Science of the Total Environment 496 (2014) 248–256

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Stakeholder engagement in dredged material management decisions Zachary A. Collier, Matthew E. Bates, Matthew D. Wood, Igor Linkov ⁎ US Army Engineer Research and Development Center, 696 Virginia Road, Concord, MA 01742, United States

H I G H L I G H T S • • • • •

Stakeholder engagement can facilitate dredged material management decisions. Multi-criteria decision analysis can be used with groups to frame these problems. This approach was used with stakeholders in Long Island Sound. Participatory model building led to shared understanding of dredging issues. Focusing on values rather than management alternatives aided in consensus building.

a r t i c l e

i n f o

Article history: Received 18 March 2014 Received in revised form 11 July 2014 Accepted 13 July 2014 Available online xxxx Editor: Simon Pollard Keywords: Stakeholder engagement Group decision making Sediment management Collaborative model building Multi-criteria decision analysis Long Island Sound

a b s t r a c t Dredging and disposal issues often become controversial with local stakeholders because of their competing interests. These interests tend to manifest themselves in stakeholders holding onto entrenched positions, and deadlock can result without a methodology to move the stakeholder group past the status quo. However, these situations can be represented as multi-stakeholder, multi-criteria decision problems. In this paper, we describe a case study in which multi-criteria decision analysis was implemented in a multi-stakeholder setting in order to generate recommendations on dredged material placement for Long Island Sound's Dredged Material Management Plan. A working-group of representatives from various stakeholder organizations was formed and consulted to help prioritize sediment placement sites for each dredging center in the region by collaboratively building a multi-criteria decision model. The resulting model framed the problem as several alternatives, criteria, sub-criteria, and metrics relevant to stakeholder interests in the Long Island Sound region. An elicitation of values, represented as criteria weights, was then conducted. Results show that in general, stakeholders tended to agree that all criteria were at least somewhat important, and on average there was strong agreement on the order of preferences among the diverse groups of stakeholders. By developing the decision model iteratively with stakeholders as a group and soliciting their preferences, the process sought to increase stakeholder involvement at the front-end of the prioritization process and lead to increased knowledge and consensus regarding the importance of site-specific criteria. Published by Elsevier B.V.

1. Introduction Ensuring navigation throughout the waterways of the United States is one of the missions of the US Army Corps of Engineers (USACE). Navigational dredging is required to sustain the operation of coastal infrastructure and facilitate commerce. In the process, millions of cubic yards of sediment are dredged (235 million cubic yards in 2012) and must be placed or otherwise managed (USACE, 2013). However, sediment management and remediation projects are often performed in complicated political environments where stakeholders are sensitive to different decision paths and actively engaged in championing for or ⁎ Corresponding author at: 696 Virginia Road, Concord, MA 01742, United States. Tel.: +1 978 318 8197. E-mail address: [email protected] (I. Linkov).

http://dx.doi.org/10.1016/j.scitotenv.2014.07.044 0048-9697/Published by Elsevier B.V.

against specific project alternatives (Rogers et al., 2013). Management of contaminated sediments, in particular, can prove challenging in a multi-stakeholder setting, since sites may be contaminated from industrial activities that occurred in the distant past and from multiple sources — the industry entities may no longer exist or responsibility may be difficult to pinpoint, yet the current stakeholders in the community must select a course of action (Sparrevik et al., 2011a). Inviting the active participation of relevant stakeholder groups at the forefront of the decision process can avoid later conflict, but only if all parties feel that their views are being accurately and meaningfully incorporated into the process and if all groups feel that they have a fair say in the final recommendation. Effective stakeholder engagement has been shown to be an invaluable component for the successful design and execution of policies and services in environmental management as well as other domains.

