Citation
A new institutionalist examination of Colorado's community-based land trusts

Material Information

Title:
A new institutionalist examination of Colorado's community-based land trusts
Creator:
Wunderlich, Karl Arthur
Publication Date:
Language:
English
Physical Description:
210 leaves : ;

Subjects

Subjects / Keywords:
Land trusts -- Colorado ( lcsh )
Institutional economics ( lcsh )
Transaction costs ( lcsh )
Institutional economics ( fast )
Land trusts ( fast )
Transaction costs ( fast )
Colorado ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 197-210).
General Note:
School of Public Affairs
Statement of Responsibility:
by Karl Arthur Wunderlich.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
54436527 ( OCLC )
ocm54436527

Full Text
A NEW INSTITUTIONALIST EXAMINATION OF COLORADO'S
COMMUNITY-BASED LAND TRUSTS
B.S., Virginia Polytechnic Institute and State University, 1984
M.S., Colorado State University, 1988
M.A., University of Colorado, 1993
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Affairs
2002
by
Karl Arthur Wunderlich


2002 by Karl Arthur Wunderlich
All rights reserved.


This thesis for the Doctor of Philosophy
degree by
Karl Arthur Wunderlich
has been approved
by
IS ^
Date
in


Wunderlich, Karl Arthur (Ph.D., Public Affairs)
A New Institutionalist Examination of Colorado's Community-Based Land
Trusts
Thesis directed by Professor Peter deLeon
ABSTRACT
This dissertation examines factors that dictate effectiveness among
Colorado's community-based land trusts. Land trusts are defined as nonprofit
organizations whose primary mission is to protect lands held in private
ownership. Protecting natural amenities that are valued by the public on land
that remains in private ownership make land trusts unique in the area of natural
resource management.
The research is framed within a new institutional economics framework
that argues that institutions generally evolve in manners that minimize
transaction costs. Property is an institution, a collection of formal and informal
rules defined in terms of how society wishes to manage its resources.
Organizations function within the rules dictated by an institution. Land trusts
are organizations that are emerging as a potent solution to protecting valuable
natural assets that exist within private property. Therefore, land trusts warrant
IV


examination in terms of their ability to minimize transaction costs or,
alternatively, to summon transaction resources.
These transaction resources, categorized as social capital, concentric
regimes, and political entrepreneurs, are hypothesized as contributing to
effectiveness among land trusts. A survey of twenty-seven land trusts in
Colorado was examined in terms of their relative ability to convert these
transaction resources into effective land trust actions. The single dependent
variable, effectiveness, was defined as the number of activities completed per
year. Activities include conservation easements, fee-simple purchases, and
third-party transfers. The data were analyzed using traditional descriptive and
inferential techniques as well as a combinatorial technique, fuzzy set qualitative
comparative analysis.
Results indicated significant associations between the dependent
variable, effectiveness, and independent variables within the categories of social
capital and concentric regimes. Weak or nonexistent associations were found
with regard to political entrepreneurs. Other exogenous factors were examined
as well. Effective land trusts were found where public sector open space
programs were absent. There was no significant association between
v


effectiveness and measures of threat from increasing population, housing
development, or loss of agricultural lands. Conclusions suggest that land trusts
that have active community education and outreach programs, do not actively
solicit landowners to donate easements, and receive external government
support are most effective.
This abstract accurately represents the content of the candidate's thesis. I
recommend its publication.
SignedJ
Peter deLeon
vi


ACKNOWLEDGMENT
This may well be one of the most enjoyable parts of any dissertation. It
suggests that there quite possibly may be an end to this particular stage of a
seemingly endless process, it allows me to write in the first person, and, best of
all, I can recognize all those that have provided the support, guidance, reflection,
and inspiration without which this effort would have been nothing but a passing
thought. First, I would like to thank all of the members of Colorado's land trust
community that gave me their time, answered endless questions, and assured me
that caring about Colorado's landscapes is a worthy sentiment. I hope the
relationship continues. I would also like to recognize the financial support of
the Tim Wirth Chair in Environment and Community Development, Resources
for the Future, and the Lincoln Institute of Land Policy.
My dissertation committee has provided unique support and each
deserves special recognition. From my first semester, Peter deLeon set the
standard, guided me through the process, and occasionally, when needed, pulled
sharply on the chain. A very special thank you to an intellectual inspiration and
mentor. Linda deLeon gave the best advise of all: "write about something for
which you are passionate." I did. Jae Moon patiently kept an eye on my


analytic detail.. .and I suspect I tested his vision. Jane Ellen Hamilton
volunteered her time from driving air over the state overseeing Colorado's land
trusts to keep me grounded, answer obscure questions at what always seemed to
be busy times, and always found a moment for words of encouragement. Edella
Schlager offered both theoretical and analytical guidance with a confident,
gentle, firmness that only comes to those who are truly genuine. Finally, special
thanks to Toddi Steelman whose research on community and natural resource
policy provided the early seeds for this dissertation. Her support, only exceeded
by her patience, allowed me to pursue the ideas within these pages. This is a
wonderful committee. I got lucky.
Special gratitude and appreciation must go to those people that know me
in my highs and lows, when I have shined and when I have been more than a
little tarnished. I must begin by thanking Phyllis Resnick whose friendship over
a decade of disturbingly similar graduate school paths has blossomed into so
much more. My gratitude for your patience, sacrifice, and support through this
period cannot be put into words. Thank you for being such a special part of my
life. And, lest I forget, thanks for the help with those Kuhn-Tucker conditions!


Finally, a special thank you for my family. My parents established the
highest priority in our family to education. Then, they backed it up with years, a
lifetime really, of support and example to both my sister and myself. My
mother's tireless work ethic still astounds me. To say she sets an intimidating
standard is no exaggeration. My father is one of only a handful of truly
powerful intellects that I have had the privilege to know. His thoughts inspire
me. I write this not out of familial bias but because he has proved it for decades
in spoken word, written word, and, occasionally, by not saying anything at all.
This stage of my education has been a long haul. Thanks, again.
Thanks to you all.


CONTENTS
Tables.......................................................xiii
CHAPTER
1. INTRODUCTION............................................... 1
Dissertation Overview.........................;.......6
2. LAND TRUSTS AND CONSERVATION EASEMENTS.......................8
Status and Description of Land Trusts.....................8
History of Land Trusts................................11
Conservation Easements...................................15
Conclusion...............................................20
3. REVIEW OF LITERATURE................:....................22
Open Space as Common Pool Resource and Public Good.......22
Loss of Open Space as a Collective Action Problem........24
Property as a Continuum................................ 28
Institutions, Organizations, and Institutional Evolution.31
The Role of Transaction Costs........................... 37
Transaction Resources................................... 42
Social Capital.....................................44
x


Concentric Regimes.................................48
Political Entrepreneurs............................51
Resource-Specific Factors..........................53
Summary...................................................54
4. RESEARCH DESIGN AND ANALYTIC METHODS........................58
Variables.................................................60
Research Hypotheses.......................................62
Data Collection..........................................68
Combinatorial Analysis....................................70
Qualitative Comparative Analysis..........................74
Fuzzy Set Qualitative Comparative Analysis................80
Fuzzy Analysis of Land Trust Effectiveness............... 83
Summary...................................................88
5. RESULTS..................................................... 90
A Picture of Community Land Trusts in Colorado............91
Social Capital and External Support.......................95
Exogenous Factors.........................................98
Statistical Inference and Tests of Significance..........100
Combinatorial Analysis...................................109
Testing for Necessity....................................114
xi


Testing for Sufficiency...........................118
6. CONCLUSIONS AND IMPLICATIONS....................!....123
Descriptive Results and Tests of Significance.....126
Combinatorial Relationships..................... 137
Reflections on Contributing Theory................143
Social Capital..............................146
Concentric Regimes and Co-Management........150
Political Entrepreneurs.................... 152
Research Limitations..............................154
Applications for Land Trusts......................156
Future Research...................................158
Last Thoughts.....................................163
APPENDIX
A. INTERVIEW INSTRUMENT................................167
B. CONTACT LETTER......................................182
C. CONSENT FORM........................................185
D. FUZZY SET QUALITATIVE COMPARATIVE
ANALYSIS OUTPUT......................................188
WORKS CITED...................................................197
Xll


TABLES
Table
2.1 Types of Land Protected by Land Trusts Across the United States......9
2.2 Land Trust Activities...................................:............11
2.3 Number of Land Trusts................................................13
4.1 Hypothetical Truth Table Showing Collapse of Military Regimes........79
5.1 Descriptive Statistics...............................................91
5.2 Colorado Community Land Trust Activity in 2000.......................93
5.3 Primary Goals of Colorado's Community Land Trusts.................;... 94
5.4 Results of Bivariate Regressions as Tests of Significance...........103
5.5 Independent Variables Selected to Explain Effectiveness.............108
5.6 Fuzzy Set Data Matrix...............................................114
5.7 Test of Necessity................................................. 116
5.8 Test of Necessity with Negated Dependent Variable...................118
xiii


CHAPTER 1
INTRODUCTION
The institutions that manage natural and cultural resources reflect the
ideologies of how we as a society choose to govern ourselves. As Neil Komesar
(1994, 3) declares, it is about these decision making processes. More
particularly, it is about the way we choose to allocate authority. It is about
deciding who decides. In an age of devolving government responsibility, this
question is all the more relevant. Tightening federal budgets and the doctrine
(and practice) of New Federalism combined with an increasing awareness of
"place" among communities has led to more localized efforts of environmental
protection (Lester 1990). Institutions such as private property or state-based
governance have been advanced by various ideological camps that attempt to
manage natural resources but none alone seems adequate at achieving the
multiple objectives of promoting long-term sustainability of the resource,
engaging community level action, and surviving the social, economic, legal, and
technical complexities inherent in a neo-liberal, capitalist, and technically
advanced culture. Yet, these are precisely the conditions that must be met to
successfully manage the long-term sustainability of natural and cultural heritage
in this country.
1


This dissertation examines factors that dictate the effectiveness of .
community-based land trusts in Colorado. Land trusts are non-profit
organizations committed to the protection of open space, habitat, biotic
resources, and cultural amenities. Private, legally incorporated nonprofit
organizations, land trusts hold and manage community land conservation
easements, offering an alternative to government ownership" (Poole 1993, 69).
They come in all shapes and sizes. Some are national organizations with state or
local chapters, large budgets and membership drives. Others are run from the
kitchen table of a founding citizen. This dissertation focuses specifically on
local or community-based land trusts as opposed to regional or national land
trusts. The latter trusts tend to have more extensive resources, greater public
recognition, and a more global perspective. Community-based land trusts are
especially worthy of study because these organizations play a recognized role
within a specific community to resolve a natural resource problem (Endicott
1993). To what degree do community-based land trusts contribute to
operationalizing local knowledge and values and creating, if ever so slightly,
what Fisher (2000, 28) calls a "return to grass roots democracy"? Understanding
factors that may lead to the effectiveness of these rapidly emerging mechanisms
for natural resource management is critical to assessing the juxtaposition of roles
2


between citizen participation, the state, and the broad institution of property with
regard to environmental protection and ecosystem sustainability.
Land trusts are compelling for a number of other reasons. While some
land trusts actually purchase land to protect it from development or other uses
deemed undesirable, by far the fastest growing tool used by land trusts is the
conservation easement. A conservation easement "is a legally binding
agreement that permanently restricts the development and future use of the land
to ensure protection of its conservation values" (Gustanski 2000, 9). The
conservation easement separates rights of conservation from the broad,
possessory interest associated with fee-simple ownership. A non-profit
organization (the land trust), acting in the public interest, can hold the rights of
development to a property without taking full ownership of the land. The
extensive use of the conservation easement is what makes the explosion of land
trusts a particularly research-worthy evolution in the protection of land and
natural resources. By preserving land with conservation easements, land trusts
are broadening and changing one of the fundamental institutions for managing
natural resources, the institution of property.
3


