are represented in
the IG isocurves, whereas shared, more community based outputs like less chemical and
sediment loadings to nearby water sources, enhancement of ecosystems, and long-term
farm sustainability are represented in the IM isocurves.
32
Figure 3.2 also demonstrates that, due to the jointness and non-separability of the
dual interests, a farmer is unable to pick a certain level of self-interest without
simultaneously choosing a level of shared other-interest. Yet, the tendency within the
conservation literature is to treat the choice behavior of farmers as separable independent
effects involving mere tradeoffs (Sautter et al, 2008). Thus, the two interests are
implicitly modeled as separate components in the literature, and not like the interrelated
paths 0G and 0M.
As noted in section 3.1, a farmer that pursues path 0G, or, in the extreme case the
vertical axis, is described as Homo economicus, and is assumed to have his or her
behavior arise only out of self-interested tendencies. In contrast to a farmer that pursues
only self-interest, though, metaeconomics proposes that a farmer may also wish to pursue
the shared other-interest on expansion path 0M. A person that pursues this path can be
characterized as Homo sociologicus, akin to the nature of human behavior presumed in
standard sociology. A farmer portrayed in this fashion is assumed to be motivated in
their behavior by empathetic-sympathetic tendencies such as roles in the farming
community, interdependence and identifying with place and others, and community
norms and traditions. Pursuing these community oriented tendencies would again
maximize outcomes (i.e. maximize profit and the utility it can buy), but it is maximized
in the other-interest domain by achieving shared community oriented goals at point C.
By settling upon point C in the space, a farmer uses many conservation techniques and
will use relatively small amounts of industrial inputs. Control over farming processes is
also desired less at this point. Also, drawing upon philosophy, we find that farmers
maximizing the shared other-interest are choosing to buy more completely into a
33
conservation ethic, which is all about “feeling with” or being “in sympathy with”
(Solomon, 2007, p. 64) other conservation farmers and downstream water users valuing
higher quality water. Without first identifying with and walking in the same space as
others, though, a shared conservation ethic could not evolve.
As noted, the conservation literature assumes that the self-interest and shared
other-interest tendencies are considered independently by the individual decision maker.
Metaeconomics, meanwhile, proffers that both dispositions must be considered jointly
and simultaneously when making a decision. Instead of choosing to maximize either the
self-interest or shared other-interest tendency, metaeconomics posits that the individual
strives to integrate both interests on path 0Z in order to make a choice that satisfies both
domains and provides peace of mind to the decision maker. This is represented by point
B in the space. At point B, the conflict between the self-interest and shared other-interest
within the individual has been resolved and integrated into an own-interest that considers
both the self and others.
Closer examination of point B also leads one to recognize that metaeconomics
allows for the individual farmer to engage in self-sacrifice in both domains of interest. At
this point, we find that the farmer has settled upon the intersection of IG2, IM
2, and the
budget/capital constraint RoR
o. Yet, if a farmer intended to maximize while acting on
the self-interest tendency or other-interest tendency, he or she would orient themselves to
the intersection of the budget constraint with either isoquant IG3 or IM
3, respectively. By
locating at point B, farmers are choosing to give up a little in both domains. This is to
say, the producer may give up some profit in order to install conservation measures on
their farms and do the right thing for the environment, yet they also give up some of the
34
outcomes from pursuing less in the way of shared other-interest in order to earn enough
profit to remain viable. Thus, point B becomes a type of “satisficing” choice for the
farmer, and as noted, provides a certain peace of mind. We have a kind of Homo
satisficicus state of being on path 0Z.
Not only does metaeconomics make the role of self-sacrifice explicit when
describing choice behavior, it also describes how control and self-control play an
important role in decision making. For example, a farmer moving along path 0G will
likely use intensive tillage practices on crop acres in order to help facilitate deeper root
penetration, help maintain soil fertility, and destroy weeds. In contrast, a person moving
along path 0M is much more willing to use reduced or conservation tillage systems on his
or her farm and to give more control to the natural ecological system. A farmer on this
path is also more likely to give in to more control coming out of regulations, or other
kinds of external controls (i.e. landlords). Thus, we find that a farmer on path 0M may
integrate the need for less control over his or her farm processes with the desire to give
more control to the shared other-interest (Sautter et al, 2008). This integration results in
the farmer helping to eliminate soil erosion while also buying into a conservation or
sustainability ethic.
It follows that making the role and need for self-control explicit is also an
important feature of metaeconomics. Farmers that are on the satisficing path 0Z will be
tempted by both inward desires and outside influences (i.e. the “Freedom to Farm” Act)
to move back to path 0G and maximize the egoistic-hedonistic interest, or they may
ultimately succumb to social and community norms and the conservation ethic exhibited
by path 0M and attempt to maximize the empathetic-sympathetic interest. So, self-
35
control is needed by the individual in order to act independently and with courage in
order to achieve satisfactory outcomes in the two parts of the individual‟s own-interest as
demonstrated by path 0Z. Thus, as noted by Sautter et al (2008), “the preference for
control and the ability to take control also temper the self-interest in moving toward path
0Z.”
Finally, it is important to note that metaeconomics posits that the isoquant sets IG
and IM as well as the expansion paths 0G and 0M tend to be in the subconscious of the
individual and are not often considered. While paths 0G and 0M may frame the space
that cognitive and conscious individual thought occurs within, the paths themselves come
to be through “instinct” or intuition (Kahneman, 2003). It is also likely that even 0Z,
once cognitively considered, may become part of an individual‟s intuition. This is
especially true if the decision to be made is one that occurs on a routine basis. In fact,
this is a major theme woven throughout a recent piece written by McCown (2005). In his
writing, McCown, who is building upon the idea of phenomenology in decision making
submitted by Schutz and Luckmann (1973), notes that “in normal routine activity,
commitments are made, and action is taken without conscious deliberation” (p. 22,
emphasis added); so, commitments to the shared interest represented in 0M lead to
tempering the pursuit of self-interest 0G, leading to routine on path 0Z. An example of
this type of decision in the farming context is determining what type of tillage system to
use on crop acres. While a farmer must consciously decide what type of tillage system to
use when starting his or her farming operation, once the decision has been made it
becomes much easier to rely on intuition and subconscious feelings about past tillage
decisions when making choices.
36
3.4 Mathematics of Metaeconomics
While the metaeconomic model of decision making provides a more accurate
reflection of actual choice behavior, it is also a more complex model than that of standard
microeconomics. By allowing for two joint and non-separable interests to motivate
behavior, metaeconomics has effectively changed the mathematical assumptions of
conventional microeconomics. However, the mathematics is workable, and brings
precision to metaeconomic theory that is on par with that achieved in mathematical
microeconomics.
In the case of farmers living in the Blue River/Tuttle Creek Lake watershed, the
implementation of conservation practices on farms in the region can be represented as an
input in the individual producer‟s production function. Therefore, the mathematical
representation of metaeconomics in this case is an elaborated version of the individual
firm presented by standard microeconomics. In order to demonstrate the elaborated
production mathematics, we draw upon Lynne (2006b). For mathematics concerning the
consumer case, see Lynne (2006a).
Metaeconomics proposes two production functions, with one representing the
production of a product of commercial interest that can be sold in markets in the pursuit
of self-interest (i.e. corn, soybeans, etc.). The other production function represents the
more subjective result from producing a product in a way that satisfies the shared sense
of producing the product in the “right way” (i.e. use conservation measures that reduce
non-point surface water pollution, farming in a more organic manner, etc). There is self-
interest in producing a crop, in that the crop can be taken to market and sold for profit, as
37
represented in the IG production function. However, we must account for the fact that
suppliers also react with empathy to the needs, wants, and desires of their potential
buyers and others (e.g. downstream water users), which is represented in the IM
production function. So, there are always two production functions that must be
accounted for:
(10) IG = IG(X1, X2)
(11) IM = IM(X1, X2)
This is a form of the multi-ware production process presented by Frisch (1965, pp. 269-
281), a powerful jointness model form that has had virtually no success in making its way
into mainstream economic thought (for one attempt to bring it into the literature, though,
see Lynne, 1988). Frisch cited two examples of the multi-ware production process, with
the most prominent example being that of wool and mutton production. Wool and
mutton are non-separable and jointly produced outputs in which the sheep itself, along
with the environmental and social system that the sheep is raised within, largely
determines the relative amounts of each output produced in an “unconscious” way
(Kruse, 2003). This type of production function is represented in Figure 3.2 as two
overlapping sets of isoquants, with the IG set representing the egoistic-hedonistic
tendencies and the IM set representing the empathetic-sympathetic tendencies that also
motivate production and supply behavior. The overlapping isoquants demonstrate that
there is little to no substitution between the more objective market recognized egoistic-
hedonistic output and the more subjective empathetic-sympathetic output. This is the
main feature of a joint, multi-ware production process.
38
To illustrate the metaeconomic model, the following form of an objective function
that balances the egoistic-hedonistic and empathetic-sympathetic tendencies is assumed
(although there are many other possibilities):
(12) 2 21 2 1 2 1 1 1 2( , ) ( , ) ( )GG M MpI X X I X X I I R r X r X
where rj refers to the input prices paid for attributes Xj by the individual farmer; and p is
the market generated price for the egoistic interest in providing the product or crop.
Also, notice that there is a subjective element to cost represented in κ1 and κ2; there is
also a subjective payoff, τ, from empathizing, or “walking-in-the-shoes,” of both the
consumer and input supplier. Notice, too, that both outputs arise from the same cost R
such that the joint cost cannot be allocated.
Next, we obtain the first order conditions (FOC):
(13) 1
1 1
1 1
( ) ( ) setMG
M GI I
p I I rX X X
(14) 2 2
2 2 2
( ) ( ) setM
M GIG I
p I I rX X X
(15) 1 1 1 1 2 2set0R r X r X
The least cost condition shows:
(16) 1 1 1 1
2 2 2 2
( )( ) ( )( )
( )( ) ( )( )
M G G M
M G G M
p I I X I I X r
p I I X I I X r
Finding the expansion path 0Z in Figure 3.2 from equation (16):
(17) 2 2 1 1 2 2( , , , , , )G MX X r r p I I R
Intriguingly, notice how the microeconomics expansion path 0G is the default case, when
τ = 0; γ = 0; ι = κ1 = κ2 = 1; this traditional path does not include either integration or
39
balancing of the egoistic-hedonistic and empathetic-sympathetic interests. In fact, path
0G only considers egocentric tendencies, as empathy and sympathy are not considered
motives for behavior in standard microeconomics. Also, notice the peculiarity of
equation (17), in that product and input prices as well as the IG and IM interest variables
all impact the expansion path. The derived demand function for an input is similarly
influenced (i.e. the individual firm has tempered both the self-interest and other-interest)
along some path 0Z:
(18) 1 1 1 1 2 2( , , , , , , )D D G MX X r r p I I R N
We now see from equation (18) that, unlike the conventional microeconomics case,
empathy (IM), as well as egocentric tendencies (IG), affect input demand (and also product
demand, if this was a product demand curve). The fact that the IG and IM interest
variables impact input demand also suggests the need to measure these interests, either
directly or indirectly, when trying to explain behavior. Kahneman has called for similar
research, suggesting that it is possible to measure an individual‟s experienced utility
(Kahneman, Wakker, and Sarin, 1997). Based on this insight, then, various proxies for
the IG and IM interests will be used in explaining conservation behavior in the Blue
River/Tuttle Creek watershed.
By inserting equation (17) and then equation (18) into the objective function (12),
we can also derive the ego-empathy frontier represented in Figure 3.3. Staying on the
budget line RRo we can trace the curve in Figure 3.3 represented in:
(19) , 1 1 2 2( , , , , )G MI I r r p R
With equation (19), we can now define and explore the mutual benefit associated with
reciprocity (Cory, 2006a; cited in Lynne, 2006b) by examining the extent to which there
40
is a “bulge” in the curves, i.e. the extent to which the distances across the isocurves
increases as we move along path 0Z.
Figure 3.3 also allows for the examination of the ego/empathy ratio, a concept
that has been emphasized by Cory (2006a). This occurrence is represented in the
derivative dIG/dIM, calculated from equation (19) and given by:
(20) //
/M
G MG M
G
dI IT
dI I
We see, then, that the dynamic balance between ego and empathy occurs somewhere in
the region AC where TG/M = -1 (Cory, 2006b). As Cory (2004) notes, though, it is not the
particular value of TG/M that is of importance, but rather the orientation and divergence on
either side of this point. So, a person with an orientation toward the other-interest (TG/M <
-1) would be observed exhibiting behavior that contributes to the provision of public
goods and participating in activities like crop rotations and farm conservation strategies.
