MANAGING GREAT LAKES WATER AS A COMMON POOL RESOURCE: BARRIERS
TO AND OPPORTUNITIES FOR COLLECTIVE ACTION AMONG IRRIGATORS IN
SOUTHWEST MICHIGAN
By
Rachel N. Ford

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Agricultural, Food, and Resource Economics
2012

ABSTRACT
MANAGING GREAT LAKES WATER AS A COMMON POOL RESOURCE: BARRIERS
TO AND OPPORTUNITIES FOR COLLECTIVE ACTION AMONG IRRIGATORS IN
SOUTHWEST MICHIGAN
By
Rachel Natalie Ford

In response to the Great Lakes-St. Lawrence River Basin Water Resources Compact, which
protects the region against diversions of water outside of the Great Lakes- St. Lawrence River
Basin, the State of Michigan has created a water use program to be implemented by the
Michigan Department of Environmental Quality (DEQ). The program regulates large-quantity
water withdrawals (LQWs) of greater than 100,000 gallons per day to prevent adverse resource
impacts (ARIs), defined as reduction of the index flow in streams and subsequent decline of fish
populations. The rivers and their corresponding watersheds in the state were segmented into over
5,000 catchment areas to be used as management units for the identification and mitigation of
ARIs. An assessment process consisting of an online screening tool and a site-specific review
uses data on aquifer yields, stream index flow, and fish population response to categorize each
new or increased LQW according to the risk of ARI. Should an ARI be detected in a catchment,
all registered and permitted large-quantity water appropriators potentially will be convened into a
water user committee to create a management plan for reducing water use in the area. The
dynamics of these negotiations are unknown. This research explores a case study of agricultural
irrigators in Southwest Michigan for conditions that influence transaction costs of negotiating
and monitoring rules of a water reduction management plan. The identification of these potential
barriers and opportunities for collective action results in suggestions for the DEQ to reduce
transaction costs and improve the efficacy of the water use program in Michigan.

For my mother, Dr. Yvonne L. Ford, Ph.D. RN

iii

ACKNOWLEDGEMENTS
I would like to thank my advisor Dr. Pat Norris for her guidance in the research process
from data collection to outlines to writing and revising; without your help and encouragement I
might have given up long ago. I would also like to thank my other committee members, Dr.
David Schweikhardt and Dr. David Lusch, for your advice and recommendations.
Many people helped in gathering and interpreting data including Howard Reeves of the
U.S. Geological Survey, Jeremiah Asher of the Institute for Water Research at Michigan State
University, Andy LeBaron of the Michigan Department of Environmental Quality, Dr. David
Lusch of Michigan State University, and Lyndon Kelley of Michigan State University Extension.
Thank you, Rebecca Mino, for allowing me to use your research paper as a guide for qualitative
analysis. I would also like to acknowledge and thank all of the agricultural irrigators in
Southwest Michigan who participated in focus groups and individual interviews.
I also need to thank my parents who have supported me in everything I do. Jenny Kim
and Cathy Finnie and all other friends who have helped to relieve stress also deserve thanks.
This research was made possible through the Gordon and Norma Guyer and Gary L.
Seevers Endowment in Natural Resource Conservation. The institutional review board number
associated with this research is IRB# x11-231.

iv

TABLE OF CONTENTS
LIST OF TABLES……………………………………………………………………………... vii
LIST OF FIGURES……………………………………………………………………………. viii
CHAPTER ONE: INTRODUCTION TO THE PROBLEM ………………………………….. 1
1.1 The Great Lakes-St. Lawrence River Basin Water Resources Compact…………... 1
1.2 Michigan’s Water Use Program…………………………………………………….2
1.3 The Water Withdrawal Assessment Process………………………………………..3
1.31 Fish Response Curves and Risk of Adverse Resource Impact……………. 4
1.32 Catchments………………………………………………………………... 6
1.33 Water Rights in Michigan…………………………………………….............. 8
1.4 Great Lakes Water as a Common Pool Resource……….......................................... 9
1.5 The Research............................................................................................................. 11
CHAPTER TWO: COMMON POOL RESOURCE PROBLEMS……………………………. 13
2.1 Models of Appropriator Behavior………………………………………………….. 13
2.11 The Logic of Collective Action……………………………………………. 14
2.12 The Tragedy of the Commons…………………………………………………. 15
2.13 Prisoner’s Dilemma Game ………………………………………………. 15
2.14 The Assurance Problem………………………………………………………… 16
2.2 Institutional Change………………………………………………………………... 18
2.3 Framework for Evaluating the Case Study................................................................ 19
2.4 Discussion of Evaluation Framework……………………………………………… 22
CHAPTER THREE: THE CASE STUDY AREA….................................................................. 32
3.1 General Characteristics of Location…………………………………………….. …32
3.11 Agricultural Economy of the Case Study Area………………………………. 33
3.12 Water Availability……………………………………………………………….. 35
3.2 Characteristics of Agricultural Appropriators……………………………………... 40
3.21 Agricultural Statistics…………………………………………………………… 41
3.22 Qualitative Data...........................................................................................44
3.3 Qualitative Research……………………………………………………………….. 47
3.4 Topics………………………………………………………………………………. 48
3.41 Frustration……………………………………………………………………….. 48
3.42 Distrust……………………………………………………………………………. 50
3.43 Heterogeneity……………………………………………………………………. 52
3.44 Relationships…………………………………………………………………….. 56
3.45 Anti-Diversion……………………………………………………………………. 57
3.46 Summary………………………………………………………………………….. 57
CHAPTER FOUR: ANALYSIS AND RESULTS…………………………………………….. 59
4.1 Definition of the Resource…………………………………………………………. 59
4.2 Analysis……………………………………………………………………………. 60
4.3 Summary of Findings………………………………………………………………. 82
v

CHAPTER FIVE: CONCLUSIONS…………………………………………………………... 85
5.1 Opportunities………………………………………………………………………..85
5.2 Barriers……………………………………………………………………………... 86
5.3 Suggestions for the DEQ…………………………………………………………... 87
5.4 Other Large Quantity Water Users………………………………………………… 89
APPENDICES…………………………………………………………………………………. 91
Appendix A: Recruitment Script………………………………………………………. 92
Appendix B: Focus Group Script………………………………………………………. 94
Appendix C: Individual Interview Script………………………………………………. 100
Appendix D: References to Transcripts for Topics in Chapter Three…………………. 103
REFERENCES……………………………………………………………………………….... 104

vi

LIST OF TABLES
Table 1. Relationship between percentage reduction in stream flow and risk zone…………… 5
Table 2. Prisoner’s Dilemma Payoff Schedule……………………………………………........ 16
Table 3. 2007 Average and median size of farms in acres by county…………………………. 34
Table 4. 2007 Irrigated acreage in case study area…………………………………………….. 34
Table 5. 2007 data for number of farms with principal operators who are full owners, part
owners, and tenants by county…………………………………………………………………. 41
Table 6. 2007 data for number of farms with principal operators whose occupation is
farming or other by county…………………………………………………………………….. 42
Table 7. 2008 data on barriers to making improvements for reduced energy use or water
conservation by number of irrigated farms in Michigan………………………………………. 43
Table 8. 2008 data on number of irrigated farms and acreage in Michigan that received
payments from organizations for irrigation and/or drainage improvements…………………... 44
Table 9. Barriers, opportunities and irrelevant conditions for collective action according
to evaluation framework and data from Branch, Cass, Kalamazoo, and
St. Joseph Counties…………………………………………………………………………….. 83

vii

LIST OF FIGURES
Figure 1. Relationship between reduction in stream flow, impact on thriving
and characteristic fish populations, and subjectively determined risk zones………………….. 5
Figure 2. Number of registered and permitted appropriators in each catchment as of
August 2011……………………………………………………………………………………. 7
Figure 3. Framework for Institutional Change………………………………………………… 19
Figure 4. Evaluation Framework………………………………………………………………. 21
Figure 5. Map of the case study area…………………………………………………………... 32
Figure 6. Michigan’s top commodities, 2009………………………………………………….. 33
Figure 7. Yield from bedrock aquifers…………………………………………………………. 36
Figure 8. Case Study area from bedrock aquifer yield………………………………………… 37
Figure 9. Yield from glacial aquifers…………………………………………………………... 38
Figure 10. Case Study area from glacial aquifer yield map……………………………………. 39
Figure 11. Number of registered or permitted large quantity water users by catchment
in case study area………………………………………………………………………………. 40
Figure 12. Catchments in the case study area………………………………………………….. 61
Figure 13. Comparison of well simulation in A) MODFLOW model with B) WWAT
analytical model………………………………………………………………………………... 63
Figure 14. Frequencies of pumping capacity registered in WWAT since July 9, 2009 in
the counties of Branch, Cass, Kalamazoo, and St. Joseph…………………………………….. 71

viii

Chapter One: Introduction to the Problem

1.1 The Great Lakes – St. Lawrence River Basin Water Resources Compact
Sandra Postel, in Pillar of Sand, details the influence of irrigation on the construction and
demise of civilizations from 4000 B.C. to present. Her writing is the story of irrigation, both the
benefits humans can derive from it and the immense social and environmental destruction it can
cause (Postel 1999). “Water scarcity is now the single biggest threat to global food production”
(p. 6), Postel wrote while describing how costly engineering projects have been used to
transform dry areas into agricultural lands. Similarly in, The Great Lakes Water Wars, Peter
Annin (2006) describes the use of human infrastructure to divert water from areas of abundance
to areas of scarcity. Both works emphasize the danger this introduces as an environmental hazard
and as a source of conflict between regions that share water resources. Like roads, water
engineering projects create new environmental and economic connections between regions that
are otherwise separated. The formation of new groups of water users creates possibilities for
conflict that arises out of the incompatible use of a resource. While Postel took a global survey of
water conflict, Annin focused on the Great Lakes region where 18% of the world’s surface
freshwater resides (Annin 2006).
According to Annin, there has been little limit to human imagination in proposed water
diversion schemes from the Great Lakes to other parts of the continent. In response to these
diversion schemes, a series of regional agreements have been evolving to legitimately protect the
Lakes from external interests. The current, and most conservative document, is the 2005 Great
Lakes-St. Lawrence River Basin Sustainable Water Resource Agreement (Agreement). This
Agreement between the eight U.S. states and two Canadian provinces within the Great Lakes

1

Basin (Basin), committed the parties to “protect, conserve, and restore the Waters of the Great
Lakes” (Council of Great Lakes Governors 2005a: Section 101a). On the American side of the
border is the legally binding Great Lakes- St. Lawrence River Basin Water Resources Compact
(Compact) (Council of Great Lakes Governors 2005b). The goals and standards of the Compact
are to be incorporated into the water resource management program of each state, and, the
progress of each is to be reported to the Council of Great Lakes Governors (Council of Great
Lakes Governors 2005b: Section 4.2.1). Each state (Indiana, Illinois, Minnesota, Wisconsin,
New York, Pennsylvania, Ohio, and Michigan) will prohibit diversions that do not meet stringent
exceptions standards (Council of Great Lakes Governors 2005b: Sections 4.8, 4.9), will monitor
large quantity water withdrawals (LQWs) of more than 100,000 gallons per day within the Basin
(Council of Great Lakes Governors 2005b: Section 4.1), and will implement programs to
promote the efficient use of water (Council of Great Lakes Governors 2005b: Section 4.2).
These water conservation programs provide a prohibition of diversions that is non-discriminatory
and does not give the region an economic advantage (Council of Great Lakes Governors 2001).
The ability to legally refuse a diversion outside of the Basin on the basis of protection,
conservation, and restoration is dependent on parties within the Basin making efforts to protect,
conserve, and restore.

1.2 Michigan’s Water Use Program
The State of Michigan, with only a small piece of the Upper Peninsula lying outside the
Basin, has a great interest in the protection of The Great Lakes for recreation and economic
purposes (Annin 2006); state legislation has supported this interest with the creation of a water
use program to be administered by the Department of Environmental Quality (DEQ). The state’s

2

Natural Resources and Environmental Protection Act states, “The Great Lakes are a binational
public treasure and are held in trust by the Great Lakes states and provinces” (MCL 324.501a).
Public Act 148 (PA 148) in 2003 created the Groundwater Conservation Advisory Council
(Council) to report on sustainable management of the water resources in the state; PA 148 also
mandated the creation of a groundwater inventory and accompanying publicly accessible maps
(GWIM) to inform the Council’s report (State of Michigan GWIM 2005). In 2006, PA 34
reconvened the Council to design an assessment process for authorizing large quantity water
withdrawals in response to the interstate Compact, which would become national law in 2008
(Groundwater Conservation Advisory Council 2007). The Council was charged with creating an
assessment tool by identifying indicators of water resource sustainability and determining how a
potential withdrawal would affect those indicators (Groundwater Conservation Advisory Council
2007). A technical committee was formed to create this assessment process, which consists of
two types of review to authorize requested withdrawals. The first is an online tool called the
Water Withdrawal Assessment Tool (WWAT), which screens withdrawals for potentially nonsustainable pumping rates. If requested, LQWs that are rejected by the WWAT undergo a second,
site-specific review by the DEQ using more precise local data.

1.3 The Water Withdrawal Assessment Process
The assessment process relates surface and groundwater withdrawal to reduction in
stream flow and subsequent ecological impacts. The WWAT was created as a conservative tool
accessed online to evaluate proposed withdrawals of more than 100,000 gallons per day in terms
of their risk of ecological impact (Hamilton and Seelbach 2011). An unacceptable level of
ecological impact is labeled an Adverse Resource Impact (ARI) in the Compact and was defined

3

by fish population decline due to stream flow reduction on February 1, 2009 (MCL 324.32701).
The WWAT has an online interface that uses resource user input on quantity, duration, and
location of the withdrawal as variables in an analytical model (a series of linked equations that
represent a system with interacting variables), and a GIS that relates aquifer characteristics with
stream flow and stream flow with fish population response at each location (Reeves et al. 2009).
The response of fish populations to reductions in stream flow is used to indicate an ARI.

1.31 Fish Response Curves and Risk of Adverse Resource Impact
Streams across the state were classified by the Council into 11 different categories based
on size and temperature. The amount of stream flow reduction from the baseline index flow, the
median flow during the lowest-flow-month of the year (typically August), that will cause an ARI
varies with each stream type due to the stream’s characteristic fish populations and their
resilience to temperature and flow changes (Hamilton and Seelbach 2010). By studying specie
density in streams with varying temperatures and flows across the state, a model for fish
responses to habitat changes due to withdrawal was built (Zorn et al. 2008). Curves relating
stream flow reduction and fish population change were created from this model for each type of
stream. An example of a fish response curve is presented in Figure 1. The “thriving” fish curve
shows how the density of fish populations declines while the “characteristic” fish curve shows
the decline in both number of fish and number of species (Groundwater Conservation Advisory
Council 2007). ARI risk zone thresholds were set by considering ecological impact and societal
needs; a zone D rating is considered an ARI and therefore does not pass the assessment process
(Hamilton and Seelbach 2010).

4

Figure 1. Relationship between reduction in stream flow, impact on thriving and characteristic
fish populations, and subjectively determined risk zones (Source: Hamilton and Seelbach 2010).

Table 1 shows the relationship between percent reduction in stream flow and the risk
zones of each of the eleven classifications of streams.

Cold……………....
Cold-Transitional...
Cool……………...
Warm…………….

A-B
14%
…
6
10

Stream
B-C
14%
4
15
18

ARI
20%
4
25
24

Small river
A-B
B-C
ARI
10.5% 10.5%
21%
…
2
2
15
19
25
8
13
17

Large River
A-B
B-C
ARI
…
…
…
…
3%
3%
14%
19
25
10
16
22

Table 1. Relationship between percentage reduction in stream flow and risk zone (Source:
Hamilton and Seelbach 2010).

Each zone in the assessment process corresponds to a regulatory decision; LQWs
evaluated to be in zones A and B contain little risk for causing an ARI and are registered without
a site-specific review. A withdrawal in a cold-transitional stream of any size that is evaluated to
cause a zone B impact may not be registered without a site-specific review (MCL 324.32706b).
Cold-transitional LQWs in zone B and withdrawals in zone C are registered through a site5

specific review only if corresponding conservation measures are implemented (MCL
324.32706c), but withdrawals evaluated to cause a zone D impact are highly likely to cause an
ARI and are not allowed (MCL 324.32706b).

1.32 Catchments
For the evaluation process, the hydrologic system in the state was broken into catchment
areas, which are stream segments and their drainage areas. These segments were created by
identifying 10,000 unique units by hydrologic and ecologic characteristics (Reeves et al. 2009).
Absorbing the smallest segments into neighboring segments with similar properties and
incorporating associated drainage areas resulted in approximately 5,000 catchment areas as
management units to which risk zones are assigned for the evaluation of individual and
cumulative impacts of proposed LQWs (MCL 324.32706a). Catchments may have multiple users
(appropriators) with different purposes for their withdrawals. Figure 2 shows the number of
1

registered and permitted appropriators per catchment across the state as of August 2011.