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

Reyers et al. (2009) argue that these planning domains with a strong scientific or technical component are at their core social processes that require effective participation and learning. This engagement and learning, in turn, promote acceptance and adoption of resulting policy plans (Hadorn et al., 2006). Support for this point of view has been found in such environmental domains as catchment management (Allen et al., 2011), water resource and ecological planning (Stevenson et al., 2012), and land stewardship (Cocklin et al., 2007). For instance, Allen et al. (2011) reports that concerted stakeholder engagement efforts resulted in increased collaboration and cooperation. Stevenson et al. (2012) show that this increased collaboration, in turn, can lead to improved ecosystem management. Stakeholder engagement efforts have also produced beneficial outcomes in other domains including public health (Gustavsen and Hanson, 2009; Menon et al., 2007; Sibbald et al., 2009), transportation planning (Forrester, 2009), and corporate social responsibility efforts (Kannabiran, 2009), among others. These efforts are successful in part because they help decision makers to consider factors that might otherwise be neglected during the design process (Cagan and Vogel, 2002). For instance, Sibbald et al. (2009) report that stakeholders that were surveyed as part of an initiative to improve health system priority setting focused on themes of inclusiveness and education for end users, and made no mention of the health outcomes that were the focus of the academics and decision makers who were surveyed. Sparrevik et al. (2011a) evaluated the factors that affected the contentious dredging project in Oslo Harbor, Norway, in which the dredging of contaminated sediments precipitated protests and civil disobedience. They point to the stakeholders' perceptions of transparency and controllability in the decision making process as strong factors affecting their risk perceptions and thus strong negative reactions. This objection to the nature of the process as opposed to the decision itself is supported by behavioral decision research, which shows that both individuals and groups face difficulties in complex decision environments where uncertainties and value judgments must be made, leading groups to revert to established, entrenched positions (McDaniels et al., 1999). To improve this, the National Research Council (NRC) recommends that for contaminated sediment management decisions, “early involvement of stakeholders is important for heading off disagreements and for building consensus” (NRC, 1997). However, not all stakeholder engagement plans are equally effective, and it is not sufficient to merely inform stakeholders throughout the decision making process, one must actively engage them in decision making (Oen et al., 2010; Hermans et al., 2007). Moreover, the NRC recommends structured decision analysis as a method to balance risks, costs, and benefits of various sediment management strategies (NRC, 1997). The conceptual decision models that lie at the heart of decision analytic approaches are a way to represent the shared social reality that reflects the problem at hand (Phillips, 1984). Multi-criteria decision analysis (MCDA) is one such framework for establishing common understanding among disparate stakeholder groups and guiding the process of stakeholder preference elicitation (Keeney and Raiffa, 1976; Linkov and Moberg, 2011). MCDA and related approaches have been used as a vehicle for promoting stakeholder engagement and participation in public policy development, and have been shown to effectively synthesize and address concerns, preferences, and aspirations from disparate stakeholder groups (Kiker et al., 2005). These approaches often use decision analysis or other modeling tools as a way to focus discussions about complex systems across stakeholder groups. Decision conferencing (Phillips, 1984, 2006; Phillips and Bana e Costa, 2007) uses decision analysis methods like MCDA, but focuses more on developing a shared understanding of each others' values, a commitment to the problem, and on-the-fly exploration of how different alternatives may be valued by the group of stakeholders. Valuefocused thinking (Arvai et al., 2001; Keeney, 2009) uses the MCDA structure to facilitate discussion, but starts with the question of what is valuable about the decision rather than the value of the alternatives

249

given the decision. Mediated modeling (van den Belt, 2004) and shared vision planning (Creighton, 2010) use system dynamic models that represent the interactions among system entities and evaluate the extent to which each alternative alters the system and how those changes provide value to involved stakeholders. Stakeholder engagement initiatives that promote effective communication (Fischhoff et al., 2011; Fischhoff and Scheufele, 2013) and are grounded in decision analytic methods like MCDA have demonstrated effectiveness in incorporating a wide range of objectives and stakeholders into the decision making process across a variety of environmental decision-making applications (Estevez et al., 2013). In the sediment management domain, approaches based on or related to MCDA have been applied to several case studies. Sparrevik et al. (2011b) and Kim et al. (2009) involved stakeholder groups in the sediment management process in Norway and South Korea, respectively, but the stakeholder involvement in the MCDA modeling was mainly limited to identification of alternatives and criteria weighting. MCDA was used for managing sediments in the New York/New Jersey Harbor, but in this case, experts were involved instead of a broader set of stakeholders (Linkov et al., 2006, 2007; Yatsalo et al., 2007). Seager et al. (2006) engaged stakeholders in the entire management process of contaminated sediment in the Cocheco River in Dover, New Hampshire. However, only four criteria (human habitat, ecological habitat, environmental quality, and cost) were evaluated in this stakeholder-driven decision model. In this paper, we seek to build upon these existing approaches of MCDA for sediment management responsive to the calls for structured decision analysis by the NRC (1997). We propose to do this by exploring a case study which included a broader set of stakeholders throughout the process of criteria and alternative identification, as well as elicitation of criteria weights. Herein, we describe the multi-criteria, multistakeholder engagement process carried out in Long Island Sound (LIS) in greater detail. First, the specific stakeholder engagement methods and their foundation in MCDA are described, followed by the results gained through working group feedback and stakeholder elicitation. Finally, discussions and recommendations for future work are explored. 2. Methods 2.1. Study area Long Island Sound (LIS), in the northeastern United States, is an important body of water for shipping, recreation, and the environment (Fig. 1). Over the next 30 years, the dredging centers located in the LIS region are projected to produce approximately 38.5 million cubic yards of dredged material (Long Island Sound Dredged Material Management Plan Working Group, 2011a). The local economy of LIS is dependent upon navigation in and around the area. Economic activities include marine transportation (e.g., cargo vessels and chartered fishing services), commercial fishing and seafood industries, recreational boating, ferry-dependent tourism, and the Naval Submarine Base New London (Long Island Sound Dredged Material Management Plan Working Group, 2011b). Together, these activities contribute to an economic output of $9.4 billion per year, including 55,720 jobs and $1.6 billion in annual federal and state tax revenues (Long Island Sound Dredged Material Management Plan Working Group, 2011b). The development of a dredged material management plan (DMMP) for LIS was requested after the US Environmental Protection Agency (EPA) designated two open water dredged material disposal sites in LIS. The current project is motivated in part by a previous multi-year, multi-million dollar failed effort to establish a LIS DMMP because of public backlash over concerns about impacts of open water placement of potentially contaminated dredged material. The overall goal of the LIS DMMP is to develop a comprehensive dredged material

