Comparative network theory

Material Information

Comparative network theory an interpretive study of the Canada-United States Pacific Salmon treaty
Mingus, Matthew Scott
Publication Date:
Physical Description:
xv, 215 leaves : illustrations ; 28 cm


Subjects / Keywords:
Pacific Salmon Treaty Act of 1985 (United States) ( fast )
Interorganizational relations ( lcsh )
Social sciences -- Network analysis ( lcsh )
Interorganizational relations ( fast )
Social sciences -- Network analysis ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 197-215).
General Note:
School of Public Affairs
Statement of Responsibility:
by Matthew Scott Mingus.

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:
42618299 ( OCLC )
LD1190.P86 1999d M56 ( lcc )

Full Text
Matthew Scott Mingus
B.A., University of Denver, 1988
M.P A., University of Victoria, 1991
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Administration

1999 by Matthew Scott Mingus
All rights reserved.

This thesis for the Doctor of Philosophy
degree by
Matthew Scott Mingus
has been approved

Mingus, Matthew Scott (Ph.D., Public Administration)
Comparative Network Theory: An Interpretive Study of the Canada-United States
Pacific Salmon Treaty
Thesis directed by Richard J. Stillman II
Sparse research has compared the network systems of public administrators
between nations or between subnational units of government, although networks are
seemingly as important to managing cross-border resources as individual agencies or
diplomatic relations once were. This dissertation advances the field of comparative
network theory by developing written descriptions of networks, comparing these
network systems, and proposing the Comparative Network Framework (CNF) as a tool
to assess network compatibility when subnets form supranets.
The international allocation of Pacific salmon between Canada and the United
States is a protracted dispute that has defied resolution in spite of the 1985 Pacific
Salmon Treaty (PST). The equity principle recognizes that nations have the right to
benefits commensurate with the value of their production of fish species even though
these species spend much of their lives migrating throughout the ocean (i.e.,
anadromous salmon and steelhead trout species). An analysis of archival documents,
face-to-face interviews, and further rounds of input from the interviewees and

additional informants compliment one another in the interpretive, hermeneutic
methodology that is utilized for this thesis.
The theoretical frame utilized in this study is network theory, broadly
interpreted. An expansive review of network theory is included and identifies six
models of networks from the public administration literature: policy networks, network
management, advocacy coalitions, intergovernmental networks, interorganizational
networks, and issue networks. This review assists in the interpretation of concepts that
emerged as the written descriptions of the Oregon, Washington, Alaska, and British
Columbia subnational networks were prepared and revised. The Marsh and Rhodes
Typology is initially used to compare these subnational networks to determine if
significant differences or similarities account for some of the inability to reach
agreement on the perplexing equity principle. Next, the CNF is developed to explain
some of the notable incompatibilities with regard to the PST and, more generally, to
predict the likelihood of compatibility when subnetworks form a supraneL This
analysis suggests that several flaws with the PST could have been anticipated and
possibly even compensated for in the design of the Pacific Salmon Commission.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.

I dedicate this thesis to my daughter, Rene Elizabeth Mingus, for her ceaseless
affection and her constant reminders that life is more fun outdoors than indoors.

I appreciate the faculty of the Graduate School of Public Affairs, particularly
Richard Stillman, Linda deLeon, and Peter deLeon, for allowing me to push myself
hard through my PhD. program and for providing support after those occasional
academic and personal crises. This thesis and my degree would be unrealized goals
if not for their moral, academic, and financial support.
My entire dissertation committee has been remarkable for identifying conceptual
inconsistencies in my drafts while Professors Stillman and Steelman have provided
detailed editorial feedback. I accept responsibility for any remaining flaws, but they
must be acknowledged for greatly improving the quality of this final product.
Acknowledgements are also due to:
the informants who each provided some of their valuable time, often grudgingly,
but always in the hope that I might have something useful to say at the end of
the day;
the Graduate School of the University of Colorado at Denver for providing
partial support for my fieldwork through the Graduate Research Opportunities
Grant program;
the Graduate School of Public Affairs for providing financial support to allow
my doctoral committee to meet in one place at one point in time; and
the School of Public Affairs and Administration at Western Michigan University
for providing a comfortable space and much encouragement as I completed
putting my thoughts into words.
Lastly, the support of my wife, Dr. Tabitha Mingus, has been unfailing when called
upon. Her editorial comments were appreciated, but her emotional support was far
more valuable. She will always be Dr. Mingus the 1st.

Figures xiii
Tables xiv
Acronyms xv
1. INTRODUCTION................................................1
Thesis Overview...........................................1
Research Questions.......................................10
Practical and Theoretical Significance...................12
2. REVIEW OF THE LITERATURE...................................15
Six Models of Networks...................................17
Policy Networks......................................19
Network Management...................................20
Advocacy Coalitions..................................21
Intergovernmental Networks...........................22
Intel-organizational Networks........................24
Issue Networks.......................................25
Typologies of Networks...................................27

Rhodes Typology........................................31
Wilks and Wright Typology..............................33
Marsh and Rhodes Typology..............................34
U.S.-Canadian Diplomatic and Administrative Relations........38
Different Political Systems............................38
Comparative Research on Networks.......................40
High Degree of Informality.............................42
Three Other Issues.....................................44
Key Issues from this U.S.-Canadian Literature..........47
Chapter Summary..............................................49
3. METHODOLOGICAL CONSIDERATIONS..................................51
Interpretive Research and the Hermeneutic Circle.............53
Data Collection and Analysis.................................61
Archival Research......................................61
Field Interviews.......................................63
Review of Drafts by Interviewees and Additional Informants ...69
Modifications to the Original Plan...........................70
Strengths and Limitations....................................72
Remark on the Iterative Process and Degree of Agreement......75
Overall Network Structure....................................80

State of Alaska................................................84
Perspective on Equity....................................84
State Structure and PSC Representation...................86
Local and Statewide Organizations........................88
Tribal Connections.......................................90
National Government and Other States.....................90
Congressional Relations..................................92
Across the Border........................................94
British Columbia...............................................95
Perspective on Equity....................................95
Provincial Structure and PSC Representation..............98
Local and Provincial Organizations......................101
First Nations Connections...............................104
National Government.....................................105
National Parliament.....................................107
Across the Border.......................................108
State of Washington...........................................109
Perspective on Equity...................................109
State Structure and PSC Representation..................112
Local and Statewide Organizations.......................115

Tribal Connections
National Government and Other States...................122
Congressional Relations................................124
Across the Border......................................125
State of Oregon..............................................127
Perspective on Equity..................................127
State Structure and PSC Representation.................128
Local and Statewide Organizations......................131
Tribal Connections.....................................132
National Government and Other States...................133
Congressional Relations................................133
Across the Border......................................134
Five Overall Themes..........................................135
Internal Consensus Hardens U.S. Negotiating Position...135
Power Imbalance in the Supranet........................137
Vociferous Collection of Competing User Groups.......139
Design Flaws with 1997 Stakeholders Process...........141
Marketing to Potential Network Participants............144
5. FROM SUBNET TO SUPRANET........................................147
Globalization and the Network Heuristic......................150
Subnet Comparisons Using the Marsh and Rhodes Typology.......152

Supranets and the Comparative Network Framework (CNF).161
Questions Regarding Supranet Micro Compatibility.........166
A Final Commentary: Politics, Public Servants, and Balance.175
A. ITEMS FROM ARCHIVAL RESEARCH................................181
Locations for Document Collection..........................182
Alphabetical List of Collected Documents...................182
OF NETWORKS..................................................192

1.1 General Migratory Pattern of Pacific Salmon..................4
1.2 Area Covered by the Pacific Salmon Treaty of 1985............7
4.1 Bilateral Network on Pacific Salmon.........................81
4.2 State of Alaska Network.....................................87
4.3 Province of British Columbia Network.......................100
4.4 State of Washington Network................................114
4.5 State of Oregon Network....................................129
5.1 Subnational Policy Networks Located Using Marsh and
Rhodes Typology............................................155
5.1 (cont) Subnational Policy Networks Located Using Marsh
and Rhodes Typology........................................156

2.1 General Comparisons of Six Network Models...................29
2.2 Types of Policy Networks (Marsh and Rhodes Typology)........36
3.1 Four Main Lines of Questioning..............................65
3.1 (cont) Four Main Lines of Questioning.......................66
3.2 Informant Names, Titles, and Organizations..................68
5.1 Comparative Network Framework (CNF)........................163
B. 1 Potential Distinguishing Characteristics of Policy Networks
B.2 Potential Distinguishing Characteristics of Policy Networks
B.3 Potential Distinguishing Characteristics of Policy Networks

ADF&G Alaska Department of Fish and Game
B.C. Fisheries British Columbia Ministry of Fisheries
CNF Comparative Network Framework
DFAIT Department of Foreign Affairs and International Trade (Canada) (Canadas Department of State)
DFO Department of Fisheries and Oceans (Canada)
ESA Endangered Species Act of 1973 (U.S.)
ILOS International Law of the Sea (general)
EPSFC International Pacific Salmon Fisheries Commission
NMFS National Marine Fisheries Service
NPFMC North Pacific Fisheries Management Council
ODFW Oregon Department of Fish and Wildlife
PFMC Pacific Fisheries Management Council
PSC Pacific Salmon Commission
PST Pacific Salmon Treaty
WCVI West Coast of Vancouver Island
WDFW Washington Department of Fish and Wildlife

In the long run, even in the short run when you look at this, I think we have
mismanaged where we havent had an agreement. Not on purpose, but
because we are so focused on trying to manipulate our fisheries for
advantage, or avoid disadvantage, that we lose sight of some of the signals
coming in out of either fishery monitoring or out of resources monitoring.
This is part science and part art. When you lose attention and dont read
the signals correctly you could make some major management errors.
Dennis Austin (1998b)
Washington Department of Fish and Wildlife
Thesis Overview
The Pacific Salmon Treaty (PST) has been the source of much tension
between Canada and the United States since 1994 when negotiations failed to reach
agreement on annual catch limits. Each jurisdiction struggled to maximize its
access to the available resource while often arguing, at the same time, to increase
salmon escapement in the name of conservation. This adversarial process led to a
mismanagement of the salmon resource as highlighted in the opening quote.
A breakdown over the interpretation and implementation of the PSTs equity
principle the concept that each nation should have a harvest equivalent to the
value of its production has been the primary issue in recent disputes. This issue
has been extensively studied from an economic, legal, biological, and fisheries

management viewpoint, yet precious little energy has been devoted to studying the
dispute in terms of public administration.
This thesis falls into this latter category because it (1) develops descriptions
of the networks used by senior public administrators to help determine the positions
they take at the negotiating table; (2) compares these subnational network structures
for Oregon, Washington, Alaska, and British Columbia (B.C.), with a view toward
making practical suggestions; and (3) proposes a Comparative Network Framework
(CNF) that is worthy of additional research. As such, this thesis makes a significant
contribution to the public administration literature for practitioners and theorists.
Within this dissertation networks primarily within a nation are referred to as
subnetworks or subnets while the larger networks they form when they reach across
national boundaries to interact with other subnets are referred to as supranets. The
most general statement of this thesis is that it develops a preliminary model to
predict the compatibility of subnets when they form a supranet. An exploratory
approach based on one binational treaty provides the context for this research.
Sometimes looking at an old problem in a new way can produce extremely
beneficial outcomes. For example, when America stopped mourning automobile
fatalities and started treating crashes as preventable, improved automobile safety
and highway design, stricter licensing requirements, and enforced motorcycle
helmet laws combined to save more than 100,000 lives annually (Will 1992, A10).
Such may be the case with network theory and the Pacific Salmon Treaty.

In Fisheries, Pollution, and Canadian-American Transnational Relations,
Anthony Scott (1976,234) argued that population growth and new technologies
both act to broaden the capacity to exploit resources, seldom to the benefit of the
environment. The exploitation of transboundary renewable resources such as a
particular fishery is marked by crossborder externalities while the benefits are less
controversial, the undesirable spillovers are the subject of much friction. Scott
(236) suggested that, The salmon fishery is an example of a resource that gains
greatly in value from agreements or organizations to reduce the externalities that
would be suffered by both nations.
Canada and the United States negotiated the Pacific Salmon Treaty over a
fifteen-year period and signed it in Ottawa, Ontario, on January 28, 1985, at least in
part because each nation agreed with Scotts suggestion. Unfortunately, as Dennis
Austin (1998b) of the Washington Department of Fish and Wildlife (WDFW)
stresses in the opening quote, the pressure to get our fair share of the resource is
strong and leads to mismanagement by all jurisdictions.
As illustrated in Figure 1.1, Pacific salmon are an anadromous species
meaning they undertake a lifelong migration (two to seven years long) from the
freshwater rivers of their birth, through the saltwater ocean northward toward
Alaska and westward toward Russia, and then back to the exact location of their

Fig. 1.1. General Migratory Pattern of Pacific Salmon
Source: Department of Fisheries and Oceans Canada

birth to reproduce and die.1 To the ocean fisherman, then, the location of the
salmons birthplace is largely unknown. To the conservationist and the economist
the salmons birthplace is highly relevant because most conservation and
productivity improvement measures involve protecting and enhancing the
freshwater habitat of the salmon (B.C. Ministry of Environment Lands & Parks
1997; Cone and Ridlington 1996; Emery 1996; Roos 1991). Therefore, creating
natural economic incentives to conserve and develop salmon stocks requires that the
jurisdiction investing in conservation is assured of its ability to reap the benefit by
maintaining the right to harvest fish equivalent in value to those originating in its
fresh waters and to allow adequate numbers of fish to return to spawn.
The PST is intended to provide this assurance to each nation (including
native tribes), as embodied in the principles agreed to in Article HI, including
subparagraph 1(b) that provides for each Party to receive benefits equivalent to the
production of salmon originating in its waters and subparagraph 3(a) emphasizing
the desirability in most cases of reducing interceptions (U.S. Department of State
1985). Interceptions, as one might guess, occur whenever fishermen from one
nation harvest fish that originated in the fresh waters of the other nation. This is
referred to as the equity principle and is in keeping with the International Law of
the Sea (ILOS) (Emery 1996; Hollick 1976,166).
1 Steelhead trout are also included in the PST although they account for a small portion of the fishery.
Steelhead do not always die after spawning.

