Journal of Environmental Management 92 (2011) 92e101

Contents lists available at ScienceDirect

Journal of Environmental Management

journal homepage: www.elsevier.com/locate/jenvman

A resilience-based approach for comparing expert preferences across two

large-scale coastal management programs
Jongseong Ryu a, *, Thomas M. Leschine a, Jungho Nam b, Won Keun Chang b, Karen Dyson a
a
School of Marine Affairs, University of Washington, Seattle, WA 98105-6715, USA
b
Korea Maritime Institute, 1652 Sangam-dong, Seoul 121-270, South Korea

a r t i c l e i n f o a b s t r a c t

Article history: This report proposes a method for assessing resilience-building components in coastal socialeecological
Received 27 August 2009 systems. Using the proposed model, the preferences of experts in Masan Bay (South Korea) and Puget
Received in revised form Sound (USA) are compared. A total of 30 management objectives were determined and used to build
16 July 2010
a hierarchic tree designed using the principles of the Analytic Hierarchy Process (AHP). Surveys were
Accepted 20 August 2010
Available online 9 September 2010
performed with 35 Puget Sound experts using face-to-face interviews and with 28 Masan Bay experts by
mail. The results demonstrate that the legal objective, which enables legislation, was the highest
preferred component in both regions. The knowledge translation variable was also given a high pref-
Keywords:
Resilience
erence score in both regions. An analysis of variance (ANOVA) showed that the Puget Sound experts
Coastal management significantly favored attention to education, habitat restoration and species protection objectives in
Prioritization comparison to the Masan Bay experts. The Masan Bay experts placed greater emphasis on legislation and
Multiple criteria decision analysis the type of institutional design than did the Puget Sound experts. Using cluster analysis, four distinct
Analytic hierarchy process groups of respondents were independently identified in Puget Sound and three groups were identified in
Masan Bay. One unique subgroup in the Puget Sound experts group, which was characterized by its high
preferences for habitat restoration and species protection, was not observed in Masan Bay. Demographic
variables (length of career and role in coastal issue) failed to account for the differences in groupings and
preferences in either region, except for the variable ‘favoring information source’ in the Puget Sound
group. This finding implies that the demographic information was not related to differences in group
opinions in both regions. The analysis framework presented here was effective in identifying expert
preferences regarding the overall structure and emphasis in coastal management programs. Thus, this
framework can be applied towards coastal policy development.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction
Coastal areas are attractive regions for living, recreation and
economic activity. Approximately one third of the United States’
gross domestic product (about $4.5 trillion) is generated within
coastal watershed areas (US COP, 2004). Human populations are
concentrated along coasts, and consequently, coastal ecosystems
are some of the most impacted and altered worldwide (Adger et al.,
2005). Additionally, human societies and globally linked economies
heavily rely on ecosystem services and support (MA, 2005). Given
the concentration of human activities in coastal zones, coastal
environmental issues are too complex to be addressed by existing
environmental management approaches that focus on single issues
or resources. Furthermore, many global issues, such as climate
* Corresponding author. Tel.: þ1 253 303 2203; fax: þ1 206 543 1417.
E-mail addresses: bbenthos@u.washington.edu, jsryu90@gmail.com (J. Ryu).
change and the current global economic crisis, stem from complex
underlying interdependencies, which have made it difficult to deal
with environmental problems using steady-state approaches (Folke
et al., 2005). It is now evident that environmental policies must
shift from aiming to control ecosystems towards a new paradigm of
supporting sustainable development (Costanza et al., 2000).
Emerging theories and approaches in environmental manage-
ment have stressed the importance of assessing and strengthening
resilience (Gibbs, 2009). Two aspects of resilience are considered
valuable to this new paradigm. The first is the capacity to absorb
shocks while maintaining system functions. This aspect has been
the focus in many ecological and social studies. The second
concerns the capacity for renewal, re-organization and develop-
ment, should the system collapse, and although this aspect has
been less studied, it is essential for socialeecological sustainability
(Gunderson, 2003). The capacity to adapt and shape change is an
important component of resilience in a socialeecological system.
Scientific studies and policies addressing sustainability are needed