250

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

Fig. 1. Long Island Sound study area (adapted from USACE, 2011).

management plan for the USACE that evaluates a wide array of potential placement sites and methods, and recommends practicable, implementable solutions to manage dredged material in an economically sound and environmentally acceptable manner. The LIS DMMP will include an in-depth analysis of all potential dredged material management alternatives including open-water placement, beneficial use, upland placement, and innovative treatment technologies. These alternatives can be used by dredging proponents in developing alternative

analyses for their dredging in the LIS vicinity. The process calls for Federal agencies to seek public input regarding development of the LIS DMMP. 2.2. Stakeholder engagement approach The stakeholder engagement process began with the formation of a working group, comprising individuals representing Federal, State,

Criteria Environmental Media

Ecological Receptors

Human Welfare

Economics

Sub-Criteria Aquatic

Terrestrial

Air

Fish

Birds Plants

Shell Fish

Benthic

Mammals

Health

Social

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

Short Term

Long Term

Other

Metrics ~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

~~~~~ ~~~~~ ~~~~~

Alternative Placement Sites No Action

Upland Placement

Open Water

Innovative Technology

Suitable Fine

Suitable Sandy

Dredged Material Types Unsuitable

Fig. 2. Dredged material management decision model.

Beneficial Use

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

non-governmental organizations (NGOs), and private citizen groups from both New York and Connecticut (see Supplementary data for a list of participating organizations). Grounded in well-established best practices for stakeholder engagement, the following steps were performed: Step 1: Establish a common understanding. Together, through facilitated discussion, we explored background materials on various dredged‐material placement alternatives so that all working group members could reach a common basis of understanding. Topics included the objectives behind the DMMP, dredging needs in the LIS region, an overview of sediment quality testing, and the definition of various placement alternatives and dredged material types (i.e., suitable fine, suitable sandy, and unsuitable). Step 2: Develop a collaborative decision model. As a group, we developed a broad hierarchy of criteria, sub-criteria, and metrics for evaluating the impacts and benefits of dredged material placement. Based on insights from the field of MCDA, and following the general approach of Seager et al. (2006), the components of the decision problem were identified and organized by asking the members of the working group about what outcomes were important to them, rather than which specific management alternatives they supported. Starting at the highest level, broad criteria were identified, followed by specific sub-criteria, and then metrics with which to assess the sub-criteria. This process was conducted through group elicitation and list building in a series of working group meetings, and then subsequently refined individually through worksheets, and finalized again through elicitation at a working group meeting. Similarly, placement alternatives were elicited as a group by creating a list of alternatives and then were organized by which alternatives were feasible for a given dredged material type.

Table 1 Criterion and sub-criterion definitions.

251

Step 3: Conduct individual interviews. Once a comprehensive decision model was built, the stakeholders were asked to provide their preferences towards the importance of the identified criteria and sub-criteria. In individual interviews, working group members were instructed to respond from the perspective of the organization that they represent by quantifying preferences for and trade‐offs between these impacts and benefits of dredged‐material placement alternatives. This was accomplished in the context of establishing the relative importance of each of the previously identified criteria and sub-criteria. Prior to interviewing, a read-ahead packet was distributed to summarize the progress of the working group and instructions for the interviews. Interviews were conducted individually by telephone, during which interviewees were asked to rank the criteria (or sub-criteria) from 1 to n, where n is the number of criteria or sub-criteria. Item #1 was given 100 points. We then asked, “If item #1 was given 100 points, how many points would you give to item #2 relative to item #1?” This was repeated for all criteria or sub-criteria. In addition, the stakeholders were given the opportunity to add supplemental narrative responses along with their point scores. See the Supplementary data for a sample interview questionnaire. Step 4: Synthesize the data. The interview responses were then coalesced and summarized to show the distribution of priorities encountered in the working group. Points allocated to criteria and sub-criteria were converted into weights based on Eq. (1): S W i ¼ Xni

ð1Þ

S i¼1 i