The PST created the Pacific Salmon Commission (PSC) as a forum for joint
cooperation2 with regard to the areas specified in Figure 1.2, roughly from northern
California to the northern border of British Columbia (note that this overlaps with
the Alaskan border). Salmon form a meaningful element of the Pacific Northwest
economy, potentially worth more than $1 billion annually in direct economic
benefits, although prices are extremely low at the present time (Koberstein 1996,
11).3 This compares to a $400 billion gross regional product, which would suggest
that symbolic meaning is more significant than economic impact if not for the fact
that numerous individual towns and villages along the coast are almost entirely
dependent on the fisheries.
For B.C. this meant roughly $400 million in annual wholesale value from
1993 to 1995. In addition, hundreds of thousands of British Columbians and
visitors also engage in recreational angling contributing an estimated $1.2 billion
annually in direct and indirect benefits to the B.C. economy (B.C. Ministry of
Environment Lands & Parks 1997, 8). For comparative purposes, Alaskas total
salmon production value (excluding roe) was slightly more than $1 billion in the
2 The PSC subsumed the International Pacific Salmon Fisheries Commission (IPSFC), whose
responsibilities for joint fisheries management in the convention waters area now rest largely on
the shoulders of the Fraser River Panel (Rutter 1997,3-4). The Southern and Northern Panels
negotiate fishing regimes but lack this historical in-season fisheries management role.
3 World salmon prices have dropped by two-thirds in the past ten years, largely because of growth in
the netcage (formed) salmon industry (Canada Quarterly 1998; Glavin 1998).

peak years of 1989 and 1992, dropping to roughly $838 million in 1996 (Knapp
While catch statistics for Oregon and Washington are substantially lower
and increasingly recreational, the point is that a transboundary renewable resource
of this magnitude can produce exasperating management problems as is the case
with Pacific salmon. For example, in Summer 1997, B.C. fisherman blockaded the
Alaskan ferry Malaspina in the harbor at Prince Rupert, B.C., because of the lack of
an agreement to reduce U.S. interceptions of B.C. fish (Wood 1997). In addition,
B.C.s Premier (1997), Glen Clark, cancelled the U.S. seabed lease for the Nanoose
Torpedo Range along northern Vancouver Island in an attempt to bring the
Americans back to the negotiating table.
Perhaps the key factor complicating this issue is the disparate growth and
decline of salmon fisheries in this region. From 1974 to 1995 Alaskas catch has
grown steadily from roughly 60,000 tonnes to 450,000 tonnes; B.C.s catch bounced
all around between 40,000 to 110,000 tonnes, and has declined steadily since 1995;
and the combined Oregon and Washington catch remained below 40,000 tonnes,
nearing 10,000 tonnes in recent years. In 1996, the combined Oregon, Washington,
and B.C. catch of Pacific salmon was approximately 5% of the Alaskan catch (catch
statistics from Munro, McDorman, and McKelvey 1998,29).
4 The decline in production values has more to do with reduced prices for salmon on the world
markets than with reduced production volume. In fact, the 1996 volume exceeded that of the 1989
and 1992 peak production vcdue years (Knapp 1998, 1-2).

While fisheries management for Pacific salmon stocks has been improved
because of the PST, the equity principle has never been implemented even though
the U.S. and Canada reached agreement on the extent of interceptions in 1989
(Canada Department of Fisheries and Oceans 1996). The B.C. Fisheries Strategy
(B.C. Ministry of Agriculture Fisheries and Food 1997a, 14) claims that U.S. fishing
fleets, particularly from Alaska, have increased interceptions 50% while B.C. has
reduced them by 30%, resulting in an estimated loss to B.C. of C$60 million a year.
Bilateral agreement exists on the numbers of intercepted fish, but not on the
complex issue of how to value the catch (Smith and Steel 1997).
Govemment-to-govemment negotiations failed to reach agreement in 1994
and 1995 and then an international mediation process with New Zealand
Ambassador Christopher Beeby failed to reach an acceptable agreement in February
1996 (Hunter 1997; Westneat 1997). Next, a direct stakeholder participation
process failed to reach a deal in the Spring of 1997 (Canada 1997; U.S. Department
of State 1997). Former Environmental Protection Agency Director William
Ruckelshaus and former University of British Columbia President David Strangway
were appointed as Special Representatives to reinvigorate the stakeholders
process (Sunde 1997; U.S. Department of State 1997) and their joint report
concluded that the PSC process is fundamentally broken and the stakeholders
process should not be revisited (Strangway and Ruckleshaus 1998).

Research Questions
The recent attempt to involve stakeholders directly in the negotiation process
was the genesis of the focus on network structures found in this thesis. This
research studied the network structures of senior public administrators in British
Columbia, Washington, Oregon, and Alaska with regard to the PST and the equity
principle. Networks is essentially an approach that seeks to describe how
individuals, businesses, and interest groups participate in the public decision making
process outside the realm of electoral politics. A key point made in the next chapter
is that the informality of U.S.-Canadian relations provides an opportunity for
networks to have real power because formal political and diplomatic processes are a
less powerful influence. The role of subnational governments (states and provinces
in this case) and possibly even multinational businesses and interest groups can be
substantial as international relations shift away from high level diplomacy toward
administrative relationships.
Although the PST is a binational agreement in form, the subnational
governments are key players, especially in the U.S. where the implementing
legislation gives the states voting rights on the PSC (U.S. House 1988, section
3632). This marginalizes the national voice in the U.S. to a non-voting role (a role
of influence or persuasion rather than power or authority).
On the Canadian side, the national government has constitutional
jurisdiction; however, the 1997 Canada-British Columbia Agreement on the

Management of Pacific Salmon Fishery Issues and its precursor in the form of the
1996 Memorandum of Understanding on Fisheries Issues (B.C. Ministry of
Agriculture Fisheries and Food 1997b) make it appear that provincial influence has
become substantial on Pacific salmon issues.5 On February 18, 1998, Premier Clark
(2) established the B.C. Ministry of Fisheries (B.C. Fisheries) as a separate agency
to raise the visibility of this newfound provincial power and he appointed veteran
PSC representative Bill Lefeaux-Valentine to serve as its Deputy Minister. Soon
afterwards, he hired federal fisheries administrator Bud Graham as an Assistant
Deputy Minister.
Within this context, the primary research questions explored are:
1. Who are participants in the networks utilized by senior public
administrators in Oregon, Washington, Alaska, and British Columbia to establish
PST negotiating positions?
2. What characteristics of networks help define the relationships between
these networks?
3. How can a better understanding and comparison of these four networks
assist public administrators who seek resolution to this protracted policy dispute
(i.e., the inability to reach agreements on annual fishing regimes through the PSC
structure and process since 1994)?
5 Actions by Canadian Department of Fisheries and Oceans Minister David Anderson in June 1998
draw this assertion into question. These actions are discussed in chapter 4.

4. What can the research focused on the first three questions contribute to
the development of a field of comparative network theory?6
It is expected that a clear description of these networks (question #1) and an
understanding of how this relates to the types of networks described in the literature
(question #2) will provide a sense of how differences among network systems for
these four jurisdictions have added to the complexity of reaching agreement on the
equity principle (question #3). Meanwhile, the fourth question is an ideational or
conceptual enterprise an attempt to apply the details of the case and knowledge of
network theory to suggest a direction for comparative network theory within the
broader field of comparative public administration.
Practical and Theoretical Significance
The significance of research is often a touchy subject as some people look
for hands-on applications while others seek theoretical usefulness. This thesis
attempts to provide both, yet theoretical significance emerging from good fieldwork
is the main objective.
Network research is a relatively new and rapidly growing field of study
within public administration, and descriptions of existing networks of public
6 A fair amount is written about comparing networks; however, in this thesis comparative network
theory refers specifically to the cross-border or international aspects of this research. The Summer
1998 symposium issue of Public Administration and the closely related book Comparing Policy
Networks (Marsh 1998) are examples where there is a heavy use of the word compare, yet a
minimal focus on how networks relate to one another across borders.

administrators are difficult to find in the literature (OToole 1997,48). None could
be found with regard to the management of cross-border resources. Hugh Heclo
(1978) stated that issue networks defy description, hence the focus of the first two
research questions, which concentrate on describing the networks.
Next, the third question attempts to provide advice to practitioners based on
the first two research questions. The project would be remiss if it invested so much
time and energy meeting with members of the negotiating teams up and down the
Pacific coast and then overlooked their pressing need for advice and potential new
ideas. Nevertheless, should this advice prove ineffectual or fall on deaf ears, this
will not necessarily detract from the significance of the theoretical implications of
this thesis. The reason for this contention is that the project focuses on a fast
moving issue the tone seems to shift almost daily much of the time and so
getting the patients pulse at one point in time may not prove a good indicator upon
which to base the patients prognosis.
Finally, this thesis adds to the larger body of research on comparative public
administration.7 This is significant because comparative public administration has
waned since the 1970s when researchers gave up hope that a universal science of
public administration might be possible through comparative research (Heady
7 This thesis may also be useful to scholars in international relations theory where a stream of thought
examines network concepts by studying transgovemmental relations (between subnational units of
government), transnational relations (includes corporate and not-for-profit sectors), the impact of
diverse domestic structures, and interdependence (see Evangelista 1997; Knutsen 1997; Starr 1997).

1998). They were correct that too many variables seemingly exist for a grand
theory of public administration; however, this does not need to detract from the
usefulness of making limited comparisons. In particular, the Canada-U.S.
relationship represents the largest trading partnership in the world with the U.S.
importing $14 billion from Canada in 1997 and exporting $21 billion to Canada that
same year (Canadian Embassy 1998). These two governments interact routinely on
a wide array of policy issues, at both the political/diplomatic and
administrative/bureaucratic levels, and so it is worthwhile to examine how the
growing field of network theory applies to the extensive administrative relationships
that support this degree of economic interdependence.
In conclusion, this is a policy dispute that could be used to define
intractable. The first formal bilateral discussions on the topic were held in 1892
with agreements on the matter being signed at irregular intervals since the Bryce-
Root Treaty in 1908.8 By seeking to provide a new perspective on the issue one
that focuses on public administrators, networks, and subnational units of
government this research seeks to contribute something to the ongoing dialogue
other than another request for additional data about the problem. The concluding
chapter also develops more general ideas for the field of comparative network
theory based on the analysis of this cross-border issue.
* See Natural Resource Consultants (1986, Appendix 2) for an excellent review of this history of
joint cooperation on the Pacific salmon fishery.

Networks are structures of interdependence involving multiple
organizations or parts thereof, where one unit is not merely the formal
subordinate of the others in some larger hierarchical arrangement. . .
administrators cannot be expected to exercise decisive leverage by virtue of
their formal position. Influence in larger networks is more difficult to
document, predict, and model than it is in relatively simple two- or three-
party relationships.
Laurence J. OToole, Jr. (1997,45)
This review of the literature casts the net broadly in describing the existing
body of literature on network theory, applying OTooles definition of networks in
the opening quote, which stresses the interdependence of actors on the modem stage
and the diffuseness of their power. This information is presented in three sections:
(1) six models of networks, (2) typologies of networks, and (3) U.S.-Canadian
diplomatic and administrative relations. The first section discusses current
approaches to network theory in the public administration literature while the
second section discusses key attempts to pull this literature together into a coherent
framework for the field. The third section reviews literature on the U.S.-Canadian
relationship that is relevant to developing a focus on comparative network theory
across this particular international border. A brief summary concludes the chapter.

The Marsh and Rhodes Typology discussed in the second section represents
the current state-of-the-art in the literature with regard to thinking about the
different types of networks and how they relate to one another. This is crucial for
the present research because having an intelligent discussion of comparative
network theory will require a theoretical basis for comparison. While the Marsh and
Rhodes Typology proves useful in chapter five for making these comparisons, the
Comparative Network Framework (CNF) is introduced at that stage as an analytical
model to help determine the compatibility of networks as they interact across
national boundaries. This framework is presented at that stage because of perceived
inadequacies that arise when applying the Marsh and Rhodes Typology to cross-
border policy issues.
The flow of this chapter is that the ne..t section introduces the network
lingo while the second section focuses on conceptual attempts to harmonize,
unify, or consolidate this field of study. The third section examines literature more
directly related to the U.S-Canadian environment of the PST rather than network
theory in particular. This split focus is necessary in this literature review because
the details of the comparative environment are meaningful when applying network
theory to the particular case at hand while the broader theoretical concepts are
critical to developing the field of comparative network theory.