0301-4797/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jenvman.2010.08.020
 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101
to clarify the complex feedbacks between the interlinked social and
ecological systems and their capacity for resilience. Additionally it
is important to account for interactions across spatial and temporal
scales that secure the capacity to reorganize in the face of change
(Berkes et al., 2003; Folke et al., 2005). Resilience is essential when
building an adaptive governance system to deal with ecosystem
complexity and when delivering ecosystem services and resources
for societal development (Gunderson, 1999; Adger et al., 2005).
In practice, limited budgets restrain the number of resilience-
building projects that can be addressed simultaneously (Leach and
Pelkey, 2001). When a manager makes a resource allocation deci-
sion through a transparent, scientific process, it should in theory be
supported by multiple stakeholders. Furthermore, increasing
transparency enhances the legitimacy of the process for mediation
between experts and decision makers, allowing science and tech-
nology to be mobilized for sustainability (Cash et al., 2003).
Managers introduce their own intuition, judgment, and common
sense in the decision-making process. Therefore, using a rational
procedure can increase a manager’s confidence in decision-making,
allowing the management process to be more easily open to
stakeholders (Low et al., 2003; Hajkowicz, 2008). This strategy may
reduce conflict among multiple stakeholders and increase the
possibility of success in environmental management (Regan et al.,
2006; Pomeroy and Douvere, 2008).
This study employed surveys designed to determine the expert
preferences regarding various resilience-building techniques in
two socialeecological systems. The study tested the hypothesis that
resilience-focused management could be an essential component
of successful environmental management. The term “experts” is
used here to refer to the group of scientists and managers inter-
viewed for this study who are experts on the issues and context of
coastal management programs. The rationale for performing
surveys is that the decision makers would at some point introduce
their own intuition, judgment, or common-sense arguments into
their decision-making process rather than depending on a rational
procedure (Low et al., 2003; Hajkowicz, 2008).
This study uses multiple criteria decision analysis (MCDA) to
evaluate the relative importance of multiple objectives to
a diverse group of experts. MCDA is commonly used to provide
a transparent, structured, rigorous and objective evaluation of
multiple criteria (Hajkowicz, 2008). Several studies have indicated
the importance of including stakeholder preferences in environ-
mental management (Pomeroy and Douvere, 2008); however,
most MCDA studies have focused on the general public, and few
have focused on the opinions of scientific experts. The novelty of
this study lies in targeting experts in the area of sustainable
coastal management.
This study has two goals. The first goal is to provide a decision
model when choosing between different approaches for estab-
lishing resilience in coastal socialeecological systems. The second
is to compare expert preferences based on the proposed decision
model between two coastal regions (Masan Bay vs. Puget Sound).
These two regions were selected because they have a similar
history of environmental management and a comparable social
context, since there is a strong governmental and public commit-
ment to restoring the coastal ecosystems in both areas (Fig. 1).
There are also well acknowledged cultural and institutional
differences between South Korea and the United States.
Researchers do not often consider how cultural and institutional
differences are expressed through views on environmental
management, and quantitative cross-cultural comparisons are even
rarer. In this study, the collective opinions of experts from these
two countries were quantitatively analyzed. The goal of the study
was to compare the overall preferences and response patterns
using survey questionnaires. The implications of the overall
93
preferences and the differences between the two regions are also
discussed.
2. Geographical settings
The geography and history of environmental management in
Masan Bay and Puget Sound are briefly introduced in Fig. 1, which
serves to justify the comparability of the two regions in this study.
2.1. Masan Bay
Masan Bay is a small, semi-enclosed bay located on the southern
coast of South Korea. The bay area covers an area of 70.9 km2, and it
is 8.5 km long and measures 5 km across at its widest part. The
average tidal range is 1.3 m, and the bay has a low water-exchange
rate with the ocean in comparison to other nearby bays. In the
south, the bay is connected to the Jinhae Bay system (approxi-
mately 637 km2 in area) which includes the Masan Bay and several
other small bays. Three cities surround Masan Bay, namely Masan-
si, Changweon-si, and Jinhae-si. These three cities are highly
industrialized and had a combined population of 914,375 people as
of 2005. The population density is 3167 persons per km2, which is
6.5 times the national average population density. The Masan Bay
watershed covers approximately 263 km2 and is 43% of the size of
the three cities. The relatively small size of the watershed is due to
its steep slope towards the coast. Sixteen small streams quickly
flow into the bay due to this steep slope, supplying an average of
660,000 tons of freshwater daily.
Two major national industrial districts (Masan Free Trade Zone
and Changweon National Industrial Complex) and one large-scale
commercial harbor (Masan Harbor) are located around the Masan
Bay. These two large industrial districts were established on
reclaimed lands that were originally coastal wetlands and waters
during the 1970s and 1980s. More than 60,000 companies operate
within the watershed, which constitutes nearly 90% of the economy
of the three cities. Given its small, concentrated socio-geographical
setting, the Masan Bay was quickly and heavily polluted by a variety
of wastes, including untreated municipal and industrial sewage. As
a result, Gapo Beach was permanently closed in 1975, and shellfish
harvesting around Masan Bay has been prohibited since 1979. In
1982, the Korean government designated the Masan Bay as
a ‘Special Management Area of Coastal Pollution’. Since then, most
governmental activities have combated the pollution in the bay by
focusing on sewage treatment and dredging contaminated sedi-
ments (Lee, 1998). About 2 million cubic meters of contaminated
bottom sediments were dredged from the bay from 1990 to 1994
(Masan-si, 1994). In total, 36 million USD has been invested in this
dredging effort. Through similar concentrated efforts by the
government, the sewage treatment rate in point source reached
over 97.1% as of 2005. However, the water quality of Masan Bay was
the worst among Korean bays in 2006, which is evident that both of
these efforts failed to improve its water quality. Only 43% of
freshwater inputs from streams (283,000 out of 660,000 tons per
day), including surface runoff, are now treated prior to flowing into
the bay (MLTM, 2008). The Korean government has begun to
recognize the need for an integrated coastal management scheme
that considers various pollutant inputs and stakeholders partici-
pation in terms of their combined effect on the environmental
health of the bay. Consequently, in 2005, the local and central
governments reached an agreement to adopt a Total Maximum
Daily Load (TMDL) of pollutants approach (focusing on nitrogen) in
order to regulate the chemical oxygen demand. In the same year,
the Masan Bay Partnership was created to encourage the input and
cooperation of diverse stakeholders in the TMDL program. Imple-
mentation of the TMDL program began in 2007.
94 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101

Fig. 1. Geography and environmental history of Masan Bay and Puget Sound. Note the difference in scale in the inserted maps. The abbreviations used stand for the following. CWA:
Clean Water Act; ESA: Endangered Species Act; MMPA: Marine Mammal Protection Act; SMA: Special Management Area; TMDL: Total Maximum Daily Load; USD: US Dollars; WQ:
water quality.