where wi is the weight of criterion i, si is the score in points assigned to criterion i, and n is the number of criteria being weighted within that particular group of criteria or sub-criteria.

Table 2 Classification of placement alternatives by material type.

Environmental media

Media sustaining organisms or human activities

Alternatives

Uses

Unsuitable Suitable Suitable fine sandy

Aquatic

No action Upland placement

x x x x x

x x x x x

x

Ecological receptors

Specific biological organisms affected by dredged material placement

x x

x x

Birds Fish Shellfish Benthic Mammals Plants Other

All avian species, including shorebirds, waterfowl, and landbased birds Finfish Shellfish, including clams, lobsters, crabs, and oysters Bottom dwelling invertebrates, worms, etc. All mammalian species, land-based and marine Aquatic and non-aquatic plants, including algae Other species of concern, as appropriate

No action Shoreline CDF Upland CDF Mines & quarries Landfills In-harbor CAD cell Confined open water placement Island CDF Unconfined open water placement Near shore placement Varies by product

x x x x x x x

Air

The water areas at or around dredged material placement sites, including the ocean & large bodies of water directly affected by material placement The land areas at or around dredged material placement sites, including soil and the land surface, wetlands, grasslands, agriculture, and brownfields, and including adjacent surface and groundwater The air at or around dredged material placement sites

Human welfare

Non-monetized human values

Health

Human health effects resulting from exposure to dredged materials Social benefits and impacts arising from dredged material placement

Terrestrial

Social

Economics

Monetized costs and benefits

Short term

Monetary effects occurring within 1 year of project commencement Monetary effects occurring 1 or more years after project commencement

Long term

Aquatic placement

Innovative treatment technologies Beneficial uses

x x

Brownfields & other x redevelopments Island creation or x restoration Agriculture/aquaculture Shoreline restoration Habitat restoration/ enhancement of creation Road bed & berm material Landfill & CDF/CAD cap material Beach & dune nourishment Near shore bar placement Asphalt/cement & other aggregates

x

x

x

x

x x x

x x x

x x

x x

CDF = confined disposal facility; and CAD = confined aquatic disposal.

x x x

252

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

Table 3 Participation summary. Number of respondents:

20

By area represented

Number

Percentage

Connecticut New York Other (e.g., Federal)

15 2 3

75% 10% 15%

By primary mission

Number

Percentage

Environmental Commerce Mixed By organization sector Federal/state Local/regional NGO

5 8 7 Number 7 6 7

25% 40% 35% Percentage 35% 30% 35%

This equation normalizes the points given to a particular criterion to the total number of points given to all of the criteria at the similar level in the value hierarchy. The responses were aggregated across the entire group for each criterion and subcriterion through averaging, and categorized in a number of ways depending on the classifications of the respondents, including state, organizational mission, and organizational sector. Step 5: Communicate results. These initial results were reviewed with the working group in a facilitated discussion similar to earlier steps in the process. Working group members were reminded via a presentation of the structure of the model, how their input informed the model structure, and how individual interview data was aggregated and incorporated into the decision model. Members were encouraged to interrupt with questions of clarification and to express comments and concerns throughout the presentation. A broader discussion of the model results and implications followed. 3. Results 3.1. Model framing The structure of the decision model developed by the working group is shown in Fig. 2. Four main criteria were identified, including

environmental media, ecological receptors, human welfare, and economics. Within each of these high level criteria, the working group members identified a number of sub-criteria and metrics. Five broad categories of sediment placement alternatives were also identified. Moreover, an essential part of stakeholder engagement is to ensure that everyone is speaking a common language — therefore much of the front-end of the exercise involved reaching shared definitions of the criteria and sub-criteria (Table 1). After further discussion, the stakeholders indicated that different dredged material types existing within the LIS region would affect their decisions in terms of not only which alternatives (Table 2) are appropriate, but also their preferences regarding criteria and sub-criteria. Specifically, three dredged material types were identified: • Unsuitable — Fine material which contains a significant silt and/or clay component, that has tested as chemically and biologically unsuitable for unconfined aquatic placement and requires some means of containment, whether in-water confined aquatic disposal (CAD) cells or upland, or requires treatment before final use or placement. • Fine — Clean fine grained material with a significant silt or clay