Six Models of Networks
The generality in OTooles definition may be an attempt to incorporate
research in the past several decades on policy networks (Blom-Hansen 1997;
Daugbjerg 1997; Klijn, Koppenjan, and Termeer 1995; Miller 1994; Rhodes 1991;
Rhodes 1997), policy communities and policy complexes (Rhodes 1997; Schneider
and Ingram 1997), advocacy coalitions research (Sabatier 1988; Sabatier and
Jenkins-Smith 1993), policy issue networks (Skok 1995), intergovernmental
networks (Agranoff 1990; Blom-Hansen 1997; Gage 1990; Mandell 1990; Rhodes
1991; Skogstad 1996), interorganizational networks (Chisholm 1998; Hanf and
Scharpf 1978; Harrison and Weiss 1998; Rogers and Whetten 1982), and issue
networks (Heclo 1978; McFarland 1987).
Network theories have come about as some theorists criticized the
instrumental logic of policy analysis by applying open systems concepts (Katz and
Kahn 1966) to place organizations in their environments (Hanf and Scharpf 1978),
some sought to explain networks as an outgrowth of broader models of pluralism
and corporatism (Jordan and Schubert 1992), and others tried to understand the real-
world influences and constraints on public executives (Heclo 1977,1978). A
common conclusion was that network theory developed as the popular iron triangle9
concept in political science gave way to less regimented views of pluralism. This
9 This is the notion that bureau executives, legislative committee members and staff and specialized
interest group staff members make most significant public policy decisions in cooperation with one
another (see Freeman 1965, chapter 1).

development is in keeping with the trend toward a more flexible, collaborative, and
responsive public sector predicted by futurists starting in the 1960s (Bennis and
Slater 1968; Toffler 1970).
This section discusses policy networks, network management, advocacy
coalitions, intergovernmental networks, interorganizational networks (sometimes
termed multiorganizational networks), and issue networks, in turn. Each of these
phrases is used to describe a different model for network theory, although the six
groupings cross paths on occasion. Literature that focuses on how decisions are
made for specific policy subsectors or subsystems is included under policy
networks. This includes literature pertaining to policy communities and policy
complexes as these phrases have developed in an effort to clarify the policy network
model rather than to offer a competing model.10
Hugh Miller (1994,379) started to sift through the intricate web of terms and
concepts that has developed by stating that policy networks were more focused than
policy communities and more interdependent than issue networks where action may
not be imminent. Miller (379) also highlighted the common focus away from the
inner workings of the organization and toward the mosaic of interactions among the
sometimes diverse, sometimes narrowly interested parties engaged in the struggle
over the allocation of values.
10 Policy community actually predates policy network within the field of political science, yet
policy networks is the term that cuts across numerous disciplines (Klijn 1998,29).

Policy Networks
With regard to the policy networks approach, Anne Larason Schneider and
Helen Ingram (1997) address policy networks, policy communities and policy
complexes in their book Policy Design for Democracy. These three concepts are
placed on a continuum from (1) policy networks, in which there is a common
allegiance to scientific and professional norms as well as a shared perspective and a
common body of knowledge; to (2) policy communities, which are made up of
highly educated and trained specialists in the science and technology of a particular
policy area and are a closely knit, coherent, stable, and closed form of networks;
to (3) policy complexes, which are loosely connected networks that encompass a
large diversity of opinion and where all the members agree that utilitarian
rationality and science should govern decisions (Schneider and Ingram 1997, 156-
157). Networks, in this model, are focused on establishing and maintaining public
policy within particular policy subfields. In particular, applying scientific expertise
to help develop policy is a key attribute.
A wide body of literature including sector-specific studies and studies on
particular public management topics exists using the policy networks approach.
Sector specific applications include the environment (Nelson 1996; Sigigneau 1996;
Skogstad 1996), human services and health care policy (Agranoff 1990; Boase
1996; Chisholm 1998; Harrison and Weiss 1998), agriculture (Daugbjerg 1997), and
corporate welfare (Verdier 1995). Applications on particular management topics

include privatization (Rhodes 1997, chapter 5; Zahariadis and Allen 1995),
intergovernmental management (AgranofF 1990; Blom-Hansen 1997; Mandell
1990; Rhodes 1997, chapter 6; Skogstad 1996), and supranational entities (Nelson
1996; Rhodes 1997, chapter 7). Thus it is not difficult to see that the policy network
concept is being discussed in an expansive array of public administration subfields.
Network Management
Network management, a relative newcomer in the literature compared to the
policy networks model, is focused on how political leaders and public
administrators can manage existing networks. Managing Complex Networks
(Kickert, Klijn, and Koppenjan 1997) is an edited book with chapters that cover the
public management spectrum from game theory and systems theory to the tools
approach (i.e., focus on choice of policy instruments) and implementation processes.
In the introductory chapter the editors rigorously object to the focus of the
new public management on developing a businesslike approach to government
focusing on performance indicators, deregulation, and privatization (3). Instead,
governance is described as a directed influence of social processes while its
subfield, public management, is equated directly with network management:
A wide variety of actions from different actors has consequences for
governance. Public management differs from governance in that it
focuses on the consciously and deliberately undertaken actions of
public actors to influence societal processes (or policy processes).
Therefore, public management is governance, but not all governance

is public management... One of the major challenges with which
public management as a form of governance is confronted, is to deal
with network-like situations, that is, situations of interdependencies.
Public management should therefore be seen as network
management (Kickert, Klijn, and Koppenjan 1997, 2-3)
This view of public management as network management comes Ml circle
from the original views of policy networks that primarily focused on explaining that
government policies frequently failed because experts and self-interested groups had
special access to the policy making arena (i.e., vested interests blocking legitimate
proposals for change). Managing Complex Networks argues that such shortcomings
in the real world are not intrinsic to policy networks and can be avoided through
adequate management (Kickert, Klijn, and Koppenjan 1997,9). While the network
management model still uses the term policy networks, the focus is on actively
controlling policy networks rather than on explaining that policy networks exist or
explaining how policy networks impact policy development.
Advocacy Coalitions
The Advocacy Coalition Framework (ACF) of Paul A. Sabatier and Hank C.
Jenkins-Smith is broader than policy networks because it seeks to fuse cognitive
and political factors in order to better explain change (Sabatier and Jenkins-Smith
1993, 110, emphasis added). It does not assume agreement among participating
individuals and organizations on the scientific method, or on any particular method.
ACF postulates that policy change over time is a function of (1) the interaction of

competing advocacy coalitions within a policy subsystem; (2) changes external to
the subsystem in, for example, socioeconomic conditions or the outputs of other
policy subsystems; and (3) the effects of stable system parameters such as
constitutional requirements on the resources of various subsystem actors (5).
The punch line is that advocacy coalitions are organized around common
core beliefs and thus are relatively stable over periods of a decade or more with
learning occurring primarily in areas related to secondary beliefs. Ultimately,
Changes in core elements of public policies require the replacement of one
dominant coalition by another, and this transition is hypothesized to result primarily
from changes external to the subsystem (Sabatier and Jenkins-Smith 1993, 5-6).
ACF has not received the attention throughout the range of public administration
subfields that has been documented in the policy networks literature.
Intergovernmental Networks
Intergovernmental networks have focused on local, regional, state, and
national governments (local or subcentral and central governments in the British
version) as key actors in interdependent relationships. Myma Mandell describes
intergovernmental management networks as exhibiting high degrees of interaction,
interdependence, trust, and areas of agreement, ultimately developing beyond mere
linking and toward being held together to achieve some purpose (Agranoff 1990,
58). Working within the parameters of the law and meeting the complex

requirements of a wide array of grant programs are also distinguishing
characteristics of intergovernmental networks. In brief, this model has generally
focused more on the managerial model than the public policy development model,
and frequently the goal has been to increase contacts and communication in order to
facilitate smoother joint implementation of programs.
Gage (1990, 130) extends intergovernmental networks by presenting a
model that includes three types of concentric or nested intergovernmental networks:
issue networks, intergovernmental policy networks, and implementation networks.
This suggests that the approach of intergovernmental networks may be an
application of the other network types to a specific issue area or subfield of public
administration, namely intergovernmental management. Indeed, R.A.W. Rhodes
(for example 1981, 1991, 1997) research on policy networks has generally focused
on intergovernmental issues although placed in the context of the evolving British
unitary state system rather than a federalist system (like the U.S. and Canada). This
is true of much of the research on policy networks, which demonstrates how
interconnected the various levels of government have become in modern times.11
Nevertheless, intergovernmental network literature has focused more on
administration and implementation than the policy network model, which has
focused heavily on policy development. This intergovernmental network model
11 The Organization for Economic Co-operation and Development (1997) has produced an excellent
review of the complexity of intergovernmental relations in numerous Western nations.

may be a precursor to the network management model because of its pragmatic
attitude that networks are required for managing intergovernmental programs.
Interorganizational Networks
The approach of interorganizational or multiorganizational networks tends to
focus on managerial or implementation issues and originated in organizational
theory and development (Hanf and Scharpf 1978), often with a business
administration or generic administration focus rather than a public or governmental
focus. Decades later, examples of this approach can be found squarely in the field
of public administration, especially community development programs such as
workforce development, economic development, and community-based substance
abuse prevention (Chisholm 1998; Harrison and Weiss 1998; Mingus 1994).
A frequent identifying characteristic of this approach is that networks are
created by design to achieve specific intended purposes, whereas policy networks
literature generally assumes that networks are naturally occurring phenomena to be
explained. Intergovernmental networks might side with policy networks on this
point as Deil Wright has described intergovernmental management as a means for
understanding and coping with the system as it is (Agranoff 1990, 58).
Nevertheless, the intergovernmental network and interorganizational network
models frequently share a practical focus on program implementation and so they
have much in common.

Issue Networks
In contrast to the other approaches, issue networks assume rapidly changing,
dynamic networks that ebb and flow in a quantum-like manner.12 Heclo (1978,103)
tells us that, An issue network is a shared-knowledge group having to do with
some aspect (or, as defined by the network, some problem) of public policy. He
presents issue networks as a supplement to the iron triangle theory, which argues
that tight-knit groups of congressional committees (staff and politicians),
bureaucrats, and interest groups dominate or control public policy in Washington,
D.C., for any given topic of concern.
Heclo (1978, 103) argues that, Instead of power commensurate with
responsibility, issue networks seek influence commensurate with their
understanding of the various, complex social choices being made. In other words,
in an information society that is rapidly changing, well-informed policy activists
come from many comers and their knowledge is valued because of a strong
perceived need on the part of government to be right when making policy
decisions. Contrast this with the policy network models focus on scientific
expertise and agreement on decision-making methodologies.
The impact of this overlay of issue networks upon iron triangles is
highlighted as Heclo (119) explains that networks aggravate problems of
12 One policy network definition highlights this contrast; We define policy networks as (more or
less) stable patterns of social relations between interdependent actors, which take shape around
policy problems and/or programs (Kickert, Klijn, and Koppenjan 1997, 6, emphasis in original).

legitimacy and public disenfranchisement. Heclo presented four hypothesized
results of this power shift: (1) issue networks search out complexity in otherwise
simple problems, (2) issue networks help to process dissension rather than to reach a
workable consensus, (3) policy activists in issue networks are frequently true
believers and they zap the confidence from ostensible leaders who are generally
spread too thin, and (4) issue networks thrive on continuously analyzing alternative
courses of action rather than on suspending disbelief and accepting that something
must be done (119-121).
Key differences of this approach include that non-professionals may have
extensive influence and that the focus is on broad policy concerns rather than a
narrow policy subfield. An issue network might be concerned about protecting
consumers from the mythical all-powerful, multi-national corporation whereas a
related policy network might be focused on deregulating the natural gas industry to
create a more efficient economy.
Miller (1994,379) highlighted the contrast when he described the need for
the policy network model to replace the traditional progressivist model in public
administration (a move Heclo would likely resist):
By using the term [policy] network, I refer to social relations that are
recurring, but are less formal and bounded than social relations
institutionalized in organizational roles. Nonetheless, network
relations are sufficiently regular that sense-making, trust-building, or
value-sharing may occur.

This definition highlights a significant difference between issue networks
and policy networks, other than the reliance on a scientific paradigm. The
difference is that Heclo does not assume that trust building or value sharing occur in
issue networks whereas sense making is assumed because his definition is based on
shared knowledge. The lack of these two elements provides a plausible explanation
for Heclos (1978,124) pessimistic conclusion that, Whipsawed between cynics
and true believers, policy would always tend to evolve to levels of insolubility.
Heclo seems to have discounted or even avoided the possibility that the personal
interaction of people in issue networks might help them communicate better, learn
from one another, and alter their values over time, thereby leading to a plausible
contrasting conclusion that issue networks might help move intractable policy
issues toward closure.
Typologies of Networks
The maze of terms and models in the above section represents a critical
problem for the field of network theory. Identical terms are regularly used to
describe different concepts and the same concepts are described in different ways
because of the interdisciplinary nature of public administration. The preceding
section discusses a number of similarities and differences between the six network
models that appear in the literature but does not specifically compare and contrast
the six models.