On the other hand, those with an orientation toward the egoistic-hedonistic self-interest
(0 < TG/M < -1) would tend to live a more material lifestyle (Lynne, 2006b). So, the trade-
off balance that provides peace of mind in decision making is likely in and around TG/M =
-1. However, this is ultimately an empirical question.
Given the graphical and mathematical representations of standard and
metaeconomics production theory shown above, the task now at hand is to empirically
estimate the equivalent of Equations (9) and (18), with both representing the demand for
conservation tillage This will be done with the use of various proxy measures collected
from survey respondents in the Blue River/Tuttle Creek Lake watershed. In the end,
41
these empirical results will provide the basis for comparison of the standard
microeconomic production model and the metaeconomic model.
42
Figure 3.1. Self-interest (G) isoquant curves for conservation tillage (e) and all other
industrial inputs (d)
43
Figure 3.2. Joint Interests. Relationship between the farmer‟s pursuit of a joint self-
interest (IG) on path 0G and an internalized yet shared other-interest (IM) on path 0M with
the path 0Z showing sacrifice in both domains of interest.
44
Figure 3.3. Interests Frontier. Frontiers of interest illustrating the need for balance and the
possibility for synergy, sum greater than the sum of the parts kinds of outcomes, in the
pursuit of two joint and non-separable interests in conservation.
45
TOWARD EMPIRICAL TESTING
4.1. Physical Description of Study Area
Tuttle Creek Lake, Kansas, is a 14,000 acre reservoir located in northeast Kansas
at the lower end of the Big Blue River. The watershed consists of a total area of 9,682
square miles with approximately three-quarters of the drainage area located in soutcentral
and southeast Nebraska. The remaining drainage area is located in northeast Kansas (See
Figure 4.1). The lake, built in 1962 under the direction of the U.S. Army Corps of
Engineers (USACE), provides flood control, irrigation, water supply, recreation, fish and
wildlife management, low flow augmentation, and navigation flow supplementation to
the region (Shea, Burbach, Lynne, Martin, and Milner, 2006). Outflow from Tuttle
Creek Lake enters the Big Blue River about nine miles above its confluence with the
Smoky Hill and Republican rivers near Manhattan, Kansas. At this location, all three
rivers join together to form the Kansas River.
Land use within the Tuttle Creek Lake watershed is primrily agricultural, with
approximately 72 percent of the land used to grow corn, grain sorghum, and other crops.
Another 10 percent of the land is pastureland and another 10 percent is wooded area
(Shea et al, 2006). Herbicides are used extensively throughout the region to control
weeds.
The topography of the watershed varies widely, with slopes ranging anywhere
from 1 percent to greater than 10 percent. In general, the land within the northern and
western portions of the Big Blue and Little Blue River basins is relatively flat, with
slopes of 3 percent or less. In contrast, the remainder of the watershed exhibits extensive
soil dissection which results in slopes of 10 percent or greater (Shea et al., 2006).
46
The predominate soil types in the watershed are silty clay loams, which allow
water infiltration rates in the region to vary from moderate to very slow. Thus, most soils
possess a moderate to very high potential of transporting chemical contaminatnts to
surface water in solution, or bound to eroded soil particles.
4.2. Description of Institutional Arrangements in Study Area
The current institutional situation in the Tuttle Creek Lake watershed has given
rise to water quality problems throughout the region. The historical presumption in the
watershed has been that farmers upstream of Tuttle Creek Lake have the right to allow
chemicals and sediments to runoff and deposit into rivers and streams. Traditionally,
upstream farmers have not been obligated to be concerned with downstream water users‟
rights to clean water. Both state laws and the Federal Clean Water Act put the burden of
cleaning contaminated water on those that are currently using the water, unless the water
users can definitively show which entities are creating the water pollution; and since
agricultural runoff is a non-point source of pollution, those downstream of Tuttle Creek
Lake can not show precisely who is causing the poor water qualtiy in the Lake. So, in
effect, the downstream water users have the duty to accept substandard water quality.
This institutional setup has led to water quality problems for both the Big Blue
River and Tuttle Creek Lake. In fact, Tuttle Creek Lake is listed on the Clean Water Act
Section 303(d) list as impaired for siltation, eutrophication, atrazine, and alachlor.
Extremely large loads of suspended solids and nutrients enter the lake during spring and
summer storm events and excessive siltation has impacted the upper third of the
reservoir‟s conservation pool (Shea et al, 2006). Recent estimates have shown that
47
siltation has reduced the volume of the reservoir‟s conservation pool by 30 to 50 percent
(Barnes, 2006). There are also many observations during the period of record where
atrazine concentrations in Tuttle Creek Lake exceed both the aquatic life and public
drinking water standards of 3 μg/L as mandated by the Clean Water Act. Observations of
the lake also show phosphorous levels to be excessively high at the deep water site,
which could potentially lead to eutrophication. Records also indicate that water quality in
the Big Blue River, a major tributary of Tuttle Creek Lake, is impacted by excessive
amounts of E. Coli and fecal coliform bacteria, nutrients, and atrazine (O'Brien, 2008).
In recent years, the current institutions in the Tuttle Creek Waterhed have been
called into question by downstream water users (instututions are defined as working
rules, norms, traditions, and property relations; see Bromley, 2008). Thus, irritation has
started to build within and between the upstream agricultural producers and downstream
water users of the region. The questioning and evolution of the institutional makeup of
the watershed has been brought on for several reasons, including the desire for clean
water for recreational purposes, general concern for plants and animals that use the water
in the region, and aesthetics. However, of paramount concern to downstream water users
is the quality and quantity of potable water sources in the region. As noted earlier,
outflow from Tuttle Creek Lake helps to provide water flow to the Kansas River. Shea et
al (2006) note that approximately 50 percent of the flow in the Kansas River can be
directly attributed to supplies from Tuttle Creek Lake. This is important to note, as the
Kansas River provides drinking water to major population centers in northeast Kansas,
including Kansas City, Topeka, and Lawrence. Therefore, polluted water from Tuttle
48
Creek Lake is infiltrating the Kansas River and jeopardizing the quality of the water
supply for these areas in northeast Kansas.
Water quantity is also of concern to the region due to the fact that cities in
northeast Kansas are growing rapidly. Margaret Stafford of the Topeka Capital-Journal
(2003) reported that the population of Johnson County, Kansas had grown by 27 percent
between 1990 and 2000, and that the city of Olathe alone had grown by 47 percent
(nearly 30,000 people). Obviously, this rapid population expansion has strained the
ability of municipalities to provide water to their citizens. For example, Water District 1
in Johnson County, Kansas currently provides about 200 million gallons of water per day
to its customers, but projections show that the supply will need to grow to 330 million
gallons of water per day in the next 40 years in order to accommodate population growth
in the county (Stafford, 2003). As the population and demand for water continues to
grow in the region, Tuttle Creek Lake will become an ever more important source for
water supply. However, as cited earlier, the capacity for the reservoir to hold water has
been reduced due to siltation. Therefore, practices upstream that contribute to soil
erosion and siltation are now looked upon in a negative light by water suppliers
downstream.
In an effort to reduce water quality problems and alleviate tension between
agricultural producers and water users, the USDA has continued to expand programs and
funding for conservation programs in the agricultural midwest. The USDA Economic
Research Service (ERS) notes that policy makers have been devoting ever more attention
to conservation policies and programs that promote greater environmental quality
throughout the nation. Expenditures on conservation practices started increasing rapidly
49
around 1986. However, starting around 1995, these expenditures began increasing at a
much slower pace. Eventually, in 2001, expenditures on conservation programs in the
U.S. settled at about 2.5 billion dollars per year. The vast majority of money spent during
this time period was used in land retirment programs such as the Conservation Reserve
Program (CRP). Starting in 2002, though, a dramatic shift in both overall funding and
types of conservation programs funded was employed by policymakers. In fact, from
2001 to 2005, expenditures on conservation has increased from approximately 2.5 billion
dollars to nearly 4.5 billion dollars. Also, much more money was spent on working land
programs like the Environmental Quality Incentive Program (ERS, 2007).
While policymakers have continued to increase funding for conservation
programs, the method used to administer this funding to farmers participating in the
programs has remained relatively unchanged. Policymakers have shown a distinct
preference for administering conservation payments on a volunteer basis (ERS, 2007).
Therefore, willing participants enroll in programs at their local Farm Service Agency
(FSA) office in order to receive conservation payments. This approach, though, does not
lead to targeting of areas most succeptible to non-point agricultural pollution. It also
leads to problems in verifying that conservation measures are being administered
appropriately and that they lead to improved environmental conditions.
Over the past several years, an increasing amount of emphasis has been placed on
conservation issues in U.S. Farm Bill legislation. Yet, the main purpose of the bill is to
provide financial support and stability to those working in the agricultural field. To this
end, the U.S. Farm Bill has undergone several changes. According to ERS (2008), U.S.
farm policy had focused on price and income supports since the 1930‟s; also, until 1996,
50
farm policy relied in part on supply management in the form of acreage limits and storage
programs. However, starting in 1985 and reinforced by the so-called “Freedom to Farm
Act” of 1996, agricultural commodity policy has shifted toward a greater market-based
orientation and has relied less on government intervention (ERS, 2008).
These new policies administer commodity payments based on the calculation base
farm yields. This means that a farmer with a large base yield will receive a larger
commodity payment. Thus we find incentives created that encourage increased
agricultural production. This could potentially produce greater amounts of non-point
agricultural pollution. So, it appears that policies regarding commodity prices may
produce results counter to results that may be produced by conservation policies.
Regardless, though, both commodity and conservation policies are rooted in traditional
economic rationale and assume that farmer behavior is motivated solely by financial
incentives.
4.3. Empirical Models
As noted in Section 3.4, the task at hand is to estimate Equations (9) and (20)
using various proxy measures. Proxies for outside influences and preferences for control
in the context with which the farmer operates within are also used in empirical models.
The ultimate goal of this research is to use the information and insights gained in
order to promote increased usage of conservation measures in the four county critical
area of the Blue River watershed and influence conservation policy in the United States.
For this reason, models that are probabilistic in nature are best suited to analyze
conservation behavior in the study area. In other words, we seek to understand how
51
changes in independent variables impact the probablility of adopting conservation
technologies on working farms, with particular attention being focused on the individual
farmer‟s decision regarding tillage strategies. Therefore, models of the probit and logit
variety will be used to test conservation behavior in the four county target area.
Four logit models of the following functional forms will be used to compare the
results produced from standard microeconomic theory and metaeconomic theory:
(21) 1 0 1 2Pr(0, ) ( ) ( )i i iX R N
(22) 0 1 2 3Pr(0, 1) ( ) ( ) ( * )i i Gi Mi iX R N I I
(23) 0 1 2 3 4Pr(0, 1) ( ) ( ) ( * ) ( )i ii i M G i iX R N I I H
(24) 0 1 2 3 4Pr(0, 1) ( ) ( ) ( * ) ( ) 5( * )i ii i M G i Gi i iX R N I I H I V
where Ri = the income (as a proxy for financial and capital capacity) of the ith
farmer; Ni
= the physical characteristics of the ith
farmer‟s crop land; IGi = proxy for self-interest of
the ith
farmer; IMi = proxy for the shared other-interest of the ith
farmer; Hi = proxy for
habitual tendencies of the ith
farmer; and Vi = proxy for preference for control by ith
farmer. Notice that equation (21) represents the standard empirical derived demand
equation given in standard production microeconomics, whereas equation (22) reqresents
the empirical derived demand equation offered by metaeconomic theory. Equations (23)
and (24) build upon the metaeconomic model by adding important variables that account
for habitual tendencies and preferences for control exhibited by the individual farmer.
Results from the aforementioned logit models should assist researchers and policy
makers to pragmatically choose which theoretical model best characterizes a farmer
making the conservation technology adoption decision. While this information is
extremely valuable, it is also important to understand what factors influence the
52
conservation intensity decision (i.e. proportion of farm under conservation technology) a
farmer makes once he or she chooses to utilize conservation strategies. For this reason,
Heckman or “Heckit” empirical models (Heckman, 1979) have also been constructed in
order to better understand the conservation intensity decision.
4.4. Development of Survey Instrument and Data Collection
As previously presented in Figure 4.1, the Blue River/Tuttle Creek Lake
watershed covers a large portion of southcentral and southeast Nebraska, as well as
northeast Kansas. However, the use of natural resource assesment maps and empirical
surface water quality data allowed physical scientists from the University of Nebraska-
Lincoln to identify a critical four county area of nonpoint source runoff that may impact
Tuttle Creek Lake near the Nebraska-Kansas border (Shea et al, 2006). This critical area
includes Jefferson and Gage counties in Nebraska, as well as Washington and Marshall
counties in Kansas. Therefore, efforts to promote behavioral modification involving
conservation measures on farms have been targeted to this four county area.