Conservation easements, and the land trusts that hold them, should
engage scholarly interest for another reason as well. Land and natural amenities,
such as those protected by conservation easements, represent a public good.
Open space, riparian habitat, migration corridors and other natural amenities
share to a great degree the conditions of nonrivalry and nonexcludability that
define a public good (Comes and Sandler 1996). Often (but not always), public
goods are provided and/or managed by a public sector organization, such as a
federal, state, or local government. Private goods are owned and managed by
private sector organizations or individuals. Land trusts demonstrate a blending
of these sectors and institutions, by which citizen-driven private organizations
(albeit with non-profit status) are providing public goods by separating the rights
of conservation from possessory interest in land. Importantly, this process has
been facilitated primarily by a federal tax code that creates an incentive, and
potentially a subsidy, for private landowners to enter into conservation
easements.
So why examine land trusts in Colorado? Colorado is the state with the
fastest growing number of land trusts in the Southwest region which is the
second fastest growing region in the nation. Colorado has the tenth largest
number of land trusts in the country (Land Tmst Alliance 2000). While land
4


trusts have existed in the New England area since the turn of the 20th century,1
they are relatively new to the West where issues of congestion, sprawl, and open
space preservation have emerged only within the last few decades. What makes
an investigation of land trusts in the West unique is the surrounding political
climate that is inherently skeptical of government and staunchly supportive of
private property rights. Land trust activity often focuses on areas on the fringes
of urban development. These areas tend to be owned by long-term residents
with agricultural roots and interests, precisely those that fear an erosion of their
property rights by public sector involvement.
The intellectual framework that guides this research is found in the new
institutionalism (also called new institutional economics) literature. Since land
trusts are observed to be organizations participating in an evolution of the
institution of property, it seems reasonable to examine their effectiveness from
within a framework of new institutional economics. New institutionalism
defines institutions as formal and informal rules and organizations as the
players within those rules (North 1990). Furthermore, much of new
institutional economics argues that institutions evolve along conditions that
1 This is not entirely surprising given New Englands relative shortage of available open space
near urban areas combined with their history in strongly democratic forms of governance, e.g.
town hall meetings.
5


minimize the transaction costs of achieving a particular transaction. Therefore,
effective organizations can be examined as those that are best able to minimize
transaction costs given the rules of the property game.
This dissertation examines how land trusts go about engaging in land
preserving actions from the perspective of new institutional economics. How, in
other words, do land trusts avoid or minimize transaction costs? What
distinguishes relatively effective land trusts from those that are less effective?
The answer to these questions lies within a number of somewhat (but far from
entirely) overlapping literatures. A single theoretical perspective simply will not
do when considering the broad variation of characteristics that describe land
trusts and the communities they serve. Some mix of the communitarian
elements of common property theory and social capital theory juxtaposed with
new federalism and political entrepreneurship may help explain effectiveness
among some land trusts relative to others.
Dissertation Overview
This dissertation proceeds as follows. First, a discussion of land trusts in
general, including their history and current status and the nature of conservation
easements is provided. The next chapter presents the conceptual framework and
6


supporting literatures. Afterwards, data collection and analytic methods are
discussed in terms of explaining effectiveness among Colorados local land
trusts. Analysis and discussion of the data follows, ending with implications and
conclusions, a summary, and a peek at future research opportunities.
7


CHAPTER 2
LAND TRUSTS AND CONSERVATION EASEMENTS
Status and Description of Land Trusts
Land trusts are private, nonprofit organizations that work with property
owners to protect open land through direct, voluntary land transactions as an
alternative to government ownership of land (Poole 1993; Hocker 1996).
Cheever (1996) considers land trusts to be the most active and forward-looking
element in the national effort for environmental preservation. There are,
however, few features that are consistent across all land trusts. They vary in size
from large, national organizations such as The Nature Conservancy and The
Trust for Public Lands, to the small, community based organizations that are the
focus of this dissertation. Some have budgets in the millions of dollars. Some
function on $10,000 a year and an all-volunteer staff. Certainly, land trusts are
not a homogenous group.
Land trusts vary in mission as well. Some are committed to the
preservation of particular species of biota. Others focus on open space
preservation or public education (Table 2.1 and Table 2.2). Often these
8


missions overlap in effect, if not in intent. For example, lands protected to
support forests or habitat often protect scenic views and open space as well.
Likewise, lands set aside through conservation easements as part of land use
planning will often include public access.
Table 2.1: Types of Land Protected by Land Trusts Across the United
States
Land Type Percent
Wetlands 52
River Corridors 51
Watersheds/Water Quality 47
F armland/Ranchland 46
Nature Preserves 45
Open Space 43
Endangered Species Habitat 42
Trails 27
Historic 24
Timberland/Forests 14
Urban Lands 10
Source: Land Trust Alliance, 2000
Land trusts achieve their goals primarily through land purchases,
conservation easements, and third party transfers.2 By far, the most commonly
used tool over the last ten years is the conservation easement. By 2000, local
2 Third-party transfers occur when a land trust temporarily holds a conservation easement with
the intention of transferring that easement to another land trust or public agency. Often, land
trusts participate in third-party transfers because a public agency can not act quickly enough to
protect a threatened parcel of land. Land trusts also will participate in third-party transfers
because, in Colorado, land trusts less than two years old may not hold a conservation easement.
9


and regional land trusts across the country protected approximately 2.5 million
acres (Land Trust Alliance 2000), greater than the size of Delaware. This is up
from just 290,000 acres protected by conservation easements in 1988. In 1999,
Colorado ranked fifth in the country with the most acreage held in conservation
easements (Nudel 1999). Many larger or more financially secure land trusts
purchase lands as well. Over 1.2 million acres are actually owned through fee-
simple title by land trusts across the country. Finally, some trusts engage in
third party transfers, purchasing land for an interim period and then transferring
them to a public agency. Hocker (1996, 249) extends the range of land trust
activity suggesting that land trusts protect land by whatever means will work.
This includes acting as active participants in growth management plans,
partnering with public agencies, educating communities regarding the benefits
of land and resource protection, and providing landowners with information
regarding state and federal tax laws as they apply to conservation easements and
donated land.
Land trusts are helping define new categories of land ownership, what
Geisler and Daneker (2000) call third sector property. These revisions in
property reflect a social value that exceeds the measure of value that can be
attributed to private ownership (e.g. grazing cattle, building houses, etc.). In
10


other words, there is more to the value of land than the private returns that may
be extracted. There is a public good component, a form of social mortgage
that has served as the foundation for Judeo-Christian definitions of stewardship
for centuries (Salsich 2000). As with common property found in other
countries, third sector property partitions benefits between individuals and their
communities and are predicated on local control (Geisler and Daneker 2000,
xiv). In so doing, land trusts play a growing part in redefining land as a social
asset as well as an economic asset.
Table 2.2: Land Trust Activities
Activities Percent
Maintaining land for recreation/public access 76
Environmental education 75
Land use planning 60
Biological monitoring/research 45
Ecological restoration 44
Management of rare and endangered species o r rn . 37
Source: Land Trust Alliance, 1998
History of Land Trusts
While the number of land trusts has been growing dramatically over the
last ten to twenty years, the idea of a land trust is not new. The first land trust, 3
3 Land trust activities were not surveyed by the Land Trust Alliance in the 2000 census. Data
from 1998 were the latest available.
11


the Massachusetts Trustees of Reservations, was formed in 1891 by the state
legislature to hold small parcels of land for the enjoyment of the public (Hocker
1996). The Massachusetts state legislature held that lands owned by the
Trustees would be open to the public and exempt from taxation. Four years
later, Great Britain formed its National Trust based on the Massachusetts model.
In 1900, the Sempervirens Fund was formed to preserve redwoods in California
followed in 1901 by the Society for the Protection of New Hampshire Forests
(Hocker 1996).
Throughout the early and mid 1900s, the growth in land trusts was
modest. By the 1960s there were 130 land trusts in the United States and
approximately 430 by 1980 (Hocker 1996). The explosion in land trusts has
occurred in the last two decades, fueled by growing environmental awareness, a
concern for the consequences of losing open space, and a general discomfort
among private landowners about government environmental regulation (Hocker
1996; Cheever 1996). Certainly contributing to their growth, land trusts seem to
be politically bipartisan, a significant achievement in the arenas of property and
environmental protection. In the last ten years, there has been a 63 percent
increase in the number of land trusts in this country, with a 135 percent increase
in the number of protected acres. As of 2000, there were 1,262 land trusts in the
12


United States preserving 6.2 million acres of land, larger than the areas of Rhode
Island and Connecticut combined, through a combination of conservation
easements and fee-simple ownership (Land Trust Alliance 2000). While the
greatest number of land trusts are found in the New England states, the second
fastest growing region is the Rocky Mountains, with Colorado leading that
region (Land Trust Alliance 2000) (See Table 2.3).
Table 2.3: Number of Land Trusts
Region 2000 1990 Percent Increase
Southcentral 25 11 127
(AR, LA, OK, TX) Southwest 57 26 119
(AZ, CO, NM, UT) Southeast 115 62 118
(AL, FL, GA, KY, MS, NC, SC, TN) Pacific 139 79 76
(CA, HI, NV) Mid-Atlantic 174 105 66
(DC, DE, MD, NJ, NY, PA, VA, WV) Midwest 186 119 56
(IA, IL, IN, KS, MI, MN, MO, NE, ND, OH, SD, WI) Northwest 69 50 24
(AK, ID, MT, OR, WA, WY) Northeast (CT, MA, ME, NH, NY, RI, VT) 497 433 15
Source: Land Trust Alliance, 2000
Land trusts are generally not opposed to development but instead serve
as advocates for controlled and managed growth, opposing the worst elements of
13


sprawl. Some even argue that land trusts are dependent on sprawl for their very
existence, since without sprawl there would be very little call for their work
(Nijhuis 2000). Equally important, land trusts political palatability encourages
public officials to collaborate with land trusts and support funding programs for
local land protection initiatives. Development interests often may also find
reasons to support land trusts. Preserving open space and elements of rural
character may actually raise property values for surrounding development
(Nijhuis 2000; Runge, et al. 2000).
Cheever (1996, 2) observes that the magic of the land trust movement
has something to do with its apparent contradictions. Land trusts, according to
Cheever, are radical in the sense that they actually lock up private land, an act
rarely achieved by traditional environmental protection. Nevertheless, while
land trusts lock up private land, they often do so while maintaining private
ownership. In addition, these land preserving actions occur primarily through
private, voluntary land transactions thereby appeasing those distrustful of
government while still serving a public need. Finally, they achieve public goals
at far less cost and controversy than could be achieved by public sector
counterparts. In fact, land trusts such as the Trust for Public Land explicitly
14


facilitate and effect tenuous land transactions that would not otherwise occur
through public sector action because of time or bureaucratic impediments.
Conservation Easements
While this dissertation examines all three of land trusts land preservation
tools: conservation easements, fee-simple ownership, and third-party
transactions, conservation easements are the most common tool used by land
trusts to preserve a parcel of land.4 They may be purchased either from a land
owner or donated entirely. By 1998, almost 1.4 million acres were protected
with conservation easements by land trusts (Gustanski 2000) and by 2000 this
number rose to nearly 2.6 million acres (Land Trust Alliance 2000). The
importance of the conservation easement becomes more evident when one
considers that fee title ownership is a viable option for only the most wealthy of
land trusts. Purchasing land outright for preservation in areas that are under
development pressure is tremendously expensive given the opportunity cost of
developing these lands. Conservation easements, often donated by the
landowner, only incur the costs of identifying, negotiating, and monitoring a
particular property. Over half the land trusts in the country rely on all-volunteer
4 The use of conservation easements are not limited to land trusts. Conservation easements are
held by public entities as well. This discussion is restricted to conservation easements used by
land trusts only.
15


staffs and many have budgets of less than $10,000 (Hocker 1996). The
conservation easement makes possible land protection efforts that otherwise
would never occur.
A conservation easement is a less than fee, non-possessory interest in a
parcel of land (Gustanski 2000). Rights to a parcel of land are partial and
specified. The easement is created with the same legal formalities found with
other real estate transactions. Unlike fee-simple title to land, conservation
easements reserve specific rights to a parcel of land, not its ownership. Legal
scholars are fond of making the analogy of the rights of land ownership to a
bundle of sticks. Fee-simple title is the entire bundle but these rights can be
separated (Cheever 1996; Gustanski 2000). A conservation easement may select
only one or two sticks, say, the right to build structures or the right to graze
cattle. A conservation easement may also be open-ended such as restricting any
activities that may interfere with elk calving.
It is not clear when the oldest conservation easement was formed.
Gustanski (2000) identifies the first conservation easement as being formed in
the late 1880s in the Boston area, while Cheever (1996) suggests that the first
conservation easement was created in 1956. Regardless of their origin,
16