Table 2.1 draws from the previous discussion to make comparisons in terms
of seven meaningful categories. These seven categories are:
1. Main Focus on Policy or Administration this category examines the extent
that each model is focused on policy issues versus implementation issues.
2. Breadth of Focus (Narrow, Moderate, Broad) this category asks if the model
assumes networks are highly focused (i.e., oil policy), moderately focused (i.e.,
energy policy), or broadly focused (i.e., environmental protection).
3. Natural or Deliberate Design while most network theorists study what they
believe to be a naturally occurring form of social organization, some models
assume that networks are created and designed to accomplish specific tasks.
4. Stability of Network (High, Medium, Low) this category asks how resistant
to change networks are within each of the six models.
5. Members Mostly Professional or Amateur many network models assume that
technocrats are at the core of networks while some models assume that lay
members or the general public also play a significant role.
6. Agreement on Rational/Scientific Methodology this category asks if each
model views networks as having a cohesive approach or significant conflict
regarding what information is relevant for making decisions.
7. Focused Mainly on Facts or Values this category asks if each theoretical
model assumes that facts or values are more significant for the creation and
transformation of a particular network.

Table 2.1. General Comparisons of Six Network Models
Main Focus on Policy or Administration Breadth of FOCUS (Narrow, Moderate, Broad) Natural or Deliberate Design Stability of Network (High, Medium, Low) Members Mostly Professional or Amateur Agreement on Rational/ Scientific Methodology FOCUSEO Mainly on FACT80R Values
Policy Networks Policy Narrow Natural High Professional Yes Facts
Network Management Both Moderate Both Medium Professional Yes Facts
Advocacy Coalitions Policy Narrow Natural High Both No Values
Intergovernmental Networks Administration Narrow Both High Professional Yes Facts
Interorganlzational Networks Both Moderate Deliberate Medium Both Maybe Both
Issue Networks Policy Broad Natural Low Both No Facts

Table 2.1 reveals a strong focus of network theory on policy issues,
especially leaning toward narrow policy subfields and consisting primarily of
professionals (public administrators, technicians, scientists, etc.). While the models
vary considerably, this also leads to a strong focus on facts and a strong
commitment to the rational/scientific method. A number of the newer models
especially intergovernmental networks and network management are frequently
focused on implementation and administrative issues rather than public policy. The
advocacy coalition approach is focused again on policy, but sticks out because of its
focus on value systems as a critical determinant in public policy decisions and in
changes to the coalitions (i.e, networks).
This section on typologies takes this idea of comparing the network types a
step further by reporting on attempts to unify this theoretical field by developing a
formal typology to discuss and compare various networks. These efforts are critical
because the goal of the present research is to develop the field of comparative
network theory, which inherently requires a tool to help make meaningful.
intelligent comparisons. American public administration has been criticized for its
general inability to develop a cumulative knowledge base (Adams and White 1994;
White 1986). To help avert such criticism in the present research, this literature
review examines these existing tools and uses the strongest one as a starting point
for making network comparisons in chapter five. Strongest particularly speaks to
the ability of the typology to synthesize the bulk of this theoretical field.

Typologies in general have not flourished within American public
administration, perhaps because this pragmatic field has a love-hate relationship
with academia whereas typologies are often seen in more theoretical fields such as
political science, sociology, and linguistics. Substantial efforts to develop such
frameworks for network theory have been made in Great Britain. This section
presents the initial Rhodes Typology, the Wilks and Wright Typology, and the
Marsh and Rhodes Typology.13
Rhodes Typology
Earlier work by R.A.W. Rhodes (Rhodes 1981), a scholar of British central-
subcentral government relations, grew out of organizational theory rather than
political concepts of subgovemments. This is not too surprising because Rhodes
was studying the government as one organization (a unitary state) rather than
studying a system of more-or-less autonomous governments (a federalist state).
While Rhodes consistently used the phrase policy networks in studying
subgovemments, probably because he emphasized relationships between political
institutions, the American research on subgovemments (sometimes policy
networks, other times intergovernmental networks) has a stronger focus on
individuals as key figures in networks. Also, Rhodes focused on the sectoral level
while American research on policy networks was more focused (subsectoral or
13 Formal typologies other than these three were not found in the literature.

subsubsectorai level). Perhaps these differences are endemic to the unique
American system, as both specialization and individualism are strong characteristics
of this system (Stillman 1998,59-60).
American research on intergovernmental networks is also generally tightly
focused; however, Gage and Mandell (1990) presented a range of views (from
tightly focused to broad issue networks) in their edited book, Strategies for
Managing Intergovernmental Policies and Networks. Specifically, Gage (1990,
129-131) classified networks in the intergovernmental arena when he described
three types of concentric or nested intergovernmental networks: issue networks,
intergovernmental policy networks, and implementation networks. Issue networks
are said to be the broadest, most extensive networks while interorganizational
policy networks are more cohesive and typified by a membership that has a higher
degree of functional integration. The more tightly structured implementation
network is focused on accomplishing a collective policy goal, and these networks
have substructures for policy making, planning, service provision, and evaluation.
The Gage classification scheme is a fairly rough typology, however it is
more useful for comparing networks than James Skoks (1995) policy issue
networks or interorganizational policy networks by Gary Wamsley and Mayer
Zald (1983) because these phrases blur rather than clarify the network distinctions.14
14 This is not meant to detract from the value of their work. The current goal of comparing networks
differs from the goals of their research.

The six models of networks presented earlier could be considered a
typology as well. It has many similarities with the Rhodes typology, which
highlights five types of networks: policy/territorial community, professional
networks, intergovernmental networks, producer networks, and issue networks
(Rhodes 1997, 37-39). The difference is that Rhodes (1991,203) distinguished the
types with instability/stability, open/closed membership, large/small membership,
vertical interdependence (shared service delivery responsibility), and horizontal
articulation (ability to penetrate other networks). Issue networks, for example, are
unstable networks with a large number of members and limited vertical
interdependence while intergovernmental networks have a stable, limited
membership, limited vertical interdependence, and extensive horizontal articulation.
Wilks and Wright Typology
The Rhodes Typology came under attack because its five network types did
not appear to form a continuum as explained by Rhodes, because it focused
exclusively on structural issues and therefore left out interpersonal relations, and
because it focused at the sectoral level while much policy research seeks to explain
subsectoral policy decisions. Two new typologies emerged in the wake: the Wilks
and Wright Typology and the Marsh and Rhodes Typology. Wilkes and Wright
borrow the American focus on subsectors, arguing that it is not very useful to focus
on government-industry relations because the dynamics of such relationships vary

extensively for example, even within the chemicals sector there exist wide
variations for subsectors such as basic chemicals, pharmaceuticals, agrochemical,
paints, personal hygiene, and so forth (Rhodes 1997,40-41). A policy network
would focus on a specific subsector, frequently associated with a regulatory arena,
while the entire sector, chemicals in this example, would relate to a policy
community. Policy universe, a term Wilks and Wright introduce, would refer to the
population of potential actors that share a concern about industrial policy.
This typology received a number of criticisms, mostly aimed at its use of
terms already in popular currency for dramatically different meanings. For
example, policy communities were considered to be tightly integrated networks, yet
the Wilks and Wright Typology uses the phrase to represent loosely-integrated issue
networks (Rhodes 1997,42).15 Also, the list of factors to explain variations in
network types was quite long and the concept was difficult to extend beyond
industrial policy.
Marsh and Rhodes Typology
A decade after his original typology was published, Rhodes went back to the
drawing board, so to speak, with collaborator David Marsh. The Marsh and Rhodes
Typology was developed and is the state-of-the-art tool in the public administration
15 In equating the Wilks and Wright policy community with the term issue networks, Rhodes
appears to be using issue networks, in this case, to represent a concept much narrower than Heclos
original use of the term.

literature for comparing and contrasting networks, although it is based on
experience in unitary states and it is not focused on cross-border comparisons. Its
primary purpose is to create a common sense of the meaning of many of the terms
that have developed in this body of literature, and it has made strides in this area
including influencing Dutch researchers to adopt policy networks as the generic
term within the network management model.16 Its secondary purpose has been to
outline distinguishing characteristics of policy networks.
This typology uses the term policy networks as the overall generic term
for the field,17 varying on a continuum from policy communities where the
relationships are quite close to issue networks where relationships are distant or
loose (Rhodes 1997,43-45). In other words, two terms already in use in the
networks literature are used to represent the two ideal subtypes of policy networks.
Nevertheless, this is not within the policy networks model discussed earlier because
the typology seeks to pull that model, the issue network model, and others together
into a coherent framework for talking about networks in the public sector. Eight
variables within four categories are used to distinguish a particular policy network
as represented in Table 2.2. The four categories are membership, integration,
resources, and power.
16 In feet, Rhodes wrote the foreward for Managing Complex Networks.
17 Policy networks, therefore, distinguishes networks in public administration from other types of
networks such as local area networks, telecommunications networks, or personal networks.

Table 2.2. Types of Policy Networks (Marsh and Rhodes Typology)
Dimension Policy Community Issue Network
Size Limited number, some consciously excluded Large and open
Interests Economic and/or professional Whole range of affected interests
Frequency of Interaction Frequent and high quality among all members Contacts fluctuate in frequency and intensity
Continuity Stable membership and values Access fluctuates significantly
Consensus Members share values and accept legitimacy of the outcome Some agreement exists, conflicts are present
Distribution Within network All participants have resources to exchange Most participants are consultative, limited number have resources
Distribution Within member Organizations Hierarchical, leaders can deliver their members Participants have variable abilities to regulate actions of their members
Power Balanced enough for positive sinn game Unequal power leading to zero stun game
Source: Adapted from Rhodes 1997,44.

This typology represented a significant advancement from the proliferation
of new terms to explain the meaning of ones ideas, and intellectual arguments over
the meaning of these terms, toward a typology that allowed for the comparison of
networks that might be fairly similar or dramatically different. It is still, however, a
typology with British foundations rather than international foundations and seems to
be missing at least two aspects that are included in Gages basic classification. The
first of these is the focus on implementation. Neither ideal type in the Marsh and
Rhodes model captures the hands-on, working level or service-delivery network,
and it is difficult to imagine how something in between the two ideal types could fit
into this model. The second is the whole central-subcentral or intergovernmental
aspect that was the basis for Rhodes original research (for example 1981,1991).
This intergovernmental aspect is assumed within the Marsh and Rhodes Typology
rather than being incorporated as a distinguishing element of various networks.
While the Marsh and Rhodes Typology should be considered state-of-the-art for
network theory at this time, these weaknesses provide some of the needed
justification as the Comparative Network Framework in introduced in chapter five.
The next section of this literature review examines the literature on U.S.-
Canadian comparative administration to help provide a solid foundation for this
research within the comparative environment that impacts the PST. This section
highlights issues that impact the development of and use of networks within each
nation and the literature suggests ways that networks within the two nations might

differ from each other. In other words, the next section seeks to bridge the gap
between general network theory and the specific environment of cross-border
relations in the Pacific Northwest
U.S.-Canadian Diplomatic and Administrative Relations
Within the body of literature on network theory, little comparative research
exists. This thesis addresses this gap by comparing networks in the U.S. and
Canada and so one must also ask what the literature on U.S.-Canadian relations
indicates that might be applicable to developing comparative network theory. First,
the critical differences between the American and Canadian political systems is
discussed, then this section focuses on comparative research, the high degree of
informality in U.S.-Canadian relations, and several other significant issues that have
the potential to impact a study of these cross-border networks. The final subsection
highlights the key issues and their potential impacts on the PST networks.
Different Political Systems
Canadian and American political systems and public administration are
significantly different in that Americas founding fathers revolted against most
things British and created a federal system with clear separation of powers between
the branches of government Their Canadian counterparts opted for the federal
model yet maintained the British Parliamentary style of operations where the

executive and legislative are undeniably fused. Numerous sources document the
impact of these significant differences (Franks 1993; Holmes 1976; Lipset 1990;
Schnabel 1991; Simeon 1994; Swanson 1978; U.S. Advisory Commission on
Intergovernmental Relations 1981). Ronald L. Watts of Queens University
provides an extensive comparative overview of the two systems. Three contrasts
can be drawn from his overview that would likely impact the comparative structure
of networks (paraphrased from Watts 1993,292-315):
interest groups are powerful in the American system to the extent that
bureaucracies are frequently influential representatives of the interests of their
clientele groups whereas interest groups in Canada have less of an impact
because they have fewer useful points-of-access (e.g., Canada has a less
politicized bureaucracy, elected representatives who vote along party lines due
to party discipline, and an unelected Senate);
Canada has a more unified and powerful executive with the potential exception
of a few minority governments in its recent history whereas the legislative
branch in the U.S. regularly asserts its supremacy by defeating even major
Presidential initiatives; and
the extent of individual and interest group representation in the U.S. has led to a
democratic system that can be exceedingly slow to act whereas representation is
blocked in the Canadian system, as attested to by recent Constitutional reform