1

Appropriators withdrawing more than 100,000 gpd must register their withdrawals;
appropriators withdrawing more than 2,000,000 gpd or withdrawing more than 1,000,000 gpd for
a municipal system that is evaluated in zone C must obtain a permit (MCL 324.32723).
6

Figure 2. Number of registered and permitted appropriators in each catchment as of August 2011.
(Data on appropriators obtained from Andrew LeBaron, DEQ; catchment delineation based on
version 8 obtained from Jeremiah Asher, Institute of Water Research, Michigan State University).
For interpretation of the references to color in this and all other figures, the reader is referred to
the electronic version of this thesis.

Abutting catchments have aquifer continuity so that withdrawals in one catchment can
affect streams in neighboring catchments. An inverse distance equation is used to evaluate the
impact on neighboring streams (Reeves et al. 2009, pp 8-9):

7

f

1
d
i
i
j

1
1, n d j

(1)

Where
fi

is the fraction of the captured water attributed to catchment i,

di

is the linear distance between the proposed well and the closest stream in
catchment i,

n

is the number of adjacent catchments

and

This weighting equation allows for one withdrawal to negatively impact multiple
catchment areas and multiple appropriators.

1.33 Water Rights in Michigan
In effect, the legislation behind the water use program moves Michigan from a state
operating under the Riparian Doctrine to one operating under regulated riparianism (Dellapenna
1994). The Riparian Doctrine entitle landowners to the reasonable use of surface water within or
abutting his or her land up to an amount where that use begins to interfere with other
appropriators’ rights to reasonable water use (Lusch 2011). Many eastern states have recognized
the impact on ecosystems that multiple uses of surface and groundwater cause and in response
have created environmental standards to be met; this essentially recognizes public interest in
environmental integrity and protects that interest from appropriator overextraction (Dellapenna
1994). Because landowners are legally entitled to the reasonable use of water but ARI
8

occurrences are unlawful, an appropriator proposing a new or increased LQW that is rejected has
standing to negotiate with those who are affecting that same catchment for their reduction. The
appropriator can appeal to others to reduce their water consumption to allow for at least some of
his or her proposed LQW. If negotiation does not lead to satisfactory results for all parties,
disputes can be settled with litigation.
An ARI can be identified in two ways: the first is with the water withdrawal assessment
process, which utilizes the WWAT or a site-specific review. The second is if withdrawals have
caused stream flows to be noticeably lower than usual in the field. If the potential for an ARI is
found in the assessment process, the responsibility for preventing the ARI lies with the proposed
appropriator. He or she can reduce the proposed withdrawal, abandon the withdrawal, or
negotiate with other appropriators to reduce their water use. If detected in the field, the ARI is
the responsibility of all registered appropriators in the same catchment. The DEQ will convene
all appropriators contributing to the ARI into a water user committee to create a plan for
reduction of aggregate withdrawals (MCL 324.32725). If unregistered appropriators are
2

identified as contributing as well, then they are held responsible for reducing the impact.

1.4 Great Lakes Water as a Common Pool Resource
The situation where a group of appropriators is held responsible for an ARI is a common
pool resource (CPR) situation. A CPR situation is one where multiple users appropriate benefits
from a shared resource; these benefits are rival, meaning that the marginal cost of use is positive
and even increasing (Oakerson 1992). These costs can be time costs, monetary costs, or energy
2

All registered and permitted LQWs have a rebuttable presumption that they are not causing an
ARI (MCL 324.32722). However, in the event that an ARI is detected, the DEQ will convene a
meeting of all holders of registered and permitted LQWs presumed to contribute to the ARI and
hold them collectively responsible for mitigation (MCL 324.32725(3)).
9

costs. Appropriation of units by one user imposes a cost on other users when those units are no
longer available for others’ use and it takes more time, effort, and money to capture remaining
benefits. A CPR dilemma occurs when the collective outcome of appropriation could be
improved with alternative institutions (Gardner et al. 1990). Institutions are the formal and
informal rules that articulate relationships between people; they define a set of choices available
to economic agents via constraints and opportunities on behavior that affect incentives (Schmid
2004). Alternative institutions produce alternative choices, which result in alternative outcomes.
The first writings on CPRs assumed that the situation of multiple resource users would
always lead to the same outcome: destruction of the resource. Researchers believed that
privatization or state interference were the best options for preventing resource overexploitation.
More careful examinations of CPR situations have resulted in an extensive body of research
contributing to a theory of successful collective action. The goal in this field of research is to
identify variables leading to enduring CPR institutions that promote sustainable use of the
resource while reducing conflict between appropriators (Bromley 1992). The identification of
these variables will lead to the prediction of successful CPR management situations and will help
to create institutions that mitigate the externalities produced by CPR dilemmas (Ostrom 1990).
These institutions are successful because they lower the costs of transactions between
CPR users. All transactions between economic actors have associated costs, which are defined as
“the value of opportunities given up when any action is undertaken” where the value is
determined by each economic actor (Schmid 2004, p. 71). CPR situations are successful when
information, negotiating, monitoring, and enforcement costs are limited by characteristics of the
resource, the resource users, the institutional arrangements, and the environment external to the
CPR situation (Agrawal 2001).

10

1.5 The Research
Whether water user committees convened by the DEQ will yield effective institutions for
the collective use of water is unknown. Existing and emerging institutions affecting water use in
Michigan are likely to have a significant impact on the magnitude of transaction costs and thus
the opportunities for a successful collective management strategy. The goal of this research is to
identify characteristics that will likely lead to high transaction costs causing barriers to collective
action and characteristics that will likely lead to low transaction costs creating opportunities for
collective action in the context of Michigan’s water use program. A sample of agricultural
irrigators in four counties in Southwest Michigan will be used as a case study. The question to be
answered is: What barriers and opportunities, due to transaction costs, exist among irrigators in
Southwest Michigan that will hinder or help the goals of Michigan’s water use program?

The specific objectives of this research are to:
1. Review existing literature for the purpose of generating a framework to use for evaluating
the institutional environment and transaction costs in the case study group.
2. Explore and describe institutions that exist among irrigators in the case study.
3. Determine if these institutions create high transaction costs that result in barriers or will
lower transaction costs that result in opportunities for realizing the goals of Michigan’s
water use program.
4. To draw conclusions based on the findings about what changes might reduce transaction
costs in areas where they are high to benefit the program.

11

To meet these objectives, a case study of agricultural irrigators from four counties in
Southwest Michigan was identified as representative of agricultural water users in Michigan.
Primary data from focus groups conducted in the spring and interviews in the fall of 2011 is
supplemented with secondary data from technical reports and censuses. These four counties
contain some of the most intensely irrigated areas in the state. Agricultural irrigators are prolific
in Southwest Michigan and their particular agricultural systems cause them to use a high amount
of water relative to other parts of the state. This creates a potential for conflict over water use and
makes for an appropriate region for this analysis.
The next chapter contains a review of literature on CPR use and identifies a framework with
which to evaluate transaction costs with the case study data. Chapter Three is a description of
case study characteristics: the location, the resource, and the irrigators. The methods for
conducting focus groups and completing qualitative analysis are also presented in this chapter.
Chapter Four is an application of the framework generated in Chapter Two to describe the
institutions and their transaction costs that affect water use decisions in the case study area. The
fifth chapter draws conclusions from the analysis and identifies challenges for and possible
reduction of transaction costs to improve Michigan’s water use program.

12

Chapter Two: Common Pool Resource Problems

A common pool resource has a high cost of exclusion and is subtractable (rival), (Ostrom
and Gardner 1993). A high exclusion cost good is one where the benefits of the good are costly
to exclude others from (Schmid 2004). In a resource situation, this results in multiple users
extracting the good, leading to an increasing cost of appropriation and potential destruction of
the resource (Hardin 1968). A resource that is subtractable, or rival, has the characteristic that
any unit appropriated from the resource reduces the amount available to other users by one unit
(Oakerson 1992). The result of a high exclusion cost good that is subtractable is a large number
of users who have incentive to capture present benefits at the expense of future availability to
ensure that they benefit from the resource before it is destroyed by the group. These
appropriators have a high discount rate and no incentive to restrain their use (Ostrom 1990).
When society as a whole values the benefits from the resource with a lower discount rate than
the appropriators, then new institutions are built to force appropriators to lower extraction rates
and match societal values (Runge 1981). Hardin (1968) argued that privatization or
nationalization of a common pool resource will protect it from overexploitation. The evolution of
research on CPR institutions has gone from identification of incentives for appropriators to
overexploit to identification of characteristics of the CPR situation and institutions surrounding it
that have resulted in lasting use of the resource.

2.1 Models of Appropriator Behavior
Three related models have been used to argue that the high exclusion costs and
subtractability of a CPR will result in overexploitation. These models of appropriator behavior

13

are Olson’s The Logic of Collective Action (1971), Hardin’s “Tragedy of the Commons” (1968),
and, from game theory, the Prisoner’s Dilemma game.

2.11 The Logic of Collective Action
High exclusion costs allow users to benefit from a good without paying for it (free riding)
when the cost of the good is born by other users in the group. In The Logic of Collective Action,
Olson (1971) describes a situation where each additional user of a good creates no cost to a
group of users in terms of units lost; however, the model contains an important discussion of
behavior that, in a CPR situation, can lead to suboptimal outcomes. Olson makes the important
point that free rider behavior will reduce contributions to pay for the provision of the good
because free riders will not contribute if costs are larger than the benefits provided by their
marginal contributions. Those members will benefit from the good without bearing any of the
costs. While this can reduce the amount of the good available, in some cases to zero, in some
groups, free riding behavior will not affect the provision of the good. This is due to the actions of
a privileged group. These are members of the group who benefit so much from the existence of
the public good that, regardless of free riding activity, they pay to provide the entire good. In a
resource use situation, free riding activity occurs when one group of users restrains use to lower
extraction costs or for sustainability and another group of users do not. Those who do not restrain
their use benefit from the low extraction costs and future availability without giving up any
present gain.

14

2.12 The Tragedy of the Commons
In the “tragedy of the commons” (Hardin 1968), similar free riding behavior causes the
destruction of a CPR. Hardin’s argument begins with a field open to multiple herders who
choose how many animals to graze. Each unit of resource extracted results in a cost to the group
of lost future benefits; however, each herder gains the entire benefit of his or her obtained unit
while sharing the cost with the group of users. The marginal cost of grazing is lower than the
marginal benefit and every herder continues to add animals until the entire common plot is
destroyed. Any decision by one herder to constrain use of the resource offers opportunity for
other herders to free-ride on that herder’s sacrifice and gain for themselves. The benefits are
transferred from one herder to others. Hardin points out that an individual has no incentive to
constrain his or her use of the resource because the resource will be destroyed by the group
regardless of the individual’s actions.

2.13 Prisoner’s Dilemma Game
The decision of a resource user to constrain or exploit a resource has been modeled with
the Prisoner’s Dilemma (PD) game (Runge 1984). The PD game examines how rational actors,
faced with a certain set of payoffs, can make seemingly irrational decisions (Mueller 2003). In
the game, two prisoners are held for a crime that had no witnesses. Each player’s choice will
affect the other, but each makes a decision independent of the other. If both stay silent, they both
go to prison for one year. If one confesses and the other stays silent, the confessor is rewarded by
going free and the other is punished with five years in prison. If both confess, both are given a
two year sentence. The payoff schedule (adapted from Nicholson and Snyder 2008) is presented
in Table 2.

15

Player Two Confess

Player Two Silent

Player One Confess

(-2, -2)

(0, -5)

Player One Silent

(-5, 0)

(-1, -1)

Table 2. Prisoner’s Dilemma Payoff Schedule.

The players can maximize their joint utility by both staying silent; however, the
individual best choice is to confess because neither can know how the other player will act. If
Player One confesses, Player Two can maximize her benefits by also confessing. Faced with the
choice of a two-year sentence or a five-year sentence, she will choose a two-year sentence. If
Player One stays silent, then Player Two’s choice is between a one-year sentence by staying
silent and going free by confessing. Thus, the dominant strategy for both players is to confess.
This results in both going to prison for two years rather than the one year that would have
occurred if they both stayed silent.
In a CPR situation, an appropriator can choose to constrain his use for future
consumption but is not assured that other users of the resource will do so. If other appropriators
do not restrain their use, then the user can lose in two ways: he gives up benefits in the present
and he may lose benefits in the future if the collective extraction rate is unsustainable. If he and
others choose not to constrain their extraction to a sustainable rate, they gain present benefits at
the sacrifice of future benefits.

2.14 The Assurance Problem
These models have been persuasive in the argument to either privatize or nationalize
CPRs; however, in reality there have been many examples of appropriator organizations that are
enduring and effective in altering incentives to exploit the CPR (Ciriacy-Wantrup and Bishop
16

1975). Hardin’s (1968) tragedy is a description of a resource with no property rights associated
with it. In contrast, a CPR is a resource that belongs to a group of appropriators separate from a
general public (Bromley 1992). Like individual property rights, communal property rights can
decrease the discount rate on resource benefits for owners in comparison to open-access
situations (Ostrom 1990). A group of resource owners has incentive to coordinate management
strategies to protect future benefits from unsustainable extraction in the present and do so with
institutions that provide assurance that the group of resource users will conform to sustainable
appropriation rates (Runge 1981).
Bromley (1992) warns against assumptions that the PD game is an accurate portrayal of
resource users in any CPR situation. Rather, the external institutional settings of the game found
in the payoff schedule influence the dominant strategy of confession. By changing the payoff
schedule, the strategies of the players also change. In the PD game, each player prefers for the
other to stay silent while she confesses. In the Assurance Problem (Runge 1981), the payoff is
structured so that each player prefers the same action, with both either constraining use or both
overextracting. The appeal of fairness is apparent in experimental, repeated PD games: players
often adopt a tit-for-tat strategy where they choose silence round after round unless the other
player confesses; then they begin to confess as well (Runge 1984).
The Assurance Problem offers an alternative model to the PD game for appropriator
behavior. Other research has discounted the reliability of the PD game as a generalized CPR
appropriator behavior model. For example, Ostrom and Walker (1989) found that
communication improves coordination among CPR users, an element that is missing from the PD
game. Social capital, defined as sympathy toward another person, can alter payoff schedules
when a player internalizes the well-being of the other player (Robison et al. 2002). The outcome

17

of repeated interactions between players differs from that of a one-shot game (Wade 1988).
These factors all support the argument that externally determined payoffs and rules affect player
strategy. A CPR situation, therefore, does not inherently lead to overexploitation.

2.2 Institutional Change
In the case of a CPR dilemma, then, the institutions that influence appropriator behavior
can be manipulated to create a Pareto-Improving move. Institutions exist on three levels that are
changed via feedback loops (Kiser and Ostrom 2000). The analysis of institutional arrangements
begins at the individual or operational choice level pictured in Figure 3. Individuals make
decisions based on rules set at collective and/or constitutional levels, the outcome of which
influences changes in rules at those levels. In the case of an inferior outcome, individuals create
new rules at the higher two levels within the constitutional or collective environment created by
higher institutional arrangements. Institutional change will be effective only if individuals are
aware of the change, they adopt new strategies because of the change, and the collection of new
strategies results in a new outcome (Kiser and Ostrom 2000).

18

Figure 3. Framework for Institutional Change (Source: adapted from Sabatier 1991).

2.3 Framework for Evaluating the Case Study
Successful institutional arrangements will differ in each CPR situation, which consists of
characteristics of the resource, characteristics of the resource users, and the institutional
arrangements both surrounding use of the CPR and external to it (Agrawal 2001). Three

19

influential works that identify common traits among enduring CPR institutions were compiled by
Agrawal (2001) to create a comprehensive list of conditions leading to successful collective
action. Agrawal notes two major weaknesses in the three works by Ostrom (1990), Wade (1988),
and Baland and Platteau (1996). First, all three works focus on the collective institutional
arrangements and neglect the important external characteristics of resource systems,
appropriators, and constitutional arrangements. Second, the conditions are neither empirically
tested nor generated from independent studies. Wade (1988) draws from an earlier work of
Ostrom (Ostrom 1985), and Baland and Platteau (1996) cite both Wade (1988) and Ostrom
(1990). The interaction of one condition with another or the relative importance of one condition
over another is unknown. Agrawal (2001) fills in his identified gaps with other literature but
saves empirical testing for future research. The investigation into successful CPR institutions did
not contain any examples of regulated riparianism situations; however, the framework presents
general analysis of CPR situations that will be helpful in evaluating the case study.
Each of the conditions presented by Agrawal leads to either lower transaction costs in the
management of a CPR or an incentive to bear those transaction costs when the costs to the
resource users of participating in collective action are lower than the benefits. For this research,
the characteristics of the resource system, group, and environment external to the CPR rules will
be examined to comment on the incentives to bear transaction costs and the magnitude of those
transaction costs. However, because no formal institutional arrangements exist in the case study
area governing the use of water resources, Agrawal’s section on institutional arrangements is
omitted from the framework. Agrawal also includes interdependence in his list of conditions, but
it is also omitted from the framework here because interdependence in the use of a resource is
inherent in the characteristic of subtractability. Also, leadership, where the actions of some

20

resource users influence others, and social capital, where utility functions are interdependent
capture this condition. One condition is added to Agrawal’s list of group characteristics; an
important condition that can reduce transaction costs in the emergence of appropriator
organizations is that resource users have experience as members of other organizations whether
they are connected to use of the resource or not (Ostrom 1992). Another condition that supports
the emergence of appropriator organizations is added to Agrawal’s list of resource characteristics.
This is the existence of indicators of CPR conditions (Ostrom 1992). The evaluation framework
is presented in Figure 4.