2.2. Puget Sound
Puget Sound is an estuary containing multiple semi-enclosed
glacial fjords, bays, and islands in Washington State in the United
States. It extends from the eastern end of the Strait of Juan de Fuca
in the north to Olympia (the capital of Washington State) in the
south. The Puget Sound basin covers approximately 41,422 km2 of
watershed area and encloses 2330 km2 of water. The sound has an
average depth of 62.5 m, nearly 3700 km of shoreline, and more
than 200 islands (Gustafson et al., 2000). The tidal range is 2.4 m at
the northern end of the Sound and 4.6 m at its southern end. The
freshwater inflow is approximately 3.4 billion tons daily, and the
major sources of freshwater are the Skagit and Snohomish Rivers.
The shorelines and estuaries along the eastern side of Puget Sound
provide excellent harbors, especially Elliott Bay in Seattle and
Commencement Bay in Tacoma. The flatlands in the estuaries
adjacent to these bays provide extensive space for industrial and
agricultural development. These uplands were originally extensive
estuarine wetlands and mudflats that were filled in during the early
1900s. At least 76% of the wetlands around Puget Sound have been
eliminated, particularly in urbanized estuaries. Substantial losses of
mudflats and beaches have also occurred in the deltas of these
 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101
estuaries (Gustafson et al., 2000). This highly developed area con-
tained within the King, Snohomish, Pierce and Kitsap Counties has
become the urban and population center of the Puget Sound region.
The combined population of these four counties was over 3.6
million residents as of 2008 (PSRC, 2008a). These four counties
forecast an increase to almost 5 million by 2040 (PSRC, 2008b).
Puget Sound was once considered to have a nearly inex-
haustible resource base of timber, minerals, and fish. As the
economy grew, untreated sewage and toxic wastes routinely
began to be dumped into the Sound and its tributaries. Since 1972,
when the Clean Water Act (CWA) was introduced, progress has
been made towards controlling these wastes. The Puget Sound
Water Quality Authority, first established in 1983 and broadened
and clarified by 1985 amendments, developed the water quality
management plan (Leschine, 1990). The Sound became part of the
National Estuary Program in 1988 under Section 320 of the CWA.
Even though point source pollution is much better regulated
today, there are significant amounts of “legacy” pollution
remaining in Puget Sound. Around the Sound, 972 companies and
government facilities currently hold water pollution permits
issued by the Washington State Department of Ecology. The list
includes about 300 sand and gravel mines, over 100 boatyards,
more than 60 dairies and commercial farms, five oil refineries,
four paper mills and one of the nation’s largest aluminum
smelters (Seattle Post-Intelligencer, November 18e22, 2002).
Throughout Puget Sound, there remain concerns about the
disposal of municipal and industrial wastes, although, in contrast
with the past, most wastes are now treated prior to being dis-
charged into the Sound. Surface runoff associated with the rapid
increase in development that has occurred over the past several
decades is now seen as one of the top threats to Puget Sound (PSP,
2009). The Puget Sound Partnership was created by state legis-
lation in 2007. Under the aegis of this partnership, government
agencies, environmentalists, Indian tribes, businesses and other
groups are working to restore the health of Puget Sound by
minimizing the environmental consequences of their actions.
3. Materials and methods
3.1. The analytic hierarchy process (AHP)
The analytic hierarchy process (AHP) is one of the most widely
used methods in MCDA. It provides a framework for planning,
priority setting and resource allocation (Saaty, 1988). It was
developed by a business school professor, Thomas L. Saaty, in the
early 1970s. AHP has been applied in a wide range of areas,
including natural resources, to analyze preferences for manage-
ment objectives and alternatives. Its wide usage is likely due to the
simplicity relative to the analytical power. In this study, the AHP
technique is used to quantify an expert’s preference between
different management objectives for increasing resilience in
a coastal socialeecological system. To accomplish this, a hierarchy
of objectives is first developed, and then preference information is
elicited based on a series of pairwise comparisons between the
objectives. A detailed description of AHP and its applications can be
found in numerous articles (Leung et al., 1998; Ramanathan, 2001;
Mardle et al., 2004; Wattage and Mardle, 2005). This study follows
the four main steps outlined by Mardle et al. (2004) for applying
AHP:
1. Develop a hierarchy of interrelated decision objectives
describing the problem and create a survey format. The process
of building the hierarchical tree for this study will be detailed in
Section 3.2.
95
2. Perform a criteria-oriented pairwise comparison, which is
based on a mail/face-to-face survey, typically using a 9-point
scale. This study used the face-to-face method (Puget Sound)
and mail method (Masan Bay) to interview a total of 63 experts
(35 in Puget Sound and 28 in Masan Bay).
3. Compute local priorities based on the respondents’ relative
weights for the decision criteria and evaluate the consistency of
comparisons using the eigenvalue method (Saaty, 1988). To
determine the maximum eigenvalue, positive reciprocal
matrices of judgment were constructed and algebraically
computed, according to the methods of Saaty (1988), using
Microsoft EXCEL.
4. Aggregate local priorities using an arithmetic mean. Forman
and Peniwati (1998) have shown that the aggregation of indi-
vidual priorities will satisfy the unanimity condition (Pareto
principle) with either an arithmetic or geometric mean.
The AHP is capable of analyzing both management objectives
and alternatives, but this study did not consider management
alternatives. Consistency, a key component of the AHP, is based on
the largest eigenvalue obtained when solving the eigenvalue
equation in step 2. The consistency of the judgmental matrix can be
determined by a measure called the Consistency Ratio (CR ¼ CI/RI),
where CI is the Consistency Index and RI is the Random Index. If the
CR of the matrix is high, then the preferences are not consistent,
and thus they are not reliable. In general, a CR of 0.1 or less is
considered to be acceptable. However, after validation of each
individual’s internal consistency, responses in the 0.1e0.2 CR range
may also be accepted (Wattage and Mardle, 2005). Therefore, the
tier-2 analysis (see Section 3.2 for an explanation of the tiers)
undertaken in the results section was completed with 33 (out of
a total of 35) responses in Puget Sound and 26 (out of a total of 28)
responses in Masan Bay. For the tier-3 analysis, the number of
responses used in the analysis differed among the groups within
which aggregate individual priorities were compared.
3.2. Building decision models
The management objectives in this study were selected to
represent both the similarities and differences of the two regions. A
list of objectives was compiled by combining information from
preliminary manager interviews and a literature review. It was
arranged into a hierarchical tree to build a decision model to
evaluate expert preferences using AHP. This tree structure was then
reviewed independently by the two professors at the School of
Marine Affairs, University of Washington. The final decision model
for the survey is shown in Fig. 2.
At the top of the tree is the overarching goal, which is the
integrity and resilience of the socialeecological systems in the area.
Under this overall goal, the model contains three levels or tiers.
Three tier-1 criteria are identified as essential to achieve the goal:
1) Legal/institutional structure, 2) Social infrastructure and 3)
Action.
The legal/institutional structure includes two subcriteria (tier-
2): 1) Enabling legislation and 2) Type of institution/agency.
Legislation is further divided into two types of law (tier-3) that 1)
Protect ecosystems (conserving law), and 2) Direct and regulate
growth and development (developing law). Similarly, the institu-
tion/agency criterion is further broken down into 1) Central control,
and 2) Polycentric, which refers to voluntary citizen organizations
in partnership with the government.
The Social infrastructure category refers to components that
increase the resilience of social systems, such as social capital and
social memory (Putnam, 2000; Folke et al., 2005). There are four
main objectives under Social infrastructure: 1) Scientific research,
96 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101