proposals (e.g., elected Senate, increased public input on intergovernmental
negotiations, increased provincial powers).
The idea that groups have more points of access in the American system
than in Canada (i.e., pluralist verses elitist systems) is supported by Mary Ann E.
Steger, Nicholas P. Lovrich, Jr., John C. Pierce, and Brent S. Steel (1992,95) in
their study of the political communication systems of environmental groups. These
differences are likely to be applicable at the state and provincial level as well
because the provincial governments in Canada are modeled after the national
parliament in much the same way that Article IV, Section 4, of the U.S. Constitution
directs state governments to mirror the republican national government.
Comparative Research on Networks
A number of researchers have examined U.S.-Canadian relations with regard
to environmental concerns and a few have contemplated U.S.-Canadian fishery
issues. On the environmental side, Mimi Larsen Becker (1996) and Frederick R.
Inscho and Mary H. Durfee (1995) focus on bilateral processes to revive water
quality in the Great Lakes region while Lynton Caldwell (1985,1992) and John E.
Carroll (1983) focus more broadly on transboundary environmental issues. In
addition, Lisa J. Mitchell (1994) and Steger et al. focus on comparative issues -
pesticide decision-making processes and the political communications systems of
environmental groups, respectively. Hollick (1976) provides extensive research on

the impact of differences in bureaucratic decision-making processes on International
Law of the Sea (ILOS) negotiations. Most of this research has a tendency to focus
on the level of diplomatic efforts and none of this comparative work discusses
network theory.
Scott (1976) provides research that is closer to focusing on network theory
rather than formal, diplomatic relations as he examines transnational cooperation
among private citizens, corporations, and interest groups. Scott suggests natural
environmental resources as a newcomer to the list of types of transnational
relationships and describes a variety of relations that may impact the work of treaty
organizations. For example, he notes that the Sierra Club and Ducks Unlimited are
two American organizations that now have quite active Canadian branches.
The interconnectedness of the international scientific community through
common professional organizations and what Scott refers to as a single information-
propaganda-research network provides another example of transnational
relationships. The impact is that Canadian and American scientists often agree on
the details of important issues and on methodological approaches. Joint ownership
in commercial ventures in oil and gas exploration, mining, and other areas provide
further examples where interests other than national interests are at work. As Scott
(1976,251) states, Certainly the head offices of these corporations must be more
interested in landings than loyalties, though transfrontier differences in costs, prices,
and taxes must also affect their preferences. Finally, Canadian and American

involvement in common organizations such as the International Council for the
Exploration of the Sea and the International North Pacific Fisheries Commission
may also help the two nations discover how much common ground they have.
High Degree of Informality
The minimal focus on network theory in U.S.-Canadian research is an
unexpected gap in the literature given that important research projects in the late
1970s highlighted the degree of informality and agency-to-agency contact that
characterize the relationship. For example, in the only study of its kind Roger Frank
Swanson (1978,221-265) described in extensive detail the intergovernmental
relations between states and provinces. These relationships were divided into
agreements, understandings, and arrangements, respectively defined by decreasing
formality. In this U.S. Department of State sponsored study, Swanson (232,237)
examined currently operative processes with direct communication between state
and provincial officials on an ongoing basis and over 70% of the interactions were
categorized as arrangements, indicating that state/provincial interaction is largely
an informal affair. Maine and New Brunswick led the way in terms of most active
state-province pairings with 32 interactions while Washington and British Columbia
placed second with 22 interactions.18
18 In the east this has developed to the point where there is an annual conference of New England
Governors and Eastern Canadian Premiers (Lubin 1993).

Likewise, federal-to-federal interaction is characterized by a high degree of
informal activity. Kal J. Holsti and Thomas Allen Levy (1976) argue that while
there has been an increase in formal governmental institutions between Canada and
the United States, these institutions do not constitute the core of the bilateral
relationship. Their review of the bilateral national institutions determines that the
Decision-making processes in older commissions could be characterized as joint,
concerned with problem solving rather than bargaining; they are technical and
nonnational in the sense that differences in the commissions are seldom based on
national distinctions (Holsti and Levy 1976,286). Institutions created after World
War II are said to contain few integrative features and opt either for decision making
based on bargaining between national teams of negotiators or just for consultation.
Transgovemmental relations, defined by Robert O. Keohane and Joseph S. Nye, Jr.
(1976) as the noninstitutionalized relationships between subunits of governments
and the activities they undertake that remain reasonably immune from central
control, are said to form the core of the relationship. Elsewhere, this is called
transregional paradiplomacy and is extended beyond the U.S.-Canadian relationship
(Duchacek 1990).
On the topic of province-state relations, Holsti and Levy (1976, 303) stated:
Many policy areas are the subjects of an extensive network of
informal links and formal agreements. In some cases, cooperative
administrative ties are facilitated by joint membership in transborder
intergovernmental organizations.

Therefore, this high degree of informality from a diplomatic or
international relations point of view raises public administrators and their
networks to the core of this bilateral relationship. The point is not that networks
should be equated with the informal decision-making process rather than the formal
decision-making process. Networks should describe the entire process. The key
point is that the informality of U.S.-Canadian relations provides an opportunity for
networks to have real power because formal political and diplomatic processes are a
less powerful influence.
Three Other Issues
Three other issues emerged in this literature that must be considered when
examining networks: the general predisposition not to link issues across policy
sectors, the extent that elected representatives are bound by constituent wishes, and
the extraconstitutional means of incorporating the valid interests of all levels of
Issue Linking. The first of these issues is identified by Holsti and Levy
(1976,292) as they state that linkages are seldom made between different issue
areas or policy sectors in U.S.-Canadian relations. They clarify their reasoning in
the following passage:
But the explanation for the absence of linkages between policy
sectors or conflict issues lies, in my opinion [Holsti wrote this
section], more in the bureaucratic networks between the two

governments than in carefully reasoned strategies of bargaining:
officials in separate agencies find it virtually impossible, given their
vested interests, to give up something dear to them so that another
agency can obtain something dear to it (292)
Annette Baker Fox and Alfred O. Hero, Jr., (1976,410) echo this view of
issue linkage when they state that it is neither in the interest of the bureaucrats
where they are involved in an issue to do anything but ensure that it is strictly
channeled nor in the interest of the weaker country to permit any packaging of
issues. The important implication for the proposed research is that political rather
than bureaucratic forces probably developed the issue linking that B.C. initiated in
the summer of 1997 (e.g., blockade of Alaskan cruise ship, canceling U.S. seabed
lease for the Nanoose Torpedo Range, etc.). One might expect, therefore, to find
strong political control in the B.C. network thereby reducing administrative control.
Constituent Control. A second issue pertinent to network theory can be
derived from the research of CJE.S. Franks (1993, 6) as he detailed differences in
systems of representation, specifically in the extent that representatives must act as
their constituents would like them to act He suggested that the system of
representation in the U.S. compels more consistency in this regard because of the
structure of interest groups and political parties, the representativeness of the public
service, the use of opinion polling, and the huge gap between the wants of voters
and the abilities of government (7-9). Representation, therefore, is only direct at
election time and is modified by a series of intervening factors in between elections.

This link between constituents and their interest groups (frequently network
participants) and elected officials, and the role of politicians in the networks of
public administrator^ is thus worthy of attention.
This creates two potential considerations for public administration networks.
The first is that given the opportunity and the perceived need to do so, American
politicians might override the recommendations of public administrators in favor of
uneasy or distraught constituents. The second is that once an issue is ironed out
nationally in the U.S., the international position becomes quite rigid because the
national decision represents such a complex agreement among a wide array of
interests. This is in keeping with the earlier points by Watts that interest groups
have more power in the open American system than in the relatively closed
Canadian system and in keeping with the conclusion of Hollick (1976) with regard
to ILOS issues.
Legitimacy of Subnational Interests. The third key issue is that
extraconstitutional means of incorporating the valid interests of all levels of
government in international affairs are common in todays increasingly
interconnected world. As Brian Hocking (1993,7) reminds us,
The management of foreign relations demands cooperation between
levels of government... Within the context of federal states, as in
the domestic sphere, this has involved overcoming the constraints
imposed by constitutional norms through processes of
intergovernmental negotiation and collaboration.

In other words, states and provinces are frequently co-equals with their
national governments in the international sphere, and extraconstitutional means for
collaboration are often developed to recognize the valid interests of all levels of
governmental actors in global issues. While not particularly focused on network
theory, Hockings research suggests that local, regional, and national governments
(and subunits of these) will be a natural part of the networks for state and provincial
public administrators. The advice regarding public administration networks is not to
overlook the dynamics of administrator-to-administrator contact in developing
network descriptions and not to assume that any one government is monolithic.
Even a brief exposure to congressional hearings (U.S. House 1985, 1994,
1997a; U.S. Senate 1985a, 1985b) on the PST provides ample evidence that Native
tribes, coastal villages, the Department of Agriculture, the Department of the
Interior, the Western Area Power Administration, the State of Idaho, and numerous
other governmental actors are also quite involved in this complex issue, particularly
because the required conservation of habitat to protect salmon stocks extends inland
by 500 miles or more.
Kev Issues from this U.S.-Qnadian Literature
While networks are not the focus of much U.S.-Canadian comparative
research, this body of literature does provide useful information to consider in this

study. Specifically, four issues emerged from this literature that may illuminate the
study of these four subnational networks:
1. Interest groups appear to have more points of access in the U.S. system and so
their influence in the subnational networks is anticipated to be stronger in the
states than in B.C. This also makes the U.S. system slower to make decisions
because representation takes time.
2. Administrative level contact forms the core of the bilateral relationship and
state-provincial contact routinely occurs with little centralized control. This
supports the focus of the present research on administrative networks and
suggests that states and provinces are frequently coequals in the international
arena because extraconstitutional means of incorporating their legitimate
interests are common. Network descriptions must therefore discuss all levels of
government and agency-to-agency contact should not be overlooked.
3. Linking of diverse policy issues is not the norm in U.S.-Canadian relations
because it is not in the interests of departmental bureaucrats who are at the core
of the relationship. Recent issue linking by B.C. of non-salmon issues to the
PST negotiations, therefore, probably indicates significant political influence
within that subnational network.
4. American elected officials are more bound by the interests of their constituent
groups and so, given the opportunity, they are more likely than their Canadian
counterparts to override the decisions of public administrators if constituents

react negatively to the decisions. This provides a special status for constituents
and interest groups within the state networks because public administrators
know these constituents have the ears of politicians if necessary. This also
means that domestic agreements in the U.S., once reached, may be viewed as
fixed rather than as negotiating positions.
Chanter Summary
Networks are essentially structures of interdependence involving multiple
organizations and a key factor is that administrators cannot be expected to exercise
decisive leverage just because they represent government Six models of networks
have been distilled from the literature: policy networks, network management,
advocacy coalitions, intergovernmental networks, interorganizational networks, and
issue networks.
In an attempt to make sense of this broad array of network imagery that
developed in the literature, researchers developed typologies. These typologies for
networks started out as rough classifications systems (i.e., here are three types and
here is how they relate to one another) and became more formal as researchers
sought to ensure that their typologies encompassed the universe of potential
networks and sought to compare these networks in terms of meaningful sets of
indicators or reference points (open verses closed membership, how broad is the
scope of the policy concerns, etc.).

The Marsh and Rhodes Typology depicted in Table 22 is state-of-the-art in
this field. Two lines of reasoning support this conclusion: (1) the typology has
been developed and refined over more than a decade through a competition of ideas
in the academic literature and (2) it is more successful than its two predecessors in
its explicit attempt to encompass the whole spectrum of network models. In brief,
the Marsh and Rhodes Typology describes two ideal types of policy networks -
policy communities that are focused, coherent, and stable and issue networks that are
large, loose, and more ephemeral. Any particular policy network, within this
typological schema, can be described as points on eight continuums related to
membership, integration, resources, and power. As the most current and
comprehensive typology in the field, this Marsh and Rhodes Typology will be
utilized as a starting point for comparing the subnational networks for this research.
Anticipated weaknesses include its failures to consider federal states and to discuss
implementation concerns.
Finally, this literature review has demonstrated both a minimal focus on
comparative, cross-border research in the field of network theory and a minimal
focus on network theory in U.S.-Canadian research. It has demonstrated that
subnational units of government can serve as the focal point for these networks and
that state and provincial public administrators are expected to be significant actors
in these networks given the low level (i.e., non-diplomatic and non-national) at
which most U.S.-Canadian relations are conducted.