A total of 4,191 surveys were mailed to known farm operators in the four county
target area of the watershed. Names and addresses of operators were obtained from farm
operator lists maintained by the local county offices of the Farm Service Agency, U.S.
Department of Agriculture. According to the FSA, the population of the four county area
consists of 3,731 total operators. In the original survey mailing, operators were offered
$40 to complete the survey. A subsequent mailing of the survey commenced a few
weeks after the first mailing was complete. This mailing included a random subsample
of 460 non-respondents of the original 3,731 operators. This time, respondents were
53
offered $80 for their completed questionairres. Intriguingly, the response rate from
operators offered $80 for responses was lower (10.2 percent) than the response rate from
operators that were offered $40 for their completed surveys (15.8 percent). Overall, the
response rate from the 3,731 operators was 17.1 percent (639 survey responses). Due to
missing responses on the proposed dependent variables, 498 surveys were used for
statistical analysis.
While a survey response rate of 17.1 percent is considered respectable, some may
choose to argue that the response is too low to make generalizations about the farming
population in the four county target area. However, there is evidence to suggest that the
survey response rate may in fact be larger than 17.1 percent. First, the survey created
was intended to be administered to farm operators in the target area, as the operators are
the individuals most likely to be in the field making conservation decisions. However,
the primary investigators listed on the cover page of the administered survey received
several phone calls and e-mail correspondences from individuals that had received the
survey that do not participate in day-to-day farming operations. Thus, this antecdotal
evidence suggests that both operators as well as owners/landlords may have received the
survey. The potential exists that the FSA farm operator lists obtained were not properly
maintained and oversampling may have occurred.
There is other evidence to suggest that oversampling may have occurred. As
noted, FSA operator lists indicated that there were 3,731 operators located in the four
county target area in the watershed. Yet, according to the National Agricultural Statistics
Service (NASS) census for 2002, there is only a total of 3,184 farms in the four county
target region. Therefore, if it is assumed that each farm has one principal operator,
54
oversampling by nearly 550 individuals occurred in the sample. If we remove these 550
surveys from the overall sample, the survey response rate is 20 percent. If we further
assume that the trend of shrinking farm nubers in the target area has continued during the
period from 2002 to 2007 (census statistics from 2007 will not be released until February,
2009), it is possible that the actual response rate could be around 23 to 25 percent.
3.4. Description of Variables
Dependent Variables
Two dependent variables were created for the Blue River/Tuttle Creek Lake
watershed dataset: no01 and noratio_1. The no01 variable is a binary (0,1) variable used
to explain the adoption of conservation tillage in the logit analyses, whereas the noratio_1
variable is used in the continuous ordinary least squares (OLS) portion of Heckman
models used to explain conservation tillage intensity. Both of these variables were
created by using information obtained from Questions 2a and 2b of the survey instrument
entitled “A Soil and Chemical Management Survey of Kansas and Nebraska Farmers”
(See Appendix A).
Questions 2a and 2b use a matrix format in order to ascertain the number of acres
under various types of tillage regimes. Respondents reported the number of acres under
conventional tillage (less than 15 percent crop residue), reduced tillage (15 percent to 30
percent crop residue), and conservation tillage/no-till (greater than 30 percent crop
residue) cropping schemes. Respondents were also asked to break this information down
by dryland and irrigated acres, as well as practices on highly erodible land (HEL) and Not
Erodible land.
55
The computation of the no01 dependent variable is fairly straight forward: those
respondents that use any amount of conservation tillage/no-till received a score of 1 for
the no01 variable, and those that use no conservation tillage/no-till received a score of 0.
The computation of the noratio_1 variable, on the other hand, required a bit of simple
arithmatic. As the variable name implies, noratio_1 gives the proportion of an individual
farmer‟s land that utilizes a conservation tillage/no-till technology. This was computed
by taking the total acres of conservation tillage/no-till reported in questions 2a and 2b and
dividing this amount into the total amount of acres farmed as reported in the matricies.
Any missing values were then replaced by mean substitution. Ultimately, this provides a
continuous variable with values that range between 0 and 1.
Independent Variables
Income/Financial Capacity (Ri)
Income data from farmers in the Blue River/Tuttle Creek Lake watershed was
collected and is an important component in both the microeconomic and metaeconomic
derived demand models. The variable, named income2_1, was collected via Question 33
in the administered survey instrument. This question asked respondents to choose a
category that best described their total income from both gross farm sales and other
farm/conservation payments, again, as noted, to indicate the financial or capital capacity
of the farm. Responses were scaled such that the final income variable is reported in
thousand of dollars. Also, missing income values were treated with mean substitution. It
is hypothesized that the income variable will have a positive and significant impact upon
both the no01 and noratio_1 dependent variables.
Soil Slope (Ni)
56
The physical context of the land in production is thought to be an important
determinant in the adoption of conservation technologies. In the case of tillage strategies,
the most important physical factor appears to be soil slope (i.e. land steepness). For this
reason, soil slopes are estimated in the four county critical area of the Blue River/Tuttle
Creek Lake watershed.
In order to compute soil slope, individual survey respondents were asked to mark
an “X” on a county map in order to indicate the general location of the respondent‟s
principal farm. Then, geographic latitude and longitude coordinates of the principal
farms were determined by using the computer program 3-D Topoquads. Once the
geograpic coordinates were obtained, this information was then utilized in Geographic
Information Systems (GIS) software in order to obtain information regarding soil slope
on a particular respondent‟s principal farm.
Individuals with specialized knowledge in the GIS field were relied upon to
compute the soil slope variable. These individuals used two different data layers when
computing soil slope: a 30 meter digital elevation model (DEM) of the four county area
and a map of 2007 NASS cropland data. Using both the 30 m DEM and the cropland
data, different footprints were created for each crop scenario in the region. These
footprints were used to extract the soil slope information.
With the soil slope information in hand, zonal statistics were then used within the
GIS system in order to compute the minimum, maximum, range, and standard deviation
of the slope information. The final soil slope variable is the mean of the soil information
extracted from the GIS system. Again, all missing values were replaced via mean
substitution.
57
While physical land characteristics are considered to be important factors in the
individual conservation adoption decision, it is important to note that, for this particular
study, the soil slope variable is hypothesized to not be statistically significant when
attempting to explain a farmer‟s rationale for adopting different tillage strategies. This
hypothesis is due to the fact that the four county target area was specifically chosen
because physical science models predicted that the area exhibited physical characteristics
that lead to a great risk of transporting eroded soil and chemical runoff to the Blue River
system and, ultimately, Tuttle Creek Lake. Since the study area was chosen due to these
characteristics (which included soil steepness), it is expected that there will not be enough
variation in the values of the soil slope variable to yield it as a powerful and statistically
significant predictor of farmer conservation behavior. However, the inclusion of the soil
slope variable does in subsequent statistical models does serve as a statistical test to help
determine that the correct study area has been chosen.
58
Other-Interest*Self-Interest (IMi*IGi)
The other-interest*self-interest independent variable is the core variable of the
empirical metaeconomic model. As previously indicated, it is theorized that humans rely
upon joint, non-separable shared other-interest tendencies and self-interest tendencies
when making economic decisions. For this reason, other-interest and self-interest
indepenent variables can not be modeled separately. Thus, it has been decided that
proxies for an individual‟s shared other-interest and self-interest tendencies will be
multiplied such that both proxies are taken into account when creating a single
independent variable.
Three proxies were used to measure a survey respondent‟s orientation toward a
shared other-interest: empathy, sympathy, and empathy/others. The need for three
different shared other-interest proxies is due to the fact that other-interest dispositions in
humans evolve. For example, psychologists and neuroscientists like Decety, Michalska,
and Akitsuki (2008) have shown with functional Magnetic Resonance Imaging (fMRI)
that the programming for empathy has been “hard-wired” into the brain circuitry of
normal functioning children. Their results are consistent with previous fMRI studies
involving adults. Thus, it appears that the ability to empathize with other humans is an
innate characteristic possessed by normal humans. Therefore, it is also proposed in this
paper that all humans have the ability to project themselves into the perceived mental
state of others (i.e. “walk in the shoes of others”).
While empathy is defined as the ability to project oneself into the mental state of
others, sympathy is defined in a much different manner. While most relate sympathy to
feelings of compassion, this paper defines sympathy in much the same way as it is
59
defined by the philosopher Solomon (2007): a human‟s ability to sympathize is
characterized as the ability to buy into a specific group ethic. So, we find that humans
can indeed become “in sympathy with” particular groups and buy into specific group
ethics. This is achieved through the use of empathy. Individuals can project themselves
into the state of mind of specific groups and choose to become in sympathy with the
group in question if the group ethic and goals align with the individual‟s goals. So, we
find that the key to becoming in sympathy with particular groups is the act of empathy.
In other words, empathy can move an individual towards sympathy. However, it should
be noted that the act of empathizing does not automatically lend itself to sympathy.
Becoming in sympathy with a group is still an individual choice that can be accepted or
rejected, but empathizing does provide important information to the individual that aids
in the decision making process. So, in terms of the research at hand, it is proposed that
all inhabitants in the four county target area have the ability to empathize (albeit some
have a greater capacity than others), but it is unlikely that all inhabitants have become in
sympathy with various groups that use the Blue River watershed and Tuttle Creek Lake.
While the acts of empathizing and sympathizing occur strictly within the
individual, it cannot be denied that the opinions and lobbying of other human beings can
in fact influence an individual‟s decision making process. Therefore, the empathy/others
variable was created in order to assess which specific individuals and groups can
influence farmer conservation behavior. This is the final other-interest proxy used in
testing the metaeconomic theory in the study area.
In order to assess a survey respondent‟s capacity to empathize, the perspective
taking and fantasy subscales of the Davis Empathy Scale (a widely used and accepted
60
psychological scale), were used (Davis, 1980). Respondents were asked to respond to 13
agree/disagree items in order to evaluate their empathetic shared other-interest
tendencies. The line items were all similar to the following form, and formatted as
recommended by the Theory of Planned Behavior (Ajzen, 1991):
In the end, though, factor analysis conducted with the statistical package SPSS (16.0)
showed that 7 of the 13 line items accounted for most of the variation in the empathy
scale presented in question 22. Therefore, the mean of these 7 items (lines 5, 6, 7, 8, 11,
12, and 13) were used to assess a respondent‟s orientation toward the empathetic shared
other-interest.
While the Davis Empathy Scale is a widley used survey instrument that can assess
a person‟s empathizing ability, there is no such instrument available to measure the
degree to which a person is “in sympathy with” various group ethics. For this reason, a
new sympathy scale was created and administered to the farming population of the four
county target area in the watershed. This sympathy scale, which utilized five line items,
was tailored to the problem at hand in the Blue River/Tuttle Creek Lake watershed. The
line items were constructed as seven point agree/disagree Likert scales and all were
similar to the following form:
61
Ultimately, each respondent was asked if they could be in sympathy with three specific
groups in the region: recreational users of Tuttle Creek Lake, public water suppliers
below Tuttle Creek Dam, and finally, all others that use water from the watershed. Then,
the mean of all line items was calculated in order to determine a respondent‟s final
sympathetic other-interest disposition.
As was the case with the created sympathy scale, there is also not a generally
accepted psychological scale that can asses the degree to which respondents‟ decisions
are influenced by other people. However, the applied theory of planned behavior
(Ajzen, 1991) does provide some guidance in creating questionairres that assess human
behavior. Therefore, the empathy/others variable was constructed specifically for this
study by using information provided by the applied theory of planned behavior.
Question 19 of the survey instrument assesses the influence that others may have
on tillage decisions, with particular attention being paid to other Tuttle Creek Lake water
users, those that can have a direct impact on farming operations (i.e. Farm Service
Agency, landlords, etc.), and family members. Survey participants were asked to respond
to the following statement:
Other people may influence your thinking about farming decisions, and
may affect decisions on your farm. Please mark an “X” at a spot on each
line to indicate how likely you think it is that these people believe you
should use conservation tillage/no-till and chemical BMPs.
62
Respondents then placed an “X” on 18 different line items, with all items similar to this
form:
Statistical factor analysis of the line item responses confirmed the presence of the three
influential groups cited above. Threfore, there are three independent variables that
represent other water users of Tuttle Creek Lake (mean of line items 10, 11, 12, and 13),
those with direct ties to the farming community (mean of line items 2, 3, 4, 5, 6, 7, 8, 9,
and 18), and family members (mean of line items 1, 14, 15, 16, and 17).