1) Resource System Characteristics
a. Small size
b. Well-defined boundaries
c. Low levels of mobility
d. Possibilities of storage
e. Predictability
f. Indicators of CPR conditions
2) Group Characteristics
a. Small size
b. Clear boundaries
c. Shared norms
d. Social capital
e. Appropriate leadership
f. Heterogeneity of endowments
g. Homogeneity of interests and identities
h. Low levels of poverty
i. Past experience with organizations
1 and 2) Relationships between resource system characteristics and group characteristics
a. Overlap between user residential location and resource location
b. High levels of dependence on resource
c. Fairness in allocation of benefits
d. Low levels of user demand
e. Gradual change in levels of demand
Figure 4. Evaluation Framework (Source: Adapted from Agrawal (2001) and Ostrom (1992)).

21

3)

External Environment
a. Technology
i. Low-cost exclusion technology
ii. Time for adaptation to new technologies related to the commons
b. Low levels of articulation with external markets
c. Gradual change in articulation with external markets
d. State
i. Power of local authority
ii. Supportive sanctioning institutions
iii. External aid for conservation activities
iv. Nested levels of appropriation, provision, enforcement,
governance

Figure 4, Cont. Evaluation Framework (Source: Adapted from Agrawal (2001) and Ostrom
(1992)).

2.4 Discussion of Evaluation Framework
This framework lists conditions that affect the emergence and endurance of CPR
institutions that last over time and reduce conflict between users (Bromley 1992). There are two
types of opportunities for collective action found in the presence of these conditions. First, when
the benefit to resource users of participating in collective action is greater than the cost, they will
be more likely to bear the costs of coordination. Second, certain characteristics of the resource,
users, and institutions surrounding a CPR lead to lower transaction costs in creating and
implementing management plans. The costs are associated with information (uncertainty),
negotiating, and monitoring. In this section, each condition is discussed in preparation for the
analysis presented in the fourth chapter.

Resource system Characteristics
Small size: Monitoring user behavior and the condition of the resource will allow appropriators
complete information on compliance and resource health. Smaller resource areas are less costly

22

to monitor than larger ones because the cost of collecting information about the resource is lower
(Baland and Platteau 1996).

Well-defined boundaries: With well-defined boundaries, it is clear where the management area
begins and ends (Ostrom 1990), thereby facilitating exclusion and monitoring. The information
about the effects of resource user decisions on the resource is less costly to obtain when
boundaries are definite because the impact is contained in one area and can therefore be
measured.

Low levels of mobility, possibilities of storage: Resource mobility creates fluid boundaries and
large resource areas increasing the cost of information and monitoring because of both larger
areas and more resource users. If a resource can be stored, it will be made stationary, reducing
costs associated with risk and uncertainty (Schlager et al. 1994). When uncertainty is low,
resource users have greater assurance that they will have access to future benefits and therefore
less incentive to overexploit in the present.

Predictability, indicators of CPR conditions: Naughton-Treves and Sanderson (1995) discuss
the costs of managing resources with unknown levels of availability especially when the effects
of human development and climate change on availability are uncertain. This uncertainty reduces
assurance of future benefits and thus increases the discount rate of appropriators. When the
conditions of the CPR are unknown or unpredictable, extraction rates for sustainable use may be
unknown or fluctuating. To set appropriate rules for sustainable extraction rates, then, resource

23

users will have to bear time and monetary costs to understand the characteristics of the resource
and its response to varied extraction rates (Ostrom 1992).

Group Characteristics
Small Size: Each member of a resource user group has their own set of interests and ideas when
negotiating water management plans. The more interests represented in the group, the higher the
negotiating costs. Also, the larger the group of resource users, the higher the cost of monitoring
necessary to ensure compliance (Olson 1971).

Clear Boundaries: A clear identification of who is using the CPR and thus whose actions affect
the sustainability of the resource will result in more complete understanding of the status of the
resource to increase assurance. It is less costly to accumulate information about users for creating
a management plan and to then monitor those users. An unidentified user not participating in the
creation of or compliance with rules about CPR use could result in overextraction despite others’
restraint in use (Ostrom 1990).

Shared Norms: Social norms can create assurance in resource use. When concern for social
propriety is high, group members are likely to align their choices with the majority and expect
others to do so as well (Wade 1988). If norms regarding resource use do exist, then those norms
can act to enforce agreements about restraint when users find avoiding the costs of breaking
those norms an incentive to organize.

24

Social Capital: Social capital, defined as sympathy, causes utility functions of individuals to be
interdependent; one person’s gain provides utility for another (Robison and Ritchie 2010). In
relationships characterized by social capital, incentives to break rules regarding resource use
decrease when such actions negatively affect other users. Each resource user has the interests of
others in mind when negotiating reduction plans, resulting in a more cooperative atmosphere that
reduces costs associated with negotiation. These users will also have incentive to follow rules
created through the negotiations because they will not want to cheat their neighbors. Monitoring
is less costly when users are less likely to cheat on incentives.

Appropriate Leadership: A leader who can combine aspects of traditional management
strategies with new information on conservation, resource conditions, and threats to future
benefits can decrease the costs of negotiation and information dissemination (Baland and
Platteau 1996). These leaders will help in the propagation of new information regarding
sustainable use of the resource and can influence the decisions of other resource users thereby
reducing negotiating costs (Taylor 1987).

Heterogeneity of endowments: Appropriators with larger endowments of the resource may
have a greater incentive to promote sustainable use. Olson’s (1971) idea of the privileged group
arises when a small sub-group benefits enough from a good that they will provide the good for
an entire group. In resource use, this group would use so much of the resource that their
constraint alone, regardless of any other users’ appropriation rate, would lead to sustained
benefits over time. The privileged group has incentive to create CPR institutions among

25

themselves because they gain from future benefits provided by collective restraint (Baland and
Platteau 1996).

Homogeneity of interests and identities: Cultural differences affect users’ views about
management, conventions, and norms and, as a result, affect coordination (Runge 1992 citing
Sugden 1982). While some groups such as traditional fisheries in Japan have long- standing rules
about resource conservation, others, such as large corporate fisheries in Japan exploit it at the
rate their technology allows (Baland and Platteau 1999). The divergence of strategies caused by
differences in interest and identities creates barriers to effective collective decisions when one
strategy is in direct contrast to another. The costs of negotiating rules regarding extraction rates
are lower when interests and identities are homogeneous.

Low levels of poverty: There is no clear consensus on how poverty affects the use of a common
pool resource (Agrawal 2001). If a poor community is highly reliant on a resource, the pressure
of survival may increase discount rates until appropriation rates are higher than sustainable
(Baland and Platteau 1999). There may be little incentive to reduce rates of harvest when doing
so will result in starvation for the appropriators. For these appropriators, there is a high cost to
reducing extraction rates. They will be less likely to bear the costs of collective action, to agree
to lower extraction rates in negotiation, and to comply with those agreements. The costs of
organizing the group and negotiating effective rules for conservation are high and the cost of
ensuring compliance is also high as appropriators gain more benefit from cheating on rules than
they do from complying with them.

26

Past experience with organizations: Communities who have experience with organizations will
have lower costs in creating new organizations because they will have structures of organization
to guide them (Ostrom 1992).

Relationship between resource system characteristics and group characteristics
Overlap between user-group residential location and resource location: When appropriators
live in the same area as the CPR, they are more likely to have similar interests, be available to
work out conflicts, and have a cooperative attitude with their neighbors (Baland and Platteau
1996). The appropriators are thus more likely to trust their neighbors, creating assurance, which
can result in low transaction and monitoring costs (Robison and Ritchie 2010). Resource users
who have similar interests will be more likely to be cooperative during a water user committee
than those who do not.

High levels of dependence by group members on resource system: The conservation of a
CPR can act as risk mitigation when other income opportunities do not exist (Baland and
Platteau 1999, Wade 1988). Group members who are highly reliant on their resource, then, have
incentive to bear the costs of collective action in order to ensure their own well-being.

Fairness in allocation of benefits from common resources: When resource users have
confidence that their allocations are fair, they have less incentive to cheat on agreements about
appropriation (McKean 1992). This reduces monitoring costs by way of assurance and makes
negotiation more likely than in a situation where resource users feel allocations are unfair.

27

Low levels of user demand: When appropriator demand for a resource is high, extraction rates
to meet that demand are also high. Users will be less willing to reduce their water use because
the benefit of captured resource units is higher than the perceived cost of collective action
(Agrawal 2001).

Gradual change in levels of demand: Population and market pressures influence rates of
resource use. If these pressures cause demand to increase quickly, appropriators might respond
by increasing their extraction rates to an unsustainable level. This increase in demand creates an
incentive to cheat on rules agreed upon for collective restraint. Monitoring costs increase. Users
will also be less willing to negotiate their extraction rates because the opportunity cost of
foregoing units of resource increases (Agrawal 2001).

External environment
Technology
Low-cost exclusion technology: Technology that lowers the cost to exclude others from resource
use will result in well-defined boundaries for the resource user group. This increases the efficacy
of collective action by identifying all those users to be involved in monitoring and reducing the
3

cost of information about CPR conditions (Wade 1988).

3

In Wade’s 1988 edition of Village Republics, the condition was listed as: “the higher the costs
of exclusion technology (such as fencing) the better the chances of success (pg 215)”. Baland
and Platteau (1996) cite this same condition (pg 287). I believe this is a typing mistake because
when exclusion technology is expensive, it will be more difficult to exclude others resulting in
Hardin’s tragedy. Agrawal (2001) changes the word “higher” to “lower” without explanation.
28

Time for adaptation to new technologies related to the commons: New technologies can reduce
the cost-benefit ratio of extraction, resulting in increased extraction rates (Agrawal 2001).
Increased extraction rates occur because users find they can appropriate more benefits from the
resource at a lower cost. Thus, reducing those rates imposes higher opportunity costs on those
users causing them to be less willing to bear the costs of negotiating on collective reduction of
appropriation. A slower adaptation to new technologies allows the appropriators an opportunity
to negotiate new rules about resource use before permanent damage to the CPR occurs.

Low levels of articulation with external markets: Markets determine prices for outputs
produced through the use of resources, some of which are CPRs. When appropriator demand for
benefits from a CPR is tied closely to market price for output, an increase in price will increase
appropriator demand, leading to higher extraction rates. Appropriators have greater incentive to
continue extracting large amounts of the CPR because of the benefits they obtain from the
resource. They may be unwilling to bear the costs of collective action and then to agree to a
reduction in extraction rates.

Gradual changes in articulation with external markets: A rapid increase in articulation with
markets would be paired with a rapid increase in demand that users of a CPR might not have
anticipated. As a result, extraction rates can become too high for sustained appropriation over
time and users may be less willing to negotiate rules to lower those extraction rates because the
value of resource units are higher. The opportunity cost of constraint is higher causing less
willingness to bear the costs of coordination and higher costs in that coordination.

29

State
Power of local authority: When government authorities create new rules that weaken local
management regimes rather than give recognition to them, the CPR may be overexploited
(Ostrom 1990). Rules that are created by the local community of users have often been
successful in restraining appropriation rates because communities have more information about
the trade-offs users face in lowering extraction rates and are therefore able to create rules
appropriate for the situation. When central government rules are enforced, users may be less
willing to follow them, increasing monitoring costs (Wade 1988).

Supportive sanctioning institutions: The lack of penalties for non-compliance can result in a
failure to enforce CPR rules on appropriation rates, unless all appropriators share the same rate
of time preference (Baland and Platteau 1996). Ostrom (1990) and Wade (1988) both concluded
that sanctioning systems at the collective level can be effective, but Baland and Platteau (1996)
concluded that monitors may find high costs of reporting when facing antagonism from their
peers. Monitors from external sanctioning institutions may not experience those costs and may
therefore be more effective.

External aid to compensate local users for conservation activities: Especially in areas with high
levels of poverty and low opportunities for alternative income-generating opportunities,
foregoing use of a resource results in a loss of essential income. This illustrates a case where the
conservation goals set at the Constitutional Choice Level are at odds with the goals at the
Operational Choice Level (refer to Figure 3). Successful management strategies may require user
compensation for the income loss from restrained appropriation (Baland and Platteau 1996). This

30

would decrease the opportunity cost of restrained extraction and thus decrease the cost of
collective action.

Nested levels of appropriation, provision, enforcement, governance: CPR appropriation,
provision, enforcement, and governance rules at a local level are a sub-set of rules at higher
levels (Ostrom 1990). The articulation with rules and administrative resources at higher levels
can help CPR users to incorporate outside information and technologies into their management
strategies resulting in low negotiating, information, and monitoring costs (Ostrom 1992).

This list offers a framework for identifying high and low transaction costs that create
barriers and opportunities faced by the Michigan Department of Environmental Quality in
establishing an effective water use program.

31

Chapter Three: The Case Study Area

To apply the evaluation framework in Figure 4, data were collected in the case study area
using technical and statistical reports as well as focus groups and interviews.

3.1 General Characteristics of Location
Figure 5 is a map showing the case study location in Michigan’s Lower Peninsula.

Figure 5. Map of the case study area (shown in pink).

32

3.11 Agricultural Economy of the Case Study Area
The importance of agriculture to Michigan’s economy, the profitability of the crops
grown in the case study area and the intensity of irrigation used in growing these crops suggest
that water is a highly valuable input that contributes to both the local economy and the greater
state economy. In the state of Michigan, farming, industries related to farming, and industries
related to food production had an estimated value of $71.3 billion in 2007 (Knudsen and
Peterson 2009). The highest value crops from 2009 are shown in Figure 6.

Figure 6. Michigan’s top commodities (Source: Kleweno 2010).

The counties of Branch, Cass, Kalamazoo, and St. Joseph are all large contributors to this
agricultural economy. Corn, soy, and wheat are grown on more than 50% of the cropland in each
county, and all rank in the list of top 20 counties for the production of those crops. Also, potatoes
are important locally and are a crop that requires intense irrigation (Michigan Department of
Agriculture and Rural Development 2009).

33

Table 3 shows average farm size in acres for Michigan and each of the case study
4

counties. The county averages are about the same as the state average.

Michigan

Branch
County

Cass County

Kalamazoo
County

St. Joseph
County

179

222

235

170

209

60

70

63

40

60

Average size
of farm
(acres)
Median size
of farm
(acres)

Table 3. 2007 Average and median size of farms in acres by county (Source: 2007 Census of
Agriculture, Michigan: Volume 1, Chapter 2, Table 1).

All four counties are ranked among the top five counties in Michigan for irrigated
acreage. As shown in Table 4, St. Joseph County, ranked number one, has more than twice the
number of irrigated acres as Branch County, which is ranked number two.

Michigan Total Acreage in Irrigation: 500,428 acres
Statewide County Mean Irrigated Acreage: 6,030.40 acres
County/ State Rank
#1 St. Joseph
County
#2 Branch County
#4 Cass County
#5 Kalamazoo
County

Acres in Agriculture

Irrigated Acres
(percentage)

STDEV from
Statewide County
Mean

181,051

102,859 (56.81%)

6.86

202,382
151,852

42,923 (21.21%)
38,985 (25.67%)

2.61
2.33

115,979

31,314 (27.00%)

1.79

Table 4. 2007 Irrigated acreage in case study area (Source: USDA Census of Agriculture 2007,
Michigan: Volume 1, Chapter 2, Table 10).

4

Census data on farms reported for farms with sales of $1000 or more annually
34

3.12 Water Availability
The availability of water in Southwest Michigan is a contributing factor to the amount of
irrigation in the area. To better understand the water resources across the state, the DEQ was
given the task of creating a groundwater inventory and series of maps with passage of PA 148 of
2003 (State of Michigan GWIM 2006). The resulting maps showed aquifer yields in both
bedrock and glacial deposits and estimated the differences in aquifer characteristics throughout
the depth of deposits. In the following maps, yield is defined by the pumping rate that will
reduce standing water levels in wells by 50% if they are pumped continuously for 100
consecutive days. This measure incorporates how much water is stored in the aquifer, how well
the aquifer material transmits water, and how deep the aquifer layer is (State of Michigan GWIM
2006). Figure 7 is a map of bedrock aquifer yields.

35

Figure 7. Yield from bedrock aquifers (Source: State of Michigan GWIM 2006).