component, relatively free of contaminants or from pre-industrial or glacial deposits, suitable for aquatic or upland placement or beneficial uses such as habitat creation. • Sandy — Clean sandy material suitable for beach or near shore bar nourishment, aquatic placement, or a wide range of in-water or upland beneficial uses. 3.2. Quantitative interview results For the individual interviews, 29 organizations were invited to participate, and 20 organizations responded (approximately 69%), each with one participant representing their respective organization. Table 3 gives an overview of the participants, classified by geographic location, mission, and organizational sector. Figs. 3 and 4 show the weights calculated from Eq. (1), derived from the individual interviewing. The error bars in Fig. 3 represent one standard error about the mean. Table 4 shows the summary statistics. A 3 (sediment type) × 4 (main criteria) repeated-measure factorial ANOVA using stakeholder as the within-subject factor identified a significant main effect of the main criteria, F(3,19) = 5.659, p b 0.001, but did not identify an effect of sediment type or an interaction between the main criteria and sediment type. Follow-up pairwise t-tests comparing the different levels of the main criteria with Bonferroni correction to

Fig. 3. Average main criteria weights.

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

253

Fig. 4. Average sub-criteria weights.

control for type-I error showed that scores for economics were higher than scores for other criteria among the stakeholders (p = 0.006, 0.013, 0.001, for ecological receptors, environmental media, & human welfare, respectively). However no criterion ever scores below a 20% importance. 3.3. Response from the working group & qualitative process results Stakeholders noted several insights from the interview process and presented results. With respect to the interviews, they noted that the process of completing the pre-interview worksheet helped interviewees to introspect within their organizations and consider the importance of the factors identified in the decision model with respect to their organization's mission and goals. Often, this exercise promoted perspective-taking for the position of other working group members. Individuals realized that while their organizational missions did not explicitly consider economics, for example, they realized that their organizations did not completely ignore economics and indeed such a position would have been perverse for many organizations. This acknowledgment helped to promote perspective-taking. With respect to the presented interview results, working group members were surprised to see the relative similarity of their ratings to each other. While individuals acknowledged that some variation in scores still existed, several expressed that the presented range of scores represented considerably more agreement than they were anticipating. Indeed, the few participants that rated one criterion highly at the expense of others were perceived by others to be ‘gaming the system’ and therefore uncooperative. 4. Discussion Results from the interviews show that economics was the preferred criterion, with environmental media, ecological receptors, and human welfare all scoring approximately equally (just above 20%). This is partially due to a small number of respondents who assigned all or nearly all of their weight to that criterion at the expense of the others (Fig. 5). This skewed the average weight for economics upwards (hence the large standard error). However, the environmental media and ecological receptors criteria, taken together to broadly represent

environmental considerations, account for a combined weight of 47.2%. In this case, environmental considerations outweigh economics, and come close to being equivalent to the combined weights of economics and human welfare (52.7%). Further inspection shows that when organizational mission is accounted for (environmentally focused, commerce focused, or mixed), broadly environmental focused organizations tended to weigh environmental media, ecological receptors, and human welfare highly, while the commerce focused organizations tended to weigh Economics most highly (Fig. 6). Human health was the highest weighted sub-criterion for the environmentally focused stakeholder groups, while short term economics was the highest ranked sub-criterion for the commerce focused stakeholder groups. However, regardless of the few outliers, it is important to note that when aggregated, all of the four main criteria ended up being weighted very closely to one another, despite the variety of stakeholder groups surveyed. This seems to show that most stakeholders understood and Table 4 Results summary — Weights averaged across all material types. Main criteria

Average %

Median %

Std. dev. %

Range %

Ecological receptors Economics Environmental media Human welfare

23.3 30.7 23.9 22.0

24.6 25.3 25.3 24.1

8.7 19.4 7.1 8.7

37.1 82.6 31.2 43.5

Sub-criteria

Average %

Median %

Std. dev. %

Range %