While three proxies were needed to evaluate the evolution of shared other-
interests of farmers in our sample, only one proxy was needed to assess a farmer‟s
orientation toward self-interest tendencies. The proxy used for this study was the selfism
scale created by Phares and Erskine (1984). This scale has been tested routinely in
psychological disciplines, and it has been deemed as reliable in assessing narcissistic
(selfish) tendencies within individuals.
Question 24 of the survey instrument used in the four county target area
administered the selfism scale. Respondents were asked to give their opinions regarding
14 different line items that assessed selfish tendencies. All line items were similar to the
following form:
63
Although 14 line items were given in question 24, statistical factor analysis indicated that
8 of these items accounted for nearly all of the variance of the scale. So, based on the
information provided in the factor analysis, the mean of those 8 line items (lines 5, 7, 8,
10, 11, 12, 13, 14) were used to compute the final selfism variable.
Once the final shared other-interest and self-interest variables were computed, the
final other-interest*self-interest (IMi*IGi) metaeconomic independent variables could be
created. In order to create these variables, the results from the selfism scale were
reversed such that the two multiplicants would be of the same magnitude and direction.
Then, the results from each other-interest proxy were multiplied by the results of the self-
interest proxy. The result was five independent variables that could be used in three
separate tests of metaeconomic theory (test of empathy, test of sympathy, test of
empathy/others). It is hypothesized that these five variables will have a positive impact
on the probability of a farmer using conservation tillage techniques on his or her farm.
Habit (Ri)
Metaeconomics suggests that most farmers making operating decisions run
largely on emotion or sub-conscious feelings about farming strategies that have worked
in the past. Intriguingly, this assertion is supported by empirical research conducted by
Kahneman (2003). In his work, Kahneman determined that humans in general rely on
“intuition” or emotion in their decision making processes. In fact, he determined that
64
“effortless thought is the norm” in the everyday lives of humans. So, based on this
empirical contribution to the behavioral economic and psychological economic literature,
it seems prudent to add a measure of habit to the empirical metaeconomic model.
In metaeconomic terms, if someone is on a path 0Z, it is more likely that path will
be maintained through time. In effect, it is proposed that consciously cognitive, rational
calculation and consideration of using more conservation tillage happening at some
earlier time simply leads to underlying, less than cognitive feelings reflected in habitual
tendencies. These internalized past decisions then guide decisions made today (Sautter
et. al, 2008). Thus, we find that once farmers put into practice a new technology, they
become rather reluctant to switch back to practices used in the past.
Habitual tendencies in relation to conservation tillage strategies were measured in
the four county target area by asking the following question: Is the percentage of your
farm under conservation tillage/no-till less, the same, or more than 3 years ago?
Responses were recorded on a seven-point Likert scale. It is expected that this variable
will have a positive and significant impact on the probability of farmers using
conservation tillage on farms in the Tuttle Creek Lake watershed.
Self-Interest*Control (IGi*IVi)
As noted earlier, meateconomic theory proposes that a farmer‟s preference for
control over his or her farming operations can make a large impact upon conservation
technologies used on individual farms. For this reason, farmers in the four county target
area of the Blue River/Tuttle Creek Lake watershed were asked to respond to several line
items that assess a person‟s views in regard to control over specific farm processes.
65
Question 15 of the composed survey instrument administers the control scale.
The line items were administered with a seven point Likert scale, and all 18 items took
the following form:
Notice that respondents were essentially marking an “X” on a continuum that measured
whether he or she perceives complete control over conservation practices and their
consequences or if he or she feels that they have no control over conservation practices
and their consequences. This is very similar in nature to the idea of autonomous versus
heteronomous control presented by Angyal (1967), in which autonomous control is
represented as internal self-control and heteronomous control is represented as control
exerted upon the individual by others or the environment in which the individual resides.
It is hypothesized, then, that those who feel that they can use conservation tillage
strategies and still maintain a great amount of autonomous control over farming processes
will be more likely to use conservation tillage on individual farms. In contrast, those that
believe using conservation tillage technologies reduce a farmer‟s autonomous control
over farming processes will be less likely to use conservation tillage strategies.
Statistical factor analysis confirmed that Question 15 assesses three different
types of control: control over specific farm practices (farm control; mean of lines 1, 2, 4,
5, 6, 7, 8, 10, 11, 13, and 15), control exerted upon respondents by others (other control;
mean of lines 9, 12, 16, 17, and 18), and preferences for control over nature (nature
control; mean of lines 3 and 14). With these components determined, the results from the
selfism scale (see above) were then multiplied by the results of the separate control
66
factors in order to derive the final control variables used in the metaeconomic models.
All missing values were replaced by mean substitution. It is hypothesized that self-
interest tendencies reinforce preferences for control over farm processes, and thus the
final control variables are expected to have a significant and negative impact on the
probability of using conservation tillage on farms located in the Blue River/Tuttle Creek
Lake watershed.
67
Figure 4.1. Blue River/Tuttle Creek Lake Watershed and four county critical area
Nebraska
Kansas
Big BlueRiver
Little BlueRiver
Tuttle CreekLake
Jefferson
Gage
Washington
Marshall
68
RESULTS AND DISCUSSION
5.1. Summary Descriptive Statistics
The following tables summarize the results of the data collected from the survey
instrument sent to the four county target area in the Blue River/Tuttle Creek Lake
watershed. Each table lists the mean response to each specific line item from the
questionnaire for those that chose to respond the seven-point Likert scales. For the
purpose of providing the results of summary descriptive statistics, all non-responses and
responses of “Do Not Know” and “Does Not Apply” were excluded. Please see
Appendix A for reference to question numbers.
Table 1: Mean Responses to Question 24 (Selfism Scale)
Mean Std. Deviation
slffirst 3.40 1.728
slfnobody 3.61 1.739
slfhoard 3.55 1.581
slfworry 3.24 1.602
slfsell 3.72 1.532
slfworth 2.72 1.356
slfaggr 3.21 1.487
slfahead 2.81 1.634
69
Table 2: Mean Responses to Question 22 (Davis Empathy Scale)
Mean Std. Deviation
empimag 5.32 1.114
empdisagree 5.47 1.127
empthink 5.32 1.119
empsure 4.73 1.194
empcritic 4.96 1.183
empstory 5.07 1.089
empshoe 4.52 1.261
Table 3: Mean Responses to Question 23 (Sympathy Scale)
Mean Std. Deviation
sympsamecnty 4.74 1.250
sympothrcnty 4.68 1.257
sympoutside 4.41 1.322
sympwatsupl 4.94 1.239
sympbelow 4.90 1.216
Table 4: Mean Responses to Question 19 (Empathy/Others Water Users Scale)
Mean Std. Deviation
infwatsupl 4.38 1.880
infrecr 4.09 1.918
inffrmblw 4.15 1.765
infenv 4.74 1.881
70
Table 5: Mean Responses to Question 19 (Empathy/Others Farm Entity Scale)
Mean Std. Deviation
infsupl 5.04 1.392
infagron 5.18 1.481
infcommod 3.95 1.650
inflndr 4.36 1.608
inffrnd 4.69 1.466
infext 5.16 1.489
infdealer 3.72 1.449
inffsa 4.90 1.605
Table 6: Mean Responses to Question 19 (Empathy/Others Family Scale)
Mean Std. Deviation
infchild 4.66 2.157
infparent 4.30 2.609
infrelat 4.23 2.085
infspouse 4.88 2.110
Table 7: Mean Responses to Question 15 (Farm Control Scale)
Mean Std.
Deviation
ctrlstable 3.04 1.491
ctrlreg 2.79 1.555
ctrlself 3.45 1.719
ctrlerosion 2.90 1.559
ctrlweeds 3.10 1.592
ctrlinside 3.50 1.277
ctrlperm 3.29 1.822
ctrlme 3.39 1.448
ctrldate 3.91 1.726
ctrlappear 4.11 1.876
71
Table 8: Mean Responses to Question 15 (Other Control Scale)
Mean Std.
Deviation
ctrllendr 4.36 1.669
ctrlwateruser 4.00 1.541
ctrllndlord 3.75 1.612
ctrlother 4.48 1.486
ctrlsupply 3.87 1.651
Table 9: Mean Responses to Question 15 (Nature Control Scale)
Mean Std.
Deviation
ctrlpownat 4.99 1.849
ctrlnature 5.03 1.935
The tables presented above provide summary descriptive statistics for survey line
items used to compute the selfism, empathy, sympathy, empathy/others, and control
variables that were described in detail in section 4.4. However, these table do not
account for the theorized interactions between selfism and empathy/sympathy (self-
interest*shared other-interest) and selfism and control. Therefore, summary descriptive
statistics for the final variables used in the analysis of tillage behavior among farmers in
the Blue River/Tuttle Creek Lake watershed are presented below.
Table 10: Mean Results of Final Selfism Variable
N Mean Std.
Deviation Missing
selfism 496 3.29 1.180 0.40%
72
Table 11: Mean Results of Final Empathy/Sympathy Variables
N Mean Std.
Deviation Missing
empathy 495 5.06 0.807 0.60%
sympathy 496 4.73 1.088 0.40%
infwatuser 490 4.34 1.707 1.61%
inffarm 491 4.63 1.074 1.41%
inffamily 491 4.52 1.816 1.41%
Table 12: Mean Results for Final Control Variables
N Mean Std.
Deviation Missing
farmctrl 488 3.37 1.008 2.01%
ctrlothers 487 4.08 1.008 2.21%
natctrl 487 5.01 1.701 2.21%
Table 13: Mean Results for Final Selfism*Empathy/Sympathy Variables
N Mean Std.
Deviation Missing
selfemp 488 24.07 8.080 2.01%
self2symp 488 22.65 8.463 2.01%
slfuserinf 482 20.75 10.403 3.21%
slffarminf 483 21.92 7.923 3.01%
slffamilyinf 483 21.42 10.519 3.01%
Table 14: Mean Results for Final Selfism*Control Variables
N Mean Std.
Deviation Missing
slffarm 479 15.64 5.627 3.82%
slfothers 478 19.34 6.776 4.02%
slfnat 478 23.60 10.007 4.02%
Table 15: Mean Results for Final Income Variable
73
N Mean Std.
Deviation Missing
income2 475 158.91 211.413 4.62%
Table 16: Mean Results for Final Soil Slope Variable
N Mean Std.
Deviation Missing
slope 498 3.11 0.648 0.00%
Table 17: Mean Results for Final Habit Variable
N Mean Std.
Deviation Missing
notilchng 497 3.85 2.132 0.20%
Inspection of the tables presented above provides some interesting insights into
the psychological makeup of the respondents in the four county target area. First, notice
that the final selfism scale (Table 10) indicates that self-interest tendencies are in fact
present within farmers in the watershed. However, the mean score of the scale (3.29) is
much less than might be predicted using the standard framing of the problem of adoption
using traditional microeconomics. In fact, a microeconomics frame would suggest that
the mean score of the selfism scale would be much closer to six or seven, and, being
exactly true to the theory, everyone would need to answer seven Instead, what we find is
that the final selfism score shows that respondents are actually much closer to selfless,
rather than selfish.
In addition to selfish tendencies being present within respondents in the
watershed, survey results also indicate that shared other-interest tendencies in the form of
empathy and sympathy also exist within the region (Table 11). Intriguingly, a
74
comparison of the mean results of self-interest tendencies and shared other-interest
tendencies in the watershed actually show that shared other-interest tendencies occur at a
greater magnitude than self-interest tendencies. This finding places traditional,
microeconomic based renditions of farmer behavior in question.
Closer inspection of Table 11 also provides some critical information regarding
metaeconomic theory. Previously, it was noted that metaeconomics theorizes that all
individuals are born with an innate ability to empathize with other individuals. The
ability to empathize, then, can ultimately lead an individual to become “in sympathy
with” the ethic and goals of a particular group of people. It was carefully noted, though,
that the ability to empathize does not necessarily lead to sympathy.
Information given in Table 11 shows that this idea may in fact be plausible.
Notice that the mean score of the final empathy scale is 5.06 units. Given that empathy
was measured with a seven-point Likert scale, it is obvious that respondents clearly have
the ability to empathize with other individuals. Comparing the empathy scale with the
scale that measures sympathy, though, indicates that there is a great amount of variability
in the respondents‟ ability to sympathize with groups that use Tuttle Creek Lake. First,
the mean score of the sympathy scale was 4.73 units, a result that is lower than the mean
score of the final empathy scale. Second, and more importantly, the standard deviation of
the sympathy scale was 1.088 units. This result is higher than the standard deviation of
the empathy scale, which produced a result of 0.807 units. Based upon these numbers,
then, we can reasonably speculate that all respondents in the four county target area have
the ability to empathize (albeit at different capacities), whereas not all respondents have
become in sympathy with the ideals and goals of other users downstream that rely upon
75
Tuttle Creek Lake. This result can also be concluded by inspecting the distribution of
responses to the empathy and sympathy scales. These distributions are provided in
Figure 5.1 and 5.2. Taking all of this information together shows that the end result is in
line with ideas put forward by metaeconomic theory.