conservation easements were relatively unknown and unused until the late 1970s
and 1980s. As part of the Tax Reform Act of 1976, Congress passed the
Historic Structures Tax Act that, for the first time, codified the deductibility of
conservation and historic preservation easement donations (Small 2000). This
provision (section 170(h)) allowed a donor to claim an income tax deduction for
a thirty-year donation of a conservation easement to a qualified recipient. A
year later, the Act was amended to require the donation to be written in
perpetuity, a condition common to most conservation easements today.
Interestingly, the Historic Structures Tax Act had a five year sunset
provision such that the conditions enabling conservation easements would
disappear by 1981 pending any further congressional action. In 1980, the
Historic Structures Act was amended to eliminate the sunset provision and the
tax incentive for conservation donations became permanent. Throughout the
early 1980s, the Internal Revenue Service issued a number of rulings all
favoring individual taxpayers with regard to income tax donations for
conservation easements. Then, in 1997 as part of the Taxpayer Relief Act, a
modified version of the American Farm and Ranch Protection Act was signed
into law including a provision (section 2031 (c)) that exempted land held under
conservation easement from estate tax (Small 2000). Now, if land donated
17


meets the requirements of section 170(h), the landowner is eligible for an
income tax donation. Furthermore, if the landowner dies, an additional
percentage of the value of the donated land may be deducted in addition to the
reduction in value already attributable to the easement (Small 2000). The
combination of these two laws has given land trusts an incredibly attractive tool
for private land conservation and can be attributed to a great degree for the
sudden and continued rise in conservation easements.
There is no federal statute determining how conservation easements are
written or enforced and what they may protect. Instead, there are differences as
well as consistencies across states regarding conservation easements. The
passage of the Uniform Conservation Easement Act in 1981 (UCEA) provided a
consistent set of guidelines for states to follow in forming and administering
conservation easements. The Act was intended to design a conservation
easement that would be flexible enough to fit particular ecological conditions
and landowner circumstances (Squires 2000). Nevertheless, there are no
requirements to follow the UCEA. Twenty-two states have adopted the UCEA
standards. Most other states, Colorado included, have designed their own
statutes, defining how conservation easements may be created and conveyed and
who may own them. These statutes tend to have elements of the UCEA while
18


reflecting the priorities of those states in terms of protecting particular lands and
using private property to do so (Squires 2000). Colorado, for example, requires
a land trust to be in existence for two years prior to accepting an easement. This
reflects the states concern that a conservation organization be serious about the
perpetual5 nature of conservation easements and not use the easement as simply
a short term tax shelter. Interestingly, notes Mayo (2000), no state has a
legislative requirement requiring a land trust to demonstrate minimum net worth
or other form of financial wherewithal to serve as a steward of land. While
serious about the tax implications of conservation easements, states seem
relatively cavalier regarding the financial viability of the land trusts that hold
those conservation easements.
Land trusts and conservation easements are revolutionary tools in land
protection. During earlier periods of environmental protection, property rights
to land remained fundamentally unchanged but were subject to burdens
imposed by local, state, or federal regulations. A landowner remained full
owner of his land but was not allowed to engage in particular activities such as
5 Only three states, California, Florida, and Hawaii, require conservation easements to be written
in perpetuity. All other states allow land trusts the flexibility to create easements for a specified
period of time (Mayo 2000). Nevertheless, the tax advantages of conservation easements
disappear when they are written for periods less than perpetuity.
19


filling wetlands or damaging an endangered species habitat (Cheever 1996).
Conservation easements actually alter the nature of land ownership by dividing
up that bundle of rights and creating property rights in conservation. In doing
so, the exclusionary quality of private property is revised to include the
qualities of restraint and responsibility that characterize common or shared
property (Cheever 1996, 7). Land trusts are at the forefront of this expansion
of what was once a far more simplistic and restrictive conception of American
land ownership and natural resource protection.
Conclusion
The emergence of land trusts, the evolution of the means by which
natural resources are managed, and the conditions that dictate effectiveness
among these land trusts should strike the curiosity of policy professionals,
natural resource managers, and, not the least, members of the land trust
community. The following chapter will introduce the conceptual framework of
this investigation and the contributing literature. This dissertation examines
property as an evolving institution and land trusts as organizations symbiotically
responding to that evolution in property. The institutional approach offers the
opportunity to examine this novel manner of natural resource management from
a holistic perspective, including social, political, and economic phenomena.
20


This, in turn, avoids predetermined bias that would impose disciplinary myopia
at far too early a stage in theory development. The chapter will also argue that
truly to understand what makes some land trusts relatively more effective than
others from an institutional perspective will require contributions from a number
of literatures or schools of thought. These disciplines have made contributions to
local management of natural resources in the past. This dissertation attempts to
coalesce their efforts.
21


CHAPTER 3
REVIEW OF LITERATURE
This chapter begins by introducing the idea of open space as a common
pool resource as well as a public good. This may seem initially contradictory by
the general definitions of common pool resources and public goods but in fact
open space shares characteristics of both (Ostrom, Gardner, and Walker 1994;
Fausold and Lilieholm 1996). The chapter then discusses the blurring of public
and private property that emerges as land trusts use mechanisms, especially
conservation easements, to protect open space Next, the framework for
examining factors that determine effectiveness among land trusts will be
introduced. Finally, the contributing literatures in common property theory, co-
management, social capital, and political entrepreneurship that inform the
explanatory variables for effectiveness will be discussed.
Open Space as Common Pool Resource and Public Good
Common pool resources (CPRs) may be defined as natural or man-made
resource systems that are sufficiently large as to make excluding potential
beneficiaries costly although not impossible (Ostrom 1990). Well-published
examples include lobster fisheries (Acheson 1990), underground aquifers
22


(Blomquist 1987 and 1992), grazing commons (Netting in Ostrom 1990), and
surfing (Rider 1998). Open space shares some characteristics of CPRs.
Specifically, excluding individuals from consuming and permanently
eliminating, open space is difficult. Private agricultural landowners may hold
title to parcels of open space such as range lands or hay meadows but many of
these landowners face tremendous opportunity costs from encroaching urban
and suburban development. Furthermore, estate taxes based on assessed value
can become extremely burdensome when agricultural land is inherited.
Consequently, farmers and ranchers owning lands on the fringes of development
are under tremendous pressure to sell or subdivide their land. Open space, a
resource common to all, faces difficulties in preventing its consumption
(through development) because it may be privately traded.
Open space also shares some characteristics with public goods. When
land is formally protected as open space, it then adopts the conditions of
nonrivalry and nonexcludability that define a public good (Comes and Sandler
1996). Riparian habitat, wildlife corridors, and scenic viewsheds can be enjoyed
by all (nonexcludability) and each individuals enjoyment does not interfere
with the enjoyment of others (nonrivalry). Interestingly, a great deal of open
space, which provides numerous public goods, is privately owned. Open space
23


and other similar natural amenities lie on a cusp between common-pool
resources and public goods.
Loss of Open Space as a Collective Action Problem
Collective action is the behavior of a group of individuals, generally to
their collective benefit. Collective action problems occur when individuals
within the group recognize that there are superior gains to be had from acting
out of concert with the group interest (Olson 1965; Hardin 1982; Taylor 1987;
Ostrom 1990). Collective actions problems, according to Taylor and Singleton
(1993, 196), exist where rational individual action by each of the members of
some set of actors can lead to an inefficient or Pareto-inferior outcome. This
behavior of individuals acting contrary to societys best interest is called free-
riding (Olson 1965). Individuals free-ride when the benefits of extraction or use
to one individual exceed the cost to that individual, primarily because those
costs are borne by the collective (Runge 1992).
The collective action problem traditionally is associated with the work of
Mancur Olson (1965). Collective action problems, according to Olson, can be
problems of either provision (a public good problem) or consumption (a
common pool resource problem) or both. Group size affects provision because
24


as a group gets larger, monitoring individual behavior becomes more difficult
(Taylor 1987). Protecting valuable resources becomes provisionally more
difficult when large numbers of people wish to enjoy the resource. Like
contributions to public radio or charities, individuals feel that they may shirk
their contribution to a pool if they know that large numbers of other people may
contribute and a sense of anonymity exists.
The problem of protecting resources such as open space, valuable
habitat, or rural amenities is a problem of consumption. Consider any one of a
number of rural areas in America that lie on an urban fringe, areas that are
valuable because they have open, pleasing views while close to urban amenities.
As more individuals demonstrate interest in the area, they each become less
likely to regulate their own use, instead opting to get it while the getting is
good and building homes on the edge of these rural landscapes. There are no
restrictions to prevent problems such as congestion. In this case, free-rider
behavior encourages individuals to consume a nonsustainable amount of the
resource since the marginal cost of each individuals behavior (purchasing rural
land and building subdivisions) is far less than the marginal benefit (selling
subdivisions). This is rather similar to the traditional example of an unregulated
grazing commons (Hardin 1968) where each herder will graze as many animals
25


as possible, knowing that the costs associated with overgrazing are minimal
relative to the gains to his herd.
Many resources, including CPRs, are subject to collective action
problems and free-riding. Fisheries (Acheson 1990), unconfined aquifers
(Blomquist 1987 and 1990; Somma 1997) and other renewable resources have a
rate of extraction that allows continued use but, without rules and enforcement,
the potential for an individual to act selfishly and over-extract looms large.
Solutions to collective action problems include the familiar institutions of
private property rights where individuals are assigned rights of ownership and
act in their own interest (Demsetz 1967) and public ownership where a central
authority oversees social well-being (Ophuls 1973). In addition, common
property offers communal ownership and control where exclusion still occurs
but benefits accrue to a specified group (Ostrom 1990 and 1992; Bromley 1992).
Open space and wildlife habitat similarly suffer the threat of collective
action problems. Residential home and commercial retail developers have a
rational self-interest in developing open space. Homes that border open space
have been shown to be valued 32 percent higher than homes 3,200 feet away
(Correll, et al. 1978). Consider a beautiful valley blessed with undesignated
26


open space. Each individual wants to build his or her house to enjoy the view
but the similar actions of all individuals spoil the view for all. Social benefits
fall prey to private, self-interested action.
As discussed above, solving collective action problems generally reduces
to solutions of private or public ownership (McCay 1996). Indeed, some
counties in Colorado, such as Boulder County, Jefferson County, and Pitkin
County, have open space programs that purchase and manage land with tax
revenues.6 Collectively, residents may vote to preserve open space.
Unfortunately, open space programs are not always politically palatable or
financially feasible in some communities. The other solution, private property
ownership, is really no solution at all. Rising property values in areas facing
expanding growth and the threat of estate taxes as landowners face their own
mortality convince private property owners to sell or subdivide for residential
and commercial development.
Discomfort with the binary options of public and private property has, in
part, motivated the research in common property theory (Ostrom 1990; Bromley
6 Boulder County's open space program manages over 61,000 acres with revenues collected from
property taxes, sales taxes, and bond proceeds.
27


1992). Virtually every environmental problem indicates a failure to create
institutions that internalize externalities or, in other words, a failure to fully
specify property rights (McKean 1996). Common property can help reduce the
costs of enforcement through shared policing of the resource and encourage its
care by those with vested interests in the long term well-being of the resource.
Commonly owned property provides alternatives to public or private solutions.
In addition, there is a large continuum that potentially lies between these binary
choices that includes, but is not limited to, common property. Land trusts and
their most common tool, the conservation easement, lie within this continuum as
well.
Property as a Continuum
Property is a legal concept rooted in social relations. It refers not simply
to objects but to the relations between individuals and society that govern access
and use of material objects (Wiebe et al.1996). Property has been traditionally
organized in this country as either public or private. It is a national paradox of
sorts. Public lands are shrouded with pride, mystique, and national identity
while private land ownership is held in an almost defiant conceit that [it] is the
highest and best use of land and the centerpiece of American civil liberties
(Geisler 2000, 65). Yet Geisler cautions against the natural tendency to view
28


property ownership under this binary classification. The face of property
ownership is changing, or at least deepening in complexity. There is a growing,
or at least increasingly recognized sharing of property interests. Public interests
in the form of natural resource, habitat, open space, and cultural protection are
finding opportunities for expression on lands held privately and often
individually. Transitions are forming between the maturing social values of
environmental protection, community, democracy and sense of place (Etzioni
1983; Kemmis 1990; Brandenberg and Carroll 1995) and the stalwart American
ethics of individualism and private land ownership (Decker 2001; Moore 2001).
Land trusts, using conservation easements in particular, are facilitating
public benefits through open space protection while maintaining private
ownership of land. Property mutations such as conservation easements actively
balance economic values and social values. Geisler and Daneker (2000) call
these types of property mutations third-sector property and democratic
property, where social equity concerns are given equal standing with individual
interests. Characteristics of these property mutations include,
Landownership is neither public nor private but a hybrid. Ownership
rights are held among multiple interests.
29