Elements of success of adaptive coastal management in Puget Sound and Masan Bay

A B C
Legal/institutional Social infrastructure Actions
(social capital & social memory)

a b a b c d
Legislation Institution/agency Cleanup & Habitat Species Sewage
removal restoration recovery and treatment
protection
a Conserving a Central
law control
a b c d
b Scientific Educating actor Stakeholders Knowledge
b Developing Polycentric
research groups participating in translation
law
a
a a a Regular workshops
Natural Managers Planning only
b
b b b Participatory research
Social General public Implementation only
c
c c Boundary management
Interdisciplinary NGOs
c Both planning and
d implementation
Politicians

Fig. 2. A decision model to evaluate the expert preferences regarding management objectives in two coastal areas (Masan Bay and Puget Sound). Each alphabetical character outside
each box symbolizes the objectives in Fig. 3.

2) Education, 3) Stakeholders participation and 4) Knowledge
translation. Scientific research is further divided into three fields of
study: 1) Natural science, 2) Social science, and 3) Interdisciplinary
research. Education refers to efforts directed at four targets:
1) Managers; 2) General public; 3) Non-governmental organiza-
tions (NGOs); and 4) Elected officials. The stakeholders’ participa-
tion is further broken down into three components according to the
extent of participation in different stages of a project: 1) Manage-
ment planning only; 2) Implementation only; and 3) Both planning
and implementation. Knowledge translation refers to the integra-
tion of knowledge and systems to communicate and mediate
among experts, stakeholders and policy makers. Knowledge
translation is categorized in three specified components: 1) Regular
workshops; 2) Participatory research by multiple users; and 3)
Presence of a boundary organization. Workshops are used to
facilitate communication among experts, managers and decision
makers. Participatory research refers to joint research performed
by multiple users that encourages discussion between these groups
and promotes understanding among them. A “boundary organi-
zation” is an organization created through legislation as part of an
existing agency that acts as an intermediary between science and
policy (Cash et al., 2003)
The Actions category encompasses ongoing governmental
projects to restore the Masan Bay and Puget Sound ecosystems.
There are four types of projects in this category: 1) Cleanup and
removal; 2) Habitat restoration; 3) Species recovery and protection;
and 4) Sewage treatment. Cleanup and removal projects include
dredging contaminated sediments and removing garbage and dis-
carded fish nets. Habitat restoration involves removing dikes and
levees and restoring riverine and coastal habitats for salmon. Species
recovery and protection involves recovering endangered species,
such as Orca and Chinook (King) salmon and encouraging the
recovery of shellfish and migratory birds. Eelgrass recovery is also
included in the species recovery category. Although some overlap
exists between the habitat restoration and species protection cate-
gories, they were intentionally separated to reflect hypothesized
cultural distinctions between Korea and the United States.
Based on this hierarchic decision model, all of the respondents
were requested to answer the question, “In your opinion, what
should be the relative importance of the given objectives to achieve
the overarching goal, from the perspective of policy makers to
allocate budgets and resources?” Scores were computed using paired
comparisons among objectives of the same tier under the superior
objective common to the group of lower-tier objectives. In total, 33
pairwise comparisons were made to evaluate expert preferences
between objectives using the 9-point AHP scale. In addition, the
following demographic information was collected: 1) number of
years of experience; 2) role in coastal issues (scientist vs. manager, 3-
point scale); and 3) personal preferences regarding information
sources (public values vs. scientific literature, 9-point scale).
3.3. Quantitative analysis
A total of 63 people were surveyed, including 35 in Puget Sound
and 28 in Masan Bay. The respondents included elite technical
representatives, scientists, and managers with science backgrounds.
All of the respondents were familiar with coastal management
programs due to their roles in these programs.
Aggregated priorities were calculated using the arithmetic mean
of the individual priorities, according to the recommendation by
Forman and Peniwati (1998). One-way analysis of variance
(ANOVA) was used to compare the mean differences between the
two regions. To classify the respondents according to their prefer-
ence patterns, a cluster analysis was performed using the Pearson
product-moment correlation similarity based on preference scores
with no data transformation. Cluster analysis can identify groups of
respondents that prefer certain management objectives based on
the similarity of their relative scores. Once the groups were deter-
mined, representative objectives were selected to characterize each
group using group statistics. A mean value >0.2 of the relative
preference score indicated a strong representative objective. To
determine the extent to which the background information
collected could explain the differences in the clusters, one-way
ANOVA tests and pairwise comparisons were performed. When the
ANOVA results indicated that the demographic group means
(number of years of experience, role in coastal issues, or preference
of information sources) were significantly different each other (i.e.,
p < 0.05), Fisher’s modified Least Significant Difference (LSD) test
 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101
was used to perform a post-hoc pairwise comparison among the
clusters. The LSD test is the least stringent and most powerful
among the commonly used multiple pairwise comparison tests,
and its use minimizes the chances of false positives (type I errors).
4. Results
4.1. Comparison of aggregated priorities
The management preferences of the coastal planners in the
study areas (Fig. 3) were aggregated priorities determined from the
average of all the individual responses for each area across the
coastal management objectives shown in Fig. 2. The technical
experts in the two regions showed several significant differences in
their preferences for management objectives, as elaborated below.
The relative importance of tier-1 objectives can be inferred by
summing the tier-2 scores for each tier-1 objective (Fig. 3a). In
Puget Sound group, the Social infrastructure objectives (0.42) were
the highest concern, followed by Action objectives (0.34) and Legal/
institutional objectives (0.24). However, the Masan Bay experts had