Finally, the tables presented above also show that missing values in the final
dataset used to draw conclusions regarding farmer behavior in the watershed is not a
pervasive problem. As expected, the variable with the greatest amount of missing data
was the variable that measured a farmer‟s financial capacity to implement the technology
as represented by gross income. However, the amount of missing data for this variable
only measured 4.62 percent. Given that this and all other variables have missing data at a
rate less that 5 percent, the use of mean substitution to replace the missing values in final
behavioral models appears appropriate.
5.2. Correlations
While the summary descriptive statistics for the region provided above provide
interesting insights into the psychological makeup of farmers in the region, it also seems
that an examination of correlations between each of the scale variables used may also
yield some useful information in moving forward with an empirical test of metaeconomic
theory. Keeping this in mind, a correlation table that includes all behavioral variable
proxies used in the survey of the four county target area of the Blue River/Tuttle Creek
Lake watershed is presented below in Table 18.
76
Table 18: Correlations Between Various Behavioral Proxies
income slope selfism empathy sympathy infwatuser inffarm inffamily notilchng farmctrl ctrlothers natctrl
income 1.000 -0.149a -0.074 -0.016 -0.022 0.065 0.020 0.036 0.206
a -0.131
a 0.096
b -0.003
slope -0.149a 1.000 0.085 -0.034 0.007 -0.078 -0.011 0.067 -0.021 0.042 0.035 -0.041
selfism -0.074 0.085 1.000 -0.234a -0.225
a -0.166
a -0.101
b -0.057 -0.075 0.199
a -0.048 -0.003
empathy -0.016 -0.034 -0.234a 1.000 0.421
a 0.186
a 0.185
a 0.128
a 0.075 -0.200
a -0.118
a -0.018
sympathy -0.022 0.007 -0.225a 0.421
a 1.000 0.356
a 0.231
a 0.063 -0.064 -0.189
a -0.084 -0.153
a
infwatuser 0.065 -0.078 -0.166a 0.186
a 0.356
a 1.000 0.566
a 0.303
a 0.036 -0.151
a -0.099
b -0.072
inffarm 0.020 -0.011 -0.101b 0.185
a 0.231
a 0.566
a 1.000 0.340
a 0.042 -0.154 -0.160
a -0.077
inffamily 0.036 0.067 -0.057 0.128a 0.063 0.303
a 0.340
a 1.000 -0.067 -0.001 0.016 0.077
notilchng 0.206a -0.021 -0.075 0.075 -0.064 0.036 0.042 0.042 1.000 -0.063 0.016 0.023
farmctrl -0.231a 0.042 0.199
a -0.200
a -0.189
a -0.151
a -0.154
a -0.154
a -0.063 1.000 0.104
b 0.112
b
ctrlothers 0.096a 0.035 -0.048 -0.118
a -0.084 -0.099
b -0.160
a -0.160
a 0.016 0.104
b 1.000 0.079
natctrl -0.003 -0.041 -0.003 -0.018 -0.153a -0.072 -0.077 0.077 0.023 0.112
b 0.079 1.000
a = p<0.01, b = p<0.05
77
78
At first glance of Table 18, one should notice that there is a fair amount of
correlation exhibited between the items used in the survey of farmers situated above
Tuttle Creek Lake. While this would pose a problem in traditional microeconomic
frames, the correlation in the context of metaeconomic theory actually helps to explain
psychological tendencies and overall behavior exhibited by farmers in the watershed.
It has been noted several times that metaeconomic theory proffers that self-
interest and shared other-interest tendencies within humans are joint and non-separable.
So, one would expect that there would be a negative and significant correlation between
self-interest and shared other-interest proxies, as self-interest would pull an individual
toward maximum profit on path 0G and shared other-interest would pull the individual
toward maximizing shared other-interest(s) on path 0M. Table 18 provides just this type
of information. Notice that the selfism proxy is negatively correlated with both the
empathy and sympathy proxies. In addition, the water user and farm entity portions of
the empathy/others scale are also negatively correlated with the selfism scale.
Importantly, all of these correlations are significant at either the p<0.01 or p<0.05 level.
Therefore, these results lend credence to the idea that self-interest and shared other-
interest tendencies within the human brain are perhaps biologically intertwined and non-
separable.
In addition to the negative correlation exhibited by the self-interest and other-
interest proxies, we also find significant positive correlation between self-interest
tendencies and proxies intended to measure a farmer‟s desire for autonomous control
over his or her farming operations (farmctrl variable). Again, this result was predicted by
78
78
using metaeconomic theory. As specified earlier, it is thought that preferences for
autonomous control are reinforced by selfish tendencies. In other words, a person that
wants complete control over outcomes in his or her life is not likely to let preferences and
desires from others influence their decisions. Thus, we would expect this type of person
to be on the self-interest maximizing path 0G and let self-interest tendencies dominate
their decision making processes. Thus, we would indeed expect a positive and significant
correlation between the selfism scale and the scale used as a proxy to measure desire for
control over farm processes. We would also expect significant negative correlations
between the farm control scale and all proxies used to measure shared other-interest
tendencies within respondents, which Table 18 also shows to be true.
Finally, it should also be noted that significant and positive correlations exist
between all proxies used to measure an individual respondent‟s orientation toward a
shared other-interest. This helps to validate that each scale is in fact measuring the
intended psychological phenomenon and provides an even greater basis for using three
separate tests of the metaeconomic model, with each using a different other-interest
proxy.
5.3. Results of Logit Test of Microeconomic and Metaeconomic Theory
As noted, Logit models were used to empirically estimate the equivalent of
equations (21) through (24). The following three tables summarize the results.
79
Table 19: Logistic Estimation of No-Till Adoption Decision (Empathy Other-Interest Proxy)
Variable
Role of
Capital
Adding
Tempered Self
Adding
Habitual Tendency
Adding Selfism
Reinforced Control
Constant 1.005 -0.730 -1.823c -1.078
Income 0.006a 0.005
a 0.004
a 0.004
a
Slope -0.033 0.016 0.038 0.008
Empathy*Selfism
0.070
a 0.066
a 0.090
a
Habit
0.383
a 0.371
a
Selfism*Farm Control
-0.089
a
Selfism*Other Control
0.013
Selfism*Nature Control
0.002
-2 Log Likelihood 442.134 422.482 384.954 374.016
χ2 (Block) 31.474a 19.651
a 37.529
a 10.938
b
χ2 (Model) 31.474a 51.125
a 88.653
a 99.592
a
Nagelkerke R2 .100 .159 .266 .295
Percentage Correct:
0 0 2.2 23.1 28.6
1 100 99.8 96.8 95.8
Overall 81.7 81.9 83.3 83.5
Df 2 3 4 7
Note: a p<.01,
b p<.02,
c p<.05,
d p<.10
80
Table 20: Logistic Estimation of No-Till Adoption Decision (Sympathy Other-Interest Proxy)
Variable
Role of
Capital
Adding
Tempered Self
Adding
Habitual Tendency
Adding Selfism
Reinforced Control
Constant 1.005 0.122 -1.157 -0.787
Income 0.006a 0.006
a 0.004
a 0.004
a
Slope -0.033 -0.012 0.004 -0.022
Sympathy*Selfism
0.037
b 0.041
b 0.043
c
Habit
0.401
a 0.391
a
Selfism*Farm Control
-0.082
a
Selfism*Other Control
0.035
Selfism*Nature Control
0.017
-2 Log Likelihood 442.134 435.479 393.881 384.343
χ2 (Block) 31.474a 6.655
b 41.598
a 9.538
c
χ2 (Model) 31.474a 38.129
a 79.727
a 89.624
a
Nagelkerke R2 .100 .120 .241 .267
Percentage Correct:
0 0 0 16.5 22.0
1 100 100 96.1 96.8
Overall 81.7 81.7 81.5 83.1
Df 2 3 4 7
Note: a p<.01,
b p<.02,
c p<.05,
d p<.10
81
Table 21: Logistic Estimation of No-Till Adoption Decision (Empathy/Others Other-Interest Proxy)
Variable
Role of
Capital
Adding
Tempered Self
Adding
Habitual Tendency
Adding Selfism
Reinforced Control
Constant 1.005 -0.734 -1.894c -1.176
Income 0.006a 0.005
a 0.004
a 0.004
a
Slope -0.033 0.006 0.020 -0.022
Water User Empathy*Selfism
0.013 0.014 0.016
Farm Entity Empathy*Selfism
0.090
a 0.084
a 0.093
a
Family Empathy*Selfism
-0.021 -0.016 -0.010
Habit
0.381
a 0.374
a
Selfism*Farm Control
-0.094
a
Selfism*Other Control
0.022
Selfism*Nature Control
0.009
-2 Log Likelihood 442.134 413.540 376.536 365.692
χ2 (Block) 31.474a 28.593
a 37.005
a 10.844
b
χ2 (Model) 31.474a 60.067
a 97.072
a 107.916
a
Nagelkerke R2 .100 .185 .289 .317
Percentage Correct:
0 0 6.6 22.0 27.5
1 100 99.3 96.3 96.3
Overall 81.7 82.3 82.7 83.7
Df 2 5 6 9
Note: a p<.01,
b p<.02,
c p<.05,
d <.10
82
81
In all three models presented above, the column labeled “Role of Capital”
is representative of equation (21) and represents the standard empirical derived
demand function; the column labeled “Adding Tempered Self” is representative
of equation (22) and represents the most basic metaeconomic empirical derived
demand function; the column labeled “Adding Habitual Tendency” is
representative of equation (23) and adds the aforementioned habit variable to the
metaeconomic derived demand function; and finally, the column labeled “Adding
Selfism Reinforced Control” is representative of equation (24) and adds the three
previously mentioned control variables to the metaeconomic derived demand
function.
Examination of the results from all three models provides some very
intriguing insights into what motivates the conservation tillage adoption decision
among farmers in the four county target area above Tuttle Creek Lake. First, take
note of the results presented in the column labeled “Role of Capital” in all three
logit models. This is the empirical derived demand model described in
microeconomic-based production economics. As microeconomics would suggest,
we find that income (i.e. financial capacity) is a significant variable that helps to
explain a farmer‟s decision to adopt no-till and conservation tillage technologies.
The chi-square statistic for this model also shows the overall model to be
significant in explaining tillage behavior. While the model is significant, though,
it should be noted that the coefficient on the income variable indicates that an
increase in income actually has a very small impact on a farmer‟s tillage decision.
83
81
In fact, a one thousand dollar increase in gross income only increases the odds of
a farmer adopting no-till and conservation tillage strategies by 0.06 percent (i.e.
less than one percent). Also note that the microeconomic model does a poor job
of predicting which respondents do not use no-till and conservation tillage
technologies.
While the microeconomic model predicts conservation tillage behavior
reasonably well, it does not compare favorably to the empirical derived demand
model proffered by metaeconomic theory. In fact, we find that regardless of the
shared other-interest proxy used (empathy, sympathy, or empathy/others), the
metaeconomic derived demand model predicts conservation tillage behavior much
better than the microeconomic-based derived demand model.
The basic metaeconomic derived demand model is presented in Tables 19,
20, and 21 under the column labeled “Adding Tempered Self.” Notice that in all
three tables, the income variable remains significant, just as in the
microeconomics model. Yet, we also find that the metaeconomic variables (self-
interest*shared other-interest) contribute significantly to understanding farmer
tillage behavior. In Table 21, we find that the empathy*selfism variable is a
significant predictor when attempting to understand tillage behavior in the
watershed. The chi-square (block) statistic also shows that adding the tempered
self-interest variable improves the overall model fit. Also, we find that the R-
square statistic increased from 0.10 to 0.159.
84
81
Table 21, which presents the results of the metaeconomic derived demand
model that uses sympathy as a proxy for an individual‟s shared other-interest
tendency, tells much the same story as the results provided in Table 20. Again,
income is a significant variable in the individual tillage decision. However, like
the model that uses empathy as a proxy, the sympathy*selfism variable is also a
significant factor in predicting the tillage decision. Also, the magnitude of the
coefficient associated with the sympathy*selfism variable is larger than the
coefficient associated with an individual‟s income level. Finally, we again see
that the chi-square (block) statistic indicates that the addition of the
sympathy*selfism variable improves overall model fit, and the Nagelkerke R-
Square statistic increases from 0.10 to 0.12.