The propertys primary function is to meet broadly defined social
needs as well as individual needs, not merely to increase individual
wealth.
The locus of control over property, though sometimes tiered, is
primarily local (Geisler and Daneker 2000, xiv).
Through the institution of property, society conducts and organizes the
use of assets such as land and natural resources. Property, in this context, is the
collection of rules, not the object (Challen 2000). Property rights regulate
behavior with respect to objects of value. Property rights are the social
institutions that define or delimit the range of privileges granted to individuals
for specific assets, such as parcels of land or water (Libecap 1989, 1). They
help society define what is scarce and what is valuable. Property, notes
Bromley (1992, 4), is a social instrument, and particular property regimes are
chosen for particular social purposes. "Rights, obligations, and rules embody
people's expectations about the environment" (Hanna and Jentoft 1996, 35).
Land trusts holding rights to conservation through conservation easements on
private property for the public benefit represent an evolving social expression of
the value placed by society on open space and the manner in which we choose to
protect those lands. To understand this evolution, it is critical to examine
30


property as an institution and to examine the organizations that function within
that institution.
Institutions, Organizations, and Institutional Evolution
Institutions are the rules of the game and the humanly devised
constraints that shape human interaction (North 1990, 5). They dictate how
society chooses to conduct and organize itself beyond purely rational action.
Institutions, as collections of formal and informal rules, reflect the values of
society in terms of what we manage and how we choose to manage it (March
and Olsen 1995). Rules are "socially constructed, publicly known, anticipated,
and accepted" (March and Olsen 1995, 30). Rights are the product of rules and
are the "enforceable authority to undertake particular actions related to a specific
domain" (Ostrom and Schlager 1996,130). The study of institutional change is
rooted in an alternative to neoclassical economics that, while still grounded in a
rational choice model, argues that institutions evolve along path dependent lines
that seek to minimize transaction costs (Coase 1937; Barzel 1989; North 1990;
Furobotn and Richter 1991 and 1991a; Williamson 1991; Williamson 1996;
Challen 2000).
31


There are many interpretations of institutions (Mathews 1986) but the
most common, and the interpretation used here, is associated with Coase (1960).
Coase argued that any institution or system of property rights is capable of
leading to Pareto efficient conditions provided it is a complete system where all
rights to all the benefits from all scarce resources are imputed to someone and
are tradeable (Mathews 1986, 904). Importantly, Coase held out the caveat that
such a complete system, or institution, was impossible due to transaction costs
and that some institutions are more conducive to Pareto efficiency than others.
The general body of literature that is discussed here is often referred to
as new institutionalism but an initial clarification of the roots of new
institutionalism helps set the stage for further discussion. To begin with, why is
this institutionalism new? What, then, is "old" institutionalism? Traditional
neoclassical economics takes the existence of institutions as given, fixed, and
exogenous to standard models of production and resource allocation (Mathews
1986; Lowndes 1996). In other words, institutions do not vary so they are not
studied. Old institutionalism criticized neoclassical economics reliance on
oversimplifying economic systems. Late nineteenth and early twentieth century
economists such as Wesley Mitchell (1967) and John Commons (1934) drew on
sociology, politics, and law to create a multidisciplinary, descriptive model of
32


economic exchange that recognized that the choice of institutions had a notable
effect on exchange and resource allocation (Groenewegen, et al. 1995). They
saw the economy as complex and evolving rather than mechanistic (Colander
1996). Possibly because it lacked a theoretical foundation and relied on
descriptive and non-generalizable empiricism, old institutionalism simply gave
way to the expanding program of mathematical economics in the middle 1900s
(Foss 1995; Colander 1996; Challen 2000).
Where old institutional economics was hostile to orthodox neo-
classical economics, new institutional economics7 argues that institutions can be
analyzed within the neoclassical approach. The fundamental hypothesis,
according to Lowndes (1996, 184) is that institutions exist where their benefits
exceed the costs involved in creating and maintaining them. Consequently,
7 There is a noticeable lack of consistency surrounding terms such as new institutionalism and
new institutional economics and neoinstitutionalism. For example, Eggertson (1990) defines
new institutional economics as rejecting the rational choice underpinnings of neoclassical
economics and neoinstitutionalism as retaining neoclassical economics' rational choice
foundation but Eggertson (1996) defines the "new economics of institutions" as bringing the
study of institutions "directly into economic analysis." Similar to Eggertson (1996), Mathews
(1986) refers to this same line of reasoning as "the economics of institutions." Alternatively,
Colander (1996) refers to "new institutionalism" as the logical alternative to well-structured
market conditions found in neoclassical economics; similar to Eggertson's (1996) new
economics of institutions or Eggertson's (1990) neoinstitutionalism. Confused? This
dissertation adheres to the term "new institutional economics" and refers to a line of reasoning
that includes rational choice among individuals and firms but acknowledges the existence of
transaction costs and the need for institutions to evolve toward forms that minimize those
transaction costs.
33


new institutional economics accepts the general tenets of marginalism but
expands the neoclassical program to include asymmetric information and costs
of monitoring and enforcing agreements (Eggertsson 1990). These additional
complexities are highlighted by Coase (1960).
In order to carry out a market transaction it is necessary to
discover who it is that one wishes to deal with, to inform people
that one wishes to deal and on what terms, to conduct
negotiations leading up to a bargain, to draw up the contract, ...
and so on (Coase 1960, 15).
These additional complexities are called transaction costs and amount to
friction, or indirect costs, in a system of exchange. Eggertsson (1990,14)
defines transaction costs as the costs that arise when individuals exchange
rights to economic assets and enforce their exclusive rights. Systems without
this friction are those found in the general equilibrium models of neoclassical
economics but, unfortunately, are rarely found elsewhere. Exchange in the real
world is replete with friction. Transaction costs amount to a collapse of the
assumption within neoclassical microeconomics of full and perfect information
and instantaneous and costless transactions. Without full and perfect
information, economic agents must search out others with whom to trade,
bargain with them, risk deception and other acts of guile, monitor contracts, and,
if necessary, enforce sanctions.
34


The general welfare model of social efficiency assumes the nonexistence
of transaction costs. The process of exchange naturally leads to social optima or
Pareto efficiency. Unfortunately, real world exchanges often contain a variety
of the transaction costs mentioned above. The general assertion within new
institutionalism is that institutions that most effectively lower transaction costs
tend to be preferred to those that are less effective (Williamson 1991; Lubell et
al. 1999). The existence of money is probably the simplest example of an
institution designed to economize on the costs of transacting. The transaction
costs of barter are far greater than those associated with a common currency that
facilitates trade. Consequently, the institution of money has replaced the
o
institution of barter. This also explains the absence of money in general
equilibrium models of neoclassical microeconomics since, again, neoclassical
theory has no room for transaction costs and so cannot explain institutional
arrangements.
As noted earlier, institutions are collections of formal and informal rules.
Institutions are separate and distinct from organizations. If institutions are the
rules of the game, organizations are the players (North 1993). Organizations are 8
8 Even today, the institution of money is evolving as credit often replaces cash and forms of
electronic monetary exchange are emerging.
35


resource pools. They come into existence when individuals place their
resources (skill, money, etc.) in a central location (Foss 1995). "Organizations
are social groups, people who work together" (Eggertsson 1996, 161).
Organizations include political bodies (city councils, government agencies),
economic bodies (individuals, firms, non-profit organizations), social bodies
(church groups, soccer teams), and so forth (North 1990, 4).
Early institutionalist writing (e.g., North and Thomas 1973) argued that
institutions determine economic performance. Since neoclassical economic
performance is grounded in the behavior of the individual, it was further argued
that institutions influenced individuals. Modern new institutional economics
recognizes a symbiotic relationship between organizations and institutions
(North 1990). Organizations respond as entrepreneurs to the incentives and
conditions set forth by institutions. As these organizations form and evolve, they
then spark additional evolution within their institutional universe. In the course
of attempts to accomplish their objectives, [organizations] are a major agent of
institutional change (North 1990, 5; emphasis added). As the institution of
property evolves in response to changing social needs and values, the
organizations that best serve these needs emerge. Land trusts are emerging
quickly as community-based organizations serving a public interest in land
36


preservation, but all land trusts are not alike. What features differ among them,
and what land trusts will be more effective than others? New institutional
economics would suggest that those that best address the transaction costs of
land preserving activity would be most effective.
The Role of Transaction Costs
Transaction costs are the costs of engaging in exchange, including
negotiating, monitoring, and enforcing contracts (Coase 1960, Furubotn and
Richter 1991). Rules that regulate and enforce the use of a resource also cause
transaction costs (Feeny 1998). The costliness of information is the key to the
costs of transacting, notes North (1990,27), which consist of the costs of
measuring the valuable attributes of what is being exchanged and the costs of
protecting rights and policing and enforcing agreements. Resources must be
expended to obtain and consider alternatives, negotiate contracts, and to monitor
and enforce those contracts (Furubotn and Richter 1991). Real estate
transactions involve realtors, attorneys, assessors, surveyors and many others
whose sole purpose is to overcome incomplete information. These transaction
costs occur because agents in an exchange do not have full information
regarding issues such as the nature of the object being exchanged, another
37


agent's intentions, or the degree to which rights to that object must be enforced.9
At times, these transaction costs can be sufficiently large so as to prevent a
transaction from taking place (Coase 1960). Therefore, some benefit-cost
calculus must take place to determine if a transaction will take place.
Williamson (1998, 75) refers to this as the discriminating alignment
hypothesis where transactions align with governance structures that differ in
their cost and competence, so as to effect a transaction-cost economizing result.
If there were no transaction costs, the choice of alternative institutions would be
irrelevant. But transaction costs do exist and new institutional economics
examines institutions in terms of their ability to economize on transaction costs
(Foss 1995).
Dahlman (1979) asserts that transaction costs are simply due to imperfect
information. As a broad definition, this is basically correct. Greater detail is
provided by Barzel (1989, 2) who asserts that transaction costs are the "costs
associated with the transfer, capture, and protection of rights." Barzel (1989, 2)
9 The stereotype of the used car deal provides an example. The buyer does not share the same
information as the seller regarding the condition of the vehicle, does not know if the seller is
honest or duplicitous, and, for that matter, does not know if the title is even valid. What is
missing in this case? Full information. Transaction costs on the part of the buyer (e.g., paying
for a mechanics opinion regarding the vehicles condition) and the state (e.g., enforcing laws
against deceptive sales techniques such as turning back an odometer) add to the costs of the
transaction.
38


goes further suggesting that "when rights are perfectly delineated...the
(relevant) costs of transaction must be zero." Efforts to identify transaction
costs in greater detail have been found in investigations of fisheries management
(Libecap 1989), property rights (Barzel 1989), and irrigation systems (Tang
1992). Wallis and North (1986) actually attempt to measure the total
contribution of transaction costs in the United States economy.10 Their
conclusions demonstrate that the impact of transaction costs on gross national
product has increased over the last century and by 1970 included virtually half
of gross national product. Their point, for purposes here, is that transaction
costs have a significant impact on exchange in a modem economy and the
institutions that emerge and evolve to guide those transactions, by whatever
means, do so because of the influence of transaction costs.
Foss (1995) argues that transaction costs account for more than the
existence of firms as originally described by Coase (1937; 1960). Transaction
costs also can explain the size and scope of the public sector. Many problems
on the public agenda have high transaction costs. In particular, public resource
10 Wallis and North (1986, 98) provide an excellent example of transaction costs faced by both
buyers and sellers in the purchase of a house While too lengthy to include here, suffice to say
the costs of attorneys, realtors, title, etc. amount to just some of the transaction costs in this type
of exchange. This is particularly relevant when considering the similarity to exchanges that take
place with land trusts.
39


allocation (e.g. public lands) and public regulation (e.g. protecting wildlife
habitat) have high transaction costs. Institutional choice and evolution favor
conditions that minimize transaction costs (Foss 1995; Williamson 1998).
Consequently, in an era where government responsibility is being devolved to
more local levels (Nelson 1995) and social values articulate an increasing desire
for local management of local conservation issues, organizations such as land
trusts that efficiently serve public interests may emerge.
The ability of institutions to minimize transaction costs guides their
emergence and evolution. As such, the relation between change in transaction
costs and change in institutions is analogous to change in production costs and
change in products or processes (Mathews 1986). Nevertheless, it is important
to recognize that guidance is not synonymous with guarantee. Cursory
interpretations of the role of transaction costs may suggest that institutions
seamlessly gravitate toward those that most efficiently minimize transaction
costs as suggested by Williamson (1985): a carryover, of sorts, from the
optimizing conditions inherited from neoclassical economics. Such
interpretations are derived from institutional examinations of firms (Coase 1937;
Williamson 1985). More recent interpretations of new institutional economics
acknowledge that inefficient institutions can persist. Later work from North
40