a different order of managerial preferences. Their most important
management objectives were Legal/institutional (0.43), followed by
Social infrastructure (0.35) and then Action (0.23). The mean values
between the two regions differed for Legal/institutional (p < 0.01)
and Action objectives (p < 0.05), but not for Social infrastructure
objectives. These results indicate that experts in the two regions
have significantly different preferences for the broad coastal
management objectives.
There were even more dramatic differences in management
preferences between the two regions in the tier-2 objectives. The
preferences for five out of ten tier-2 objectives differed significantly
between the two regions (Fig. 3). The Puget Sound experts accorded
more importance to Education (p < 0.05), Habitat restoration
(p < 0.01) and Species protection (p < 0.01) objectives than did the
Masan Bay experts. The Masan Bay experts showed a higher
proclivity for Legislation (p < 0.05) and Institutional (p < 0.05)
objectives than did the Puget Sound experts.
Legislation was the number one objective in both regions. For
example, the mean preference value was 0.27 in the Masan Bay
group, implying that more than a quarter of total program effort
should be invested into legislation. In addition, the second highest
value in the Masan Bay group was given to the Institutional
objective (0.16). Taken together, the Legal/institutional objective in
tier-1 is the most important component for coastal management in
Masan Bay. In the Puget Sound group, the second highest prefer-
ence was for the Knowledge translation objective (0.15), which was
given a similar aggregated score in the Masan Bay group (0.12).
Further comparisons between the two regions in the tier-3
analysis provide more detail about the response patterns (Fig. 3b).
There were significant differences between the two regions in
seven of the seventeen tier-3 objectives. Under the Legislation
objectives (tier-2), the Puget Sound experts favored laws for regu-
lating and directing development (Developing law; p < 0.01). In
contrast, the Masan Bay experts preferred laws that conserve and
protect endangered species and habitats (Conserving law;
p < 0.01). The Puget Sound experts gave higher scores for educating
the general public, stakeholder participation in both the planning
and implementation stages, and holding regular workshops than
the Masan Bay experts (p < 0.05 in all three cases). On the other
hand, the Masan Bay experts showed significantly higher prefer-
ence for the “planning only” tier-3 option under stakeholder
participation and for “boundary management” under the Know-
ledge translation objective (p < 0.01 in both cases).
Regardless of the significant differences between the two
regions, experts from both areas placed the highest preference on
97
Legislation. Knowledge translation and Sewage treatment were
also important to both regions. Experts from both regions placed
lowest importance on actions involving Cleanup and dredging (0.05
in Puget Sound, 0.04 in Masan Bay). Among the tier-3 objectives,
both groups preferred a Polycentric objective over a Central insti-
tution objective. Additionally, experts from both regions gave
higher scores to Natural and Interdisciplinary sciences than Social
science and considered managers and politicians the most impor-
tant targets for education. Stakeholder participation in Both plan-
ning and implementation was preferred over the ‘Planning only’ or
the ‘Implementation only’ objectives for both regions. In summary,
there were similar response patterns in both regions for three of the
tier-2 objectives and five of the tier-3 objectives. Therefore, despite
having several different preferences between the two regions, the
two groups also had many similar preferences for coastal
management.
4.2. Analysis of response patterns
In order to classify groups of individuals according to their
response patterns, a cluster analysis was performed. Respondents
who showed similar opinions were grouped into the same cluster.
The 33 valid responses from the Puget Sound experts were grouped
into four clusters, and the 26 responses from the Masan Bay experts
were classified into three clusters. The clustering was done using
the similarity of the Pearson product-moment correlation, based on
the relative importance scores for the tier-2 objectives (Fig. 4). Each
resulting group was characterized by strong representative
components, which were defined as average preference scores of
over 0.2 for the group (Table 1).
In the Masan Bay group, three clusters were defined (Fig. 4).
Cluster A included the 15% of respondents who showed an equally
strong preference for Legislation and Institutions. Cluster B
comprised 19% of respondents and was characterized by high
scores in Research and Knowledge translation objectives under
Social infrastructure objective. Cluster C included the 65% of
respondents who favored Stakeholder participation and reliance on
Sewage treatment.
In the Puget Sound group, Cluster A included the 24% of
respondents who exhibited a strong preference for Legislation.
Cluster B included the 39% of respondents who favored Knowledge
translation. Cluster C included the 21% of respondents who showed
a strong preference for Habitat restoration and Species protection
under Action objective. The remaining 15% of the respondents in
Puget Sound formed Cluster D, which showed a particular prefer-
ence for Sewage treatment relative to other clusters.
In the Masan Bay group, three clusters were similar to the
clusters found for the Puget Sound group (A, B, D); however, Cluster
C was unique to the Puget Sound group. Interestingly, the largest
proportion of Puget Sound respondents (13 out of 33) selected
Knowledge translation as their highest priority, but the largest
proportion of respondents from Masan Bay (17 out of 26) favored
Stakeholder participation and reliance on Sewage treatment.
To understand the extent to which demographic components
could explain the clustering, group averages of the respondents’
demographic data were compared by ANOVA and the LSD test as
a post-hoc comparison. The demographic data included the number
of years of experience, role in coastal issues, and preference of
information sources. The ANOVA results showed that the preference
of information sources was the only significantly different variable
among the clusters of Puget Sound (p < 0.05). Subsequently, the LSD
test reported a significant difference of group means between clus-
ters A and D, and clusters A and C. In the Puget Sound group, the
number of years of experience and role in coastal issues was not
related to the group clustering. For the Masan Bay experts, there was
98 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101