Interpretation seeks to understand the meanings that actors attach to their
social situations, to their own actions, and to the actions of others. The
logic of interpretation follows the hermeneutic circle in which meaning
emerges in recognizing relationships and patterns . Interpretation is
more common and necessary than explanation and prediction.. . Malang
sense of one's situation is different from answering questions about what
happened in the past or what might happen in the future.
Jay D. White (1986, 16, 18)
In his classic article On the Growth and Knowledge in Public
Administration, White (1986, 15) highlights three modes of research in the field of
public administration positive, interpretive, and critical. Interpretive research,
described in the opening quote, is now commonplace and acceptable in the applied
field of public administration (Balfour and Mesaros 1994; Box 1992; White 1986).
An interpretive methodology was preferable for this thesis because the
essential question involved understanding who helps build the positions put forth by
political and career administrators at the negotiating table. Surveys and more
structured approaches might not have worked in this situation because (a) very few
individuals held the answers to the key questions and a low response rate would be
devastating to the reliability and validity of the study, (b) the negotiation process

and virtually all related documents were not available for public access, and (c) even
the broad structure of the networks was not predictable prior to the data collection
process. An interpretive research design allowed enough researcher flexibility to
overcome these factors. Specifically, the approach avoided the low response rate
inherent in a survey approach by allowing a persistent interviewer to take a more
personal approach and the lack of publicly available documents was overcome in
part by speaking directly to people who have read and even written some of the
critical documents.
Given the focus on network theory and the relative dearth of comparative
work on networks, an interpretive approach also allowed this research to be
exploratory and descriptive in nature (i.e., to approach the research with a broad
theoretical perspective but without a simple confirm or refute attitude). This is in
keeping with the primary research questions as noted in the first chapter:
1. Who are participants in the networks utilized by senior public
administrators in Oregon, Washington, Alaska^ and British
Columbia to establish PST negotiating positions?
2. What characteristics of networks help define the relationships
between these networks?
3. How can a better understanding and comparison of these four
networks assist public administrators who seek resolution to this
protracted policy dispute (i.e., the inability to reach agreement on
annual fishing regimes through the PSC structure and process since
4. What can the research focused on the first three questions contribute
to the development of a field of comparative network theory?

These research questions served to narrow the focus of this research
substantially, yet allowed the interpretive process to remain open rather than
focused on testing specific hypotheses. Such hypotheses would have been mere
guesses given the lack of research in this field of comparative network theory and
the minimal focus on networks related to the PSC structure and process.
The methodology utilized is described in the following five sections:
interpretive research and the hermeneutic circle, data collection and analysis,
modifications to the original plan, strengths and limitations, and a remark on the
iterative process and degree of agreement
Interpretive Research and the Hermeneutic Circle
This section provides theoretical background for the methodology. In
particular, a hermeneutic approach has been utilized and there is good reason to
believe such an approach would be more useful than standard qualitative approaches
(such as participant observation, focus groups, or one-time interviews) to answer the
four research questions. In brief, the power of this approach as contrasted with
personal interviews is that multiple contacts are made with the interviewees to
confirm and elaborate on the findings, which are simultaneously bolstered with
information from archival documents, current documents, and additional
informants. Facts are thus sought out through this triangulated agreement.

Participant observation is simply not feasible in this situation. Gaining
entree is often the most difficult part of participant observation (Jorgensen 1989,
chapter 3) and would likely be impossible concerning international negotiations on
the PST because the negotiations and most related materials, meetings, and
conferences are closed to the public. The U.S. courts have upheld the need to
maintain secrecy in much of this process and so participants in the U.S. section are
subject to security clearance procedures and promises of confidentiality (Cantillon
In terms of interviews, a standard series of interviews followed by
transcribing and analyzing the interviews would represent a rough first cut at
describing an extremely complex issue. The fact that the interviewees were
unfamiliar to the interviewer would have complicated this situation because
valuable time was also utilized to establish the relationships. In addition, the most
informed individuals about the research questions are negotiators and fishery
managers with significant negotiating experience. Many of them can evade
questions with considerable skill, thus potentially reducing the usefulness of any
methodology that might not provide an adequate opportunity for follow up
questions. The above process is merely step one in the methodology utilized for this

Hermeneutics, in contrast, focuses on iteration after iteration of
interpretation,19 constantly incorporating newly discovered information- This
method allowed the researcher to make a rough first cut and then gain additional
information from newly discovered sources (people and documents) and the original
interviewees. This cycle of interpretation, gathering additional data and viewpoints,
and reinterpretation is termed the hermeneutic circle by Hans-Georg Gadamer
(Diesing 1991,109; Gadamer 1989). While the research on hermeneutics does not
speak to an optimum number of iterations, the final section of this chapter speaks to
this issue with regard to the extent of agreement between the informants and the
At this juncture, it should be noted that utilizing hermeneutics as a
methodological technique is not the same as using hermeneutics in an attempt to
develop better public policy as discussed by John Dryzek (1982, 1990) or Lance
Noe and John Donahue (1998). The necessary distinction is that this thesis makes
no attempt to involve policy consumers, user groups, or any other group of
interested parties into the process of crafting useful policy solutions. In point of
fact, Dryzeks (1982,321) suggested approach might be useful because the equity
19 In some contexts hermeneutics involves establishing logical null hypotheses and alternative
hypotheses, and then attempting to falsify (see Popper 1959) die logical null hypotheses. Inability to
falsify leads to tentative support for the alternative hypotheses (for example, see dissertation by Sides
1995). In other contexts the hermeneutic circle refers to the process that the interpreter and the
reader go through in the effort to reach agreement, the process of continually referring to other
readings in an effort to ground the reading in question (Taylor 1971,6). These approaches differ
from that taken in the present thesis and this is to be expected according to sociologist Norman
Den yin (1989) who argues that each user lends unique interpretations to his or her methods.

principle dispute appears to fit his criteria of the residual set where traditional policy
analysis has little to offer a pluralistic decision process made up of a multiplicity
of actors and interests which is not producing manifestly good outcomes. While
this research helps place the equity dispute into Dryzeks residual category, crafting
a workable resolution to the conflict was beyond the scope of this project
Hermeneutics assumes that, consciously or otherwise, we all come to the
table with certain biases, prejudices, or preconceived notions. Hermeneutics
disavows all pretense of objectivity (Sides 1995,9) and thus approaches inter-
subjectivity by holding back ones prejudices long enough to understand the
prejudices of others and then making judgments about the competing prejudices.
The words of Gadamer (1989,299) might clarify this process,
[The hermeneutic researcher] will make conscious the prejudices
governing our understanding, so that the text, as anothers meaning,
can be isolated and valued on its own. Foregrounding a prejudice
clearly requires suspending its validity for us. For as long as our
mind is influenced by prejudice, we do not consider it a judgment
... For what leads to understanding must be something that has
asserted itself in its own separate validity. Understanding begins, as
we have already said above, when something addresses us. This is
the first condition of hermeneutics. We now know what this
requires, namely the fundamental suspension of our own prejudices.
Gadamer (1989,299) continues by focusing attention on the need to decide between
competing prejudices,
If a prejudice becomes questionable in view of what another person
or texts says to us, this does not mean that it is simply set aside and
the text or the other person accepted as valid in its place. Rather,
historical objectivism shows its naivete in accepting this disregarding

of ourselves as what actually happens. In fact our own prejudice is
properly brought into play by being put at risk. Only by being given
full play is it able to experience the others claim to truth and make it
possible for him to have full play himself.
Hermeneutics, therefore, is the process of systematically clarifying the
meaning of texts or other human constructs similar to texts such as verbal
communications without assuming that an objective reality a reality over and
above the existence of the researcher will surface as the true meaning. In his
Ph.D. dissertation, Andrew Davison (1995,3-4) makes a remarkably clear statement
on this topic:
I have jettisoned a frame within which objectivity about these matters
is even desirable. Rather than an objective statement, my account is
more faithfully conceived of as a contribution to, and interpretation
within, an ongoing dialogue ...
As in the opening quote by White, interpretation is about the clarification of
meaning. Meaning, according to Charles Taylor (1971,11), involves three
unavoidable elements:
a) Meaning is for a subject: it is not the meaning of the situation in
vacuo, but its meaning for a subject, a specific subject, a group of
subjects, or perhaps what its meaning is for the human subject as
b) Meaning is of something; that is, we can distinguish between a
given element situation, action, or whatever and its meaning.
... And thirdly,
c) Things only have meaning in a field, that is, in relation to the
meanings of other things. This means that there is no such thing
as a single, unrelated meaningful element; and it means that
changes in the other meanings in the field can involve changes in
the given element

Hermeneutics is the process of choosing between various interpretations and
incorporating new information and new prejudices to form newer, more accurate
and complete interpretations. Taylor might ask, Accurate and complete for
whom? The answer is not clear. While the present study strives to prove useful
for the informants and their colleagues, the primary focus is on understanding and
advancing the field of comparative network theory. This might make the audience
for this research the group of scholars and practitioners interested in cross-national
comparisons of networks. In actuality, the audience for the research may change
depending on which research question is considered. The audience is practitioners
in the first two questions and network theorists in the fourth question. Both
audiences are relevant for the third research question. The literature review, for
example, is both to help theorists understand the outcome of the fourth research
question and to help practitioners grasp the basic concepts of network theory.
On Taylors other two points, there is no doubt in this case that the existing
networks and the words and pictures used to describe them are separable and that
the meanings are closely tied together. This is why one would expect the four
network descriptions to be integrated to some degree, and why each description
would not be based solely on comments from informants within the jurisdiction
being described. Similarly, each new reading on network theory should contribute
in its own way to the distinguishing characteristics of policy networks suggested in

Appendix B. An example may help stress the circularity inherent in the process of
understanding meaning:
An emotion term like shame can only be explained by reference to
other concepts which in turn cannot be understood without reference
to shame. To understand these concepts we have to be in on a certain
experience, we have to understand a certain language, not just of
words, but also a certain language of mutual action and
communication, by which we blame, exhort, admire, esteem each
other. In the end we are in on this because we grow up in the ambit
of certain common meanings. (Taylor 1971,13)
The end product of a hermeneutical process may approach localized truth
if understanding20 by the relevant parties is achieved; however, an objective truth
will certainly not be the end product and localized truth may remain an elusive goal.
From the standpoint of positivist research, this leaves validity wanting. In How
Does Social Science Work? Reflections on Practice, Paul Diesing (1991, ill)
comments on this by saying that the argument that each successive understanding
is a better understanding is too optimistic, but that Gadamers maxim that each
successive understanding is different seems too pessimistic.
While hermeneutics generally refers to the interpretation of historical texts
such as the Bible or a series of minutes from Cabinet meetings on a particular issue,
the proposed method also involves the generation of new texts as proposed by
Danny L. Balfour and William Mesaros (1994). In other words, while hermeneutics
as analysis of texts has been faulted for lacking the detail afforded by the use of
20 Understanding is more relevant to hermeneutics than consensus. The subjects and interpreter may
have the same understanding of a text without necessarily agreeing on the content or ideas in the text

qualitative methods (Sides 1995), the method used herein has created rich new
texts using qualitative methods and then applied the concept of the hermeneutic
circle to refine and clarify the new texts. This is critical because most existing texts
with regard to the PST are based on highly staged events with carefully prepared
official presentations, such as hearings before Congress or Parliament, news
releases, staged speaking engagements, and the like. They would have served as a
poor starting point for an attempt to understand these networks. This is a key reason
why the methodological approach fits the research project at hand.
The approach of generating new texts allowed the researcher to gather
information in a more informal manner and thus produce a closer approximation of
reality than is afforded in such official documents and proceedings. Applying the
hermeneutic circle was designed to help the researcher avoid blatant errors and
misinterpretations while hopefully achieving some degree of localized truth. One
way to think of this iterative process is the metaphor fusing of horizons (Gadamer
1989). In essence, the horizons of the informants are pulled into the horizon of the
researcher to the extent possible through verbal and written exchange of a limited
duration. As the researcher comes to understand the viewpoint and prejudices of the
informants, understanding takes place.
Finally, there is no space outside of the process of understanding (i.e., above
history) to explore the researchers prejudices (Gadamer 1989,295-296). In other
words, the abandonment of objectivity described earlier applies throughout the

process. Prejudices are illuminated as they are confronted or provoked, and so
prejudices that are shared by all the informants may exist as blind spots in the
research. Indeed, prejudices that were simply not part of these exchanges for
whatever reason may produce such blind spots.
Data Collection and Analysis
Data to answer the first three research questions was collected in three
stages, although the first two stages overlapped in time. Archival research refers to
the collection of documents pertaining to the equity principle from a wide array of
sources, and this took place before, during, and after the time span when the field
interviews were conducted. Lastly, as described above, draft network descriptions
and diagrams were circulated for discussion with the interviewees and additional
informants. The literature review and additional sources on network theory served
as the primary source of data with regard to the fourth research question.
Archival Research
Appendix A provides a listing of resources that were obtained from an array
of libraries and archives in the Pacific Northwest, including the PSC library in
Vancouver, B.C. Roughly 80 hours of actual library time was involved in the
collection of more than 100 documents, and countless more hours in reviewing
these documents.