Ecological receptors — benthic Ecological receptors — birds Ecological receptors — fish Ecological receptors — mammals Ecological receptors — other Ecological receptors — plants Ecological receptors — shellfish Economic — long term Economic — short term Environmental media — air Environmental media — aquatic Environmental media — terrestrial Human welfare — health Human welfare — social

4.4 2.9 3.8 2.7 2.1 3.3 4.1 14.4 16.4 5.8 10.0 8.1 12.0 10.0

4.1 3.3 3.9 3.0 2.0 3.6 3.7 12.3 12.5 6.2 9.1 8.3 12.5 11.2

1.9 1.5 1.7 1.3 1.9 1.7 1.8 9.1 12.5 2.4 4.2 3.1 4.6 4.9

7.3 6.7 6.8 4.4 7.2 10.3 8.6 51.3 58.9 10.6 21.8 12.7 26.4 20.8

Note: Percentages may not sum up to 100% due to rounding.

254

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

Individual and Average Main-Criteria Weights 100%

90%

Suitable Fine Suitable Sandy

Unsuitable 80%

Averages

70%

Weight

60%

50%

40%

30%

20%

10%

0% Ecological Receptors

Economics

Environmental Media

Human Welfare

Fig. 5. Individual weights placed on main criteria.

accepted that no criterion was singularly important and that coupled social-environmental systems are complex and interdependent, and therefore many criteria are necessary for the wellbeing of the LIS area as a whole. This is reflected in the way that the stakeholders agreed to categorize the criteria – roughly approximating the “triple bottom line” of sustainability – environment, society, and economics (Elkington, 1998). The decision to conduct individual interviews to elicit criteria weights, rather than collaboratively as a group, was made for several reasons. First, while beneficial to construct the decision model as a group, elicitation of organizational values was conducted individually so that every stakeholder had a chance to provide feedback in a way that was anonymous to the rest of the group. This was important to establish a “safe” environment for stakeholders to speak their minds without influence or judgment from others. Moreover, gathering individual responses and then aggregating the results allowed for statistical analysis of the set of responses and identification of trends and outliers. In addition, it should be noted that careful selection of criteria is important when building a MCDA model. Belton and Stewart (2002) identify several desirable characteristics of criteria, including value relevance, understandability, measurability, judgmental independence,

operationality, a balance between simplicity and complexity, and conciseness. In particular, non-redundancy (the avoidance of “double counting” a particular criterion and thus increasing its relative importance) is a critical consideration when identifying a criteria set (Belton and Stewart, 2002). In light of these concerns, it is helpful to have a facilitator that is knowledgeable in decision modeling to guide the stakeholder group through the criteria identification and model building process. Finally, it was further found that instead of focusing on specific alternatives, it was beneficial to shift the focus of the discussion of the engagement process on values. Values in this context can be defined as “concepts or beliefs, about desirable end states or behaviors that transcend specific situations, guide selection or evaluation of behavior and events, and are ordered by relative importance” (Schwartz and Bilsky, 1987). Values were expressed by the stakeholders in terms of their enumerated criteria and sub-criteria. By discussing values, working group participants were challenged to think about reasons why they supported one position over another as opposed to the positions themselves, consistent with a value focused thinking (Keeney, 2009) approach to problem framing.

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

255

Main Criteria Weights-By Organizational Mission 40%

35%

30%

Weight

25%

Environmental 20%

Commerce Mixed

15%

10%

5%

0% Ecological Receptors

Economics

Environmental Media

Human Welfare

Subcriteria Weights-By Organizational Mission 20.00% 18.00% 16.00%

Weight

14.00% 12.00% 10.00% 8.00% 6.00%

Environmental

4.00%

Commerce

2.00%

Mixed

0.00%

Fig. 6. Elicited weights based on organizational mission.

While we acknowledge that these espoused theories of value from stakeholders may differ from how these values are expressed in-use (Argyris and Schon, 1974), the current work was conducted in the context of action research (e.g., Baum et al., 2006) and it is therefore difficult to evaluate the extent to which the disconnect between espoused and in-use theory occurred in the present context. This is an interesting question for stakeholder elicitation, especially where MCDA or other decision-analytic techniques are applied. However, such an investigation is beyond the scope of the current work. 5. Conclusions Decision-analytic tools can help streamline sediment-management decisions, add rigor to the decision process and transparently incorporate divergent stakeholder views. There are several important benefits that the MCDA process brings to stakeholder elicitation that may not be readily available through traditional approaches. An MCDA approach is transparent — interested parties can access and understand all assumptions leading to the final prioritization. This is especially important