Fig. 3. Comparison of overall relative importance scores between Masan Bay and Puget Sound. Comparison of scores for the a) tier-2 management objectives and b) tier-3
management objectives are shown. The number of samples in each comparison is displayed in parenthesis. Each alphabetical character represents the corresponding objective in
a hierarchical manner as shown in the legend. Significant differences in the relative preference scores between the two regions are denoted by *(p < 0.05) and **(p < 0.01) in the
legend, based on the ANOVA test. Relative preference (%) is calculated using the mean of the responses for the group.

no significant relationship between demographic data and clus-
tering. Therefore, it is unlikely that the demographic backgrounds of
the respondents impacted their preferences in this study.
5. Discussion
In the last few decades, scientists have had to confront the fact
that technologically-based environmental management cannot
control the variability of complicated ecologicalesocial systems
(Berkes et al., 2003). An alternative approach to this complexity is to
transform preexisting management approaches, based on linear
models, towards building socialeecological resilience in systems
(Adger, 2000; Adger et al., 2005). The present study revealed that
South Korea and the United States share preferences for resilience-
building components. Legislation, Knowledge translation and
Sewage treatment objectives received the highest scores from both
regions’ experts. Although these components are not necessarily
applicable to every country, adaptation of these resilience-building
components could generally be valuable in coastal systems.
The aims of the present research were to quantify and compare
expert preferences regarding coastal management in South Korea
and the United States. One of the major differences between these
 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101 99

Fig. 4. Two dendrograms for clustering the respondents from Masan Bay and Puget Sound. The dendrograms were used to cluster the a) 26 respondents in Masan Bay and b) 33
respondents in Puget Sound by group average linking of the Pearson product-moment similarity based on the relative importance scores of the tier-2 objectives with no data
transformation.

two countries that emerged in this study was the relative impor-
tance given to legal concerns. The strong inclination to favor
legislation in Masan Bay may be due to the lack of robust legislation
for coastal management in South Korea (Hong, 1995). Moreover, the
two regions showed opposite preferences for conservation and
development laws. This discrepancy likely reflects difference in the
social context of the two regions. Masan Bay has a long history of
coastal development, and environmentally-friendly legal institu-
tions have only recently begun to emerge to restore the bay. As
a result, Korean culture seems to rely more heavily on laws that
encourage conservation rather than laws that restrict development
for environmental protection. On the other hand, the Puget Sound
experts favored laws focusing on development over conservation.
Coastal areas in the United States are protected by several conser-
vation-based Acts, including the CWA, the Endangered Species Act
(ESA) and the Marine Mammal Protection Act (MMPA), which were
all passed in the early 1970s. The Puget Sound respondents seemed
to consider development management laws, such as the federal
Coastal Zone Management Act and the state’s Shoreline Manage-
ment Act, to be the ones with the greatest relevance to current
environmental concerns.
The cluster analysis results indicated another major difference
between the two cultures, since Cluster C was unique to Puget
Sound. In Puget Sound, 21% of the experts (Cluster C) underscored
the importance of habitat restoration and species protection efforts
in comparison to the Masan Bay experts, who assigned lower scores
to these activities. This difference may have resulted from the long
history of salmon restoration efforts in Puget Sound over the past
few decades, whereas concern for the protection of individual
marine species is generally absent in South Korea. In addition, Orcas
and Chinook salmon have long been regarded as iconic species that
have been protected under the ESA and MMPA in Puget Sound. These
environmental laws protect diverse habitats for species of social
importance, and as a result marine environments for these species
have been protected in Puget Sound. The Puget Sound ecosystem is
better off now than it was in the 1970s. In contrast, the Masan Bay
respondents are less concerned about species and habitat issues,
which have received less attention than pollution itself.
Interestingly, the Puget Sound respondents assigned more
importance to education and had a stronger preference for species
and habitat protection. This tendency might reflect a more bottom-
up viewpoint in Puget Sound and a more top-down viewpoint in
Masan Bay. The United States is accustomed to working closely with
a wider range of stakeholders in management processes. On the
other hand, the South Korean coastal management is generally
driven by the government, which unintentionally limits stake-
holder participation in the early stage of the management process.
The in-person research interviews included the AHP-based
survey and a follow up semi-structured interview. Taken together,
these interviews revealed some interesting results. First, the

Table 1
Classification of tier-2 management objectives based on the average preference of responses for the cluster.