(1990, 1993) and Furubotn and Richter (1993) emphasize stability as well as
efficiency in institutions (Lowndes 1996). Institutions, according to North, still
seek to economize on transaction costs but are equally influenced by cultural
norms, codes of conduct, and conventions (North 1990). Institutions, then,
evolve within cultural, social, and political constraints. If land trusts are truly
emerging as a player in the institution of propertys evolution, it is because they,
as organizations, gamer the resources to address both the formal (transaction
costs) and informal (social and cultural) components of change within the
institution of property.
Organizations evolve in competition for the most efficient forms for
more specialized transactions and can thus be viewed as initiators, or a primary
source, of institutional change (North 1993). For the purposes of this
dissertation, land trusts may be evolving as the most efficient manner to manage
local natural resources due to their ability to lower realized transaction costs. As
North (1993, 20) points out, "a decline in the perceived effectiveness of existing
organizations will weaken their ability to maintain and support the existing
institutional structure." State intervention may be appropriate for large scale
protection (e.g. wilderness status) and private ownership may be appropriate for
specific productive uses of land (e.g. crop or livestock production) but local
41


environmental concerns and stewardship may be most effectively addressed by
third sector solutions such as those offered by local land trusts.
Transaction Resources
As noted above, those institutions that accomplish economic exchanges
most efficiently are those that prevail (Williamson 1985; Lowndes 1996) within
current norms, values, and codes of conduct (North 1990). The means by which
an institution accomplishes exchange may be considered transaction resources.
Land trusts are emerging as an increasingly successful tool for the local
management of natural resources. This may occur in part because they are able
to acquire and use transaction resources more effectively than other
organizations, such as government or private land ownership, and in part
because they demonstrate compatibility with local norms and values. This
suggests an organic mix of factors that economize on transaction costs and
factors that generate social and cultural acceptance in the areas served by local
land trusts.
Taylor and Singleton (1993, 199) propose that these transaction
resources are embodied in the idea of community, which is composed of
stable, multiplex, and direct relations and shared beliefs and preferences. The
42


notion of community has long been known to be qrore than mere geography. A
mere collection of people dwelling in houses somewhat near together or within
easy reach of one another does not constitute a community (Butterfield 1919,
143). Features of community include direct communication, human interplay,
and collective undertaking with collective responsibility (Butterfield 1919).
Taylor and Singleton emphasize the virtue of communities such that it is not so
much that the groups which manage endogenous solutions have more of the
relevant resources to expend than others but, rather, that they possess certain
characteristics in virtue of which the transaction costs of an endogenous
solution are relatively low, so that they do not have to expend large quantities of
transactions resources (Taylor and Singleton 1993, 199; author's emphasis).
Many of these relations are informal, the complement to Norths (1990) factors
for institutional stability. Nee (1998) observes, informal norms arise out of
networks and are reinforced by means of ongoing social relationships (Nee
1998, 86-7). This suggests that the relationships a land trust holds with its
community such as its history and reputation may contribute to its capacity to
meet or minimize transaction costs.
43


Social Capital
When local groups and community-based organizations are able to
resolve problems relatively endogenously, they draw on, and subsequently
reinforce, elements of social capital as transaction resources. Social capital
includes obligations, information channels, and social norms (Coleman 1988;
Coleman 1990). Coleman (1990) compares social capital with physical and
human capital. Physical capital is created by making changes in materials so as
to form tools that facilitate production, human capital is created by changing
persons so as to give them skills and capabilities that make them able to act in
new ways such as education (Coleman 1990, 304). Social capital consists of
the relations among individuals, households, and other organizations that
facilitate action. If physical and human capital are the machinery of modem
exchange and transaction costs are the friction, then social capital is the grease.
Social capital enhances the benefits of investment in physical and human
capital (Putnam 1993b, 36).
Social capital may be dissected into various aspects or forms. Stiglitz
(1999) divides social capital into three component parts. First, social capital is
composed of "cognitive aptitudes and predispositions" (Stiglitz 1999, 60). In
other words, individuals in a community share a common understanding and
44


vernacular of a problem and its solutions. Second, social capital is composed of
a collection of networks or groups in which individuals are socialized. This is
reminiscent of Putnam's (1993b; 2000) networks as well. Third, social capital is
an aggregation of reputations because positive reputations reduce transaction
costs and help break barriers to relations engaged in exchange.
Stiglitz's forms of social capital can fall into one or both of two broad
categories described by Putnam (2000). Putnam (2000,22) argues that the most
important distinctions within social capital are "bonding" and "bridging" social
capital. Bonding social capital is inward looking, such as when an ethnic group
or political group reinforce each others' convictions for a cause or agenda.
Bridging social capital reaches out to others outside of a community to create
links with others that can support a group's mission much like an individual
networks with others to enhance his job performance.
Social capital is built when partnerships among community members,
citizen groups, and coalitions form (Putnam 1993a; Putnam 1993b) and is
grounded in trust among members (Putnam 1995; Smith 1997). Social capital is
the capacity of members of a community to associate with each other based on
the degree to which those members share norms and values (Fukuyama 1995).
45


Trust allows formal and informal networks to form which facilitate
coordination and communication, amplify relations, and thus allow dilemmas of
collective action to be resolved (Putnam 1995, 67; also, Fukuyama 1995)
Putnams research on social capital in both Italy (Putnam 1993a) and the United
States (Putnam 1995) indicates that the quality of public life and the
performance of social institutions are positively related to the existence of social
capital. Furthermore, Putnam (1993a, 166) implicitly argues that social capital
reduces transaction costs by enabling agents more efficiently to surmount
problems of opportunism and shirking and discourages collective action
problems.
While social capital is clearly prominent in the social science literature,
there is far from complete agreement regarding the virtues of social capital and
much of the resistance comes from scholars in mainstream economics. Solow
(2000, 6) suggests that social capital is "an attempt to gain conviction from a bad
analogy." Instead, the features others refer to as elements of social capital such
as trust, cooperation, and networks, are only behavioral patterns to Solow.
Physical capital, he observes, has a measurable rate of return and a rate of
depreciation. Social capital does not demonstrate this. Arrow (2000, 4) extends
Solow's concern by noting that capital has three components: (1) extension in
46


time, (2) deliberate sacrifice in the present for future benefit (investment), (3)
alienability. While the first component may have applications to social capital
(it exists over a period of time), Arrow argues that the second simply does not
apply to social capita since he finds "no consensus.. .for adding something
called 'social capital', to other forms of capital" (author's emphasis). In other
words, people do not invest in social capital to advance their productive
capacity. This seems somewhat narrow, if not simply incorrect, since it is
certainly not hard to imagine individuals cooperating to gain greater returns to
other investments and, in fact, Ostrom (2000) argues this point explicitly.
Finally, while Arrow (2000) provides no commentary on the third component,
his concern has some standing as social capital represents interactions between
individuals, not something that is owned by one individual.
Ostrom (2000) rescues social capital from Solow's and Arrow's critique
by acknowledging their concerns without rejecting social capital. In contrast to
physical capital, Ostrom observes that social capital appreciates rather than
depreciates as does its physical metaphor. She also notes that social capital has
significant policy implications as a mechanism to resolve collective action
problems by reducing asymmetries and raising cooperation among participants
in a community or other group. This last observation stands in contrast to
47


Arrow's assertion that individuals or groups do not invest in social capital to
advance their productive capacity. So far, the gist of the social capital critique
seems to surround the use of a phrase, rather than a substantial critique of its
contribution to economic life. As Sobel (2002, 145) concludes, "A vague
keyword is not sufficient reason to condemn a promising line of research."
Community-based land trusts provide an excellent opportunity to
examine the role of social capital. These land trusts are potentially more
effective when they employ greater degrees of social capital. Engaging
citizenry, especially prominent members of the community, may generate
legitimacy for the land trust. Land trusts depend on community support and
participation. Shared values and cultural experiences may enhance social capital
between land trusts and the community they serve. Longevity may even solidify
social capital (Lubell et al. 1999). For example, those land trusts that have been
in existence longer than others may be more effective simply because they have
established positive reputations and trust within the community.
Concentric Regimes
Contrary to Taylor and Singletons (1993) hypothesis, it is also possible
that effective community-based organizations may have the capacity to meet
48


transaction costs, not just avoid them (Ostrom 1992). Taylor and Singleton
(1993,198) focus specifically on what they call "wholly endogenous solutions"
to collective action problems. While interesting in its own right and instructive
for this investigation, their thesis is insufficient for fully examining the
complexities of community-based organizations such a found with land trusts.
Volunteerism, public and private grant-awarding organizations, and donations
may all assist land trusts in paying for the transaction costs that are impediments
to successful solutions to the problems described by Taylor and Singleton.
Collectively, this dissertation refers to these types of groups or formal
organizations as concentric regimes. Concentric regimes are groups or
organizations that engage with a land trust and, by some mechanism, support
that land trust's effectiveness by addressing transaction costs. I disagree with
their assumption that it is primarily the attributes of community that affect
transaction costs, argues Ostrom (1992, 349) regarding Taylor and Singletons
assertion above. It is just as important to see how larger political regimes may
facilitate the capacity of individuals to achieve agreements as it is to see the
importance of community in facilitating solutions to these problems (Ostrom
1992, 349).
49


The literature in co-management can help inform the role of concentric
regimes. "Comanagement, where power is actually shared, promises an
institutional response to the 'commons' problem, which is essentially the
question of how private interests can better intermesh with collective interests"
(McCay 1996, 120). Most of the literature surrounding co-management
involves the relationship between community action and state government
(Pinkerton 1994, Baland and Plateau 1996; Memon and Selsky 1998).
Nevertheless, in an era where government responsibility is being devolved to
more local levels and community-based action on natural resource issues is
growing, it may be valuable to expand this definition of co-management to
include concentric groups that are private and quasi-govemmental. Memon and
Selsky (1998, 599) describe the capacity of co-management to achieve a
commons discourse where comanaging stakeholders have the opportunities to
express and learn new perspectives and interests.
50


Political Entrepreneurs
Specific individuals, or political entrepreneurs11 that promote the agenda
of a land trust may also influence its effectiveness. Kingdon (1995) observes
that when examining case studies (of policy arenas, in Kingdon's case) one can
usually identify a particular person who was central in promoting a particular
policy or agenda. In fact, Kingdon identifies 15 out of 23 policy case studies
where an entrepreneur was important to the advancement of a policy or
program. Taylor and Singleton (1993) propose that when communities are not
able to reach a purely endogenous solution to a collective action problem,
solutions still may emerge due to a political entrepreneur. According to the
authors, "The political entrepreneur accomplishes this -can only accomplish this
- in part by strengthening community in the group" (Taylor and Singleton 1993,
203) (authors' emphasis). The entrepreneur may arise from within or from
outside of the community (Kingdon 1995). The defining characteristic of the
entrepreneur is his or her willingness to invest time, energy, reputation and
personal finances in hopes of seeing some form of return (Kingdon 1995).
Regardless, the entrepreneur's function is to facilitate, not impose, a solution
11 Also called policy entrepreneurs (Kingdon 1995; Mintrom 1997) and public entrepreneurs
(Schneider et al. 1995) depending on the environment in which the entrepreneur operates. Their
characteristics are sufficiently similar to be integrated here. For the purposes used here, they
will be considered synonymous.
51