for continued understanding of the DMMP over the coming decades. Also, the MCDA can greatly contribute to early stage planning, with the goal to utilize stakeholder values to inform prioritization of specific dredging and placement sites. Organizational values can be assumed to be relatively stable (Campbell, 2004), and therefore the weightings can continue to be applied as the data in the region develops over time. Lastly, the MCDA approach is perceived as fair to all involved. Each organization is allowed to express their concerns and potentially influence the site prioritization through individual interviews; this can be achieved without intervention from other participants, while preserving the participatory intent of the overall working group process by reconvening the working group to review the interview results and eliciting feedback. A fair and transparent stakeholder engagement process, aided by a skilled facilitator, allows groups to develop a shared understanding of sometimes disparate views in a way that lets all participants voice their preferences and concerns without any one participant dominating the discussion. This and other forms of production blocking, a reduction in the amount of information shared from taking turns with other

256

Z.A. Collier et al. / Science of the Total Environment 496 (2014) 248–256

speakers, have been shown to negatively influence the diversity of views and ideas expressed by groups especially as the number of participants grows to the size of this working group (Diehl and Strobe, 1987, 1991; Kerr and Tindale, 2004). Moreover, the use of formal decision analytic tools, particularly multi-criteria decision analysis, is a robust process by which problems can be framed, stakeholder views elicited, and relative rankings of alternatives assessed with the goal of providing collective insight into the dynamics of complex decision problems. Conflict of interest The authors declare no conflict of interest. Acknowledgments This work was supported by the US Army Corps of Engineers, New England District and Dredging Operations and Environmental Research Program. Permission was granted by the USACE Chief of Engineers to publish this material. The views and opinions expressed in this paper are those of the individual authors and not those of the US Army, or other sponsor organizations. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2014.07.044. References Allen W, Fenemor A, Kilvington M, Harmsworth G, Young R, Deans N, et al. Building collaboration and learning in integrated catchment management: the importance of social process and multiple engagement approaches. NZ J Mar Freshwater Res 2011;45(3):525–39. Argryis C, Schon DA. Theory in practice: increasing professional effectiveness. San Francisco: Jossey-Bass Publishers; 1974. Arvai JL, Gregory R, McDaniels TL. Testing a structured decision approach: value-focused thinking for deliberate risk communication. Risk Anal 2001;21(6):1065–76. Baum F, MacDougall C, Smith D. Participatory action research. J Epidemiol Community Health 2006;60(10):854–7. Belton V, Stewart T. Multiple criteria decision analysis: an integrated approach. Boston: Springer; 2002. Cagan J, Vogel CM. Creating breakthrough products. Upper Saddle River, NJ: FT Press; 2002. Campbell CR. A longitudinal study of one organization's culture: do values endure? Mid-Am J Bus 2004;19(2):41–51. Cocklin C, Mautner N, Dibden J. Public policy, private landholders: perspectives on policy mechanisms for sustainable land management. J Environ Manage 2007;85:986–98. Creighton JL. How to conduct a shared vision planning process. IWR Report 10-R-6. Alexandria, VA: US Army Institute for Water Resources; 2010. Diehl M, Stroebe W. Productivity loss in brainstorming groups: toward the solution of a riddle. J Pers Soc Psychol 1987;53:497–509. Diehl M, Stroebe W. Productivity loss in idea-generating groups: tracking down the blocking effect. J Pers Soc Psychol 1991;61:392–403. Elkington J. Partnerships from cannibals with forks: the triple bottom line of 21st-century business. Environ Qual Manag 1998;8(1):37–51. Estevez RA, Walshe T, Burgman MA. Capturing social impacts for decision-making: a multicriteria decision analysis perspective. Divers Distrib 2013;19:608–16. Fischhoff B, Scheufele DA. The science of science communication. Proc Natl Acad Sci 2013; 110(s3):14031–2. Fischhoff B, Brewer NT, Downs JS. Communicating risks and benefits: an evidence-based user's guide. Silver Spring, MD: US Food and Drug Administration; 2011. Forrester J. Improved partnership working for local authority transport planning. Eur J Transp Infrastruct Res 2009;9(3):314–30. Gustavsen K, Hanson C. Progress in public–private partnerships to fight neglected diseases. Health Aff 2009;28(6):1745–9. Hadorn GH, Bradley D, Pohl C, Rist S, Weismann U. Implications of transdisciplinarity for sustainability research. Ecol Econ 2006;60:119–28. Hermans C, Erickson J, Noordewier T, Sheldon A, Kline M. Collaborative environmental planning in river management: an application of multicriteria decision analysis in the White River Watershed in Vermont. J Environ Manage 2007;84:534–46.