Cluster Strong (>0.2) Intermediate (0.2e0.1) Weak (<0.1) N (%)

Puget Sound
A Legislation Institution Others 8 (24)
B Translation Stakeholders, Education, Others 13 (38)
Research, Legislation
C Species, Habitat Research, Education, Sewage Others 7 (21)
D Sewage Cleanup, Habitat, Species Others 5 (15)
Masan Bay
A Legislation, Institution Sewage Others 4 (15)
B Research, Translation Legislation Others 5 (19)
C Stakeholders, Sewage Translation, Research Others 17 (65)
100 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101
development of quantitative data on the attitudes of expert groups
is innovative in marine and coastal management. Second, there are
advantages in using personal surveys to obtain useful information.
The concerns noted by Leung et al. (1998) over vital interactions
between the participants being lost when using mailed surveys
did not arise in this study. Qualitative information was also
gathered during the interview to complement and help interpret
the AHP survey data, including the respondent’s main role in
supporting the coastal management. Third, the development of
links in ideas between groups and between sectors is a particu-
larly interesting strength of the face-to-face surveys. The biases of
the interest groups are made more explicit during the interview.
Overall, the interviewees felt that the relevance of the survey was
very high. All those who participated showed enthusiasm about
the aims of the survey and the research, and showed an interest in
seeing the results and participating in future activities related to
this study.
The decision model presented in this study is just one applica-
tion of the presently employed approach. The management objec-
tives included in this model might not represent all possible
influences on the resilience and integrity of the socialeecological
system in the two regions. For example, there are jurisdictional
issues between local and state governments in Puget Sound, which
were not reflected in the ‘type of institution’ objective as we
defined it. A few respondents pointed out that ambiguity in some
objectives made their judgments more difficult. For example,
habitat restoration and species protection could not be clearly
distinguished in the view of some respondents. One respondent felt
that the tier-1 objectives were too interrelated for him to assign
ratings of relative importance. Our data analysis and interpretation
have drawn on the respondents’ comments and feedback obtained
during the interview. However, it was necessary to maintain the
same hierarchical structure throughout the survey for reliable
comparisons to be made between the two regions. An advantage of
AHP is that additional objectives may easily be added to the model
without changing the basic principles of estimation and the
calculation procedures, which facilitates future studies that may
provide additional comparative results.
6. Conclusions
Determining the relative importance of management objectives
is essential for successful management for several reasons. First,
preferences based on existing information are a powerful means for
evaluating the alignment of current policies with what needs to be
done in order to establish more sustainable coastal areas. Second,
determining relative priorities is important for developing appro-
priate strategies, since there are multiple, often conflicting, objec-
tives in coastal management. Finally, this study highlights the
different viewpoints of experts in two regions in the prioritization
of coastal management.
The analysis presented in this paper does not require complex
models and assumptions. The method used here is effective in
revealing expert preferences and can be used in policy develop-
ment. In particular, it could help policy makers have confidence in
their decisions, and create credibility and legitimacy from various
stakeholders.
In this cross-cultural comparison, we found both similar and
differing responses between the two regions. The results suggest
that experts in both regions strongly perceive the importance of
Social infrastructures, which are being underinvested in the current
coastal management of both countries. Improvement of Social
infrastructures is a strong candidate for building resilience in social
systems. Our cluster analysis revealed three expert subgroups in
the both regions characterized by similar preferences (Legislation,
Knowledge translation, Sewage treatment), as well as one group
favoring habitat and species protection that was only identified in
Puget Sound.
This study joins a growing body of research that addresses
building socialeecological resilience in a multi-disciplinary
analytical framework. Such a framework of analysis can offer useful
information to the management process, potentially assisting in
policy-making. Finally, linking efforts across disciplines can
enhance the quality of coastal management despite the complexity
and uncertainty of the ecological systems.
Acknowledgements
The authors wish to express sincere appreciation to four anon-
ymous reviewers for their valuable reviews, and the Washington
Sea Grant (New Models for Solving Future Policy and Management
Challenges in the Nearshore and Estuarine Environment) and the
Korean Ministry of Land, Transportation and Marine Affairs
(Establishing Total Pollution Load Management System (TPLMS) in
Masan Bay of Korea, 2007) for research funding. We especially
thank Professors E. Miles, D. Fluharty and T. Klinger (School of