(Taylor and Singleton 1993). These efforts may help to coalesce local public
opinion or gamer support from external organizations and government.
Political entrepreneurs are also those who seek some form of policy
change through identifying problems, networking, setting the terms of the
political debate, and building coalitions (Mintrom 1997). Political entrepreneurs
may assist in forming the building blocks that provide a foundation for creating
social capital. Prominent members of the community, local elected officials, or
politically connected individuals may serve as political entrepreneurs by
capitalizing on their political connections to establish supportive networks
within their particular communities (Kingdon 1984; Schneider et al. 1995).
These individuals have the ability to gamer advantages by establishing formal
and informal contractual arrangements with the owners of inputs that economize
on transaction costs (Schneider et al. 1995). Schneider et al. (1995) suggest that
while, according to North (1990), institutions are devised to reduce transaction
costs, enforcement and compliance costs of existing institutions provide
incentives for political entrepreneurs to seek new reforms that likewise reduce
these transaction costs. In other words, political entrepreneurs are able to
connect with members of the community in ways that gamer support and trust
so as to further the mission of, in this case, their land trust. This is echoed by
52


Bianco and Bates (1990), who add that while leaders or entrepreneurs with
strong positive reputations within the community have the capacity to initiate
cooperative behavior among actors (local residents or local government), the
role of entrepreneurs is less clear in terms of sustaining that cooperative
behavior. Possibly, political entrepreneurs may lay the ground work for
establishing social capital, but are not capable of providing longevity to that
social capital.
Resource-Specific Factors
Finally, there are potential factors to effectiveness that are not specific to
transaction costs. To examine institutional change, one must also examine the
"relationships among variables characterizing the resource" (Ostrom and
Schlager 1996, 142). Most important may be the relative scarcity of
undeveloped land. All other things being equal, as more land is developed, the
remaining undeveloped parcels grow in value. Is the effectiveness of land trusts
in part a function of remaining open space? Are land trusts more effective when
their community is faced with a greater sense of threat? Fausold and Lilieholm
(1996) cite declining open space in growing metropolitan regions, fragmentation
of farmland, the aging of typical landowners of large tracts, and growing
concern for ecosystem management as factors that are stimulating land
53


preservation efforts. The institutional economics framework assumes that land
trusts are assumed to be boundedly rational (Eggertsson 1996) in that they only
engage in activities where the benefits (loosely defined to include both market
and nonmarket) exceed the costs. If the scarcity of open space increases the
value of what remains, the benefits to land preserving action increases positively
with scarcity. Therefore, it is possible that effectiveness is directly related to
scarcity of open space.
Summary
This literature review establishes the logical steps and supporting
literature for examining the effectiveness of community-based land trusts in
Colorado. It begins by establishing the loss of natural amenities, such as open
space, as a collective-action problem; a problem where individual, self-
interested action leads to the detriment of the collective. The institution of
property serves to resolve collective action problems by according ownership
and the rights of use, exclusion, and sale to the owners of a resource. In the
United States, we have become accustomed to the binary classes of ownership,
public and private. While these binary classes of natural resource ownership
have been a mainstay of environmental stewardship, forms of third-sector
property are emerging that integrate public values with private ownership. Land
54


trusts are the fastest growing and most prominent class of organization that are
effecting this rising prominence of third-sector property.
The sudden emergence of land trusts speaks to a class of organizations
that are able to minimize the transaction costs of realizing public values on
private land. The growth of land trusts and the commensurate growth of acres
held under conservation easements demonstrates that public values can
cohabitate with private land ownership. Keeping with the new institutionalism
framework, transaction costs are taken as given in this dissertation and therefore
there is no attempt to identify, catalog or quantify transactions costs. Instead,
and more importantly, this dissertation looks to the transaction resources that
allow land trusts such a prominent role in land protection. Transaction resources
are the means to either meet or avoid transaction costs. Social capital, through
education, trust, reputation, and reciprocity, may allow land trusts the ability to
establish transactional relations with less resistance than other organizations
(e.g., local government) using less effective means of property ownership (e.g.,
condemnation or purchase). Land trusts may also prove to be effective in using
relations with other levels of governance to acquire the financial or political
support to further their activities. Land trusts may also take advantage of
political entrepreneurs that serve as barrier busters by introducing new and
55


innovative land protection methods or bringing to a land trust a pre-established
confidence with their community. In so doing, they may lower the transaction
costs of identifying properties and negotiating deals that would prove to be far
more burdensome for organizations that are not so well integrated with a
community.
To summarize, this research is an examination of the factors that dictate
the effectiveness of a community-based land trust. These factors are called
transaction resources and serve to either meet or avoid transaction costs. The
idea of transaction costs is an element of the new institutionalism literature that
provides explanations for the emergence, form, and change of the formal and
informal rules that guide social and economic behavior. Land trusts are
particularly noteworthy organizations in that they are able to function with
increasing prominence in the seams between public and private ownership of
land, the area called third-sector property. Nevertheless, all land trusts are far
from alike. Some have significantly higher levels of land protection activity
than others and it is very possible that the transaction resources discussed in this
chapter will lend insights as to the relative effectiveness of these fast emerging,
community-level land protection organizations.
56


Having organized the literatures that direct this research to answer how
community-based land trusts are effective, the next step is to organize an
analytic plan and establish metrics that are derived from these literatures. The
following chapter provides the analytic framework. It will describe the choices
of dependent and independent variables, data collection instrument and other
secondary sources of data, and methods of analysis.
57


CHAPTER 4
RESEARCH DESIGN AND ANALYTIC METHODS
This chapter begins with a brief discussion of the analytic design and the
dependent variable, followed by identifying the general hypotheses to be tested.
Next, a discussion of the research population is presented along with the primary
data collection instrument. Finally, a discussion concerning the logic of
qualitative comparative analysis (QCA) is provided. The analysis of data takes
place in three stages. First, descriptive data are provided to create a picture of
the population of community-based land trusts in Colorado. Descriptive data
allow readers to compare features of Colorado land trusts with land trusts in
other states or regions and address to what degree the results presented here can
be generalized or tested further. Description is an important part of social
research. Social research looks for causes or explanations for particular events
or outcomes. It is hard to develop explanations before we know something
about the world that is being investigated (King, Keohane, and Verba 1994).
This dissertation examines the entire population of Colorado land trusts.
In the next step, simple bivariate regression is used to test the
significance of independent variables that correlate with effectiveness among
58


land trusts. It should be emphasized that these techniques are employed only to
test for significance. According to Fox (1998, 147), we often carry out
significance tests.. .in order to assess the likelihood that a relationship found in
population data is due to chance or random processes of any kind. Given the
absence of a testable model and low degrees of freedom resulting from relatively
small N (n=27), multivariate regressions is inappropriate and not employed.
Furthermore, causal inferences established from small-N multivariate
regressions would be suspect. Another method of inferring causality must be
explored.
In the last step, fuzzy set qualitative comparative analysis, a method to
explore the combinatorial relationships of factors leading to land trust
effectiveness, is employed (Ragin 1987; Ragin 2000). This method tests for
multiple causal relationships that lead to effectiveness among the population of
land trusts in Colorado. It is a method of analysis that is relatively new to social
science and particularly new to research in policy studies. It recognizes that
multiple combinatorial relationships may explain a single outcome within a
population. Unlike traditional quantitative inferential techniques that are highly
reductionist in terms of their explanatory models, this method welcomes
59


diversity in causal relationships. A discussion of this combinatorial analysis
follows the section on data collection.
Variables
The dependent variable throughout this dissertation is land trust
effectiveness measured as the number of actions (easements, fee-simple
ownership, and third-party transfers) per year of a land trusts existence. This
metric has been chosen for specific reasons. First, it is measurable across land
trusts. While many land trusts may vary in their mission, they all must either
hold conservation easements or fee-simple title or participate in a third party
transfer. Second, these actions are the most tangible product of land trusts. In
fact, one of the most unique features of land trusts is their ability to hold
conservation easements. Finally, the factors that determine whether or not an
easement or purchase will occur are independent of the number of acres. The
transaction costs associated with effecting an easement or purchase are not
necessarily proportionate to the physical size of the property. Location,
competing uses, and the opportunity cost of land protection are as likely to
influence the transaction costs of any one land trust activity. Therefore, the size
of a protected parcel is less relevant as a measure of effectiveness than the action
itself. From the standpoint of this research, there is no reason to assume that
60


numbers of acres protected are in themselves a measure of land trust
effectiveness.
The term effectiveness is contentious since, without a specified
definition, what one land trust may consider effective (e.g., preserving an
endangered species) may be quite different than another land trusts idea of
effectiveness (e.g., maintaining agricultural lifestyles). It is important to
remember that this dissertation is, in part, motivated by change in the institution
of property that guides the use, access, and rights of exchange for natural
resources, and, furthermore, how land trusts function as part of this evolution.
Across all land trusts, the measure that best represents their role in reestablishing
public values in the use, access, and exchange of land and other resources is the
number of activities they undertake.12 Other researchers may wish to undertake
other interpretations of land trust effectiveness based on other motivating
questions.
12 Certainly, it is easy to imagine each land trust arguing that their measure of effectiveness is
unique to their mission and, to a great degree, they would be correct. Nevertheless, this
dissertation examines factors, across land trusts, that dictate effectiveness. In order to establish
comparability, a single, common measure must be chosen and that is the number of activities
accomplished.
61


Research Hypotheses
As discussed in the literature review, this dissertation draws on a number
of parallel literatures to explain effectiveness among land trusts. The need for
drawing on parallel literatures lies in the role of land trusts in creating public
goods on private property. As discussed above, land trusts are relaxing these
once polar positions of property. The form of property that emerges is not
purely private, public, or common but the question remains: what is it about
these particular organizations that explain their success in evolving our
conception of property? Research into land trust effectiveness must step back
and examine potential contributions from a number of bodies of theory.
The process of drawing on a number of potential contributions and examining
them in a combinatorial fashion requires a hybrid approach to the typical
research question/hypothesis approach. This dissertation seeks to advance an
explanation for land trust effectiveness but is not guided by a previously
established model such as found with common property (e.g., Ostrom 1990).
Land trust effectiveness may, instead, be explained by an array of contributing
literatures as presented in the literature review, above. To some degree, it
follows the broad, searching style of grounded theory where the researcher
builds toward theory rather than tests a theory (Creswell 1994; Strauss and
62