Finally, Table 22 presents the metaeconomic derived demand model when
the empathy/others variables are used as proxies for an individual‟s shared other-
interest tendency. Yet again, we find that this model tells much the same story as
the models presented in Tables 20 and 21. Again, income is a significant variable
in the tillage decision, but the farm entity*selfism coefficient is also significant
and greater in magnitude than the income coefficient. Somewhat surprisingly,
though, this model shows that both other lake users and family members do not
appear to impact the conservation tillage decision in farmers residing above Tuttle
Creek Lake. Despite these surprising results, the metaeconomic model presented
in Table 22 still yields a better fitting model than the standard production
economics derived demand model. This is evidenced by the significant chi-
85
81
square (block) statistic and an increase in the Nagelkerke R-square statistic from
0.10 to 0.185.
The results presented in Tables 20, 21, and 22 show that the
metaeconomic derived demand model yields a better description of what
motivates tillage behavior in the watershed than the microeconomics model.
However, it is theorized that the basic metaeconomic model can be further refined
and improved by adding variables that account for individual habitual tendencies
and preferences for control. The columns labeled “Adding Habitual Tendency”
and “Adding Selfism Reinforced Control” provide the results when proxies for
these two phenomena are added to the metaeconomic model.
Inspection of Tables 20, 21, and 22 show that adding individual habitual
tendencies does in fact improve the basic metaeconomic derived demand model,
regardless of which shared other-interest proxy is used. In all three cases, the
habit variable coefficients are significant, large in magnitude, and in the
hypothesized positive direction. In addition to this, the income and shared other-
interest variables all remain significant in all three models. The addition of a
habit variable also substantially improves the model fit of the metaeconomic
derived demand model. All three chi-square (block) statistics are significant, and
in all three instances we find considerable increases in the Nagelkerke R-square
statistics. This suggests that subconscious feelings about tillage decisions made in
the past play a great part in tillage decisions that are made today or in the future.
86
81
Finally, we also find that the preference for control variables presented in
the columns labeled “Adding Selfism Reinforced Control” also help to refine and
improve the basic metaeconomic derived demand model. Regardless of the
shared other-interest proxy used, the selfism*farm control variable becomes a
significant predictor when attempting to understand tillage behavior. The variable
is also in the hypothesized negative direction. Again, as when adding the habit
variable, the addition of the selfism reinforced control variables contributes
significantly to the overall model fit, as evidenced by the significant chi-square
(block) statistic in all three cases. The Nagelkerke R-square statistics also
increase with the addition of the control variables to the metaeconomic model. It
should be noted, though, that only the selfism*farm control variable is significant
in the model. This indicates that an individual‟s preferences for control over
nature and attitudes toward control exerted on their farms by others are not
important in the tillage adoption decision.
In sum, the results presented in Tables 20, 21, and 22 indicate that the
refined metaeconomic model that includes habitual tendencies and preferences for
control is vastly superior at predicting tillage behavior in the Blue River/Tuttle
Creek Lake watershed than the standard microeconomics model. This
metaeconomic model gives the largest Nagelkerke R-square statistics, best overall
model fit, and yields the greatest percentage of correct 0,1 dependent variable
predictions. Thus, it appears that new economic models that account for the
87
81
psychological dispositions described here should be created in order to truly
understand the conservation adoption decision made by farmers.
5.4. Results of Heckman Test of Conservation Tillage Intensity
While the results above give a fairly clear depiction of the several different
factors that can motivate farmers in the Blue River/Tuttle Creek watershed to
adopt conservation tillage strategies, it is also important to attempt to understand
what factors influence the conservation tillage intensity decision made by farmers
in the watershed. Given that the refined metaeconomic derived demand model
did the best job in predicting the adoption decision made by farmers, these same
variables were used in constructing Heckman models from the data obtained from
the region. The results of the three models, which were constructed with the
software program Shazam, are presented below.
88
Table 22: Probit Estimation of Individual No-Till Adoption Decision (Empathy Other-Interest Proxy)
Variable Coefficient Standard Error T-Stat
Constant -0.461 0.494 -0.933
Income 0.00163 0.000550 2.96a
Slope -0.0150 0.121 -0.125
Empathy*Selfism 0.0501 0.0133 3.76a
Habit 0.217 0.365 5.95a
Selfism*Farm Control -0.0471 0.0152 -3.10a
Selfism*Other Control 0.00361 0.0161 0.225
Selfism*Nature Control 0.00130 0.00931 0.140
Log Likelihood -236.80 - -
Cragg-Uhler R2 .29 - -
% Right Predictions 0.84
a p<.01,
b p<.02,
c p<.05,
d p<.10
N = 498
89
Table 23: Semi-Log Estimation of Individual No-Till Intensity (Empathy Other-Interest Proxy)
Variable Coefficient Standard Error T-Stat
Constant -0.234 0.177 -1.32
Income 0.000225 0.000127 1.77d
Slope -0.0409 0.0404 -1.01
Empathy*Selfism 0.00182 0.00421 0.431
Habit -0.00616 0.0134 -0.462
Selfism*Farm Control -0.0156 0.00534 -2.92a
Selfism*Other Control 0.0100 0.00496 2.02c
Selfism*Nature Control 0.00400 0.00308 1.30
Adjusted R2 .03 - -
a p<.01,
b p<.02,
c p<.05,
d p<.10
N = 407
IMR = 0.070
90
Table 24: Probit Estimation of Individual No-Till Adoption Decision (Sympathy Other-Interest Proxy)
Variable Coefficient Standard Error T-Stat
Constant -0.308 0.488 -0.630
Income 0.00160 0.000543 2.95a
Slope -0.0338 0.119 -0.283
Sympathy*Selfism 0.0259 0.0111 2.34b
Habit 0.225 0.0364 6.16a
Selfism*Farm Control -0.0438 0.0150 -2.92a
Selfism*Other Control 0.0159 0.0155 1.03
Selfism*Nature Control 0.00929 0.00870 1.07
Log Likelihood -236.80 - -
Cragg-Uhler R2 .26 - -
% Right Predictions 0.83
a p<.01,
b p<.02,
c p<.05,
d p<.10
N = 498
91
Table 25: Semi-Log Estimation of Individual No-Till Intensity (Sympathy Other-Interest Proxy)
Variable Coefficient Standard Error T-Stat
Constant -0.201 0.175 -1.15
Income 0.000218 0.000127 1.72d
Slope -0.0437 0.0403 -1.08
Sympathy*Selfism -0.00119 0.00366 -0.326
Habit -0.00561 0.0133 -0.421
Selfism*Farm Control -0.0151 0.00531 -2.85a
Selfism*Other Control 0.0116 0.00493 2.34a
Selfism*Nature Control 0.00440 0.00302 1.46
Adjusted R2 .03
a p<.01,
b p<.02,
c p<.05,
d p<.10
N = 407
IMR = 0.382
92
Table 26: Probit Estimation of Individual No-Till Adoption Decision (Empathy/Others Other-Interest Proxy)
Variable Coefficient Standard Error T-Stat
Constant -0.609 0.505 -1.21
Income 0.00175 0.000574 3.05a
Slope -0.0199 0.123 -0.162
Water User*Selfism 0.00678 0.0118 0.573
Farm Entities*Selfism 0.0542 0.0169 3.21a
Family*Selfism -0.00610 0.00993 -0.615
Habit 0.211 0.0367 5.76a
Selfism*Farm Control -0.0468 0.0152 -3.09a
Selfism*Other Control 0.00840 0.0155 0.543
Selfism*Nature Control 0.00603 0.00910 0.661
Log Likelihood -236.80 - -
Cragg-Uhler R2 .31 - -
% Right Predictions 0.83
a p<.01,
b p<.02,
c p<.05,
d p<.10
N = 498
93
Table 27: Semi-Log Estimation of Individual No-Till Intensity (Empathy/Others Other-Interest Proxy)
Variable Coefficient Standard Error T-Stat
Constant -0.251 0.178 -1.41
Income 0.000229 0.000127 1.80d
Slope -0.0353 0.0406 -0.869
Water User*Selfism 0.00325 0.00384 0.847
Farm Entities*Selfism 0.00136 0.00559 0.243
Family*Selfism -0.00352 0.00340 -1.04
Habit -0.00752 0.0134 -0.560
Selfism*Farm Control -0.0158 0.00534 -2.96a
Selfism*Other Control 0.0107 0.00473 2.27c
Selfism*Nature Control 0.00458 0.00307 1.49
Adjusted R2 .03 - -
a p<.01,
b p<.02,
c p<.05,
d p<.10
N = 407
IMR = 0.211
94
93
Notice that the Heckman model is a two stage process. First, a probit estimation
of the (0,1) adoption decision must be constructed. Then, using the information obtained
in the probit analysis, a statistic called the Inverse Mills Ratio is computed and used as a
variable in the second step standard OLS estimation. In the OLS estimation, all
observations with a “0” on the dependent variable are removed, such that the estimate is
applied only to those that use conservation tillage in varying intensities. The inverse
mills ratio, then, is meant to correct for the resulting bias in the estimation due to the
elimination of all data associated with those respondents that do not use conservation
tillage technologies.
The probit estimation is similar in nature to the logit model. In fact, the only real
difference is that the probit estimation assumes a normally distributed error term, whereas
a logit model assumes that the error term is distributed in a logistical manner. Given the
similarities in the modeling techniques, it would be expected that the results from the
probit estimations used in the Heckman technique above should be quite similar to the
results produced by the logit model estimation described earlier. Inspection of Tables 23,
25, and 27 show this to be exactly the case. In fact, the probit estimation conducted as a
part of the overall Heckman technique tells the exact same story as the logit models
presented earlier. This is to say, the probit estimation, regardless of which shared other-
interest proxy is used, indicates that the shared other-interest*selfism variable, the habit
variable, the selfism*farm control variable, and the income variable are all significant
predictors in the conservation tillage adoption decision. So, given the similarities in the
results between the probit and logit estimations, we assert with relative confidence that
95
93
metaeconomic theory provides a robust and consistent derived demand model that helps
to explain the tillage adoption decision exhibited by farmers in the four county target area
above Tuttle Creek Lake.
While the metaeconomic variables described above tell a consistent story
regarding the conservation tillage adoption decision, the conservation tillage intensity
decision appears to be influenced by different factors. Inspection of Tables 24, 26, and
28 shows that the only variables that are consistently significant in predicting the tillage
intensity decision are selfism*farm control and selfism*other control. Notice that the
income variable, too, is significant in all models. However, it is only significant at the
p<0.10 level, a cutoff value that is determined to be too high for determining significance
levels in behavioral economics research.
One should take note that the coefficient associated with the selfism*farm control
is negative, while the coefficient associated with the selfism*other control variable is
positive. This result, though, is not entirely unexpected. Farm control is measured as the
degree to which the respondent perceives having autonomous control over his or her
farming processes while using conservation tillage strategies. This variable was coded
for statistical analysis in such a way that larger numbers indicated less perceived
autonomous control over farming processes while using conservation tillage strategies.
So, a one unit increase in this proxy should impact conservation tillage intensity in a
negative manner. Thus, the negative coefficient on the final selfism*farm control
variable makes sense in this context.
96
93
The positive coefficient on the selfism*other control also makes a great deal of
sense. This result is due to how the other control variable was coded for statistical
analysis purposes. The variable was coded in such a way that larger numbers indicate
that others have no heteronomous control over the conservation tillage intensity decision.
If the respondent perceives that others cannot exert heteronomous control over the
conservation tillage intensity decision, it can be reasonably inferred that the respondent
must feel that he or she maintains autonomous control over the situation. Thus, we
would expect that a one unit increase in this variable would yield a positive result upon
the intensity of conservation tillage used on an individual farm.
While the direction of the selfism*farm control and selfism*other control can be
reasonably explained, one must also take the results in Tables 24, 26, and 28 with a grain
of salt. First, the selfism*other control variable is not a significant variable when
sympathy is used as a shared other-interest proxy. Also, the model fits for the standard
OLS portions of the Heckman models presented above are poor, with R-squared values of
no greater than 0.03 units. Therefore, it appears that follow-up work should be conducted
in the watershed in order to better determine what motivates the conservation tillage
intensity decision.
97
93
Figure 5.1. Distribution of mean empathy scale responses
Figure 5.2. Distribution of mean sympathy scale responses
1.00 2.00 3.00 4.00 5.00 6.00 7.00
empathy
0
5
10
15
20
25
30
35
40
Co
un
t
1.00 2.00 3.00 4.00 5.00 6.00 7.00
sympathy
0
10
20
30
40
50
60
70
80
90
Co
un
t
98
93
CONCLUSIONS, IMPLICATIONS, AND RECOMMENDATIONS
The overall goal of this research was to more fully understand farmer
conservation behavior, and in the process, perhaps discover some information that could
prove to be useful in solving the pollution problems in the Blue River and Tuttle Creek
Lake. It was also thought that results from this research could be used to help improve
conservation policies administered throughout the United States.