Figure 8 is the map of bedrock aquifer yield for the counties in the case study. With the
exception of a very small portion of both Kalamazoo and Branch Counties, there are no
underlying bedrock aquifers that would serve as a potential source of water in the case study area.
36

Figure 8. Case Study area from bedrock aquifer yield map (Source: State of Michigan GWIM
2006).

Figures 9 and 10 are maps of glacial aquifer yield. Figure 9 shows the entire state and
Figure 10 shows the case study counties. These maps show that the counties have an abundance
of water relative to much of the state but also some areas of thin aquifer material adding to
heterogeneity in water availability in the region.

37

Figure 9. Yield from glacial aquifers (Source: State of Michigan GWIM 2006).

38

Figure 10. Case Study area from glacial aquifer yield map (Source: State of Michigan GWIM
2006).

The area, then, has a natural abundance of water in glacial aquifers allowing for
production of high value, irrigated crops. For the 2007 Census of Agriculture, Michigan farmers
on 577 farms were questioned about reasons for discontinuing irrigation. Only one farm had a
surface water shortage and only one farm had a groundwater shortage. It is unclear whether they
are the same farm (2008 Farm and Ranch Irrigation Survey Table 43).
Other non-agricultural sectors also make use of this abundance including municipal
systems, industrial operations, and golf course irrigation. Figure 11 provides a map showing the
number of large quantity water users by catchment in the case study area. These include all large
quantity (100,000 gallons per day or larger) water users that have registered their LQWs or have
been permitted by the DEQ (data obtained from Andrew LeBaron, DEQ). Two catchments
contain 16 registered users, though many of the smaller catchments contain two or fewer.

39

Figure 11. Number of registered or permitted large quantity water users by catchment in case
study area (Data on appropriators obtained from Andrew LeBaron, DEQ; catchment delineation
based on version 8 obtained from Jeremiah Asher, Institute of Water Research, Michigan State
University).

3.2 Characteristics of Agricultural Appropriators
The description of agricultural irrigator characteristics is derived from two sources of
data. The first source is the USDA Census of Agriculture for the year 2007 and the second is
three focus groups and two interviews conducted in the year 2011.

40

3.21 Agricultural Statistics
Table 5 shows statistics on principal operators and whether they own, partially own
(meaning they rent some land that they farm), or rent the farm they work on. The majority of the
operators are full owners of their farms.

Number of
farms with
principal
operators
who are…
Full owners
Part owners
Tenants
Number of
farms

Michigan

Branch
County

Cass County

Kalamazoo
County

St. Joseph
County

39,311
(70.18%)
14,703
(26.25%)
2,000
(3.57%)

840
(74.40%)
252
(22.32%)
37
(3.28%)

553
(68.19%)
232
(28.61%)
26
(3.21%)

615
(72.01%)
198
(23.19%)
41
(4.80%)

727
(70.38%)
255
(24.69%)
51
(4.94%)

56,014

1,129

811

854

1,033

Table 5. 2007 data for number of Farms with principal operators who are full owners, part
owners, and tenants by county. Percentages in parenthesis are percentage of total number of
farms (Source: 2007 Census of Agriculture, Michigan: Volume 1, Chapter 2, Table 46).

Table 6 shows data on primary occupation of the principal operator. The majority of
farmers in the case study area have off-farm jobs to supplement their farming activities.

41

Number of
farms
where
occupation
of principal
operator
is…
Farming
Other
Number of
farms

Michigan

Branch
County

Cass County

Kalamazoo
County

St. Joseph
County

24,795
(44.27%)
31,219
(55.73%)

420
(37.20%)
709
(62.80%)

343
(42.29%)
468
(57.71%)

398
(46.60%)
456
(53.40%)

405
(39.21%)
628
(60.79%)

56,014

1,129

811

854

1,033

Table 6. 2007 data for number of farms with principal operators whose occupation is farming or
other by county. Percentages in parenthesis are percentage of total number of farms (Source:
2007 Census of Agriculture, Michigan: Volume 1, Chapter 2, Table 46).

Table 7 shows barriers to making improvements for conservation of energy or water on
farms using irrigation across the State of Michigan. The most common responses were first, that
the investments were not a priority and second, that benefits of the improvements did not cover
costs.

42

Total
Investigating
improvements
not a priority
Risk of
reduced yield
or poor crop
quality
Physical
field/crop
condition limit
system
improvements
Benefits of
improvements
do not cover
instillation
costs
Cannot
finance
improvements
Landlord will
not share in
cost
Uncertainty
about future
availability of
water
Will not be
farming long
enough to
justify
improvements
Other

Number of Farms
1,316

Acreage
749,210

585
(44.45%)

74,986
(10.01%)

234
(17.78%)

30,599
(4.08%)

88
(6.69%)

26,081
(3.48%)

310
(23.56%)

61,477
(8.21%)

256
(19.45%)

37,433
(5.00%)

52
(3.95%)

12,249
(1.63%)

31
(2.36%)

10,395
(1.39%)

50
(3.80%)

14,644
(1.95%)

117
(8.89%)

33,737
(4.50%)

Table 7. 2008 data on barriers to making improvements for reduced energy use or water
conservation for a sample of irrigated farms in Michigan. Percentages refer to percentage of total
responses. Respondents could choose more than one option (Source: Source: 2007 Census of
Agriculture 2008 Farm and Ranch Irrigation Survey, Table 41).

43

To help cover the cost of improvements in irrigation, cost-sharing programs are available
to irrigators in Michigan. Table 8 shows the payments received for irrigation and/or drainage
improvements to farms in the state by irrigated acreage surveyed across the state.

USDA
programs
Irrigated
acreage

Non USDA
Federal
programs

186,290
(38.37%)

(D)

State, local,
or district
programs
11,546
(2.38%)

other
8,628
(1.78%)

Table 8. 2008 data on irrigated acreage in Michigan that received payments from organizations
for irrigation and/or drainage improvements. Farms may receive funding from multiple sources,
so acreage listed under one program may also be listed under another. Percentages refer to
percentage of irrigated acreage surveyed in Michigan (Source: 2007 Census of Agriculture 2008
Farm and Ranch Irrigation Survey, Table 39).

3.22 Qualitative Data
To understand the relationships among irrigators and between irrigators and the DEQ,
three focus groups were conducted in the spring of 2011, and follow up interviews with selected
participants were held in the fall of 2011.
Focus groups convene people into a non-judgmental and non-threatening environment to
discuss a topic of interest to the researcher. In the correct setting, participants will open up and
share their opinions, thoughts, and experiences comfortably (Krueger and Casey 2009). In a
group setting, participants discuss with their peers the questions posed, allowing different ideas
and experiences to be shared. The interaction between participants may cause them to think of
things they might not otherwise, potentially creating a richer data set than an individual interview
might. An initial three groups are held and the common ideas that arise across these groups are
identified. If new information is still being shared after the third group, then more groups are
held until the set of information reaches saturation (Krueger and Casey 2009). It is then assumed
44

that all available information has been exposed. A successful group will consist of enough
people that a variety of ideas are shared, but not so many that participants do not have
opportunities to speak. Krueger and Casey (2009) suggest that a range of 5 to 10 people is
appropriate. While participants may feel pressure to conform to others’ ideas or existing
relationships cause information to be withheld, the method offers a way to gather diverse
viewpoints and create a theory about the common experience for irrigators.
The focus group participants were selected from a list of registered agricultural irrigators

5

from the Michigan Department of Agriculture and Rural Development. Because this is not a
quantitative analysis, the sample of irrigators need not be random. However, a selection process
was developed. The original list was paired down to include only irrigators in the counties of
Branch, Cass, Kalamazoo, and St. Joseph. The number of participants sampled from each county
was proportional to the number of registered irrigators from each county in the entire list. The
result was 16 participants from Branch, 15 from Cass, 20 from Kalamazoo, and 41 from St.
Joseph County. The recruitment calls proceeded with a goal of fair representation from each
county. Compensation of $50 was promised to each participant to offset the costs of participating;
however, many people did not have the time or the interest to participate. Quite quickly, the
sampling strategy devolved into calling people regardless of their county so that a sufficient
number would be recruited. The calls were met with some animosity, some confusion, and a little
enthusiasm, but mostly went unanswered. A script was read for each of the calls to ensure
accurate and consistent information (see Appendix A), and those who answered were invited to
attend one of three focus groups, all at the same location on three different dates in the spring of
2011. The dates were March 28, March 31, and April 21. The location was the St. Joseph County
5

All irrigators with capacity of 100,000 gallons per day or greater are required to annually report
their withdrawals to the Michigan Department of Agriculture (PA 148 of 2003).
45

Intermediate School District Building in Centreville because of its central location in the case
study area.
Seven individuals participated in the first focus group and the second and third each had
five participants. The discussions lasted from one hour to one and a half hours and were recorded.
A transcriptionist produced transcripts of each discussion. In each group, participants showed a
diversity of knowledge about the water use program, willingness to share, and suspicion about
the research. Most of the participants expressed that they enjoyed farming, and, though many of
them knew each other, no antagonism among participants was apparent. That there was a fair
amount of comity in the groups, however, should not lead to a conclusion that the farmers in the
case study area all are amicable or agree with the ideas shared. The resulting data from the
groups show that there are instances of animosity within the community of irrigators. Also, the
sample was from a subset of irrigators in compliance with registration and reporting
6

requirements; reports to MDARD in 2004 accounted for only 69.7% of irrigated acreage
reported by the National Agricultural Statistics Service in 2002 (State of Michigan GWIM 2005).
The opinions and experiences of irrigators who are not registered and/or do not report use were
not represented in the data collection.
The questions asked were aimed at learning about relationships between irrigators,
experiences with reducing water use, relationships between irrigators and state agencies, and
issues that could arise in a cooperative setting. The script used in the focus groups can be found
in Appendix B.

6

Irrigators report annual water use to Michigan Department of Agriculture and Rural
Development but register their withdrawals for prevention of ARIs with Michigan Department of
Environmental Quality.
46

3.3 Qualitative Research
The resulting transcripts were read to identify common themes and ideas using a method
for analyzing qualitative data developed from Taylor and Bogdan (1998). The authors divide the
process into three steps: discovery, coding, and discounting. Discovery entails reading and
rereading data while developing generalizations about the content. Coding is the process of
categorizing data to support, refine, or reject ideas generated in the discovery step; and
discounting involves interpreting data in the context in which it was collected. Group dynamics
will influence what information is shared; participants may feel pressure from their peers to
withhold information or to express agreement with ideas that they do not truly agree with
(Krueger and Casey 2009). In an evaluation of information on ecosystem services gained from
both focus groups and individual interviews, Kaplowitz and Hoehn (2001) found that the two
methods yield different but complementary results and concluded that the environment in which
data is collected will play an important role in what is shared. Recognizing the influence of the
researcher, the relationship between the participants, and the difference between information
given by a single participant and information given from multiple participants will help to insure
that data is interpreted in the proper context (Taylor and Bogdan 1998).
To follow the method of Taylor and Bogdan (1998), the transcripts were read several
times to gain familiarity and to identify common ideas. The transcripts were identified to the
extent possible by listening to the recordings and distinguishing one farmer’s comments from
another. This was helpful when similar ideas were found throughout the transcripts because it
clarified whether it was one person restating something he had said earlier or another person who
had the same idea. Ideas that were stated by two or more people in a focus group generally were
followed by other participants expressing agreement.

47

Individual interviews with two focus group participants were conducted in the fall of
2011 to clarify common ideas found in the focus groups and to gather missing information
needed for evaluating the case study in terms of the evaluation framework presented in Figure 4.
The questions asked in the individual interviews are presented in Appendix C.
Notes were taken during individual interviews and one was recorded; the other
participant asked not to be recorded. From the focus group transcripts, the notes from the
individual interviews, and all recorded discussions, several topics emerged as common themes or
ideas that characterize this group of irrigators.

3.4 Topics
The topics are presented in the following discussion using quotes from the focus groups
and descriptions of conversations. References to the transcripts for each topic can be found in
Appendix D.

3.41 Frustration
Two stories were shared that illustrated the frustration farmers have had with state or
federal regulations that constrain their operations. The first story was shared by a farmer who
was prevented from digging a pond because his land was part of a “valuable wetland.” The
second story was of a farmer whose attempt to get permission to burn down a barn was denied
due to air quality regulations. He expressed his frustration by saying, “I hate [the] DEQ”, and
“they come at us from the standpoint that we’re guilty.” He felt that his efforts to work within
state regulations went unnoticed when no one came on-site to evaluate the situation before they

48

forbade his actions. Two participants in the second group talked about the wetlands regulations
stopping them from either draining their ground or removing tree stumps from their land. One of
those participants said, “Everyone in here who farms is fiercely independent and one thing that
galls us more than anything else is to have someone come in and tell us what we can and can’t
do.” These past interactions with state agencies have influenced views on future regulations that
in their view are likely unnecessarily strict as well. In the first group, one participant suggested
that saving fish at the expense of his crop returns was unreasonable. Because the farmers do not
see any unreasonable environmental destruction caused by their operations and have no
experience with problems related to the overuse of water, the state’s efforts seem to be overly
cautious. If water scarcity were to become a problem, some participants said they would want to
keep state agencies out of efforts to conserve water, leaving it to the local community of water
users.
The state is not their only opposing force; the participants also noted public criticism.
Because irrigation infrastructure is very visible, it is easy for local water level reductions to be
blamed on farmers. One participant described a lake that had noticeable fluctuations and in a low
period the recreational community, thinking the agricultural use of water was a permanent tax on
the system, wanted to drill a well and pump water into the lake. Before that could happen, the
lake recovered and a few years later, it was at a record high. Another participant told of exposed
land that his grandfather walked across in the 1930s that is currently under water and expressed
that the intensification of agricultural operations could not possibly be causing local changes in
surface water levels. One group was adamant that recorded data has never shown water level
changes. In a different group, several participants said since Michigan gets precipitation for most
of the year and irrigation occurs only in three months, it is “ludicrous” that they get blamed for

49

water shortages. Even though the irrigators do not often see evidence of their impact on local
water levels, there is an example of an effort made in response to neighbors’ complaints. One of
the participants had, in cooperation with two other neighboring farmers, changed his crops and
drilled some new wells because non-farming neighbors had complained of low water levels in
their wells. The farmer believed it was probably an issue of inadequate wells but wanted to avoid
conflict with his neighbors.

3.42 Distrust
A reason for animosity toward the state is that case-by-case regulatory decisions affecting
farmers are based on models, which are simplifications of reality and therefore contain a certain
amount of inaccuracy. The farmers cannot connect what they see in the field with what the
regulatory agencies are claiming from the office. Wetland protection areas, according to a
participant, result from “just look[ing] at a map and … circling stuff.” The same sentiments are
reflected in the procedure of regulating water withdrawals. One participant expressed the view
that the WWAT, as the decision-making tool, is “too coarse” for actual use. Concerns about the
WWAT have been circulating in the community: “up around Schoolcraft they had a problem and
they said no more wells. Everybody was getting [zone] D – when you go in that system you get
A, B, C, D. D you don’t get any water. All they did is oop – we’ll just raise the limit. They had
a meeting - we’ll just raise the limit. They have no clue how much water is underneath and
everywhere, so they don’t know exactly what our recharge … is.” Similarly, another farmer in
the same focus group talked about the unreliability of the data used in building the tool: “Some
people are saying that perhaps they totally underestimated the amount of what the aquifer is able

50

to provide in here.” This was repeated when another participant said the more extensive data
used for site specific reviews shows that there is more water available than the WWAT portrays.
This illustrated the distrust participants have in the WWAT as a part of the water
regulation process. They know that it inaccurately portrays the availability of water and have a
perception that the amount of data used to build it is too limited. One participant said he thought
that the whole model was built around five data points and that the entire state was regulated on
just that data set. There is also a concern in the type of data put into the WWAT. The tool
requires the registrant to enter data for infrastructure capacity, which is how much the farmers
could pump all day every day throughout the summer, rather than how much they actually pump
throughout the summer. These two values can vary greatly, but since the actual water to be
pumped is unknown at the time the well is registered and constructed, even by the farmers
themselves, potential effects on the ecosystem are uncertain. One participant expressed the desire
for more accurate data before he would agree to reduce withdrawals: “They’d have to
statistically prove to me that I’m really creating an adverse impact. And they’d have to show me
some pretty widespread data to make me believe it and right now they don’t have anything.”
Similarly, commenting on the fact that not everyone in the area has registered their withdrawals,
a participant in the third focus group pointed out that the data are inaccurate in their
incompleteness: “you don’t have any data to come to us with…if you don’t have all the
information it’s either… all in or all out.” One participant told how the WWAT grievances have
inspired an irrigation association-funded study for a precise look at local water availability.