Marine Affairs, University of Washington) for useful discussions on
the ideas presented in this paper. We also wish to thank all the
participants who were surveyed in this study in Masan Bay and
Puget Sound.
References
Adger, W.N., 2000. Social and ecological resilience: are they related? Progress in
Human Geography 24 (3), 347e364.
Adger, W.N., Hughes, T.P., Folke, C., Carperter, S.R., Rockstrom, J., 2005. Sociale
ecological resilience to coastal disasters. Science 309, 1036e1039.
Berkes, F., Colding, J., Folke, C., 2003. Introduction. In: Berkes, F., Colding, J., Folke, C.
(Eds.), Navigating SocialeEcological Systems: Building Resilience for
Complexity and Change. Cambridge University Press, Cambridge, UK, pp. 1e29.
Cash, D.W., Clark, W.C., Alcock, F., Dickson, N.M., Eckley, N., Guston, D.H., Jager, J.,
Mitchell, R.B., 2003. Knowledge systems for sustainable development.
Proceedings of the National Academy of Sciences of the United States of
America 100, 8086e8091.
Costanza, R., Daly, H.E., Folke, C., Hawken, P., Holling, C.S., McMichael, T.,
Pimentel, D., Rapport, D.J., 2000. Managing our environmental portfolio.
BioScience 50, 149e155.
Folke, C., Hahn, T., Olsson, P., Norberg, J., 2005. Adaptive governance of sociale
ecological systems. Annual Review of Environment and Resources 30, 441e473.
Forman, E., Peniwati, K., 1998. Aggregating individual judgments and priorities with
the analytic hierarchy process. European Journal of Operational Research 108,
165e169.
Gibbs, M.T., 2009. Resilience: what is it and what does it mean for marine policy-
makers? Marine Policy 33, 322e331.
Gunderson, L.H., 1999. Resilience, flexibility and adaptive management: antidotes
for spurious certitude? Conservation Ecology 3 (1), 7. Available at: http://www.
consecol.org/vol3/iss1/art7.
Gunderson, L.H., 2003. Adaptive dancing: interactions between social resilience and
ecological crises. In: Berkes, F., Colding, J., Folke, C. (Eds.), Navigating Sociale
Ecological Systems: Building Resilience for Complexity and Change. Cambridge
University Press, Cambridge, UK, pp. 33e52.
Gustafson, R.G., Lenarz, W.H., McCain, B.B., Schmitt, C.C., Grant, W.S., Builder, T.L.,
Methot, R.D., 2000. Status Review of Pacific Hake, Pacific Cod, and Walleye
Pollock from Puget Sound. Dept. Commer., NOAA Tech. Memo, Washington, U.S.,
NMFS-NWFSC-44, 275 pp.
Hajkowicz, S.A., 2008. Supporting multi-stakeholder environmental decisions.
Journal of Environmental Management 88, 607e614.
Hong, S.-Y., 1995. Marine policy in the Republic of Korea. Marine Policy 19, 97e113.
Leach, W.D., Pelkey, N.W., 2001. Making watershed partnerships work: a review of
the empirical literature. Journal of Water Resources Planning and Management
127, 378e385.
Lee, J., 1998. Policy issues and management framework of Chinhae bay, Republic of
Korea. Ocean and Coastal Management 38, 161e178.
Leschine, T.M., 1990. Setting the agenda for estuarine water quality management:
lessons from Puget Sound. Ocean and Shoreline Management 13, 295e313.
Leung, P., Muraoka, J., Nakamoto, S.T., Pooley, S., 1998. Evaluating fisheries
management options in Hawaii using analytic hierarchy process (AHP). Fish-
eries Research 36, 171e183.
Low, B., Ostrom, E., Simon, C., Wilson, J., 2003. Redundancy and diversity: do they
influence optimal management? In: Berkes, F., Colding, J., Folke, C. (Eds.),
 J. Ryu et al. / Journal of Environmental Management 92 (2011) 92e101
Navigating Socialeecological Systems: Building Resilience for Complexity and
Change. Cambridge University Press, Cambridge, UK, pp. 83e114.
Mardle, S., Pascoe, S., Herrero, I., 2004. Management objective importance in fish-
eries: an evaluation using the analytic hierarchy process (AHP). Environmental
Management 33, 1e11.
Masan-si, 1994. Monitoring of Masan Bay during sediments dredging, (1990e1994).
Millennium Ecosystem Assessment (MA), 2005. Synthesis. Island Press, Wash-
ingtonDC. Available at: http://www.MAweb.org.
Ministry of Land, Transportation and Marine affairs (MLTM), 2008. Basic plan for
coastal TMDL in Masan Bay, a special management area.
Pomeroy, R., Douvere, F., 2008. The engagement of stakeholders in the marine
spatial planning process. Marine Policy 32, 816e822.
Puget Sound Partnership (PSP), 2009. Puget sound action agenda, updated may
27th 2009. Available at: http://www.psp.wa.gov/aa_action_agenda.php.
Puget Sound Regional Council (PSRC), 2008a. Puget Sound trends: population of
cities and towns. Available at: http://www.psrc.org/publications/pubs/trends/
d3oct08.pdf.
101
Puget Sound Regional Council (PSRC), 2008b. Vision 2040: final environmental
impact statement. Available at: http://www.psrc.org/projects/vision/pubs/feis/
index.htm.
Putnam, R.D., 2000. Bowling Alone: The Collapse and Revival of American
Community. Simon & Schuster, New York.
Ramanathan, R., 2001. A note on the use of the analytic hierarchy process for envi-
ronmental impact assessment. Journal of Environmental Management 63, 27e35.
Regan, H.M., Colyvan, M., Markovchick-Nicholls, L., 2006. A formal model for
consensus and negotiation in environmental management. Journal of Envi-
ronmental Management 80, 167e176.
Saaty, T.L., 1988. Multicriteria Decision Making: The Analytic Hierarchy Process. Eta
Services Ltd., Suffolk, UK.
US Commission on Ocean Policy (US COP), 2004. An Ocean Blueprint for the 21st
Century (Final report, Washington DC).
Wattage, P., Mardle, S., 2005. Stakeholder preferences towards conservation versus
development for a wetland in Sri Lanka. Journal of Environmental Management
77, 122e132.