This study was very unique in nature. This research could be classified as an
economic study, and includes all the elements one would expect to find in an empirical
study using derived demand theory as presented in microeconomic-based production
economics. However, the research conducted also sought to go beyond and transcend
(the notion of “meta”) the traditional economic framework, and thus metaeconomic
theory (which includes elements from psychology, sociology, and other social sciences)
was also used to test the motivations for farmer conservation behavior.
Given the exceptional set of circumstances used in this study, the results produced
from the research are also very unique in nature. For instance, the results indicate that a
farmer‟s income/financial capacity is an important factor in the conservation tillage
adoption decision faced by farmers in the Blue River/Tuttle Creek Lake watershed. This
result makes intuitive sense, as most conservation practices are not inherently profitable
and there is some level of cost associated with purchasing new tractors, planters, and
other equipment that must be used in order to farm using conservation tillage strategies.
What makes the result truly remarkable, though, is that increases in income/financial
99
93
capacity, in an absolute sense, actually have a very small (albeit significant) impact upon
the conservation tillage decision. In fact, our results showed that a one thousand dollar
increase in a farmer‟s income increases the odds of conservation tillage adoption by less
than one percent. This result is completely counter to the idea that substantial increases
in income are needed in order to induce farmers to engage in conservation tillage
activities.
In addition to our findings concerning the role of income in farmer tillage
behavior, other psychological variables included in this research make the results
distinctive and somewhat ground breaking. Most prominent in our research is the fact
that both self-interest tendencies and shared other-interest tendencies play a role in the
conservation tillage adoption decision made by farmers in the Blue River/Tuttle Creek
Lake watershed. This research first showed that proxies used to measure self-interest and
shared other-interest tendencies are significantly and negatively correlated. This lends
credence to the idea that ego and empathy are joint, non-separable, and in constant
conflict as proposed by metaeconomic theory. Thus, it also appears likely that individual
decisions cannot be made without considering both self-interest and shared-other interest
tendencies together. In metaeconomic terms, then, it appears as though a decision maker
must consider both conflicting tendencies and integrate them into one decision or
behavior on path 0Z.
Logit models created with survey data collected in the region also show the
importance of the self-interest and shared other-interest interaction. In all models, the
shared other-interest*self-interest variable proved to be significant when attempting to
100
93
predict conservation adoption in the watershed. Since the selfism scale was ultimately
reversed when creating this interaction variable, the result is really a measure of how a
person is oriented towards the shared other-interest in farming. The results ultimately
show, then, that those farmers who are less selfish in nature are likely currently using
conservation tillage practices and are much more likely to continue using the technology
in the future.
In addition to the results concerning the shared other-interest*self-interest
interaction, the survey data compiled and models created also indicated that a farmer‟s
preferences regarding autonomous and heteronomous control are also very important
factors in the conservation tillage adoption decision. The logit and Heckman models
created considered three different exploratory control variables: control over farm
processes, control exerted by others, and preferences for control over nature. The results
of the logit models consistently show that farm control is a significant variable. This
result shows that if a person believes that they can use conservation tillage techniques and
still keep complete autonomous control over their farm, they will be more likely to use
conservation techniques. However, if a farmer perceives a loss of control over their farm
by using conservation tillage, the odds of conservation tillage adoption significantly
deteriorate. In addition to the results from the logit models, the Heckman models
produced show that both farm control and other control impact the conservation tillage
intensity decision. So, contrary to predictions and theory offered in microeconomics, it
appears that these psychological context variables really do play an important role in
individual decision making. It should be noted, though, that preferences for control over
101
93
nature was not deemed to be major factor in the conservation tillage adoption or intensity
decision.
Finally, this research also concluded that a farmer‟s habitual tendencies play a
major role in the tillage adoption decision. In fact, in terms of coefficient magnitude, the
habit variable proved to have the greatest impact upon the conservation adoption
decision. This is to say, a one unit increase in the habit proxy yields nearly a forty
percent increase in the odds of conservation tillage adoption. So, it appears that those
that have used conservation tillage strategies in the past are much more likely to continue
using it in the future. However, it must be cautioned that the opposite situation may also
apply. This is to say, if a farmer has not been convinced of the benefits of using
conservation tillage techniques and continues to use intensive tillage technologies, he or
she is more likely to rely on subconscious feelings and intuition and continue to use
intensive tillage practices. In terms of policy, then, it seems imperative to use a mix of
education and financial incentives in order to help convince intensive tillage farmers of
the benefits of conservation tillage technologies. Once these intensive tillage users have
converted to conservation tillage techniques, they will then be much more likely to
continue using the technology.
The results from this study indicate that a single over-arching conservation policy
administered on a volunteer basis is not likely to be successful in reducing agricultural
non-point pollution. This conclusion can be drawn because our survey results show that
farmers are very heterogeneous in their psychological and economic motivations. This is
102
93
starkly different than the homogenous Homo economicus assumed in standard economic
theory.
The results show that farmers vary in the degree to which they are influenced by
both self-interest and other-interest tendencies. Those that are more self-interested in
nature are most likely influenced by profit considerations. Therefore, financial incentive
programs may help encourage these farmers to engage in conservation practices.
However, by only targeting these self-interested individuals through the use of financial
instruments, a large subset of the farming population that is largely influenced by a
psychological orientation toward the shared other-interest are most likely not going to
participate in the incentive programs. These other-interested individuals are most likely
to participate in programs that emphasize the communal side of farming in a way that can
help them identify with others. In other words, they enjoy a connection with others that
comes from being identified as a “conservation farmer,” and being in unity with other
such producers.
While there are some in the farming community that are motivated to engage in
conservation activities by the extremes of either self-interest or shared other-interest, our
results show that most individual farmers will be motivated by a complex mix of self-
interest and other-interest on the metaeconomic satisficing path 0Z. So, it appears that
the best conservation policies are those that can emphasize both self- and other-interest.
While financial incentives may help those who are beginning to use conservation
tillage measures, it is unlikely that increases in income alone are enough to sustain usage
of the technology over long periods of time. Therefore, education programs will continue
103
93
to play a role in helping farmers move toward using conservation technologies.
Education programs can especially stimulate a farmer‟s self-interest tendencies.
Extension educators can show that conservation tillage may result in using less fuel,
helping soil retain moisture, and in producing higher yields. So, while a focus on the
financial side remains important, the results of this study also point to the need to help
farmers see the possibility of using conservation measures for their own sake and lead to
more sustainable use of the techniques. In this case, we also find that educational
programs can stimulate a farmer‟s shared other-interest tendencies, showing that
conservation tillage technologies help conserve soil quantity and quality, protect
farmland for future generations, and prevent soil and chemical runoff to nearby rivers and
streams. So, educational programs can help to show farmers that conservation
technologies are good for both self-interest and other-interest tendencies. Extension
educators could design programs specifically focused on building and enhancing a
community ethic within the farming population to become a “conservation farmer” and
help establish a unity with the cause (sympathy) of using tillage strategies that improve
water quality. As more small groups of farmers…even including the morning coffee
event at the local diner…become engaged in the process, we can reasonably expect that
this sympathy can also be built through word of mouth, and through the networks of
farmers connecting with other small groups throughout the community. In fact, this
shared ethic needs to spread to others beyond the farmers themselves, to include
equipment, fertilizer, herbicide and see dealers; the landlords, and others considered
earlier herein in the list of those who may influence and be influenced by farmers. As the
104
93
unity with the cause of a less polluted lake evolves and spreads, can reasonably expect
that a synergistic path 0Z will emerge and the use of conservation tillage technologies
will be much more consistently applied.
While this research has implications for farm level conservation policy and local
implementation of programs and practices, it also has implications for the water quality
conflict that is beginning to emerge between upstream farmers and downstream water
users in the Blue River/Tuttle Creek watershed. The disagreement in the watershed, while
not highly publicized at this point in time, is a real phenomenon that can not be ignored,
especially when one considers that any type of resolution regarding water quality in the Blue
River and Tuttle Creek Lake will have an impact on both upstream and downstream
stakeholders. However, if we begin to work toward a solution to the dispute now, it is
possible that the conflict can be resolved without mandates and regulation from State and/or
Federal governments. The key to this type of agreement would be through an expression of
empathy-sympathy on the part of the both upstream producers (and other upstream residents)
and downstream water users/residents.
We look especially to the notion of sufficient reason for a new vision and change as
framed in Bromley (2006) for guidance in this situation. In turn, in order to understand the
notion of sufficient reason in the context of the metaeconomic behavior found in the
farming population, we must first step back and think about the underlying philosophical
basis for institutional and behavioral change, and policy change more generally, and for
economic theorizing about said change. Bromley (2006) argues real change is rooted in
pragmatism, and a kind of volitional pragmatism in particular, not in utilitarianism.
Pragmatism is also in the foundation of metaeconomics. This means that life is really
105
93
about “incessant doing” and that policy change is an inherent part of this “doing,” and we
do the best we can (Bromley, 2006, p. 137). It also means that as members of a
democratic society and market economy, we are able to be freely practical in our choices,
going beyond mere calculations of optimums. Thinking back to the metaecconomic
figure presented in Chapter 3, we can see that this situation is easily depicted: individuals
do not generally respond only to the price ratio and self-interest maximization at point A,
but rather temper their interests in satisficing, as represented in the kinds of real choices
illustrated at point B; point B cannot be “calculated” as can point A. Indeed, we do not
even seek the maximums at point A or C, but rather do our best at point B.
Bromley (2006, pp. 51-62) clarifies that institutions are composed of three
entities: norms and traditions, working rules, and private property relations, with private
property relations being especially important in policy construction and change. All
condition and temper behavior, in that individuals are embedded in, through internalizing,
these institutions, which are represented in the various metaeconomic paths 0M. As
policy evolves, private property regimes are modified, as some individuals receive new
duties (restraints), while others receive new rights (liberation) and expansion of
individual action (Bromley, 2006, p. 23). In effect, some are given the opportunity to act
on their newly awarded path OG; others are required to move away from it. Ultimately,
then, it is changes to these institutions that bring about overall change in policy, and
behavioral change on a new path 0Z.
In thinking about how the metaeconomics and sufficient reason frames merge, we
find that modifications to institutions occur through expressions of empathy and
106
93
sympathy. This is to say, those engaged in conflict can be freely practical in decision
making by empathizing and negotiating with each other. If all affected parties are willing
to walk in the shoes of the others, we will generally see the emergence of ever more
sympathy with the common cause, in this case a less polluted lake, with changes in
institutions and policy that bring about peaceful conflict resolution. More specifically, in
the case of stakeholders in the Blue River/Tuttle Creek Lake watershed, then, if we find
ample amounts of empathy/sympathy in the region, we would expect that a peaceful
resolution to the water quality conflict is possible.
As shown previously, survey results from farmers in the region above Tuttle
Creek Lake showed that empathy and sympathy are both present in the region. The
relatively high mean scores of both the empathy and sympathy measures show that
farmers upstream of Tuttle Creek Lake have the capacity to walk in the shoes of
downstream Lake users and engage in the policy process such that the irritation caused by
the water quality and quantity conflict may subside. Therefore, a memorandum of
agreement concerning water quality in the region indeed seems possible between the two
impacted parties.
In order for the water quality conflict to truly subside, though, downstream water
users must also be willing to engage in the act of empathizing with upstream farmers in
order to understand the difficulties faced by those farmers in preventing erosion and
chemical runoff. Also, this expression of empathy would, as in the case of the farmers,
also have to result in sympathy for the shared cause of a less polluted lake.Unfortunately,
we do not at this time have data concerning public water suppliers and residents
107
93
downstream of Tuttle Creek Lake. Therefore, more research needs to be conducted
concerning the role that empathetic and sympathetic tendencies of both downstream and
upstream water users might have on shaping future policy decisions in the watershed.
Ultimately, though, such knowledge and this approach could have a substantive impact
on the policy process and the potential resolution of the water quality conflict in the Blue
River/Tuttle Creek Lake watershed.
While substative conclusions can be drawn from this research, there are certain
limitations. The first issue that may need to be addressed concerns model form. Proxies
for self-interest, other-interest, and preferences for control were all used to create specific
interaction terms in both the logit and Heckman models. However, no linear effects for
the interactions were created in the model. This approach can be defended based upon
the theorectical representation of the brain. However, this form should continue to be
evaluated. It seems that the use of Structural Equation Modeling (SEM) could prove to
be useful to confirm the relationships between self- and other-interest, and self-interest
and preferences for control. This work, though, has been left for another study.