51

3.43 Heterogeneity
Though the farmers in Southwest Michigan are, in many ways, a distinct group of water
users, there were many elements of heterogeneity that could affect how well they work together.
In a negotiation setting, before anything could be decided, they would have to disclose all the
very different elements that go into each farm’s operations. These include investments, crops,
soil conditions, production contracts, and conservation measures. Compliance with required
water use reporting and registration also varies.
When asked about reducing water to create availability for new water users, a participant
said, “I think it would be pretty hard to want to go along with that. I mean, we all saw irrigation
as a valuable tool a long time ago when we made the investment and we’re doing it and it would
be hard to reduce your use for that reason, honestly.” He made the point again later on in the
focus group saying that there are farmers who choose not to irrigate and that those farmers use
completely different management strategies because they do not have the infrastructure
investment. During an individual interview, a farmer said that farmers choose to irrigate only
because it makes them competitive, not because they enjoy the extra work or money spent.
Several other participants also said they would have to consider how reducing water would affect
their financial returns before agreeing to a water reduction plan.
Even within the group of irrigating farmers, there are different levels of investment that
create differences in management decision sets. Some have put in large amounts of money to
create highly efficient systems while others may have old infrastructure. Farmers who have made
large investments may be unwilling to reduce their use for the benefit of someone with an
inefficient system. There are also different ways to irrigate, making comparisons across farms an
endeavor with high information costs.

52

The second factor is the type of crop each farmer is growing. While Southwest Michigan
is dominated by corn, soybeans, and potatoes, the participants explained that there are
differences between farms in terms of what is grown and how much water each crop needs for
healthy production. Not only is the amount of water variable across crops, but the time at which
crops need water is also variable. Corn, potatoes, and green beans take more water than other
crops, and seed corn production needs more water than commercial corn. Some crops are also
more expensive to grow and yield more returns than others.
These differences in investments and expected yields create a challenge for negotiations.
Part of this challenge is in sharing information and processing it to develop a water reduction
plan that is equitable. Some farmers can absorb production changes while others are financially
less stable. The interview participants said there is some financial disparity across the group of
irrigators, though it is not a wide gap. The differences are mainly in how long a farm has been in
a particular family, which affects debt, equity, and income levels. A participant said that, as an
example, a young farmer might have a more difficult time reducing inputs than more established
farmers; doing so might put them “out of business or worse.”
Soil types can change the efficiency of water use. “Different grounds… require different
amounts [of water]”; “you get a field that’s a little heavier, it holds the water, you get the field
that’s real sandy and you can put an inch on every day and the ground’s just moist out there.”
Another discussion about the physical heterogeneity in farmland was in regard to aquifer yield.
Figures 9 and 10 show that the case study area, with relatively abundant yields compared with
the rest of the state, has pockets of low yield. The participants had experience with these
heterogeneities even on their own land; wells placed only fields apart had very different wateryielding capabilities.

53

Probably one of the most important factors brought up by farmers in all focus groups
were production contracts with agricultural companies. Crops grown under contract include seed
corn, potatoes, and green beans, which all require more water than other crops in the area. These
companies, facing the principal-agent problem where the agent’s goals do not necessarily match
up with the principal’s goals (Nicholson and Snyder 2008), create tournament contracts. These
contracts are renewed each year based on past performance and create a highly competitive
atmosphere between growers. Research in St. Joseph County has shown that these types of
contracts result in over-application of inputs such as nitrogen (Preckel et al. 2000). By extension,
water is also overused. During one individual interview, a farmer described that the application
of nitrogen to potato crops requires the use of water. Even if the crop has had adequate rain, if
nitrogen levels are low when a leaf sample is taken, the irrigation system still has to be used to
apply nitrogen. He explained that potato contracts are even more competitive than seed corn
contracts and potatoes are the most water demanding crop in the area. Focus group participants
confirmed the competitiveness of those contracts and how marginal differences can easily
change a farmer’s outcome. The companies make decisions about how much water should be
applied and push the farmers to meet those conditions. Not only do individual farmers lose their
contracts in the face of low production, but entire areas can lose the presence of a company. One
farmer told how a seed corn company had set up in his community and then, right around the
time for planting in April, they suddenly took the operation further north. While this farmer said
the community recovered quite quickly, others think that these companies are very important to
the local economy (the “seed corn plants, they pump so much money into this economy, I mean
we’re very fortunate … go to any other part of the world and they don’t have that”) and are
aware that water regulations and other political barriers could prompt them to leave. While an

54

individual interview revealed that most farmers are raising seed corn under contracts, those who
are not involved in contracts are not under the same pressures as those who are.
The decision to employ conservation measures represents another division in irrigator
water use. Several of the participants described ways in which they conserve water to save
money and to reduce unnecessary use. Two farmers choose to irrigate at night; two others were
in programs that measured the moisture levels in their soils and helped them to reduce their
consumption to the minimum needed. In contrast to those who are consciously lowering their
water use are “neighbors” who constantly run their pumps for months at a time. The subject was
brought up in the individual interviews and both interviewees said that it is difficult to judge
someone else’s water use. They might have more pressure through their contracts or they might
have different ideas about how much water is needed. Regardless, the fact that some choose to
find out how to reduce their water use and others do not suggests that there is no norm regarding
how water should be used or how much is acceptable to use.
Some farmers believe that, because they own the land, they own rights to the water and
should not have to report to anyone, while others recognize state ownership and comply with
reporting laws to secure rights for future use. The participants were aware of the low reporting
compliance rate: “there’s a law that says everyone has to register. But there’s plenty of people
that aren’t and I would tell you there’s 40% of them probably that aren’t registered at all.” In a
different group, the estimation was that 30% were not registered. Not all participants knew about
the low compliance rate; one said that he thought everyone had registered because it was the law.
Those who comply do so either because it’s required of them or because they believe it is
important to register to secure that water for future farmers. Others “beat their fists on the table
and you couldn’t make them do it.”

55

3.44 Relationships
Many participants agreed that they would rather work out water conservation strategies
with their neighbors than have state interference; however, not all relationships with neighbors
seemed positive. Some participants said they thought their local community of water users would
be willing to work with each other and some thought it would be difficult to refuse a request to
work with their neighbors. One specifically said, “I would stand up and shout that farmers in St.
Joe County would work together.” On the other hand, most said that their willingness to work
together would be entirely dependent on who was asking. There were concerns that it would be
difficult to get everyone to work together and that state enforcement would probably be required.
A further complication is that some renters do not live in the area and it might be costly to
guarantee their presence at a meeting. When asked in an individual interview if relationships
between farmers who rent and farmers who own differ from those among farmers who own, a
farmer said that they are different and described a situation where renters started renting land that
changed agreements instated generations prior. That kind of event would affect a group’s ability
to work together. However, the other interviewed farmer said that the renters had the same
business mindset as those who do not rent so willingness to work together would not differ much
between the two sets of farmers.
Though social capital differs between individuals, it is clear that the participants think of
themselves as one group of water users distinct from other large water users. There were several
times throughout the focus groups when participants brought up water users who were not
irrigators. One farmer thought water scarcity problems would likely come from the
municipalities and others in the non-farming community, and another said that when all water
users were asked to get together, “everybody better bring their gloves. Because you’re going to

56

have just that, you’ll have a handful of us and a lot of them.” A third farmer in the same group
thought other water users would be less cooperative than the farmers because the public would
not want to give up the water for their lawns and golf courses. One participant stated that farmers
are the best stewards of the land, suggesting that the problems were actually stemming from
other groups of water users.

3.45 Anti-Diversion
While the participants were agitated by new regulations that could affect their operations,
two participants from different groups expressed their concern and frustration with diverting
Great Lakes water out of the Basin for the use of outside states. One farmer described a
conversation with someone in Oklahoma who believed that Great Lakes water should be a
market good. The participant said, “You know, our states own this water, not the Federal
government. And I don’t want the Federal government to come take it from us” and described
how farmers in Oklahoma buy water to wash their barns, a waste in his eyes. The other
participant said there was “no excuse” for the diversion of Lake Michigan water out through the
Chicago River, that bottled water was ridiculous, and that the waters of the Great Lakes need to
be protected.

3.46 Summary
These recurring themes and ideas show that the farmers have a shared sense of purpose:
they will defend themselves from blame and unnecessary regulations. However, there are also
many differences among them in terms of how this defense is carried out. Some choose to ignore
the regulations and carry on as though they did not exist; some choose to comply but grumble;

57

and some choose to add to the conversation by finding new information. Other differences
highlight how much water a farmer is able to give up financially due to biological characteristics
of the crops, physical characteristics of the soil and aquifers, and management strategies due to
contractual arrangements. Though they have definite concerns about state interference with their
operations, some also have a concern for keeping Great Lakes water within the basin and wish to
protect that resource. That many participants introduced themselves by describing their love of
farming, importance of irrigation, and even enjoyment of the recreational opportunities the water
resources of this state provide is evidence that the Great Lakes are an important part of their lives
in both work and leisure.

58

Chapter Four: Analysis and Results

In this chapter, the evaluation framework is applied to the case study area data. Barriers
to successful collective action are identified as those characteristics that result in high
opportunity costs to coordination, and high transaction costs in the act of coordinating. Should an
individual resource user or a group of resource users choose not to mitigate the ARI, they may
face litigation and all of the costs associated with that alternative. The choice to bear or not bear
the costs of collective action is dependent on the perceived benefit to cost ratio of litigation
versus the perceived benefit to cost ratio of negotiation. If the benefits of participating in
collective action are perceived to be higher than the costs, then users are more willing to
participate, which, in turn results in lower negotiating and monitoring costs than with a situation
where users do not perceive those benefits. Certain conditions result in high transaction costs
which leads to a lower likelihood that users will participate in collective action, choosing to bear
the costs of litigation instead. If resource users perceive collective action as beneficial, then the
gains from collective action are larger than the costs and users will choose to coordinate rather
than face each other in court.

4.1 Definition of the Resource
The purpose of the statewide program is to manage the waters of the entire state of
Michigan in compliance with the Compact (MCL 324.32702). This task was made manageable
by partitioning the network of rivers and streams into catchment areas and linking those
catchments to a GIS and analytical equation for reviewing proposed withdrawals. Catchments
are the smallest unit of watershed that the DEQ monitors for ARIs. Users of one catchment can

59

affect resources in an abutting catchment, creating larger groups of users residing in multiple
catchments. The appropriators who are identified as contributing to the adverse resource impact
will be convened into a water user committee to create a plan for mitigating their collective
impact. The following analysis corresponds to these appropriators and their related catchments.

4.2 Analysis
Each condition in the evaluation framework will be examined using data discussed in
Chapter Three and other relevant data from the focus groups and interviews. The unit of analysis
is the transaction cost.

Resource System Characteristics
The first set of conditions in the framework relates to characteristics of the catchments
found in the case study area.

Small size: A map of the case study counties and their catchments is presented in Figure 12.
Because each catchment is, by definition, ecologically and hydrologically distinct from its
abutting catchments, it has a clear boundary for purposes of management (Reeves et al. 2009).
The size and shape of each catchment varies; the size ranges from 0.0970 square miles to
approximately 4,103 square miles in the case study area (version 8 catchment delineation
obtained from Jeremiah Asher, Institute of Water Research, Michigan State University). The
larger catchments will have higher monitoring costs than the smaller catchments and many of the
larger catchments contain more resource users, resulting in higher likelihoods of ARIs in the

60

larger catchments. The size of the larger catchments could contribute to high transaction costs in
collective action.

Figure 12. Catchments in the case study area (Source: delineation based on version 8 obtained
from Jeremiah Asher, Institute of Water Research, Michigan State University).

Well-defined boundaries: The boundaries of a stream affected by an ARI are established in the
withdrawal assessment process as catchment areas. When an ARI occurs, the specified area of
management will result in lower information and monitoring costs than if boundaries were not
defined.

Low levels of mobility, possibilities of storage: Case study appropriators withdraw water from
both surface and groundwater sources. According to Schlager et al. (1994), who identified
storage as an important element in successful collective action, groundwater is a stored resource.
61

In reality, groundwater flows slowly with the gradient of the water table or pressure surface;
however, for the purpose of this analysis, the water in aquifers has relatively low levels of
mobility. Surface water, while mobile, can be pumped into ponds creating a stored resource. The
mobility of these resources, either surface water or groundwater, will not add monitoring costs to
the process of collective reduction in water use.

Predictability: Several focus group participants discussed their perceptions of the unreliability
of attempts to quantify the amount of water in aquifers and streams and their frustration with
policies based on models. There is evidence that they believe that a program to mitigate water
shortages in streams is unnecessary because those shortages are natural and not human caused.
Many farmers, according to the census data presented in Chapter Three, do not consider
conservation strategies a priority, have never experienced a groundwater shortage, and do not
consider the costs of implementing conservation measures to be covered by the benefits. A
sizeable portion of these farmers may think that water will always be available in their areas and
that the predictions of stream response to withdrawals by the DEQ’s assessment process are
incorrect.
In a sense, they are not wrong; 100% accuracy of predictions of withdrawal impact on
water dependent resources is impossible. However, the assessment process adds to knowledge of
how the water system functions and uses available data to estimate ecosystem response. Updates
to the data sets are anticipated (Groundwater Conservation Advisory Council 2007). Reeves et
al. (2009) tested the analytical model (based on inverse distance from well to stream) used in the
WWAT against a USGS-developed groundwater flow model called MODFLOW. Hypothetical
wells were placed across a grid in a catchment in Kalamazoo County in both models and the

62

resulting stream depletions were mapped. The maps from each model are presented in Figure 13.
Panel A is the MODFLOW results and panel B the analytical model results.

Figure 13. Comparison of well simulation in A) MODFLOW model with B) WWAT analytical
model (Source: Reeves et al. 2009).

The stream flow depletion in each model is similar, though MODFLOW shows higher
reductions. Reeves et al. (2009, p.12) called the results a “reasonable match” and concluded that
they support the use of the analytical model as a screening tool. When collective water
withdrawals cause an ARI, a site-specific review using more precise local data is employed to
63

overcome some of the inaccuracy of WWAT. The predictability of the resource, while not
complete, is assessed from detailed data and models that are tested for accuracy.
The assessment process for withdrawals is labeled reasonable by scientists who work for
government agencies; however, it is the appropriators who must take the information given to
them by the assessment process and use it to create CPR institutions that will mitigate an ARI. If
they do not trust the data, they will be less willing to bear the costs of collective action; as one
focus group participant said, “They’d have to statistically prove to me that I’m really creating an
adverse impact. And they’d have to show me some pretty widespread data to make me believe it
and right now they don’t have anything.” Other participants discussed times when water levels
were temporarily low and rose again without any reduction of their withdrawals, so any ARI
detection in the field could be attributed to natural fluctuations.
This distrust in DEQ’s ability to predict the effect of water withdrawals on stream flow
could create costs to convene users. If the appropriators are unable to acknowledge the problem
they are asked to fix, transaction and monitoring costs may be high enough that they will be
unwilling to bear the costs of collective action and may choose to face litigation as an alternative.

Indicators of CPR conditions: The indicator used to detect an ARI is exceedence of a
percentage reduction of the index flow in streams. Table 1 in Chapter 1 shows the percentage of
stream flow reduction that will cause enough damage to fish populations to merit action by the
DEQ. In cold-transitional rivers, ARIs occur with as little as a 2% reduction in the index flow.
While some ARIs can be detected by sight, others will be detected by models and datasets. In
many cases, the WWAT or site-specific review, both managed by DEQ, will be used to indicate
64

whether a stream or river is healthy. This means that the power to decide whether the CPR is
being managed according to standards set by Part 327 of Act 451 rests with the DEQ, which uses
a tool that the appropriators do not trust. These three elements reduce the likelihood that resource
users will bear the costs of collective action. The appropriators do not set the standards, are not
able to determine whether the standards are being met, and do not trust how the resource is
evaluated in terms of the standards. Some appropriators might not even agree with evaluating the
resource by the status of its fish populations: one participant said, “I’d hate the State of Michigan
to come in and tell me I can’t pump water anymore based on adverse impact that I’m [having] on
fish.” When appropriators do not agree with the purpose for reduction, they may find that
collective action is too costly to participate in. If they do choose to participate in the water user
committee, negotiating and monitoring costs may be high due to the view that bearing the costs
of reducing water use is higher than the perceived benefits from an increased stream flow.

Group Characteristics
The second group of conditions relates to characteristics of the resource users within the
affected catchments.

Small size: Each catchment has a designated number of appropriators who affect stream flow. If
an ARI occurs in that catchment, then all registered users will be asked to participate in a water
user committee. Because catchments have between 1 and 16 registered users in them (see Figure
11), the number of participants in a water user committee varies. Effects on multiple catchments
as described in equation (1) can draw appropriators from outside of the catchment in a water user
committee resulting in a larger number of participants. Group size in the case study area has the
65

potential to increase negotiating and monitoring costs. However, not all groups will necessarily
have large transaction costs because of group size; some may be small enough to lower those
costs.