Collection of documents was also attempted at each of the relevant
state/provincial departments, however they were all quite concerned about
document security and did not allow public access to their files and libraries. They
frequently pointed the researcher toward another organization or individual for
access to such information, often to no avail. This was not unexpected given the
closed nature of the negotiation process. What was not anticipated was the lack of
archival information available from the time the PST was negotiated and
implemented. (For example, the Oregon and British Columbia archives did not
contain departmental documents more recent than the 1920s.) In fact, the majority
of documents from archives pertained to communication between department
directors and Governors and correspondence from constituents to the Governors.
The documents that were obtained, nevertheless, are extensive. They serve as
background for this research as well as to confirm and/or challenge the primary
research data (i.e., the particular comments and beliefs of the informants).
The archival research and field interviews overlapped in time as the
researcher took two multi-purpose trips to the Pacific Northwest Because of the
flexible interpretive approach to the interviews, information from the ongoing
archival research and from the ongoing interviews continually fed into the process
for the next interview. In this way if interviews and documents both confirmed a
particular fact, questions related to the particular issue might be avoided in future
interviews or might be increasingly detailed in future interviews to obtain a greater

depth of information. The weeks in between the two trips for the field research
were especially busy to facilitate a review of the information that had been obtained
and to plan how to make the most out of the time available during the second trip.
Field Interviews
Nine individuals were interviewed; two each from Oregon, Washington,
Alaska, and British Columbia, and one from the U.S. Section Coordinators Office
at the National Marine Fisheries Service. As discussed later, this is roughly a 50%
sample of the appropriate senior officials within these four jurisdictions. In
addition, the informants represented a wealth of experience in relation to the PSC,
including two commissioners, two alternate commissioners, a member from each of
the three panels, and numerous technical committee members.
An active interviewing approach was utilized for the field interviews,
meaning that interviews were viewed as co-creative rather than as an extraction of
data from the interviewee. This approach fits nicely with the hermeneutic approach
because it assumes that interviewer objectivity is a myth and that, knowingly or
otherwise, everything the interviewer does shapes the responses of the interviewee
(Holstein and Gubrium 1995). Interviews in this light are interactional events:
(The interviewer] does not tell respondents what to say, but offers
them pertinent ways of conceptualizing issues and making
connections, pertinence being partly defined by the research topic
and partly by the substantive horizons of ongoing responses____by
conceiving of interviews as ineluctably collaborative, we can

recognize how the interviewers shaped these conversations without
rejecting the final products as somehow defiled or tainted. (39,50)
This particular approach is similar to that of Herbert Rubin and Irene Rubin
in Qualitative Interviewing: The Art of Hearing Data. They state,
Qualitative interviewing design is flexible, iterative, and continuous,
rather than prepared in advance and locked in stone__design takes
shape gradually, as the researcher listens and hears the meaning with
the data. Concerns that appear important at the beginning of the
research may seem less vital later, and points that seemed
unimportant when the study began may turn out to be valuable.
(Rubin and Rubin 1995,43)
This active, co-creative approach to interviewing was consequential for this
particular study because the researcher sought to have each interviewee identify key
network participants and their roles, interests, and desired outcomes in the equity
principle dispute. Interviewees were also asked to consider what the other network
participants might be thinking. The researcher personally conducted the interviews
in all cases and Table 3.1 identifies the four main lines of questioning. Interviews
were conducted at the interviewees place of business or in hotel meeting areas and
restaurants where the interviewees were engaged in PSC negotiations. The
interviews were conducted in person and were recorded on audiotape.21
21 A 239-page supplement to this dissertation containing transcripts of these personal interviews is
available from the author.

Table 3.1. Four Main Lines of Questioning
Line of Questioning Main Goal Probes/Follow Up
1. Please describe the background and meaning of the Equity Principle, specifically to highlight the position and concerns of (Oregon/Washington/ Alaska/British Columbia). To understand how differently the four jurisdictions interpret the basic meaning of this part of the PST and to see the nature of how they describe the dispute in current terms rather than in 13 year old treaty language. Likely follow-up questions and probes include: (a) asking for definitions of key terms that the interviewee uses, (b) asking for sources of written documentation of official positions, (c) asking how the dispute has changed over time, (d) asking how much policy on the equity principle is made at the political level versus the administrative level, and (e) asking how important the states/provinces position is versus the position of the national government.
2. What individuals and organizations do you have regular contact with to help determine your states/provinces policy with regard to implementing the equity principle? To develop a list of the players in the network surrounding the equity principle (may include advisory groups and other committees that serve as a more structured part of the networks). The interviewer will have a preliminary list of network participants from reading government documents and so the most common probes for additional information will likely involve: (a) asking the meaning of acronyms, (b) offering suggested participants to see if they are regular players in the networks, (c) making it clear that regular contacts is not meant to imply that they are frequent contacts or that the government would not form a position without input from them, and (d) asking follow-up questions to get a sense of what the interest of each contact is with regard to the equity principle in particular rather than fishing or the PST in general.

Table 3.1. (cont.) Four Main Lines of Questioning
Line of Questioning Main Goal Probes/Follow Up
3. Of the individuals and organizations we have been discussing, which are the most influential in helping to shape policy surrounding the equity principle? To develop a sense of how the participants interact and the environments in which the subnational governments come into contact with the different network participants. Probes will focus on who the most significant actors are versus the fringe players. We may even try to sort them into three groupings most important interests, significant interests and minor interests. Probes will also try to get a sense of the reason for the level of importance. Is it economic power? Number of people they represent? Because the groups have relevant expertise or information? Because they have a clear jurisdictional role (like the Treaty Tribes and national government)? Because they represent specific communities? Etc.
4. Who was involved in the stakeholders process that the U.S. and Canada initiated in early 1997? This will be a brief part of the interview and is intended to help the researcher understand how the stakeholders process drew upon the existing network structures. Probes will primarily target getting any publicly available documents on the stakeholders process as well as the general attitude of the interviewee regarding the process.

The interviewees were selected on the basis of three critieria: (1) they must
have been currently involved in the PSC process and have been involved in that
process for a substantial number of years, (2) they must have been involved at a
level where they interacted with numerous other participants in the process rather
than primarily feeding technical data into the process, and (3) they must have
directly represented one of the four subnational governments under study. The one
exception to these criteria was an interview with Dave Cantillon at the U.S. Section
Coordinators Office within the National Marine Fisheries service (NMFS). He
clearly met the first two criteria but represents the federal government rather than
one of the subnational governments. Numerous other interviewees suggested that
someone at the coordinators office should be interviewed and so he was added to
the schedule of interviews. The goal was to interview at least two senior individuals
within each of the four jurisdictions and this goal was achieved.
Interviewing eight individuals (excludes NMFS interview) represents
roughly a 50% sampling of this pool of individuals because each jurisdiction has a
lead political staff member and two or three public servants tasked extensively to
PST policy issues. Table 3.2 lists the interviewees as well as two other groups of
informants those informal conversations (i.e., not tape recorded and not following
the lines of questioning in Table 3.1) were held with on one or more occasions and
those who were not interviewed but provided feedback on the draft network
descriptions that were developed from the interviews and archival research.

Table 3.2. Informant Names, Titles, and Organizations
Name Title Organization
Rollie Rousseau^ Alternate Commissioner PSC
Bumie Bohn"' Harvest Manager & Assistant Chief, Marine Resource Program ODFW
Roy HemingwayT Energy and Salmon Advisor Governors Office
Bruce Crawford"' Assistant Director, Fish Management Program WDFW
Dennis Austin"' Policy Coordinator, Interjurisdictional Resources WDFW
Curt Smitch+ Special Assistant for Natural Resources Governors Executive Policy Office
Dan SweckerT Senator Executive Director Washington State Senate Washington Fishgrowers
Kevin Duffy' Planning & Development Program Manager Division of Commercial Fisheries, ADF&G
Dr. Jeff Koenings"' Special Assistant to the Commissioner ADF&G
Dennis Brown"' Special Advisor Office of the Premier and Cabinet Office
Larry Neilson' Commercial Fisheries Analyst B.C. Fisheries
Bud Graham+ Assistant Deputy Minister B.C. Fisheries
Dave Cantillon"' U.S. Section Coordinators Office NMFS
John CoonT Fishery Management Coordinator (Salmon) PFMC
Deborah DaoustT Political, Cultural and Communications Officer Canadian Consulate General, Seattle
Teri TaritaT Librarian PSC
* formal interview and additional feedback (tape recorded, followed Table 3.1 outline)
+ feedback on written descriptions (via facsimile, telephone, and electronic mail)
^ informal conversations (not recorded, not following Table 3.1 outline)

The interviews took place in May 1998 and June 1998 during two trips by
the researcher to the Pacific Northwest At this tune it was becoming increasingly
clear to the interviewees and the general public that an agreement for the 1998
salmon fishing season would not be reached. As identified in Table 3.1, the
interviews were fairly unstructured and a specific list of questions was not prepared.
Part of the reason for this was that the official position of the interviewee was
expected to drive each interview in a slightly different direction and the other part is
that each state/pro vince was expected to have a very different mix of interests with
regard to the equity principle. The transcripts provide evidence that the
interviewees spoke far more than the interviewer and this can be considered a mark
of success in these interviews.
Review of Drafts bv Interviewees and Additional Informants
Applying the hermeneutic circle to developing the written network
descriptions called for copious follow-up contact via telephone, electronic mail, and
facsimile. In addition, as indicated in Table 3.2, the circle of informants was
broadened at this stage to include additional senior-level informants.
Draft network description documents were prepared based on analysis of the
interview transcripts and on documents collected through the archival research.
These written descriptions were then faxed to the interviewees and additional
informants for review. The informants were asked to provide specific input on the

draft network descriptions from October 1998 through January 1999. This also
provided an opportunity for the researcher to ask additional questions related to
comparisons of the networks between the four jurisdictions, hi most cases
comments were discussed on the telephone to allow for two-way discussion of
suggested modifications. More contact time per informant was invested at this stage
of the research than in conducting the initial interviews.
An important end product of this stage was written descriptions of the four
network structures, included as chapter four, which have validity in the minds of the
researcher and the administrators that were consulted. As Balfour and Mesaros
(1994,560) state, A valid interpretation is one that fits coherently into the context
of the research problem and is confirmed or consistently enriched by all subsequent
facts and interpretations. All the informants may not agree with all the details in
chapter four because informants frequently provided contradictory interpretations.
However, these conflicts were presented to the informants, to the extent time
allowed, in an effort to reconcile the conflicts.22
Modifications to the Original Plan
The original plan for this research called for ongoing attempts to revise
guiding hypotheses as more was learned through the field research. The more the
22 This occasionally involved identifying another source, such as a published statistic or archival
letters, to provide evidence of die conflict and then seeking additional justification from informants.

researcher learned about interpretive research and hermeneutics, the more this
appeared to be a stilted attempt to make interpretive research appear more
objective than it is, and so this approach was discarded. In fact, as highlighted
earlier, hermeneutics generally denies the existence of objectivity and strives to
incorporate the subjects subjectivity into ones own subjectivity as the research
progresses, and to force the subjects to reconcile their competing subjectivities.
The original working hypotheses helped to describe some of the researchers
preconceptions, and they were:
1. The existence and nature of issue networks around the equity
principle of the Pacific Salmon Treaty has overwhelmed public
administrators with public input, thereby reducing their confidence to
2. Significant interests from the four issue networks were overly
represented in the 1997 stakeholders process thus producing
dynamics that may not have existed had all interests been
appropriately represented.
This first guiding hypothesis would have tested Heclos hypothesized results
of the shift from iron triangles toward issue networks discussed chapter two while
the second working hypothesis suggested a starting point for analyzing the
stakeholders process, which has now been entirely abandoned (Canada Department
of Fisheries and Oceans 1998a; Strangway and Ruckleshaus 1998). The
researchers preconceptions included the belief that by looking for the broadest type
of network issue networks any of the other types might be identified as well. In
addition, the researcher assumed that direct fishing interests were overly represented

in the stakeholders process, thereby making agreement more difficult because
everyone involved would have an economic stake in the outcome.
The described methodology is quite close to the original plan except for
discarding this notion that working hypotheses would grow and take form
throughout the research process, thus being validated by the iterative process.
Strengths and T.imitations
Strengths and weaknesses frequently come in somewhat contradictory pairs.
For example, the main strength of this research is that the hermeneutic approach
allows for ongoing input from informants rather than one-shot interviews. This
presumably makes the results less dependent on events that occurred the day or
week of a particular interview and allows for a degree of validity that might not be
obtained from one-shot interviews. The related limitation is that the descriptions are
heavily dependent on the personalities and life histories of the informants. On
balance, this approach should avoid serious misinterpretations that can come with a
limited familiarity with the people involved in the research.
For example, David Barretts (1990,50) masters thesis, a cost-benefit
analysis of the PST, stated:
In a very general sense both parties seem to have achieved their
objectives in the fishery but the most significant change has been
between resource managers of each country. The level of trust has
been increased dramatically, sharing of data is commonplace and the
attitude of increasing interceptions has vanished.