Importantly, this study is another in a series of tests of metaeconomic theory, with
similar results. Metaeconmic tests began in the late 1980‟s, and continue to this day.
Intriguingly, these studies continue to find evidence of a substantive role for both the
shared other-interest and control in an individual‟s decision making process. However,
this testing needs to be expanded further in order to validate the generalizability of the
model.
108
93
While there are indeed drawbacks to the metaeconomic research conducted in the
Blue River/Tuttle Creek watershed, the fact remains that the results produced have
provided some intriguing insights into potential motivators for farmers to utilize
conservation tillage strategies in the region. It is our hope, then, that these results and
future research can help to improve conservation policy in the United States. Hopefully,
too, the improved policies can lead to the restoration of rivers, streams, and lakes to a
more natural and clean state, and conflicts regarding water quality can be resolved.
109
93
References
Ajzen, I. “The Theory of Planned Behavior.” Organizational Behavior and Human
Decision Processes 50 (1991): 179-211.
Angyal, A. Foundations for a Science of Personality. Cambridge, MA: The
Commonwealth Fund by Harvard University Press, 1967.
Artikov, I., S.J. Hoffman, G.D. Lynne, L.M. Pytlik-Zillig, Qi Hu, A.J. Tomkins, K.G.
Hubbard, M.J. Hayes, and W. Waltman. “Understanding the Influence of
Climate Forecasts on Farmer Decisions as Planned Behavior.” Journal of
Applied Meteorology and Climatology 45 (2006): 1202-1214.
Barnes, P. Kansas State University, interview by Robert Sheeder (2006).
Bromley, D.W. Sufficient Reason: Volitional Pragmatism and the Meaning of Economic
Institutions. Princeton, New Jersey: Princeton University Press, 2006.
Cain, Z., and S. Lovejoy. “History and Outlook for Farm Bill Conservation Programs.”
Choices (2004): 37-42.
Cary, J.W., and R.L. Wilkinson. “Perceived Profitability and Farmers‟ Conservation
Behaviour.” Journal of Agricultural Economics 48, 1 (1997): 13-21.
Casey, F. and G. D. Lynne. “Adoption of Water Conserving Technologies in Agriculture:
The Role of Expected Profits and the Public Interest.” In F. Casey, A. Schmitz,
S. Swinton and D. Zilberman (Editors). Flexible Incentives for the Adoption of
Environmental Technologies in Agriculture. Norwell, MA: Kluwer Academic
Publishers, 1999, 229-247.
110
93
Chouinard, H.H., T. Paterson, P.R. Wandschneider, and A.M. Ohler. “Will Farmers
Trade Profits for Stewardship? Heterogeneous Motivations for Farm Practice
Selection.” Land Economics, 84,1 (2008): 66-82.
Cooper, J.C. “Combining Actual and Contingent Behavior Data to Model Farmer
Adoption of Water Quality Protection Practices.” Journal of Agricultural and
Resource Economics 78 (1997): 30-43.
-------- and Keim, R.W. “Incentive Payments to Encourage Farmer Adoption of Water
Quality Protection Practices.” American Journal of Agricultural Economics 78
(1996): 54-64.
Cory Jr., G.A. “A Behavioral Model of the Dual Motive Approach to Behavioral
Economics and Social Exchange.” Journal of Socio-Economics, 35 (2006a): 592-
612.
--------. “Physiology and Behavioral Economics.” Chp 2 In Handbook of Contemporary
Behavioral Economics: Foundations and Developments, edited by M. Altman, pp.
24-49. Armonk, New York: M.E. Sharpe, 2006b.
--------. The Consilient Brain: The Bioneurological Basis of Economics, Society, and
Politics. New York: Kluwer Academic/Plenum Publishers, 2004.
Cutforth, L.B., C.A. Francis, G.D. Lynne, D.A. Mortensen, and K.M. Eskridge. “Factors
Affecting Farmers' Diversity Approach.” Amer. J. Altern. Agri. 16, 4 (2001):
168-176.
Davis, M.H. “A Multidimensional Approach to Individual Differences in Empathy.”
JSAS Catalog of Selected Documents in Psychology 10, 4 (1980): 85-102.
111
93
Decety, J., K.J. Michalska, and Y. Akitsuki. “Who Caused the Pain? An fMRI
Investigation of Empathy and Intentionality in Children.” Neuropsychologia 46,
11 (2008): 2607-2614.
Economic Research Service. “ERS/USDA Briefing Room-Conservation Policy:
Background.”
http://www.ers.usda.gov/Briefing/ConservationPolicy/background.htm
(accessed January 30, 2008).
Ervin, C.A. and D.E. Ervin. “Factors Affecting the Use of Soil Conservation Practices:
Hypotheses, Evidence, and Policy Implications.” Land Economics 58, 3 (1982):
277-292.
Etzioni, A. “The Case for a Multiple-Utility Conception.” Economics and Philosophy, 2,
2 (1986): 159-183.
Frisch, R. Theory of Production. Chicago: Rand McNally and Company, 1965.
Gould, B.W., W.E. Saupe, and R.M. Klemme. “Conservation Tillage: The Role of Farm
and Operator Characteristics and the Perception of Soil Erosion.” Land
Economics, 65, 2, (1989): 167-182.
Heckman, J.J. “Sample Selection Bias as a Specification Error.” Econometrica 47, 1
(1979): 153-161.
Kruse, C.E. “Explaining Farmer Behavior in Relation to Sequestering Carbon on
Nebraska Farms.” Master‟s thesis, University of Nebraska-Lincoln, 2003.
Kahneman, D., P.P. Waker, and R. Sarin. “Back to Bentham? Explorations of
Experienced Utility.” Quarterly Journal of Economics 112, May (1997): 375-405.
112
93
--------. “Maps of Bounded Rationality: Psychology for Behavioral Economics.”
American Economic Review, 93 (2003): 1449-1475.
Kalinowski, C.M., G.D. Lynne, and B. Johnson. “Recycling as a Reflection of Balanced
Self-Interest: A Test of the Metaeconomics Approach.” Environment and
Behavior, 38, 3 (2006): 333-355.
Lakoff, G. Moral Politics: What Conservatives Know that Liberals Don’t. Chicago: The
University of Chicago Press, 1996.
Lichtenberg, E. “Cost-Responsiveness of Conservation Practice Adoption: A Revealed
Preference Approach.” Journal of Agricultural and Resource Economics 29, 3
(2004): 420-435.
Lohr, L. and T.A. Park. “Utility-Consistent Discrete-Continuous Choices in Soil
Conservation.” Land Economics 71, 4 (1995): 474-490.
Link, M. 2005 Nebraska Surface Water Quality Monitoring Report. Lincoln, NE:
Nebraska Department of Environmental Quality, 2005.
Lynne, G.D. “Allocatable Fixed Inputs and Jointness in Agricultural Production:
Implications for Economic Modeling: Comment.” American Journal of
Agricultural Economics, 70, 4 (1988): 948-949.
Lynne, G.D. “Modifying the Neo-Classical Approach to Technology Adoption With
Behavioral Science Models.” J. Agr. Applied Econ. 27,1 (July, 1995): 67-80.
--------. “Divided Self Models of the Socioeconomic Person: The Metaeconomics
Approach.” Journal of Socio-Economics, 28 (1999): 267-288.
113
93
--------.“On the Economics of Subselves: Toward a Metaeconomics.” Chp. 6 In
Handbook of Contemporary Behavioral Economics: Foundations and
Developments, edited by M. Altman, pp. 99-122. Armonk, New York: M.E.
Sharpe, 2006a.
--------. “Toward a Dual Motive Metaeconomic Theory.” Journal of Socioeconomics 35
(2006b): 634-651.
-------- and C. F. Casey. “Regulation of Technology Adoption When Individuals Pursue
Multiple Utility.” Journal of Socioeconomics 27, 6 (1998): 701-719.
MacLean, P. The Triune Brain in Evolution: Role in Paleocerebral Functions. New
York: Plenum, 1990.
Mayberry, D., L. Crase, and C. Gullifer. “Categorising Farming Values as Economic,
Conservation, and Lifestyle.” Journal of Economic Psychology 26 (2005): 59-72.
McCown, R. “New Thinking About Farmer Decision Makers.” In The Farmer’s
Decision: Balancing Economic Agriculture Production with Environmental
Quality, ed. J.L. Hatfield, 11-44. Ankeny, IA: Soil and Water Conservation
Society.
National Agricultural Statistics Service. 2002 Census of Agriculture, Volume 1, Chapter
2: Kansas County Level Data.
http://www.nass.usda.gov/census/census02/volume1/ks/st20_2_008_008.pdf
(accessed October 13, 2008).
114
93
National Agricultural Statistics Service. 2002 Census of Agriculture, Volume 1, Chapter
2: Nebraska County Level Data.
http://www.nass.usda.gov/census/census02/volume1/ne/st31_2_008_008.pdf
(accessed October 13, 2008).
Nowack, P. and P. Korsching. “The Human Dimension of Soil and Water Conservation:
A Historical and Methodological Perspective.” In Advances in Soil and Water
Conservation, eds. F. Peirce and W. Frye. Chelsea, MI: Ann Arbor Press.
O‟Brien, P. “Water Quality of the Big Blue River Basin: A Brief Comparison of Local
and State Surface Water Quality.” Nebraska Department of Environmental
Quality, 2008. http://www.deq.state.ne.us/ (accessed May 21, 2008).
Phares, E.J. and N. Erskine. “The Measurement of Selfism.” Educational and
Psychological Measurement, 44 (1984): 597-608.
Sautter, J.A., N. Ovchinnikova, C. Kruse, and G.D. Lynne. “Farmers‟ Decisions
Regarding Carbon Sequestration: A Metaeconomic View.” University of
Nebraska-Lincoln Working Paper (in review with Society and Natural
Resources), 2008.
Schutz, A., and T. Luckman. The Structures of the Life-World, Vol.1. Trans. R.M. Zaner
and H. Engelhardt. Evanston, Il: Northwestern University Press, 1973.
Sen, A.K. On Ethics and Economics. New York: Basil Blackwell, 1987.
--------. “Rational Fools: A Critique of the Behavioral Foundations of Economic Theory.”
Philosophy and Public Affairs, 6 (1977): 317-344.
115
93
Shea, P.J., M. Burbach, G.D. Lynne, A. Martin, and M. Milner. “USDA National
Integrated Water Quality Project Grant Proposal.” 2006.
Sheeder, R. and G.D. Lynne. “Offsetting the Oil Addiction: Farmers Motivated to
Sequester Carbon.” Cornhusker Economics, August 22, 2007.
Smith, A. An Inquiry into the Nature and Causes of the Wealth of Nations. Ed. E.
Cannan. New York: Modern Library, 1937.
--------. The Theory of Moral Sentiments (6 ed.). Eds. D.D. Raphael and A.L. Macife.
Indianapolis, IN: Liberty Classics, 1982. Originally Published in 1759.
Solomon, R.C. True to Our Feelings: What Emotions are Really Telling Us. Oxford, New
York: Oxford University Press, 2007.
Stafford, M. “Water Plentiful in Northeast Kansas, But Delivery at Times is Difficult to
Manage.” The Topeka Capital-Journal, Online Edition. November 2, 2003.
http://www.cjonline.com/stories/110203/kan_waternek.shtml (accessed May 21,
2008).
Supalla, R.J. “Institutional Arrangements to Improve Water Quality in Irrigated
Agriculture.” Water Resources Development 19, 1 (2003): 89-99.
Wallace, B. and F. Clearfield. Stewardship, Spirituality, and Natural Resources
Conservation: A Short History. USDA-NRCS, Social Sciences Institute, ID#
T015, 1997.
Wilson, D.R. “The Evolutionary Neuroscience of Human Reciprocal Sociality: A Basic
Outline for Economists.” Journal of Socio-Economics, 35, 4 (2006): 626-633.
116
93
Wu, J., R.M. Adams, C.L. Kling, and Tanaka, K. “From Microlevel Decisions to
Landscape Changes: An Assessment of Agricultural Conservation Policies.”
American Journal of Agricultural Economics, 86, 1 (2004): 26-41.
117
93
Appendix A
Soil and Chemical Management Survey of Kansas and Nebraska Farmers
118
93
119
93
120
93
121
93
122
93
123
93
124
93
125
93
126
93
127
93
128
93
129
93
130
93
131
93
132
93
133
93
134
93
135
93
136
93
137
93
138
93
139
93