Clear boundaries: The boundaries on a group of water users for an individual appropriator are
not clear. An appropriator in one catchment can contribute to an ARI in one or more neighboring
catchments. This means that an individual appropriator can be part of multiple water user
committees. This increases negotiating costs if that appropriator must work with several groups
of appropriators. Another important factor decreasing the efficacy of the water use program is
the presence of unregistered users. Participants in focus groups commented on the high number
of unregistered water users, but were not clear about whether they would identify those users.
Without their participation, the actions of unregistered users who continue to appropriate water
from an affected stream may undermine the actions of participants to restrain use in compliance
with a water user committee plan (free ride). One participant said the solution to bringing
everyone into the discussion was to report all of their unregistered neighbors while another said
he would rather not, in order to preserve relationships. Most participants expressed, though, that
they would not be cooperative unless all users were present.
When asked in the individual interviews whether they thought farmers would report their
unregistered neighbors if a water user committee was convened, both farmers said that because it
was a matter of maintaining water use, people would report their neighbors despite possible
relationship consequences. The presence of unregistered users means that successful collective
action is threatened by the presence of free riders that decrease assurance in collective

66

compliance. In Michigan’s program, these unregistered users who are identified are treated like
new users and must register their withdrawals through the WWAT.
New users cause high negotiating costs when exercising their rights to water use in a
catchment already impacted by an ARI. Once that ARI has been mitigated by a water user
committee, any new user, whether proposing a new withdrawal or being identified as
unregistered, will face high negotiating costs in water user committees because reductions in
water use have an increasing marginal cost in forgone production. A new user petitioning an old
user who has already reduced intake must convince that old user that a further reduction in water
use will somehow result in added benefits.
The lack of assurance caused by ill-defined boundaries will decrease the likelihood that
appropriators will bear the cost of collective action. New users in catchments already taxed will
find increasing negotiating and monitoring costs due to an increasing marginal opportunity cost
of reduction with each new proposed withdrawal.

Shared norms: Heterogeneities among the appropriators were discussed in Chapter Three.
These differences can affect the ability of norms to permeate a community and to function as an
element in the decision-making process. There is no social norm surrounding water use. Some
farmers said they made an effort to conserve water to reduce their bills or to avoid overuse, while
others, participants observed, “turn the damn thing on for a month at a time.” One participant
said it would be difficult to get all users to agree on reductions in water use and that some
farmers would “get pretty nasty” about anyone asking them to. The two individuals who were
privately interviewed both agreed that there was not a common idea in the community about the
definition of appropriate use. One of the interviewed farmers attributed this to the diversity of

67

crops in the area; different crops require different amounts of water. The other interviewee said
there are many interpretations of how much water a certain crop may physically need. Some
farmers rely on their experience to decide an appropriate amount of water while others turn to
updated scientific information. There is no consensus and no movement to formulate one.
Compliance with mandatory water reporting is another area without consistent group action, as
evidenced by the estimated high rates of unregistered users and the description by some in the
focus group of “beat[ing]… fists” when talking to those users about reporting.
The idea that water is personal property tied to land ownership arose in the focus groups
and the interviews. Perhaps the lack of a social norm about water use results from views on
property rights. Those who report their water use in order to secure it in the future think that the
water does not belong to them, but to the state. This is consistent with water use legislation
identifying water as a public trust and the responsibility of the state (MCL 324.32702c). Those
who do not report because they want the state to stay out of their business are expressing that
they think the water belongs to them (a common misconception that has never been true in
Michigan – Lusch, personal communication, December 2011). These different ideas about how
water should be used and the state’s role in that use may influence the choice to participate in
collective action and will likely increase negotiating and monitoring costs should a water user
committee be convened.

Social capital: Relationships are the basis for social capital, defined as “a person’s or group’s
sympathy toward another person or group that may produce a potential benefit, advantage, and
preferential treatment for another person or group of persons beyond that expected in an
exchange relationship (Robison et al. 2010: p. 19).” Decisions to cooperate are affected by

68

relationships but are not completely determined by them. There were two examples of
cooperation shared in focus groups. One farmer reduced his water use in response to complaints
from non-agricultural neighbors that the water levels in their wells were decreasing. This
participant worked with two other irrigators to remedy the problem, but he said he was not sure it
had really been caused by irrigation. Another example was from an irrigator who moved some of
his crops to help his neighbor meet the required seed corn setbacks that the seed corn company
asked for. According to Lyndon Kelley of Michigan State University Extension, setbacks prevent
pollination pollution by requiring only non-corn crops to be grown next to crops of seed corn
(Kelley, personal communication, November 2010).
These examples might contain elements of social capital, or they might suggest a social
norm: that it is important to work with neighbors if it is reasonable to do so. A focus group
participant expressed that he wanted to keep good relations with his neighbors. Another said,
“Most of my neighbors [have] farmed around each other for long enough that we don’t have
great conflicts over stuff that you would think would be a conflict.” These cooperative
interactions can happen with or without social capital, but when several participants said their
willingness to work with one another was dependent on who was asking, it suggests that some
past encounters had created negative relationships that would hinder cooperative decision
making. Social capital is present in the community of irrigators and could enhance collective
action; however, negative relationships also exist and could hinder it. The larger the group size,
the more likely negative relationships will exist between some members, which could increase
negotiating costs in a water user committee, but the presence of social capital will decrease
transaction costs both in negotiating and monitoring. The magnitude of the effects of either

69

positive or negative relationships is unknown. For this analysis, however, the presence of social
capital will be listed as an opportunity for collective action.

Leadership: Some participants in the focus groups were more outspoken and more informed
about the 2006 legislation than others. This implies that certain individuals in the case study area
seek to fully understand the laws that affect them and to be able to comment on them
knowledgeably. Perhaps these individuals and those like them are seen as leaders in the area. In
one focus group, an individual who was not present was named as someone who was heading up
some research on water availability in the area. This person was named as if everyone in the
room knew who he was. That the individual was leading a research initiative and that his name
was known in a group pulled together because they are registered irrigators provides evidence
that there are leaders in the community. Whether they have influence was explored in the
interviews.
The interviewed farmers were asked if leadership exists, that is, people whose opinions
were more influential than others, and both said yes. One of the farmers warned against thinking
that these influential people would influence in favor of this state regulation. The strength of the
leadership and the reaches of their influence are not known, but it is present. Whether this
leadership would advance the efficacy of a water user committee in the context of Michigan’s
water use program is also unclear. The presence of leadership and their participation in
compliance with the law reduces transaction costs and monitoring costs if the leadership is able
to persuade appropriators to comply as well.

70

Heterogeneity of endowments: There is ample variation across farms in terms of soil and
aquifer conditions, infrastructure, and investments. Those who are committed to contracts face
different pressures than those who own their land and their crop, but they also expect a larger
return on investment. Differences in aquifer and soil conditions make it easier or more difficult
to make water available for plants.
While the pumping capacities of those reporting to the Michigan Department of
Agriculture (and thus sampled for the focus groups) are not disclosed, the DEQ lists the WWAT
registration capacities for new registrants on their website in gallons per minute (gpm). These are
the withdrawals that passed the review process since July 9, 2009. Figure 13 shows the
frequencies of registrations from the case study area grouped into nine ranges.

Figure 14. Frequencies of pumping capacity registered in WWAT since July 9, 2009 in the
counties of Branch, Cass, Kalamazoo, and St. Joseph. Pumping capacities reported in gallons per
minute (Source: DeYoung 2011).
71

The largest number of registrants pump in the range of 401 – 500 gpm, although there is
also a relatively large number pumping in the range of 901 – 1000 gpm. Irrigators who use nine
to ten times more water than others might benefit from a CPR institution by ensuring that large
amounts of water will be available in the future. This interest may reduce negotiating and
monitoring costs because farmers have an incentive to promote conservation in their area but
there is no strong indicator of whether these heterogeneities will affect collective action
outcomes.

Homogeneity of interests and identities: When introducing themselves, many participants
included family history of farming; most farmers’ families have been in the area for several
generations. They have a shared identity as irrigators, making them different from other farmers
and other large water users. Their interests, despite examples of differing actions in terms of
water use and compliance, might also be aligned as many expressed a desire to minimize state
interference while maximizing returns on investments. Some farmers said that collective water
reduction would require an outside party: “Farmers don’t stick together that well so you almost
[need] somebody with a big stick standing up there.” However, it was clear that most participants
felt they would rather not be regulated by the state. Their common aversion to state involvement
could motivate groups of irrigators to collaborate to create water management plans. If they do
not, the DEQ can offer a plan of their own and, if not accepted, the court system can force a plan
of action (MCL 324.32725). The costs of litigation could be higher than the costs of collective
negotiation, causing a group of resource users to agree to negotiate in order to reduce state
involvement.

72

Low levels of poverty: When asked about poverty in the interviews, the farmers said that there
are some differences in financial welfare in the case study area. However, one interviewee said
that most of the differences were between farmers who irrigate and farmers who do not irrigate.
He said that farmers irrigate because they want to profit from their operations and often do so.
The other interviewee said that the largest differences occur because of the length of time a
farmer has been in the area. Those on intergenerational farms have lower debt and more equity
than new farmers starting from scratch. Even with these differences, the farmers do not face the
same poverty as the communities that Wade (1988), Baland and Platteau (1996), and other
researchers wrote about when examining poverty as a barrier to collective action. Many have
non-farm income supplementing their farm income (Table 6). Poverty, as a characteristic that
increases discount rates and thus decreases benefits to restraining resource use for the future, is
not a strong characteristic in the case study area.

Past experience with organizations: The interviewed farmers were asked about existing
organizations that would help in the process of creating a water user committee. They listed
organizations that many irrigators belong to. The Michigan Farm Bureau is one of the largest.
Other regional groups mentioned were an irrigation association and a group promoting
responsible water use. Familiar references, then, are available for guidance and could help in
structuring a water user committee, which would reduce transaction costs in this collective action
situation.
There were also examples of cooperation with the use of shared water sources.
Participants in one focus group had heard of cases where farmers shared a surface water source
and an informal agreement about who could pump at certain times, and one had experience with

73

planning his crop rotation around his neighbors because they all knew that their shared source
could not handle water intensive crops on all three farms simultaneously. The interviewed
farmers also knew of examples of shared infrastructure but said that some people would just
refuse to be in a situation with shared capital. These examples of organization and cooperation
that already exist could reduce negotiating costs of creating similar collective action institutions.

Relationship between resource system characteristics and group characteristics
Agrawal (2001) describes the importance of several conditions arising from the joint
impacts of resource system characteristics and characteristics of CPR appropriators.

Overlap between user group location and resource location: Census data presented in Table 5
showed that around 70% of the farms in the case study area are fully owned by their principal
operators. The rest are partially owned (22 – 29%) or worked by tenants (less than 5%). The
presence of renters was a topic of discussion in one focus group. Renters target Southwest
Michigan because it has soil conditions conducive for growing crops like potatoes and tomatoes.
These are valuable crops that require substantial amounts of water and sandy soil, two
characteristics of the case study area.
Participants said renters should be a part of any discussions about water use, but one
landowner said that he would be less likely to cooperate with them than he would with his
neighbor. An interviewed farmer said relationships between farmers might be different if they
were renting because their relatively newer contracts may have been at the expense of old
agreements between local farmers. The other interviewed farmer said that those who are renting
have the same concerns as those who own land, so he does not see the relationships as any

74

different. The bad relationships could be just that: an example of one farmer hindering another’s
operations and not necessarily due to the distance between them.
Regardless of who is farming, landowners are the people holding riparian or groundwater
rights, and when water user committees are convened, it is the landowner who will be asked to
participate. Not only might a retired landowner be difficult to contact (Kelley, personal
communication, November 2010), such an individual might not be as informed as some of the
working farmers. Owners are ultimately the decision-makers in discussions about plans for
eliminating an ARI, but they may have different ideas about water use than tenants. One focus
group participant described how his landowner was angry when he reported his annual water use
to MDARD (as required by law). The renter thought it was important to do so but the landowner
disagreed.
The landowner-renter situation creates barriers to collective action when the landowner,
as decision-maker, is not physically present (e.g. lives out of state) and/or has been removed
from farm management decisions for a long period of time. Their perspective might not match a
renter’s perspective, causing problems when the renter has to implement actions to which a
landowner has agreed. One farm would be represented by two people, increasing the size of the
CPR group and increasing transaction costs. Transaction costs also increase when land is rented
to different farmers over short periods of time. New participants may not have built relationships
with their fellow water users and may not be as willing to cooperate as past participants in water
user committees.

High levels of dependence by group members on resource system: High levels of dependence
may cause appropriators to steward resources that they need to rely on for the future. During

75

introductions, many participants said a large part of what they enjoyed about Michigan was the
abundant water resources and the opportunities they have as irrigators. This could make them
more apt to reduce their water use if scarcity became an issue. Irrigation in Michigan is
supplemental, meaning that depending on the crop, farmers can operate without it. However, the
heavy reliance of the local economy on seed corn and potatoes that the participants mentioned
throughout focus groups causes a dependence on irrigation. As one participant said, “I don’t
think you could have enough water available for potatoes unless it rained an inch a day every day
of the week.”
Table 6 in Chapter Three shows that many farmers do not list farming as their primary
occupation, which suggests they rely less on agriculture for income and are therefore less
dependent on irrigation than farmers who do list farming as their primary occupation. The table
does not discriminate between irrigating farmers and non-irrigating farmers; however, one
interviewed participant said that those who irrigate are those with a primary occupation of
farming because it is a much higher investment in time. Thus, irrigators in the case study area
probably do have a high dependence on water and would have incentive to steward their resource
if they sensed scarcity, increasing the costs of not creating a collective management plan for
sustainability of the resource. When the benefits of collective action are greater than the costs,
the resource users will coordinate their decisions.

Fairness in allocation of benefits from common resources: In order for farmers to begin
irrigating, they must make an investment in infrastructure that allows them to do so. In this sense,
the allocation of water to farm is fair; they have invested in infrastructure to capture benefits.
One participant expressed that poor investments by others might make him less sympathetic to

76

cooperation: “if they’re running out of water because of a … sub-standard system right off the
bat…why should I take a loss if I went to the extra money to, to put in a good set-up and they’ve
got something that in the best of times won’t work?” The discussions about investments asserted
that those who invested in infrastructure should not have to reduce consumption for those who
have either poor investments or no investments. That idea seemed pervasive across focus groups,
although one participant allowed that the water did not belong to him personally, so while he
might not be happy about sharing his allocation of water, it might be fair to do so.
If the participants in a water user committee share the perspective that it is fair to allocate
water among all users, then there might be less incentive to cheat. However, in the examination
of social norms, it was discussed that many appropriators in the area do not share this view and
think of their rights to water as private and not contingent on others’ use and ecosystem health as
with regulated riparianism. These appropriators are less likely to be willing to participate in
collective action, and more likely to cheat on agreements for water reduction than those who
believe it is fair to share the water for multiple uses. These appropriators might also cause the
time cost of negotiations to increase when setting rules for water use in a water user committee.

Low levels of user demand: It has been established in this chapter already that appropriators in
the case study area depend on a lot of water for their operations. It has also been established that
there is an abundance of water in Southwest Michigan. While user demand may not at present be
high enough relative to the amount available to cause an ARI, the demand relative to the rest of
Michigan is very high (refer to Table 4). These appropriators have a high opportunity cost of
reducing their water use. Demand is high because there is value to water as an input and giving
up that input results in a decrease in profits. The result is that appropriators are less willing to

77

constrain their water use than they would if they had lower levels of demand causing negotiating
costs and monitoring costs to be high.

Gradual changes in levels of demand: The two interviewed farmers were asked if they could
think of any event that would cause a large increase in the demand for water. One said he could
not, while the other said only if someone changed his or her crop rotation to one that required
more water. Both thought that the quantity of water demanded in the area would not change, just
who was demanding it. All land available for agriculture is already in agriculture and much of it
is in irrigation, according to the farmers. According to the statistics presented in the third chapter,
however, the majority of agricultural land in the case study area is not irrigated. If non-irrigated
land is converted to irrigated land, an upward shift in demand could occur. This could happen if
other farming operations relocated to Southwest Michigan due to localized water shortages or
declines in water quality. An increase in production contracts would increase demand, as
evidenced by the increased intensity of irrigation following the initial introduction of contracts.
A new agricultural company could increase demand for water among existing operations leading
to higher transaction costs in negotiation as described in the section above. There is no strong
indication that demand would increase and therefore no conclusion about the effect on collective
action in the case study area.

External Environment
The last set of conditions in the evaluation framework refers to the external environment
surrounding the use of the resource.

78

Technology: low exclusion cost, time for adaptation to new technologies: The technology
used to exclude potential users was discussed in the section on fair allocation of benefits. Both
the irrigation technology and property rights in land and water divide those who can appropriate
from those who cannot. Property rights are already enforced with external institutions
eliminating the costly need to create and manage new bodies for that purpose. Irrigation
infrastructure, a high-cost investment, deters casual thoughts of appropriation. If irrigation
equipment were to suddenly become less costly due to advances in technology, then farmers in
the area could increase their use of irrigation, increasing opportunity costs of reducing water use.
They are less likely to agree to management plans for reducing water use, increasing costs of
negotiating. These appropriators are also less likely to comply with these rules, which results in
high monitoring costs.