Corbinl998). Specifically, individual variables drawn from these literatures are
tested for statistical significance with respect to their ability to explain variation
in effectiveness. Then, significant variables are advanced for inclusion in the
combinatorial analyses.
Consequently, this dissertation embodies a fundamental research
question: What are the factors, and in what combinations, that contribute to
the effectiveness of community-based land trusts in Colorado? Four
contributing hypotheses are also posed to provide a structured sequence of
analysis leading to the combinatorial analyses. These hypotheses are presented,
below. Three of the hypotheses test contributions from the literature. A fourth
tests the influence of factors exogenous to land trusts' operations and
relationships.
Hi: Higher degrees of social capital are associated with greater effectiveness
among land trusts.
This hypothesis looks to the role of social capital and the degree to
which it guides land trusts' ability to initiate and complete activities.
Organizations with high degrees of social capital will have positive reputations
with their community, a shared understanding of a problem and its solutions,
63


and well-formed networks throughout their community that reflect trust and
confidence (Stiglitz 1999). Land trusts traditionally take two approaches to
selecting or identifying properties that they may protect. They may either
actively solicit a landowner which can be expensive in terms of labor hours,
mapping, and even computer software (GIS) or they may take a passive
approach, allowing their reputation and their capacity to generally educate their
community and guide prospective landowners to the land trust. The activities
land trusts use to protect properties (conservation easements, fee-simple
ownership, and third-party transactions) are universal regardless of the means by
which the properties are identified. Nonetheless, the manner by which
community-based land trusts use to originally identify properties reflects their
relations with their community and may potentially reveal their capacity to
undertake those activities in the first place.
Similarly, land trusts that have established positive networks within the
community and especially with the easement donors may avoid monitoring costs
through volunteerism and willing compliance. Such compliance may also help
avoid the additional costs of enforcing a violated easement. In summary, this
hypothesis tests the role of a land trust's relations with its community in terms of
the number of activities the land trust completes per year. It tests whether a land
64


trust can be a means toward effective land protection as an autonomous agent or
if efforts to engage with the community through education and reputation allow
greater efficacy.
H2: Effective land trusts engage the support of concentric regimes.
Singleton and Taylor's (1992) construction of community is compelling
as an explanation for purely endogenous solutions but even casual observation
suggests that theirs is a rare, if not abstract, case. Others scholars recognize a
cooperative (Berkes et al. 1991) or nested (Ostrom 1990) role for government
and non-governmental organizations. Land trusts and other nonprofit
organizations regularly look to external support to accomplish their activities.
As Ostrom (1992, 349) observes, "it is just as important to see how larger
political regimes may facilitate the capacity of individuals to achieve agreements
as it is to see the importance of community in facilitating solutions. Clearly,
the roles and relationships that land trusts hold with their communities are
important, but it would be unwise to turn a blind eye to the critical role of
government and other supporting organizations in facilitating local action. In
fact, this question is important in understanding the precise role of community-
based action in natural resource protection.
65


tax themselves to preserve open space. The existence of open space programs
demonstrates a degree of comfort with active public sector management of
natural resource protection. Many parts of Colorado are politically conservative
with respect to land ownership and regulation, opposing formal public land
protection programs. Other parts of Colorado simply do not have sufficient tax
base to support such programs even if land protection would otherwise be
supported.
Finally, a number of exogenous variables were included to measure
sense of threat. Ostrom (1992, 345) notes that while factors such as community,
support from other regimes, and political entrepreneurs may be important to the
success of community-based groups, there is a "ceterisparibus clause"
suggesting that other factors, including the attributes of the resource to be
managed, influence effectiveness as well. Land trusts have historically emerged
as a response to a perceived threat from development. To measure this sense of
threat, variables measuring county-level changes in agricultural land conversion,
housing development, and population change are also included.
67


Data Collection
Colorado has thirty local land trusts.13 The executive directors of twenty
seven of these land trusts were administered an in-person interview during the
late summer and fall of 2000 (see Appendix A). The interview instrument was
approved by the University of Colorado at Denver's Human Subjects Research
Committee in May 2000. Three of the thirty local land trusts were not
interviewed because they could not be contacted or the land trusts were no
longer in operation. All land trusts that were contacted agreed to the interview.
An in-person interview was selected as the appropriate instrument for
two reasons. First, this research describes land trusts in Colorado and examines
factors that contribute to land trust effectiveness. Unlike a broad, descriptive
investigation of land trusts, the information that was sought often required open
ended and semi-structured questions (Rossi et al. 1983). In-person interviews
are best suited to handle questions that do not force respondents to select among
a specified group of answers. Second, there is no single underlying model to
structure this investigation. Additional or unexpected information might have
arisen that the interviewer would want to pursue. In these cases, respondents
13 Colorado also has local chapters of six national land trust organizations. These were not
included as they were outside of the research scope.
68


also take on characteristics of informants who lend new insights, often with
additional probing (Yin 1994, 84). This only could be achieved via personal
interview. Finally, some of the questions required research on the part of the
respondent. Such questions as these frequently are left unanswered in mail
surveys (Rossi et al 1983). Questions that required research by respondents
were sent in advance and reviewed when the interview took place.
Each land trust was initially contacted with a letter describing the
research project, length of interview and content, confidentiality, and university
and external funding affiliations (see Appendix B). Executive directors then
were contacted by telephone to establish interview appointments. Prior to being
administered an interview, each interviewee was presented a consent form for
signature (Appendix C). The average interview was approximately one hour
and fifteen minutes with interviews ranging from forty-five minutes to two
hours. Each interview was audio taped to corroborate data transcribed during
the interview.
The interview instrument contained both open and closed-ended
questions covering: 1) descriptive questions regarding the land trust, its mission,
age, and origins; 2) donations, membership, and staff; 3) measures of
69


community; 4) external support such as grants and pro bono services; 5) political
entrepreneurship; and 6) procedures regarding identifying, negotiating, and
monitoring and enforcing purchases and easements. The interview data were
supplemented with land use data and socioeconomic data from the Colorado
Department of Agriculture, U.S. Department of Agriculture's Natural Resource
Conservation Service, and the U.S. Bureau of the Census. Other land trust data
were provided by the Land Trust Alliance and the Colorado Coalition of Land
Trusts.
Combinatorial Analysis
There is a tension between quantitative and qualitative methods within
the social sciences. Some quantitative researchers may appear wedded to
systematic statistical analyses while some qualitatively-oriented researchers
charge that quantitative analyses, while purporting to be more sophisticated, are
in fact less comprehensible and less productive in terms of providing what
Meehan (1994,158) would call a critical assessment of the human condition
(King, Keohane and Verba 1994; White and Adams 1994). The contrast
between methods may be called causal versus telological (Ragin 2000). Much
of the tension is stylistic rather than substantive, ignoring the fundamental goal
of research: to make descriptive or explanatory inferences on the basis of
70


empirical information about the world (authors emphasis) (King, Keohane and
Verba 1994, 7).
The distinction between the methods can be summarized (though
perhaps over-simplified) as follows. Ragin (1987, 53) describes the
quantitative, or variable-oriented, approach as theory-centered and less
concerned (as compared with qualitative case-oriented methods) with
understanding specific outcomes or categories of outcomes and more
concerned with assessing the correspondence between relationships.
Hypotheses are derived from preexisting theories and tested with empirical
examination aided by multivariate statistical methods (Kangas 1994).
Complexity is eschewed in favor of generality. Cases (referred to as
observations) are homogenized.14 Causality is additive-linear (Ragin 2000) in
that if Y=f(X) then Xj and X2 explains more variation in Y than Xi alone, X^ X2,
and X3 explains more variation than Xj and X2 and so forth. Ideally, the number
of observations greatly outnumbers the explanatory variables. The quantitative,
or variable-oriented, approaches are exemplified by statistical techniques such as
14 Consider, for example, the assumption within traditional neoclassical microeconomics of the
purely rational actor. Such an homogenizing assumption presumes all individuals share a similar
preference framework (more is better, preferences are transitive). This framework guides
microeconomic theory and individuals examined within cross-sectional analyses must be simply
assumed to share these preference characteristics.
71


linear regression and much of the modeling found in econometrics.
Where quantitative methods seek to generalize over larger populations,
qualitative or case-study approaches seek detail and complexity within one or a
few cases. Case studies seek answers to how and why questions and often focus
on historical or political context for these answers (Yin 1994). In their most
conventional form, case studies view cases as configurations of aspects and
causes as conjunctural and plural (Ragin 2000,14). In other words, different
combinations of causal features may explain similar outcomes. Like
quantitative methods, qualitative methods are inferential in intent (King,
Keohane, and Verba 1994). Qualitative methods are most attractive where
theories are at early development stages, concepts are vague, and measurement
conditions fail the requirements of quantitative methods such as large-N,
homogenous cases, well-defined populations, and adequate sampling (Ragin,
Berg-Schlosser, and de Meur 2000).
Clearly, both approaches have merits as well as weaknesses. Qualitative
or case-oriented strategies are effective in identifying multiple and conjunctural
relationships stimulating what Ragin (1987, 52) calls a dialogue between ideas
and evidence." Unfortunately, qualitative case-studies quickly become
72


cumbersome in even moderately large-N situations. Quantitative methods favor
parsimonious explanations of causality providing an efficient, if somewhat
abstracted, representation of reality. Quantitative methods are weakened in
situations of even relatively small sample sizes. As degrees of freedom are lost,
confidence in variables falls. Many of the questions in policy research involve
relatively small samples or populations. There are generally more than one or
two cases but not enough to provide confident results from regressions and other
quantitative techniques. Such is the situation with an examination of land trusts
in Colorado that is limited to 27 cases.
The search for a method that integrates both has an inherent appeal.
King et al. (1994) propose imposing elements of quantitative research to
improve the robustness of qualitative research. Creswell (1994) presents
combined strategies by separately engaging both quantitative and qualitative
strategies. The method that will be discussed here fuzzy set Qualitative
Comparative Analysis (fsQCA) and its precursor, Qualitative Comparative
Analysis (QCA) are provided by Ragin (1987,2000) and integrate some of the
features of qualitative research into a quantitative method. They were selected
because they are the best fit for the nature of the data and the early stage of
theory building to which this research contributes.
73


Qualitative Comparative Analysis
Qualitative comparative analysis is the methodological embodiment of
what Ragin (1987) calls "diversity oriented research." Diversity oriented
research views cases as configurations of causal explanations and disaggregates
populations into types. This is contrary to traditional quantitative methods that
homogenize populations. QCA acknowledges that similar outcomes may be
achieved by different combinations of causal conditions. In other words, QCA
examines cases as wholes and as configurations of potential causes (Hicks
1994). Furthermore, while quantitative techniques are hindered by
homogenizing assumptions of populations, both qualitative and quantitative
techniques struggle with moderately sized data matrices (Ragin et al. 2000).
Moderate-N data quickly threaten quantitative techniques with diminished
degrees of freedom while even a few caises confound traditional interpretive
qualitative analyses. QCA accomodates data matrices that fall in this gray area.
At this point, it is important to distinguish between QCA and more
familiar typology construction techniques such as cluster analysis. Cluster
analysis is a class of analytic techniques that separate data into groups or
clusters such that within-group differences are minimized and between group
74


differences are maximized (Everitt 1993). Cluster analysis can be described as
polythetic (Everitt 1993), in that classifications are based on many
characteristics of the objects that are being investigated. In contrast, monothetic
analysis will use a single characteristic to create a separate classification.
Consider the example provided by Ragin (2000, 71). "A white, suburban,
middle-class, male professional with three children and a mortgage who votes
for the Republican party differs by only one trait from a similar individual that
votes for the Socialist party." Using cluster analysis, these two individuals
could easily be grouped in the same cluster although, as Ragin notes, being
trapped on an elevator with one could be a significantly different experience
than being trapped with the other. Monothetic analysis techniques such as QCA
allow the opportunity for a single characteristic to define a separate group.
Cluster analysis often overlooks the significance of that single characteristic.
This is an important distinction when venturing into the theory building stage of
analysis.
In addition, cluster analysis does not associate independent variables
with a dependent variable. Again, it is purely taxonomic, looking for simple
groupings of data. Cluster analysis techniques "have taken their place alongside
other exploratory data analysis techniques.. .The term exploratory is important
75


here since it explains the largely absent 'p-value,' ubiquitous in many other areas
of statistics, although there have been attempts to set some clustering techniques
in an inferential framework.. .In the main however, clustering methods are
intended largely for generating rather than testing hypotheses" (Everitt 1993, 10;
author's emphasis). The goal, then, of cluster analysis is to give structure to a
complex mass of data (Everitt 1983). Cluster analysis is taxonomic. It is not
associational. As a consequence, cluster analysis makes no explicit attempt to
explain variation in a dependent variable. QCA does explore associations
between independent and dependent variables.
QCAs advantages over probabilistic inferential techniques (e.g., least
squares regression) can be summarized as follows. First, QCA allows the
examination of combinatorial theories. Inferential techniques are quickly
bogged down with problems of multicollinearity when considering the
possibility of conjunctive relationships among independent variables (Amenta
and Poulson 1994). Second, QCA expects causal heterogeneity, the
possibility that multiple combinations of causes may produce a single outcome
(Amenta and Poulson 1994). Third, QCA examines combinatorial relationships
both in terms of the absence and presence of a factor or outcome (Ragin 1987).
In other words, the absence of a variable in combination with another variable is
76


as revealing as its presence. Fourth, and maybe the most important contribution
of the QCA approach, is its contribution to theory building. QCA avoids
speaking about causation in probabilistic terms, instead seeking only the
attributes or circumstances (or their absence) that lead to an outcome (Coverdill,
Finlay, and Martin 1994). The researcher is forced to consider the context in
which causal relationships occur, especially when contradictory outcomes
emerge. This is ideal given the stage of knowledge regarding effectiveness
among community land trusts in Colorado. For example, while there is a wealth
of prior research regarding factors such as political entrepreneurship or social
capital, it is unclear how these factors relate in a conjunctive fashion. The
initial, sifting inquiry of purely inductive case study investigation is replaced by
an intermediate stage, only slightly more deductive, that seeks causal
relationships without relinquishing complexity.
Ragin originally operationalized diversity oriented research with a
simple software application called QCA15 (Drass and Ragin 1992). QCA uses
Boolean algebra to compare cases as configurations of set memberships and for
elucidating their patterned similarities and differences (Ragin 2000,120).
151 use QCA in italics to represent the software package and QCA in regular print as an acronym
for Qualitative Comparative Analysis.
77