In hindsight this statement, especially the final point made, appears chiefly
inaccurate as interceptions climbed dramatically after 1990. It was based on a
limited review of DFOs internal files and no familiarity with the people involved.
The present thesis strives to avoid this lack of in the field knowledge,23 yet risks
making conclusions that only hold true for the informants that were selected.
Generalization is to theory (the fourth research question) rather than to some
population of fisheries analysts, treaty negotiators, or network systems.
Other stiength/limitation pairings include:
The closed nature of the PST negotiation process made access to credible
information difficult. The strength is that this was countered by face-to-face
interviewing, the promise of an opportunity to correct any oversights or
inaccuracies after the interviews, and the assurance that this research was
academic rather than oriented toward advocacy.24
The rapidly changing political environment on this issue made it difficult to
know if the pulse taken was that of an entity in the process of metamorphosis.
The strength is that this is a real-life policy issue rather than a limited thought
experiment; this also made the research more interesting and enjoyable.
The selection of senior level political and bureaucratic administrators as
informants meant that a limited amount of time was available with each
23 Arnold Meltzners (1976) 2nd deadly sin of policy analysts too far away from the field and thus
not grounded in reality.
24 No assurance of confidentiality was made to the informants.

informant. The strength was that tenacity (perhaps approximating badgering at
moments) and active follow up enabled the researcher to conduct nine face-to-
face interviews and numerous hour-long telephone conversations. Facsimile
machines and the existence of electronic mail greatly facilitated this research
process by lowering the cost and the amount of time for the fieldwork.
OToole (1997,47) emphasizes the inherent limitation of networks as he states
that networks themselves are sufficiently complex that their impact on
performance is somewhat unpredictable for all involved____[however] action
guided by the hierarchy assumption is likely to lead not just to ineffectual but to
counterproductive outcomes. In other words, mind-boggling complexity was
the norm after the initial data was collected and the researcher needed to make
tough decisions about what was significant and what could be filtered out. The
strength was that the real world is also complex and so wading through this
complexity enabled the researcher to produce a picture of reality.
The informants included three administrators from each of four subnational
jurisdictions. In itself this has filtered out some of the complexity because the
federal level and the complexity of tribal governments were placed secondary to
the main focus, although they are clearly active participants with regard to the
equity principle. Adding more focal points for networks to this study would
have been reaching for too much at one time and the focus on subnational

governments at the moment was quite adequate for the descriptive and
exploratory task at hand.
Lastly, while the reliability of qualitative studies can be quite high when they are
carefully planned and executed, care must be taken not to generalize other than
to theory. Localized truth was used earlier to describe the nature of the
written network descriptions that were prepared. This is a critical strength of the
approach, yet the PST was not selected from a sample of U.S.-Canadian disputes
or cross-border resource conflicts, or anything else. The results, therefore, are
not directly transferable to other situations. The typology suggested in chapter
five is therefore to be thought of primarily as a theoretical concept in need of
testing. This is opposite of the typical positivist approach where data are
gathered to test an existing academic theory (Marshall and Rossman 1995,142-
152; Rubin and Rubin 1995, 56-58).
Remark on the Iterative Process and Degree of Agreement
It seems particularly relevant to include a brief remark at the end of this
chapter on the iterative process utilized as an important component of the
methodology. This process clearly increased the degree of understanding between
the researcher and the informants. More than half of the informants were relatively
easy to contact and provided extensive comments on materials that were sent to
them while it was more difficult to develop an ongoing relationship with the

remaining informants. In addition, enticing new informants into the process (after
the primary interviews were completed and preliminary network descriptions were
compiled) allowed for additional senior-level input into the finalized network
descriptions. Finally, the iterative process was successful in allowing the researcher
to ask additional questions of the informants to confirm or dispute statements from
other informants and to obtain input regarding several changes that took place
during the salmon fishing season after the interviews took place in May and June
1998. Each network description went through three to five iterations based on
feedback from interviewees, reviews of the archival documents, and feedback from
additional informants. The weakest link may be the Alaska description because
fewer individuals provided feedback on this description.
While it might not be fair to say that complete agreement (i.e., localized
truth) exists between the researcher and each of the informants with regard to the
network descriptions that are presented in the next chapter, it is fair to say that this
hermeneutic process led to a far superior product than would have been obtained
through one-shot interviews. That old maxim, what you see depends on where you
sit, appears to hold true for the informants. Informants were generally willing to
provide their perspectives on the issues and most of them had a strong enough
background working on PST issues to describe the perspectives of the other
jurisdictions. Agreement on those perspectives, however, is different than having an
accurate view of the competing perspectives.

It is the mark of the trained mind never to expect more precision in the
treatment of any subject than the nature of that subject permits.
Aristotle (Nicomachean Ethics I, iii)
Aristotles words of advice were kept in mind while preparing this chapter
because gathering information about networks is an essentially fuzzy task. A key
goal of this research was to develop clear descriptions of the existing network
systems of the four subnational governments with regard to the equity principle.
Chapter four delivers on this goal and seeks to add clarity to what may appear to be
inherent fuzziness. This was difficult because some potentially useful techniques
for gathering data about networks such as participant observation or tracking
telephone calls and internal correspondence were not options because of the
closed nature of the PST negotiations. Still, these network systems are generally
hidden from public view and so this chapter presents a wealth of information
previously known only to a privileged few individuals within each jurisdiction.
First, the overall network structure is briefly discussed and then the chapter
presents four subnational networks in order of the return migration of most salmon
species Alaska, British Columbia, Washington, and then Oregon. This return

migration pattern is also indicative of the natural ability of each jurisdiction to
intercept salmon from other jurisdictions as evidenced by the actual extent of
interceptions (PSC 1993; PSC Canadian Section 1991; PSC U.S. Section 1988).
This order of presentation allows a reduction in the overlap as well because a
number of issues related to Alaskas network are important to better understand the
networks of Washington and Oregon. Finally, this chapter ends with a section that
discusses five overall themes that emerge from this research.
The description of each subnational network begins with a discussion of the
equity principle issue from the perspective of the particular jurisdiction. Each
section then proceeds to discuss the state/provincial structure and PSC
representation, local and statewide organizations, tribal/First Nation connections,
the national government, Congressional/parliamentary relations, and relations across
the border. The logic is to focus on the core of the network (i.e., the formal
state/provincial structure and primary constituent groups) first because the
interviews revealed that the fishing administrators and industry interest groups are
highly involved participants in the equity principle negotiations. In each jurisdiction
a combination of fishing administrators, state/provincial political appointees, and
representatives of constituent groups work together to determine the position that
the jurisdiction takes on equity issues.
Next, the tribal groups in Oregon and Washington have a special status in
terms of fishing right*; and so it is necessary to compare this status to the other

jurisdictions. These tribes are autonomous governments and so they are distinct
from other fishing industry groups although they clearly represent their individual
fishing industries. Tribes in the southern U.S. have the court-affirmed right to fifty
percent of the total allowable catch for salmon stocks that migrate through their
historical fishing areas and they have used this right as leverage to force changes in
Pacific salmon fishing regimes. The Alaskan and British Columbian networks are
not driven by this same dynamic with regard to native peoples.
Finally, the PST is a bilateral treaty and so it is critical to include links to the
national government (administrative and political) and links across the international
border. One recommendation stemming from the review of U.S.-Canadian
literature in chapter two was that administratnr-trwadministratnr contact should not
be overlooked. The subsections on national government look particularly at these
relationships. Next, congressional and parliamentary links are discussed. This is
significant to these network descriptions because informants in B.C., Oregon, and
Washington consistently point to Alaskas block of congressional power on fishery
issues as a key reason why design flaws in the PST have not been corrected. Lastly,
also flowing from the literature review, state-provincial and state-state contact is the
focus of the across the border subsections. While the literature review suggests a
high degree of administrative contact at this subnational level, this research by and
large suggests that administrative contact outside of the negotiating forums is not
focused on specific policy issues such as ways to resolve the equity principle debate.

The above structure for presenting the network descriptions is purposefully
consistent for each of the subnational networks so the reader is able to more easily
identify the similarities and differences in these networks. While this chapter
provides a section on five overall themes that emerge as these networks relate to one
another, the next chapter makes explicit comparisons of these four subnational
networks to determine what the field of comparative network theory can learn from
this research on the PST.
Overall Network Structure
While the overall binational network related to Pacific salmon issues was not the
focus of this study, a view of that system is necessary to see how the subnational
systems are tied together. Conversations held with individuals at the Pacific Salmon
Commission, the Pacific Fisheries Management Council, the U.S. Section
Coordinators Office at the National Marine Fisheries Service, and the Canadian
Consulate in Seattle were extremely helpful in supplementing the formal interviews
to develop a picture of the overall network (Cantillon 1998; Coon 1998; Daoust
1998; Tarita 1998). This overall network is not as straightforward as adding up the
elements described by the individual jurisdictions.
Figure 4.1 illustrates this bilateral network and the box representing the PSC
structure in this figure is carried through to the subnational network figures because
this is the common core of the network structures. In itself, the PSC structure is

Figure 4.1
Bilateral Network on Pacific Salmon

quite a complex network with the commission itself; the Fraser River, Southern, and
Northern panels; numerous technical committees with their own working groups;
and administrative committees for finance, publications, etc. This decision-making
structure has been the source of a number of studies that generally failed to fully
describe the networks surrounding the PSC structure (for example, Gould 1993;
Graham 1993; Rutter 1997). The box used to represent the PSC structure in these
figures is therefore a general representation for this decision-making process.
This bilateral network reveals that there are more actors on the U.S. side,
including active federal legislative involvement and native peoples as governmental
entities. It also depicts the controlling role of the Department of Fisheries and
Oceans (DFO) in Canada and the extent that the PSC structure is divided along
national lines. These are important structural features in that DFO is a more active
force with regard to the PST than is the National Marine Fisheries Service (NMFS)
in the U.S. NMFS coordinates the U.S. Section while DFO has veto power within
the Canadian Section. Also, the division along national lines makes the PSC
process distinctly different from early examples of U.S.-Canadian binational
organizations such as the International Joint Commission formed in 1909.25 This
difference is discussed in the literature review as Holsti and Levi (1976)
demonstrated that institutions created after World War 13 tend to opt for decision
25 This elder of bilateral institutions seeks equitable solutions based on technical studies to a variety
of problems in the boundary water areas (Holsti and Levy 1976,287).

making based on bargaining rather than following the earlier model of joint problem
solving and a non-national esprit de corps. The PSC is highly nationalized in the
same way that two political parties dominate the U.S. Congress. They may
occasionally form pacts with the enemy across the aisle to achieve a common
purpose, but their separate caucuses make the majority of decisions.
Finally, the diagram portrays that much of the PSCs role in negotiating
annual fishing regimes and long-term arrangements has been lifted up to the level of
govemment-to-govemment negotiations since 1993 because of the failure to reach
agreements within the PSC. This pulls the U.S. Department of State and the
Canadian Department of Foreign Affairs and International Trade (DFAIT) into the
disputes to a greater extent, but these departments are still not the dominant actors
on PST issues. As coordinators of their respective sections, the U.S. National
Marine Fisheries Service (NMFS) and DFO are more active and visible than the
state departments. In addition, the PSC commissioners and alternate commissioners
and the same basic support team from each subnational jurisdiction are still actively
involved when the negotiations are lifted from the PSC to the level of govemment-
to-govemment negotiations.
As will become clearer in the following discussions of the subnational
networks, the larger number of participant governments on the U.S. side of the
bilateral network is a key concern for the Canadians. They find themselves
negotiating with numerous governments although only the U.S. President and the

Canadian Prime Minister signed the PST. Past feelings of helplessness with regard
to this problem have turned toward a more active approach by Canadian Fisheries
Minister David Anderson. Specifically, Anderson started to negotiate directly with
state governors during the 1998 fishing season. As is highlighted in chapter five,
this paradiplomacy may represent a move in the right direction.
State of Alaska
Perspective on Equity
The PST has two main principles conservation and optimum production,
and benefits equivalent to production from your waters and three subprinciples -
desirability of reducing interceptions, avoiding undue disruption of traditional
fisheries, and taking into account the abundance of the stocks. Alaskas view is that
Canada has taken a strict numerical balancing approach to interceptions and has thus
ignored the later two subprinciples (Duffy and Koenings 1998). Alaska, on the
other hand, contends that fisheries such as the Noyse Island purse seine fishery26
that has been in existence for over 100 years, are largely protected by the second
principle and that when Canadian bom fish from enhancement programs on the
transboundary rivers flood into Alaskan fisheries it is impossible to avoid catching
26 The primary commercial gear types for catching Pacific salmon are purse seines, gillnets, and
reefhets. Although fishing techniques are not relevant to the current discussion, these terms surface
regularly because industry groups are formed around these competing gear types. Purse seines are
the most capital intensive and most productive of die gear types.

them while properly fishing Alaskas own salmon stocks (Duffy and Koenings
1998). In this later situation, Alaska refers to the interceptions as incidental
Two other issues are critical to Alaskas perspective on equity (Duffy and
Koenings 1998):
1. Canada overlooks the impact of the processing industry. For example,
approximately $30 million of pink salmon annually is transported from
southern southeast Alaska (SSE) to Prince Rupert in Northern B.C. for
processing. Alaska could halt this practice rather than sharing the economic
benefit with Canada.
2. Salmon are jointly produced fish they need the rivers and lakes for spawning
but they also would fail to survive without the Gulf of Alaskas marine
environment where the salmon spend much of their lives and gain the majority
of their economic value. River of origin is a fairly recent and limited view
of salmon ownership.
The solution according to this Alaskan viewpoint is to negotiate percentage
shares of the available salmon stocks rather than to hire a bunch of accountants to
do bean counting. In years of high abundance, both nations would benefit equally
and in years of low abundance both nations would need to conserve equally to
ensure the long-term sustainability of each salmon stock. Abundance-based
fishing regimes would, in this view, represent equity and would share the burden of