Low levels of articulation with external markets: The yield of a crop is directly related to
irrigation. This was made clear with discussions in focus groups on the intensity of water use that
is the nature of management decisions by farmers under production contracts. The companies
encourage these agents to irrigate because it will increase yields and, in turn, increase profits for
the growers. One of the farmers who participated in an individual interview said that the reason
people choose to irrigate is not because they like to, but because it allows them to grow more
profitable crops. Farmers who assume the risk of production contracts and/or invest in irrigation
do so because they expect returns. They face high opportunity costs of reducing water use, which
leads to high negotiating and monitoring costs.

79

Gradual changes in articulation with external markets: If crops in Southwest Michigan
became incorporated into more market products, then their value would increase and willingness
to reduce water use would decrease. Collective decisions on reduction, then, would be more
costly and the result is a barrier to collective action.

State
Power of local authority: Because DEQ expects water user committees to create institutions
internally for the collective reduction of water use, the users have power in setting and
negotiating rules among themselves. However, because ARI status is evaluated by DEQ, the
agency retains some power in the process of collective water reduction. The irrigator’s distrust
in the decisions of state agencies means that the presence of DEQ could hinder the workings of
CPR institutions.
Most of DEQ’s enforcement works through registration processes that require
documentation. If such requirements of the state do not align with the expectations of the
irrigators (indications from focus groups suggest they do not), perhaps for some degree of
autonomy in their role as resource managers, negotiating agreements acceptable to the state
could become more difficult and the costs of negotiating such plans could be high.

Supportive external sanctioning institutions: The institutions involved in supporting water
reduction measures are those of the legislation that created the definition and prohibition of an
ARI and the system that protects water and property rights. These institutions call for reductions
in water use in the case of an ARI and uphold property rights during negotiations. If a plan for
collective reduction of water were to be agreed on, the DEQ would be involved in monitoring the

80

outcome and supporting those complying with the agreement which would decrease monitoring
costs. As a result, the efficacy of collective action would increase from these institutions if
distrust of DEQ as a monitor does not cause high negotiating costs.

External aid for conservation: Southwest Michigan receives support for water conservation
programs through the Agricultural Water Enhancement Program, part of the Natural Resources
Conservation Service’s Environmental Quality Incentives Program. This is a national program
that supports conservation measures in agriculture. One focus group participant had taken
advantage of AWEP and was able to reduce his pumping rates significantly with soil moisture
monitoring while retaining yields. This program is available to irrigators in Southwest Michigan
and promotes efforts to conserve water. Table 7 in Chapter Three shows that other funding is
available for conservation projects across Michigan. Reducing barriers to conservation decreases
negotiating and monitoring costs because it becomes less costly to engage in conservation
activities. Irrigators are able to maintain their yields with less water, decreasing opportunity costs
of reducing water use. These irrigators will be more likely to participate in the creation of and
agreement to plans for collective water reduction and are more likely to follow those plans than
those who face high opportunity costs.

Nested levels of appropriation, provision, enforcement, governance: The water user
committee, as a collective action institution, is a working unit embedded in the larger water use
program in Michigan. This program is part of a larger regional compact, which is part of an
international agreement. The laws and institutions that enforce property rights can be used to
support the terms of agreements made in the water user committee by holding each appropriator

81

to a reasonable level of water use. If use is unreasonable and it hinders reasonable use by another
appropriator, it is unlawful and corresponding litigation can follow (Lusch 2011). The entire set
of laws and institutions work to protect access to the benefits of water for Michigan citizens.
Without it, there are no enforceable property rights and an open access situation results. With
legitimate rights to water and the ability to protect their access to it with legal structures that
prohibit others from overuse or diversion, farmers have incentives to bear the costs of collective
action. These incentives to participate in creating and implementing a water management plan
can create a willingness to negotiate and thus decrease the costs of negotiation. Likewise, they
can also create a willingness to comply with rules for water reduction ultimately reducing
monitoring costs.

4.3 Summary of Findings
Table 8 summarizes the findings in this chapter. Because the evaluation framework is
built on generalized conditions for successful collective action, certain conditions in each CPR
situation evaluated with the framework will have a larger impact on transaction costs and be
more important in the identification of opportunities and barriers for collective action. Table 8
shows which conditions in the case study area situation contribute to the decision to participate
in a water user committee and the magnitude of transaction costs associated with the committee’s
functions. Each condition is evaluated as either creating a barrier to collective action by resulting
in high transaction costs, or creating an opportunity for collective action by increasing the
benefits to collective action and decreasing transaction costs, or being irrelevant to the situation.

82

Resource System
Barrier: high monitoring costs
Barrier: high negotiating and monitoring
Predictability
costs
Barrier: high negotiating and monitoring
Indicators of CPR conditions
costs
Group
Barrier: high negotiating and monitoring
Size
costs
Barrier: high negotiating and monitoring
Boundaries
costs
Barrier: high negotiating and monitoring
Norms
costs
Opportunity: perceived benefits of
Social capital
collective action, low negotiating and
monitoring costs
Leadership
Opportunity: low negotiating costs
Heterogeneity of endowments
Opportunity: perceived benefits of
Homogeneity of interests/identities
collective action, low negotiating and
monitoring costs
Poverty
Opportunity: low negotiating and
Existing organizations
monitoring costs
Resource System and Group
Location
Barrier: high negotiating costs
Opportunity: perceived benefits of
Dependence
collective action, low negotiating and
monitoring costs
Barrier: high negotiating and monitoring
Fairness of allocation
costs
Barrier: high negotiating and monitoring
User demand
costs
Changes in demand
Size
Boundaries
Mobility
Storage

Table 9. Barriers, opportunities and irrelevant conditions for collective action according to
evaluation framework and data from Branch, Cass, Kalamazoo, and St. Joseph Counties.

83

External Environment
Technology
Barrier: high negotiating and monitoring
Market articulation
costs
Change in market articulation
Barrier: high negotiating and monitoring
Power of local authority
costs
Supportive institutions
Opportunity: low monitoring costs
Opportunity: low negotiating and
External aid
monitoring costs
Opportunity: increased benefits of
Nested levels
collective action, low negotiating and
monitoring costs
Indicates a condition that is not considered important for this situation

Table 9, Cont. Barriers, opportunities and irrelevant conditions for collective action according to
evaluation framework and data from Branch, Cass, Kalamazoo, and St. Joseph Counties.

84

Chapter Five: Conclusions

This analysis was completed by examining a case study area of four counties in
Southwest Michigan against conditions found by researchers to lead to successful collective
action in resource management. For this research, successful collective action refers to
institutions that lead to sustained access to benefits over time without significant conflict
between resource users (Bromley 1992). Data from focus groups and secondary sources were
analyzed to understand how transaction costs affect opportunities and barriers for success of
water user committees attempting to collectively reduce water use. The results describe a set of
characterizing conditions that, through their effects on whether water users are likely to
participate in collective action and on the transaction costs associated with that collective action,
lead to barriers or opportunities for successful collective action.

5.1 Opportunities
Water use in the community of agricultural irrigators is very important. Dependence on
water is high, which creates an incentive for irrigators to participate in programs that protect that
use for the present and the future. The water use program, as an important element in regional
efforts to prohibit diversions outside of the Basin, offers that protection. While the irrigators have
made different investments that result in different perspectives on water use reduction, most
would rather work within their agricultural community to mitigate ARIs than with the DEQ. If
they have no choice about whether or not they should, as a group, decrease water use, their
common interest in reducing state involvement may result in a higher willingness to negotiate
with their peers on water reductions. The examples of farmers working together and sharing

85

infrastructure show that some irrigators have experience in cooperative situations which, like the
presence of formal organizations in the community, can result in lower transaction costs in
negotiating and organization in a water user committee.
Social capital in the community will also lead to willingness to cooperate, resulting in
lower transaction costs in both negotiating and monitoring compliance with rules created by the
water user committees. Leadership can influence the decisions of group members and increase
willingness to cooperate. Institutions that establish rights of water users to protect their access to
benefits of surface and groundwater create incentives for irrigators to bear the cost of protecting
those rights and therefore to be present at committees making decisions on water use. A last
opportunity is found in financial assistance available for practicing conservation measures. If
farmers are faced with lower costs of implementing conservation measures, they are more likely
to do so. These opportunities lead to a higher willingness and ability to reduce water use and thus
lower transaction costs in negotiating and monitoring.

5.2 Barriers
Many conditions also could create high transaction costs resulting in barriers to collective
action. The first is that the irrigators do not trust the DEQ’s method of predicting and measuring
an ARI. If they do not perceive a problem (as measured by the DEQ’s assessment process), they
are less likely to reduce their water use, leading to high transaction costs in both negotiating and
monitoring. The misperceptions about water rights result in high transaction costs and have been
important in noncompliance with other state regulations such as annual reporting of water use.
Some irrigators see the water as their own private property that should not be managed by
anyone but them; they may not believe that reducing their water use is very fair to them. The

86

water belongs to the state, however, and is treated as such. The large size of groups and
catchment areas create negotiating and monitoring costs and the fluid boundaries of water user
committees due to new users, unregistered users, renters, and multiple ARIs influenced by a
single LQW increase negotiating and monitoring costs. As the number of water user committees
an individual appropriator must work with increases, the higher are the transaction costs of
collective action for that individual, and willingness to cooperate with any one group declines.
The presence of agricultural processing companies and the high value production
contracts they negotiate reduce willingness of irrigators to decrease water use due to the risk they
face in lost yields and thus lost contracts. The final barrier is that water user committee
participants may wish that agreements made in this community of water users be made with
unwritten contracts. The DEQ may undermine this practice by requiring written contracts or
agreements resulting in less willingness among water users to cooperate.

5.3 Suggestions for the DEQ
The barriers to collective action in this community are not insurmountable. Michigan
State University Extension (MSUE) is an organization that bridges the gap between research and
application at the farm level. MSUE specialists and educators have been building relationships
with farmers for many years and have been influential in the implementation of new farming
practices and the dissemination of ideas and information that will benefit farmers, including
information about the water use program. MSUE could be engaged to work with the irrigators on
the misconception about water rights. The state defines water as a state trust and protects it from
national level interests with implementation of the Compact requirements. The irrigators benefit
from this interest and thus the regulated riparian status of water rights in Michigan. Clarifying

87

this information and presenting it to irrigators as a reason for compliance with regulations at the
state level could be the focus of MSUE programs.
Education about the water withdrawal assessment process and users’ rights to challenge
its findings can also help to create willingness to cooperate with rules about water use. By
building trust in the WWAT, the DEQ, through MSUE, can promote it as a screening tool and
dispel the belief that it is an inaccurate decision tool. Workshops on regulated riparianism and
the assessment process as two components used to protect the waters of the state could be offered
to irrigators through MSUE. In particular, targeting leaders in the community could broaden
dissemination of the information shared.
In addition to educating irrigators on their rights and responsibilities for using water in
Michigan, the DEQ, with the help of MDARD, could make efforts to identify unregistered users.
By doing so, boundaries will be more well-defined when water user committees are formed.
Everyone who should be engaged in the negotiation process will be engaged. Once unregistered
users are identified, the DEQ can also require them to register their use as new capacity in the
WWAT and to comply with annual reporting laws. DEQ resources could be used most wisely if
the catchments targeted in looking for unregistered users are those where zone C LQWs have
already been registered. Those are the areas that are at highest risk of ARIs. Once all LQW
appropriators are identified, the DEQ and MSUE can start to build capacity for negotiations by
informing users of the status of the water resource in their catchment and who they may have to
work with in the future on collaborative efforts to reduce water use.
Because agricultural processing companies that contract with growers are so influential in
water use decisions in the area, these companies must also be informed of their responsibilities as
users of Michigan water. Many businesses are working towards sustainable practices as

88

consumers demand environmental stewardship in supply chains. DEQ could encourage the
companies to work with farmers to conserve water and support environmental health in the Great
Lakes Basin, noting that such efforts could actually benefit their businesses when that
information is shared with consumers.
It may be necessary for the DEQ to bear significant administrative costs in implementing
these suggestions. However, they will increase the likelihood of success for the water use
program by decreasing transaction costs for irrigators.

5.4 Other Large Quantity Water Users
One last element not addressed by this research is the presence of other large quantity
water users in the State of Michigan. Others include, but are not limited to, municipalities,
industries, golf courses, and ski resorts. The dynamics of water user committees which include
these types of users, in addition to agricultural irrigators, may result in even higher transaction
costs in negotiation. The irrigators discussed their feelings that they were blamed for drops in
surface water levels when there are many other water users in the state contributing to ecosystem
damage. Many participants said they expect that negotiating with these water users would be
very costly and would result in conflict. These dynamics also need to be explored if the efficacy
of the water use program as a means for conservation is to be enhanced.
The results of this research show that there will be high transaction costs in creating
required institutions for collective water management among irrigators in Southwest Michigan.
Some opportunities for reducing transaction costs exist, and the DEQ can use those to begin to
overcome barriers. However, research on dynamics of negotiations that involve all types of large

89

quantity water users in Michigan is needed to further reduce transaction costs in the
implementation of the water use program.

90

APPENDICES

91

Appendix A: Recruitment Script

Name of person ____________________
Phone number _____________________
Time called ________________________
Better time to call __________________
Hello, Mr. ________. My name is Rachel Ford and I’m a graduate student at Michigan State
University. We are working with colleagues at MSU, MI Department of Agriculture, and MI
Department of Environmental Quality to better understand what Michigan’s water use program
means for agricultural irrigators. We got your name from MDA’s list of farmers who report
irrigation water use for ____ County. We are looking for irrigators to help us with our research.

You do farm, right? And irrigate?

We are getting together a small group of farmers who irrigate crops to give us input on how
irrigators might respond to water scarcity and how irrigators might work together to overcome
water access problems. We plan to bring about seven farmers together to talk about this topic and
help us understand irrigators’ opportunities and constraints.

We are going to do this three times, so we have three possible dates in mind. We will meet at the
St. Joseph County ISD building at 6:30 p.m. and the discussion will last about two hours. We’ll
have refreshments, and we will have $50 for you as a thank-you for your time and help.

The three dates are ____, _____ and _____.
Would you be able to join us for one of those meetings?

92

NO ------- Okay. Thanks for your time. Good evening.
YES --- Great. Which of the dates would you prefer?
I’d like to send you a letter just to confirm everything. I have ____________________________
for your mailing address. Is that correct?
Great. I’ll send you the letter with the details, and we look forward to seeing you on ____ date.

93

Appendix B: Focus Group Script

Agricultural Irrigator Focus Groups: St. Joseph County ISD

Introduction: My name is Pat Norris, and I am a professor at Michigan State University. This is
Rachel Ford; Rachel is a graduate student in Agricultural, Food and Resource Economics at
MSU.

First, thank you all for agreeing to be here this evening to help us with our research.

Second, let me remind each of you that you don’t have to answer any questions that you don’t
wish to answer, and you are free to leave at any time. We don’t think that any of our questions
should make anyone uncomfortable, but of course we cannot be aware of all circumstances that
may be related to our questions.

Third, as we told you when we first contacted you, our research is part of a larger effort looking
at the implications of Michigan’s water law. We don’t have an opinion of whether the law is
good or bad or whether it is or is not needed. In fact, for the purposes of this discussion, we will
not really focus on the law until near the end. What we are really interested in, for this specific
research, is how irrigators may work with one another to manage water resources, especially in
situations where availability of water may be limited for one reason or another. That is what
most of our questions will focus on.

94

There are no right or wrong answers to any of our questions. We expect that you will have
differing points of view. Please feel free to share your point of view even if it differs from what
others have said.

We won’t include any of your names in any of our reports. We do have name tents for each of
you. That is to help me remember names during our conversation, but it will also help you. You
don’t have to direct all of your comments to me. If you want to follow up on something that
someone has said, or if you want to agree or disagree or give an example, please feel free to do
that. Feel free to have a conversation with one another about these questions. My role here is to
ask questions, listen and make sure everyone has a chance to share.

We’re interested in hearing from each of you. So if someone is talking a lot, I may ask you to
give others a chance. And if you aren’t saying much, I may try to draw you more into the
conversation. Again, you don’t have to answer, but we do want to make sure all of you have a
chance to share your ideas.

We are recording because we don’t want to risk losing any of the richness of the conversation.
We will have a professional transcriptionist produce a typed copy of the tape’s contents, and
each participant will be identified with a code number. We will not share any information that
could be connected to a specific individual during any of our analysis or in any of our reports.

95

We will finish at 8:30. Feel free to get up and get more refreshments at any time if you would
like. If you have a cell phone, please set it to vibrate, and if you need to answer it you can step
outside.

Let’s get started.

Opening Question
First, let’s find out more about each other by going around the table one at a time. Tell us your
name and one thing that you particularly appreciate or enjoy about living in Michigan.

1. Introduction Questions:
Do you use surface water, groundwater or both for irrigation?

2. Transition Questions:
For those of you using surface water, do you feel you have a good idea of how many
other irrigators rely on the same stream reach for irrigation water?

What about those of you using wells. Do you have a sense of how many other irrigators
rely on wells or surface water for irrigation on neighboring farms?