Boolean algebra uses binary data to represent conditions or states. In other
words, explanatory data must either be collected or transformed to equal either 1
(present) or 0 (absent). Likewise, dependent variables are arranged as 1 (occurs)
or 0 (does not occur). Data may be described as either present in the set or
absent from the set (Ragin 1987). When all data are collected, they are
presented in a truth table that organizes conditions, outcomes, and the number of
instances for each combination of conditions and outcomes. A hypothetical
truth table is presented in Table 4.1. This table is drawn from Ragin (1987) and
considers causes for military regime collapses. Three causes are considered: (A)
represents the existence of conflict between older and younger officers, (B)
represents death of a powerful dictator, and (C) represents CIA dissatisfaction
with the regime.16
16 The reader should note that one cause represents an internal effect, one cause represents an
external influence, and one cause focuses on the effect (in this case, death) of one individual.
This is very similar to the construction of this dissertations measures of land trust effectiveness.
78


Table 4.1: Hypothetical Truth Table Showing Collapse of Military
Regimes_______________________________________________________
Condition Regime Failure Number of Instances
A B c F
0 0 0 0 9
1 0 0 1 2
0 1 0 0 3
0 0 1 1 1
1 1 0 1 2
1 0 1 0 1
0 1 1 1 1
1 1 1 1 3
Source: Ragin (1987), p. 90.
The critical point to recognize in QCA and the table above is that both
presence and absence of an explanatory factor are important in analyzing
outcomes when considering combinatorial relationships. Notice, for example,
that regime failures occur both when there is and is not conflict between older
and younger officers. Traditional quantitative analyses would be most
concerned with the frequency of causal conditions, eliminating those instances
that are less frequent. QCA is concerned with the different combinatorial
relationships.17 For example, the observation that the existence of (A) combined
with the nonexistence of (B) and (C) leads to regime failure while the existence
of (B) combined with the nonexistence of (A) and (C) does not lead to regime
17 The researcher may wish to eliminate instances with low frequencies but that decision should
be made within the terms of that particular research program.
79


failure is particularly revealing about the combinatorial nature of these causes.
These relationships would simply not appear in traditional quantitative analysis
that looks for additive effects to explain a particular outcome such as regime
failure.
Fuzzy Set Qualitative Comparative Analysis
Unfortunately, QCA is saddled with a particular feature that limits its
analytic strength: QCAs use of binary data. Most data in social sciences are not
binary in nature. True, some data are binary (e.g., male or not male) but most
data are continuous to a degree. In other words, their membership in a set
ranges from full to non-existent with the potential for any gradation of partial
membership. Consider what it means for a cross-national examination of
political systems. A researcher may wish to identify whether or not a country is
democratic but simply identifying a country as democratic or not may seem
rather empty. How democratic is that country? Likewise, a study of health care
programs may wish to examine customer satisfaction. Attempts to quantify
satisfaction may be reduced to Yes/No responses or Likert scales but such
answers lack the richness that qualitative responses may provide. Yet, how
should the researcher manage multiple cases of customer satisfaction?
Traditional qualitative methods will quickly be exhausted with multiple cases.
80


QCA has already provided a method that accommodates diversity within
a population by allowing a combinatorial analysis. What remains is to expand
the richness of the data that make up the dependent and independent variables.
This is accomplished by the use of fuzzy sets. Fuzzy sets accomplish the same
goals as crisp sets (fully absent or fully present) but also permit membership
scores between 0 and 1. Scores of, say, 0.2 or 0.3 indicate weak membership,
more out of the set than in. Scores of 0.8 or 0.9 are more in the set than out
(Ragin 2000). Data, then, must be converted into fuzzy set values lying between
the extrema of 0 and 1. The software that will analyze fuzzy sets is fsQCA
(Drass and Ragin 1999), a revision of QCA that retains QCAs combinatorial
analysis while expanding analytic diversity within set membership.
The idea of partial, or fuzzy, membership in sets, is more in keeping with
the data encountered in policy research. While certainly some data are crisp
(e.g., male or not male), much data is fuzzy (e.g., how satisfied are you with
your job; are you politically conservative?). Fuzzy sets bring formal logic and
generally qualitative formulations together (Ragin 2000). Consider adjectives
such as far, large, or strong." These adjectives struggle to be quantified
except within the context of the population. I would be considered tall within
the population that is my immediate family. I hardly qualify as tall in a much
81


broader population! For the social sciences, fuzzy logic contributes to what
Kosko (1993,146) calls a linguistic calculus.
Fuzzy sets formally originated with an article by Zadeh (1965) that
examined vague or multivalued sets sets whose elements belonged to it to
different degrees (Kosko 1993, 19). It was cooly received within the academic
community that, in the early 1960s, was enamored with the emergence of binary
processing in computers, not to mention behavioralism and reductionism in
general. The commercial market, originally in Japan, brought fuzzy logic to
the mainstream (Kosko 1993). Fuzzy logic had its modem emergence in
engineering and commercial machine intelligence such as in cameras (adjusting
for light conditions), air conditioners (responding to changes in air temperature),
and automotive engine control (responding to changes in air temperature and
pressure and road gradient) (Kosko 1993).
Fuzzy sets may be delineated in fairly coarse fashion where the only
scores can be 0, 0.5, or 1 or they may be delineated in a smooth fashion allowing
almost continuous scoring between 0 and 1. Such scoring is determined by the 18
18 "Positivism", argues Kosko (1993, 7), "works out well for scientists and mathematicians, since
it allows only them to speak. Everyone else utters 'meaningless' statements about the world and
life and morals and beauty.
82


nature of the data and the detail sought by the researcher (Ragin 2000).
Regardless of the type of scoring, the midpoint, 0.5, is critical and the scoring
should reflect this. The midpoint represents whether a case is more in the set
than not or vice versa. This means that the researcher must establish criteria for
full membership in a set, partial membership in a set, and nonexistence in a set.
Fuzzy Analysis of Land Trust Effectiveness
Independent variables for fuzzy qualitative comparative analysis are
selected based on the outcomes from the significance tests discussed above and
their correlation coefficients with the dependent variable. This approach, called
the significance approach by Amenta and Poulson (1994, 27), allows factors
that pass muster through traditional inferential techniques and then seek to
examine them in a combinatorial fashion. In Amenta and Poulsons (1994)
opinion, the significance approach is not the ideal approach from the standpoint
of employing fsQCAs foundation of quantitative analysis of qualitative data.
Ideally, combinatorial theories are tested with fsQCA but, with land trusts, no
such theories exist. "In the best of all empirical worlds, researchers would be
able to compare two or more combinatorial theories separately or
simultaneously" (Amenta and Poulson 1994, 30). This dissertation is not in
such a world. Nevertheless, the significance approach takes advantage of
83


fsQCAs combinatorial and small-N features while directly addressing the
original research hypotheses. Therefore, it is the best approach for this
application.
The first step in assigning fuzzy membership scores is to specify the
relevant domain to be assessed (Ragin 2000). This was accomplished in the
earlier stages of the research design. The domain is simply community land
trusts in Colorado. The second step involves defining the fuzzy sets from the
concepts guiding the investigation (Ragin 2000). In traditional quantitative
analyses, dependent and independent variables are simply organized and
tabulated whereupon the appropriate statistical analyses are employed. With
fuzzy sets, the researcher must assess the degree of membership in a set, not
simply its position on a continuum, relative to other cases (Ragin 2000,167).
It is not enough to merely divide all values in a variable by the highest value,
producing values between 0 and 1. There is a relevant range within which the
researcher must assign the scores. The researcher must consider what is the
midpoint in terms of the context of the research question. The midpoint
determines whether the case is more or less in the set. For example, what is a
large staff for a community land trust? Four employees? One land trust that
may have a staff of ten may certainly be large but is the unusually large number
84


relevant? There could well be no difference between four employees and ten in
terms of being large." The context of the explanatory factors within the
relevant domain define the fuzzy sets.
The final step is to determine the type of fuzzy set that is feasible for
each factor. Some variables are dichotomous by virtue of the manner in which
the particular question was asked. While these variables are technically crisp
sets, they are entirely compatible with fuzzy sets, their measure of membership
is simply crisp. Other variables may be compatible with extremely fine
distinctions while others may only allow modest distinctions such as small,
average, large."
At this point the data is ready to be analyzed with fsQCA. fsQCA allows
up to seven causal conditions and one outcome variable. Combinatorial analysis
recognizes the importance of the existence as well as the absence of a factor. In
other words, if all seven factors are included in the analysis, there will then be
14 tests of causal necessity. Results will indicate combinations of significant
factors that dictate land trust effectiveness in terms of their combined existence
and absence. Factors that are not significant are eliminated from the results.
Likewise, fsQCA will provide analysis of negative cases, cases where land trusts
85


are not effective. These, too, will be presented in terms of existence and
absence of significant factors.
fsQCA will analyze the data for necessity and sufficiency of the factors
that may dictate land trust effectiveness in both absence and presence for each
factor. Necessity simply means that the presence (or absence) of an independent
variable is necessary for the outcome; that is, it will be present in every
combination. If fuzzy membership scores for the outcome variable are less than
or equal to membership scores for a causal variable in all cases, that variable is
considered necessary (Ragin 2000).19 Necessity is reported for individual
variables, not in combinatorial fashion but necessary variables will be present in
every combination.
Of course, there is randomness to be expected in social science data as
well as factors that may explain outcomes but are not collected (Ragin 2000).
Consequently, tests of necessity and sufficiency must include probabilistic
measures. A researcher may wish to examine whether a particular factor is
"usually necessary" or a combination is "almost always sufficient." A
19 Sometimes no one factor will emerge as necessary or sufficient. Ragin (2000, 222) remarks
that "when causation is complex, no single cause may be necessary or sufficient."
86


benchmark proportion representing these linguistic qualifiers is used. Ragin
(2000) sets the qualifier "usually" equivalent to a test proportion of 0.65 and
almost always at 0.80. "The selection of the precise benchmark is to some
extent arbitrary, but it must be made explicit by the researcher and thus open to
debate" (Ragin 2000,110).
Sufficiency tests examine which variables, or combinations of variables,
either present or absent, are sufficient to explain an outcome (Ragin 2000).
Often, multiple combinations of sufficient conditions may emerge of which
some factors may be shared between combinations and others may not.
Furthermore, some combinations may include the absence of a variable that, in
another combination, was present. Obviously, the number of combinations will
be influenced by the test proportion, declining as the test proportion becomes
greater.
QCA is a technique that originally was designed to provide a means of
analyzing the deeper and more subtle aspects of qualitative analysis in a
quantitative format. This has tremendous implications for social science
research, much of which strains under the obligations of additive/probabilistic
methods. "Qualitative methods are advantageous when "theories are vague or
87


underdeveloped" (Ragin et al. 2000, 750). The interviews of land trusts provide
a mix of quantitative and qualitative data. As an investigation of effectiveness
among land trusts, this research takes advantage of two features of QCA. First,
QCA allows a quantitative analysis of what may be called a moderate-N data
matrix. Multiple regression techniques would be immediately suspect due to
low degrees of freedom. Second, QCA is combinatorial. This is especially
valuable when there is no guiding model to test, a condition that often lends
itself to qualitative approaches. Rather, this research draws on insights from a
variety of separate literatures to inform and suggest a theoretical model for land
trust effectiveness. As a consequence, QCA provides a superior vehicle for
theory building than would an inferential technique whose goals are to simply
test existing theory and infer applications over a greater population.
Summary
This chapter has introduced the research hypotheses and discussed the
methods of data collection and analysis. Included is a discussion of QCA, a
method of assessing combinatorial causality. While this method aids in
assessing causality when faced with small-N studies and qualitative data,
probably its most important contribution is in adding formality to the process of
theory construction. Examining effectiveness among land trusts is at a
88