Kannabiran G. Sustainable stakeholder engagement through innovative supply chain strategy: an exploratory study of an Indian organization. Asian Bus Manage 2009; 8(2):205–23. Keeney RL. Value-focused thinking: a path to creative decision making. Cambridge: Harvard University Press; 2009. Keeney RL, Raiffa H. Decisions with multiple objectives: preferences and value tradeoffs. New York: Wiley; 1976. Kerr NL, Tindale RS. Group performance and decision making. Annu Rev Psychol 2004;55: 623–55. Kiker GA, Bridges TS, Varghese, Seager TP, Linkov I. Application of multicriteria decision analysis in environmental decision making. Integr Environ Assess Manag 2005; 1(2):95–108. Kim J, Kim SH, Hong GH, Suedel BC, Clarke J. Multicriteria decision analysis to assess options for managing contaminated sediments: application to Southern Busan Harbor, South Korea. Integr Environ Assess Manag 2009;6(1):61–71. Linkov I, Moberg EA. Multi-criteria decision analysis: environmental applications and case studies. Boca Raton, FL: CRC Press; 2011. Linkov I, Satterstrom FK, Kiker G, Batchelor C, Bridges T, Ferguson E. From comparative risk assessment to multi-criteria decision analysis and adaptive management: recent developments and applications. Environ Int 2006;32:1072–93. Linkov I, Satterstrom FK, Yatsalo B, Tkachuk A, Kiker GA, Kim J, et al. Comparative assessment of several multi-criteria decision analysis tools for management of contaminated sediments. In: Linkov I, et al, editors. Environmental security in harbors and coastal areas. Netherlands: Springer; 2007. Long Island Sound Dredged Material Management Plan Working Group. Dredging needs fact sheet. http://www.lisdmmp.org/MeetingMaterials/Dredging_Needs_Final.pdf, 2011. Long Island Sound Dredged Material Management Plan Working Group. Economic fact sheet. http://www.lisdmmp.org/MeetingMaterials/Economic_Final.pdf, 2011. McDaniels TL, Gregory RS, Fields D. Democratizing risk management: successful public involvement in local water management decisions. Risk Anal 1999;19(3):497–510. Menon D, Stafinski T, Martin D. Priority-setting for healthcare: who, how, and is it fair? Health Policy 2007;84:220–33. National Research Council. Contaminated sediments in ports and waterways: cleanup strategies and technologies. Washington, DC: National Academy Press; 1997. Oen AMP, Sparrevik M, Barton DN, Nagothu US, Ellen GJ, Breedlove GD, et al. Sediment and society: an approach for assessing management of contaminated sediments and stakeholder involvement in Norway. J Soils Sediments 2010;10:202–8. Phillips LD. A theory of requisite decision models. Acta Psychol (Amst) 1984;56:29–48. Phillips LD. Decision conferencing. In: Operations Research Group, Department of Management, editor. Operational research working papers. London School of Economics & Political Science; 2006. p. 3–41. Phillips LD, Bana e Costa CA. Transparent prioritization, budgeting and resource allocation with multi-criteria decision analysis and decision conferencing. Ann Orient Res 2007; 154:51–68. Reyers B, Roux DJ, Cowling RM, Ginsburg AE, Nel JL, O'Farrell P. Conservation planning as a transdisciplinary process. Conserv Biol 2009;24(4):957–65. Rogers SH, Halstead JM, Seager TP. Characterization of public and stakeholder objectives in environmental management: New Hampshire's Lamprey River. J Water Resour Plan Manag 2013;139(2):217–22. Schwartz SH, Bilsky W. Toward a universal psychological structure of human values. J Pers Soc Psychol 1987;53(3):550–62. Seager TP, Rogers SH, Gardner KH, Linkov I, Howarth R. Coupling public participation and expert judgment for assessment of innovative contaminated sediment technologies. In: Morel B, Linkov I, editors. Environmental security and environmental management: the role of risk management. Netherlands: Springer; 2006. Sibbald SL, Singer PA, Upshur R, Martin DK. Priority setting: what constitutes success? A conceptual framework for successful priority setting. BMC Health Serv Res 2009; 9(43):1–12. Sparrevik M, Ellen GJ, Dujin M. Evaluation of factors affecting stakeholder risk perception of contaminated sediment disposal in Oslo Harbor. Environ Sci Technol 2011a;45: 118–24. Sparrevik M, Barton DN, Oen AMP, Sehkar NU, Linkov I. Use of multicriteria involvement processes to enhance transparency and stakeholder participation at Bergen Harbor, Norway. Integr Environ Assess Manag 2011b;7(3):414–25. Stevenson C, Sikich SA, Gold M. Engaging Los Angeles County subsistence anglers in the California marine protected area planning process. Mar Policy 2012;36:559–63. USACE. Long Island Sound Dredged Material Management Plan: Federal, State, and Local Regulations and Programs Applicable to Dredged Material Placement in CT, NY, and RI. http://www.nae.usace.army.mil/Portals/74/docs/topics/LISDMMP/ FedStateProgramsRegulation.pdf, 2011. USACE. USACE navigation — meeting America's maritime transportation needs. http:// www.usace.army.mil/Portals/2/docs/civilworks/budget/strongpt/fy2014sp_ navigation.pdf, 2013. Van den Belt M. Mediated modeling: a system dynamics approach to environmental consensus building. Washington, DC: Island Press; 2004. Yatsalo BI, Kiker GA, Kim J, Bridges TS, Seager TP, Gardner K, et al. Application of multicriteria decision analysis tools to two contaminated sediment case studies. Integr Environ Assess Manag 2007;3(2):223–33.