How much information do you have about other kinds of large quantity withdrawals in
your area, such as municipalities or industries?

96

3. Key Questions:
For this question, think about where you draw your water from, and think about others
who are drawing water from the same source. Have you ever changed your crop rotation
so that any shared water source would meet both your needs and the needs of others?

Have you ever changed your management in any season so that you needed less water?
For example, planting an early crop of potatoes or cutting an early harvest of alfalfa.
Other examples of changing management?
Did you do it because there was less water for you? Or because you knew there
were other irrigators also relying on a stream or on groundwater?

If water levels were low, perhaps because of a drought, would you consider formally
cooperating with other irrigators to get everybody through a season?
Other irrigators would include those using surface water and groundwater.

Earlier we listed a few ways in which an irrigator might change management during a
season in order to need less water. Are there any other kinds of management changes that
you think irrigators might consider in order to use less water so that everyone has access
during a drought?

What kinds of barriers do you think would make it difficult for you and other irrigators to
work together in order to collectively use less water in times of drought?

97

Would your willingness to work with others to collectively use less water be affected by
whether you were asked by a neighbor or asked by a state agency?
Why or why not?

If the state contacts irrigators for such a purpose, only those irrigators who have
registered would be contacted since the state would have no way of knowing about other
irrigators. How do you think this would affect opportunities to collectively use less
water?

So far, my questions have focused on ways in which you might work with other irrigators to
share limited water in the event of a drought. Now I want to bring the state’s water law into the
equation.

Under the new law, riparian landowners have a right to withdraw water from a stream, and
landowners have a right to use well water. However, collectively, all water users in a watershed
cannot withdraw so much water that a stream is unable to support fish populations. That is the
state's indicator of a negative ecological impact. That means that a situation could arise in which
a new irrigator wants to withdraw water from a stream or well but so much water is already
being used that a new withdrawal would reduce stream flow too much. Technically, in such a
case the only way that new irrigator could legally withdraw water would be if others in the
watershed use less.

4. Ending Questions:

98

If a neighbor approached you and asked if you would cooperate with other irrigators to
free up water so that he or she could begin irrigating, how do you think you would
respond?
o How is this situation similar to or different from the kinds of situations we were
discussing earlier?
Are there any other points or concerns you would like to raise that you were not able to
express earlier about any of the questions?

Thank you very much for your help with this research.
You have your copy of the consent form, and it has my contact information on it. I will also give
each of you one of my business cards. If you have any questions about this research or other
questions about water use research, please feel free to call me. If it is a question that I can't
answer, I will try my best to find someone who can.

Rachel will have the $50 we promised you that you can collect on the way out. There is one final
form you need to sign to verify that you received the payment.

99

Appendix C: Individual Interview Script

Thank you for agreeing to meet with me. This will be helpful to fill in some information gaps in
my thesis, things I am curious about that were not addressed during the focus groups last spring.
Are you familiar with Michigan’s water use program that attempts to identify water withdrawals
with a high risk of negative resource impact on local streams? And when such an event occurs,
the DEQ has a responsibility to convene large water users involved to create a plan for collective
reduction of water resources?
My research centers around anticipating barriers to collective decision-making during these
water user committee meetings and successful management plans that would come out of them.
Researchers have identified conditions that contribute to successful collective management rules
and I am trying to comment on the existence of those contributors in the group of irrigators in
Southwest Michigan. The following questions will hopefully help me to fully describe the area in
terms of the list of conditions. Do you have any questions?

1. Do you think agricultural irrigators in this area see themselves as a community of water
users or as completely independent operators?

2. During our focus groups, participants talked about some strategies they used for water
conservation such as soil moisture monitoring programs and watering at night. But there
were also times when participants said their neighbors irrigated constantly for months at a

100

time, which they seemed to think was unnecessary. Do you think there is a commonly
held idea in this area about how water should be used?

3. One contributing condition to cooperative management is that there is leadership within
the group that is informed of the issues surrounding the resource and decisions about its
use. Do you think that leadership exists among irrigators in this area? Do you think there
are farmers whose opinions and influence are more respected than others?

4. Do you think there is any sense of obligation to other irrigators or examples of times
when one irrigator might rely on another irrigator? Can you think of examples of shared
infrastructure, things like that?

5. Do you think the relationships between neighboring farmers differ from those between a
local farmer and a renting farmer? How?

6. Are there large differences in terms of financial prosperity in this area?

7. Can you think of any situations that would cause a farmer to drastically increase his or
her consumption of water?

8. One of the conditions that would significantly contribute to a successful collective action
plan is that the participants have some kind of background experience with an
organization. Do you think there are organizations that exist that would help with the

101

process of creating a water user committee? Examples of these include village councils,
irrigation associations, and farmer’s associations.

9. During the focus groups we talked about unregistered users. Do you think people would
report their unregistered neighbors in order to keep their established water use? What are
some of the reasons they might do this? What are some of the reasons they wouldn’t do
this?

Well those are all of the questions I have for you. Thank you very much for helping me again,
I’ve brought you a hat for your time.

102

Appendix D: References to Transcripts for Topics in Chapter Three

Frustration: A1 477-86, A1 1198-1210, A1 1192, A1 1196, A2 1216-45, A2 573-74, A1 116869, A1 1191-92, A1 681-82, A3 457, A3 501-02, A1 696-703, A1 707-10, A3 275-81, A3 30003, A1 743, A1 733-39, A1 761-62, A2 148-59
Distrust: A1 1124-25, A1 1098-99, A2 757-61, A2 664-65, A3 598-600, A3 659-62, A3 931, A1
516-18, A3 850-54, A1 717-27, A1 1173-75, A3 515-17, A2 663-72
Heterogeneity: A3 358-59, A1 433-35, A1 586-89, A2 611-15, A3 675-81, A2 255-64, A2 27273, A3 420, A1 222-23, A2 186-87, A2 287-88, A2 360, A2 810, A2 812-14, A2 366-68, A3
422-27, A2 342-45, A3 439-40, A2 354, A2 370-72, A1 943-50, A1 965-67, A1 1042-46, A2
366-67, A2 796-99, A1 267-71, A1 606-09, A1 810-16, A1 778-79, A1 805-06, A3 693-95, A1
568-69, A2 189-90, A2 200-07, A2 216-20, A2 222-24, A2 245-47, A2 294-95, A2 313-14, A1
360-61, A2 480, A1 368, A1 384-85, A1 370
Group Identity: A2 335-36, A3 305-07, A1 323, A3 321-22, A2 148, A2 526-28, A1 423, A3
684-85, A2 416-20, A2 426-27, A2 437-41, A2 441, A2 445-47, A1 591-96, A1 637-38, A1 642,
A1 646-47, A3 705-09
Anti-Diversion: A1 847-48, A1 836-842, A2 690-96

103

REFERENCES

104

References

Agrawal, Arun. 2001. Common Property Institutions and Sustainable Governance of Resources.
World Development 29(10): 1649-1672.
Annin, Peter. 2006. The Great Lakes Water Wars. Washington D.C.: Island Press.
Baland, Jean-Marie and Jean-Philippe Platteau. 1999. The ambiguous Impact of Inequality on
Local Resource Management. World Development 27(5): 773-788.
Baland, Jean-Marie and Jean-Philippe Platteau. 1996. Halting Degradation of Natural Resources:
Is there a Role for Rural Communities? Rome: Food and Agricultural Organization of the
United Nations.
Bromley, Daniel. 1992. The Commons, Property, and Common-Property Regimes. In Making
the Commons Work: Theory, Practice, and Policy, ed. Daniel W. Bromley. San
Francisco: ICS Press.
Ciriacy-Wantrup, S.V., and Richard C. Bishop. 1975. “Common Property” as a Concept in
Natural Resources Policy. Natural Resources Journal 15: 713- 727.
Council of Great Lakes Governors. 2001. Overview of the Charter Annex: Great Lakes Charter
Annex and Future Agreements. http://www.cglg.org/projects/water/links.asp (accessed
October 2011).
Council of Great Lakes Governors. 2005a. Great Lakes-St. Lawrence River Basin Sustainable
Water Resources Agreement. http://www.cglg.org/projects/water/index.asp (accessed
October 2011).
Council of Great Lakes Governors. 2005b. Great Lakes – St. Lawrence River Basin Water
Resources Compact. http://www.cglg.org/projects/water/index.asp (accessed October
2011).
Deyoung, Dave. Large Quantity Water Withdrawal Registrant Notification Requested
Summaries. 2011. Michigan Department of Environmental Quality.
<http://www.michigan.gov/deq/0,1607,7-135-3313_3684_45331-217795--,00
.html>. (accessed November 2011).
Dellapenna, Joseph, W. 1994. The Regulated Riparian Version of the ASCE Model Water Code:
The Third Way to Allocate Water. Water Resources Bulletin, American Water Resources
Association 20:107-204.
Gardner, Roy, Elinor Ostrom, and James M. Walker. The Nature of Common-Pool Resource
Problems. 1990. Rationality and Society 2(3): 335-358.
Groundwater Conservation Advisory Council. 2007. Report to the Michigan Legislature in
response to 2006 Public Act 34. http://michigan.gov/deq/0,4561,7-135-3313_41033--,00.html (accessed October 2011).
105

Hamilton, David A., and Paul W. Seelbach. 2010. Determining Environmental Limits to
Streamflow Depletion Across Michigan. http://www.miwwat.org/wateruse/
regulations.asp (accessed November 2011).
Hamilton, David A., and Paul W. Seelbach. 2011. Michigan’s Water Withdrawal Assessment
Process and Internet Screening Tool. State of Michigan Department of Natural
Resources, Fisheries Division Special Report 55, Lansing.
Hardin, Garrett. 1968. The tragedy of the commons. Science 162: 1243-1248.
Kaplowitz, Michael D. and John P. Hoehn. 2001. Do focus groups and individual interviews
reveal the same information for natural resource valuation? Ecological Economics
36(2001): 237-247.
Kiser, Larry L, and Elinor Ostrom. 2000. The Three Worlds of Action: A Metatheoretical
Synthesis of Institutional Approaches. In Polycentric games and Institutions: readings
from the workshop in Political Theory and Policy Analysis, ed. Michael D. McGinnis,
chapter 2. The University of Michigan Press.
Kleweno, David D. 2010. Michigan 2009-2010 Highlights. National Agricultural Statistics
Service, Michigan Field Office, Michigan Department of Agriculture.
Knudsen, Bill, and H. Christopher Peterson. 2009. Second Interim Update on the Economic
Impact of Michigan’s Agri-Food and Agri-Energy System. Michigan State University
Product Center for Agriculture and Natural Resources. http://www.productcenter.msu.
edu.productcenter/research (accessed December 2011).
Krueger, Richard A. and Mary Anne Casey. 2009. Focus Groups: A Practical Guide for Applied
Research, 4th ed. Thousand Oaks, CA: SAGE Publications, Inc.
Lusch, David P. 2011. An Overview of Existing Water Law in Michigan Related to Irrigation
Water Use and Riparian Considerations. Institute of Water Research, Michigan State
University. Accessed from the author.
McKean, Margaret A. 1992. Management of Traditional Common Lands (Iriaichi) in Japan. In
Making the Commons Work:Theory, Practice, and Policy ,ed. Daniel W. Bromley. San
Francisco: ICS Press.
Michigan Department of Agriculture and Rural Development. 2009. Michigan Food and
Agricultural Systems Profiles. http://www.michigan.gov/mdard/0,4610,7-125-1568220573--,00.html (accessed October 2011).
Mueller, Dennis C. 2003. Public Choice III. Cambridge, UK: Cambridge University Press.
Naughton-Teves, Lisa and Steven Sanderson. 1995. Property, Politics and Wildlife
Conservation. World Development 23(8): 1265-1275.
Nicholson, Walter and Christopher M. Snyder. 2008. Microeconomic Theory: Basic Principles
and Extensions Tenth Edition. Mason, OH: Thomson South-Western.
106

Oakerson, Ronald J. 1992. Analyzing the Commons: A Framework. In Making the Commons
Work: Theory, Practice, and Policy ,ed. Daniel W. Bromley. San Francisco: ICS Press.
Olson, Mancur. 1971. The Logic of Collective Action. Cambridge, Mass: Harvard Press.
Ostrom, Elinor. 1985. The Rudiments of a Revised Theory of the Origins, Survival, and
Performance of Institutions for Collective Action. Working paper, Workshop in Political
Theory and Policy Analysis, Indiana University.
Ostrom, Elinor. 1990. Governing the Commons: The evolution of institutions for collective
action. Cambridge, UK: Cambridge University Press.
Ostrom, Elinor. 1992. The Rudiments of a Theory of the Origins, Survival, and Performance of
Common Property Institutions. In Making the Commons Work: Theory, Practice, and
Policy, ed. Daniel W. Bromley. San Francisco: ICS Press.
Ostrom, Elinor, and James M. Walker. 1989. Communication in a Commons: Cooperation
without External Enforcement. Digital Library of the Commons Conference Paper:
American Political Science Association Meetings, Panel on Freedom and Organization:
Atlanta, GA.
Ostrom, Elinor and Roy Gardner. 1993. Coping with Asymmetries in the Commons: SelfGoverning Irrigation Systems can work. Journal of Economic Perspectives 7(4): 93-112.
Postel, Sandra. 1999. Pillar of Sand: Can the Irrigation Miracle Last? New York: Worldwatch
Institute.
Preckel, Paul V., Gerald E. Shively, Timothy G. Baker, Mei-Chin Chu, and Jessica Eide
Burrel. 2000. Contract Incentives and Excessive Nitrogen Use in Agriculture. Journal of
Agricultural and Resource Economics 25(2): 468-484.
Reeves, Howard W., David.A. Hamilton, Paul W. Seelbach, and A. Jeremiah Asher. 2009.
Ground-water-withdrawal component of the Michigan water-withdrawal screening tool:
U.S. Geological Survey Scientific Investigations Report 2009-5003.
Robison, Lindon, A. Allan Schmid, and Marcelo E. Siles. 2002. Is Social Capital Really Capital?
Review of Social Economy 60: 1-21.
Robison, Lindon J., and Bryan K. Ritchie. 2010. Relationship Economics: The Social Capital
Paradigm and its Application to Business, Politics, and Other Transactions. Great
Britain: MPG Book Groups, UK.
Runge, Carlisle Ford.1981. Common Property Externalities: Isolation, Assurance, and Resource
Depletion in a Traditional Grazing Context. American Journal of Agricultural Economics
63:595-606.
Runge, Carlisle Ford. 1984. Institutions and the Free Rider: The Assurance Problem in
Collective Action. The Journal of Politics 46(1): 154-181.

107

Runge, C. Ford. 1992. Common Property and Collective Action in Economic Development. In
Making the Commons Work:Theory, Practice, and Policy ,ed. Daniel W. Bromley. San
Francisco: ICS Press.
Sabatier, Paul A. 1991. Toward Better Theories of the Policy Process. PS: Political Science and
Politics 24(2): 147-156.
Schlager, Edella, William Blomquist, and Shui Yan Tang. 1994. Mobile Flows, Storage, and
Self-Organized Institutions for Governing Common-Pool Resources. Land Economics
70: 294-317.
Schmid, A. Allan. 2004. Conflict and Cooperation: Institutional and Behavioral Economics.
Oxford, UK: Blackwell Publishing Ltd.
State of Michigan Public Act 148 Groundwater Inventory and Map (GWIM). 2005. Executive
Summary. http://gwmap.rsgis.msu.edu/ (accessed October 2011).
State of Michigan Public Act 148 Groundwater Inventory and Map (GWIM). 2006. Technical
Report. http://gwmap.rsgis.msu.edu/ (accessed October 2011).
Sugden, R. 1982. On the Economics of Philanthropy. Economic Journal 92: 341-50.
Taylor, Michael. 1987. The Possibility of Cooperation. Cambridge: Cambridge University Press.
Taylor, Steven J. and Robert Bogdan. 1998. Introduction to Qualitative Research Methods: A
Guidebook and Resource, 3rd ed. New York: John Wiley & Sons, Inc.
United States Department of Agriculture 2007 Census Publications: 2008 Farm and Ranch
Survey. http://www.agcensus.usda.gov/Publications/2007/Online_Highlights/Farm_and_
Ranch_Irrigation_Survey/index.asp (accessed on December 2011).
United States Department of Agriculture 2007 Census Publications, Michigan: Volume 1,
Chapter2. http://www.agcensus.usda.gov/Publications/2007/Full_Report/Volume_1,_
Chapter_2_County_Level/Michigan/index.asp (accessed November 2011).
Wade, Robert. 1988. Village Republics: Economic conditions for collective action in South
India. Oakland: ICS Press.
Zorn. T.G., Paul W. Seelbach, Edward S. Rutherford, Todd C. Wills, Su-Ting Cheng, and
Michael J. Wiley. 2008. A regional-scale habitat suitability model to assess the effects of
flow reduction on fish assemblages in Michigan streams. Michigan Department of
Natural Resources, Fisheries Research Report 2089, Ann Arbor.

108