Decision Analysis Interviews in the Collaborative Management of a Large Regulated Water Course

Decision Analysis Interviews in the Collaborative Management of a Large Regulated Water Course

Mika Marttunen1 and Raimo P. Hämäläinen2

1Finnish Environment Institute

P.O.Box 140, FIN-00251 Helsinki, FINLAND

E-mail: *

2HelsinkiUniversity of Technology

Systems Analysis Laboratory

P.O. Box 1100, FIN-02015 HUT, Finland

E-mail:

Corresponding author: Tel. +358-9-40300516, fax +358-9-40300590

Abstract

There are always conflicting goals in the management of large water courses. However, by involving stakeholders actively in the planning and decision-making processes it is possible to work together towards commonly acceptable solutions. In this article, we describe how we used interactive multi-criteria decision analysis (MCDA) in a collaborative process which aimed at an ecologically, socially and economically sustainable water course regulation policy. The stakeholders' opinions about the regulation policy options and the relative importance of their impacts were elicited by the HIPRE software. Altogether 20 personal interactive decision analysis interviews, DAIs, were carried out. Our experiences suggest that DAIs can greatly improve the quality and efficiency of the collaborative planning process. In order to gain the full benefits of the MCDA approach, the interactive use of the methods is vital. It is also essential to tightly integrate the approach into the planning and decision-making process. The project's homepages are publicly available at

Key words: Multi-criteria decision analysis, decision analysis interview method, public participation, conflict management, stakeholder values, lake regulation, sustainable management

1 Introduction

The management of natural resources is nowadays a very challenging and multifaceted task. Modern societies are more diverse, environmentally conscious, and the number of stakeholders is greater than a few decades ago (see e.g. Renn and others 1995, Senecah 2004). New legislation and directives, for instance, the EU Water Framework Directive (2000/60/EU), set more demanding requirements for environmental planning.

Different participatory approaches have been used in several hundreds of environmental planning and management projects, particularly in Europe and North America. The lessons learned provide valuable information and insights for planners and managers (see e.g. Chess and Purcell 1999, Duram and Brown 1999, Susskind and others 1999, Wondolleck and Yaffee 2000, Beierle and Cayford 2002, Connick and Innes 2003,).

Stakeholder based planning has proved to be a key to a good and successful public involvement process. Due to the positive experiences, stakeholder processes are being developed beyond the traditional format of public hearings and meetings into methods that involve relatively small groups of people in intensive and collaborative processes (Beierle 2002). Typically, these have an active involvement of stakeholders that work together to identify problems, define objectives, share information, and where possible, develop collectively acceptable solutions which can not be solved individually (see e.g. Wondolleck and Yaffee 2000, Daniels and Walker 2001, Nandalal and Simonovic 2003). The underlying philosophy is to explicitly take into account emotions and social factors because these have a crucial systemic impact on the process (see e.g. Fisher and Shapiro 2005).

The understanding of the ecological and social impacts of water course management together with the planning tools have also improved considerably. As a result, we nowadays have many mathematical modelling techniques, decision support tools and GIS applications available (see e.g. Hämäläinen and others 2001, Hämäläinen 2004, Nandalal and Simonovic 2003, Mysiak and others 2005, Mustajoki and others 2006).

In complex environmental management problems it is typical to have a large amount of impact information and limited ability of decision-makers to absorb and process it. Multi-criteria decision analysis (MCDA) provides a way to manage this extensive amount of information and diversity of opinions in environmental planning processes. The MCDA approach has been explored and applied in several water resource planning and management projects (see e.g. Marttunen and Hämäläinen 1995, Keeney and others 1996, McDaniels and others 1999, Keeney and McDaniels 1999, Hämäläinen and others 2001, Bana e Costa and others 2004, Hostmann and others 2005, Marttunen and Suomalainen 2005).

MCDA methods help to improve the quality of decisions involving multiple criteria by making the communication and choices more transparent, explicit, rational and efficient. MCDA is also a tool to enhance individual learning and to support group decisions by eliciting, understanding and managing the stakeholders' values and objectives (Gregory and Keeney 1994, Hobbs and Meier 2000, vonWinterfeldt 2001, Belton and Stewart 2002). So far, there are only few reported cases where MCDA tools have been used in a truly interactive way and linked tightly with a real decision-making process (Hämäläinen 1991, Marttunen and Hämäläinen 1995, Gregory and Failing 2002, Ananda and Herath 2003, Marttunen and Suomalainen 2005). Our first experiences from decision analysis interviews with real decision-makers date back to the mid 80's related to a nuclear power plant license decision in the Parliament of Finland (Hämäläinen 1988, 1991).

There are recent studies where stakeholders have been satisfied with the use of MCDA methods (Hostmann 2005). On the other hand, in some other cases participants have had problems in understanding the MCDA procedures (Corner and Buchanan 1997, Pykäläinen and others 1999, Bell and others. 2001, Sinkko and others 2004, Bojórquez-Tapia and others 2005). Our results show that by carrying out value tree analysis individually and interactively, it is possible to overcome many problems related to weight elicitation, and to improve the participants' understanding and acceptability of the method.

This article describes how we applied the decision analysis interview (DAI) method in the collaborative planning process which aimed to develop a sustainable regulation policy for a large regulated watercourse. The study had several objectives related both to the development of the method and to the achievement of the goals set for the collaborative process. The article is structured in a following way. First, we present our case study, the LakePäijänne regulation development project. Second, we describe why and how the DAI process was undertaken. Third, we synthesize the results of the interviews. Fourth, we analyze how the DAI method supported the group decision-making and the consensus finding process. Finally, we draw conclusions about the use of the DAI method.

2 The LakePäijänne regulation development project

LakePäijänne is the second largest lake in Finland with a surface area of 1 100 square kilometers. The lake has an extensive recreational housing development along its shores and there are tens of thousands of recreational users and recreational fishermen. The River Kymijoki with a length of 120 kilometers originates from LakePäijänne and flows into the Gulf of Finland. The river has twelve power plants which generate about ten percent of the hydropower in Finland. The fields along the banks of the river are particularly prone to flooding.

The lake regulation started in 1964 (Table 1). Initially, the primary goals were to increase hydro power production and to decrease flood damage both along the lake and the river. Since the beginning of the regulation, the recreational use of the water course has increased dramatically. Recreational users of the water course have also become much more aware of the environmental effects of the regulation. The most important concern has recently been the negative impact of the regulation on the aquatic ecosystem and the inappropriate water levels of the lake for recreational use during late spring.

Already in the mid 1990s, there was a wide consensus about the need to find opportunities to modernize the regulation policy. Provincial federations, fisheries organizations as well as the holder of the regulation license, the Ministry of Forestry and Agriculture, considered this important. As a result, a large and multi-disciplinary development project was carried out during the years 1995-1999 to re-evaluate the regulation policy of the water course. The project’s aim was to collect new information for decision-making by assessing and synthesizing the ecological, economic and social impacts of water level fluctuations, and to develop recommendations which would reconcile different and partly conflicting interests of water course users.

In the beginning of the project, there was a strong mistrust especially among the fisheries organizations towards the project and the administrative bodies responsible for it. These included the National Board of Waters and Environment which later became the Finnish Environment Institute, and the Ministry of Agriculture and Forestry, which was the license holder of the regulation at that time. In order to gain public support for the project and to improve opportunities for finding a commonly acceptable new regulation policy, we felt that a systems intelligent (SI) approach (Saarinen and Hämäläinen 2004) was clearly needed and therefore an open and participatory planning process was launched. In the SI approach one acknowledges the fact that we, in this case the project and stakeholders in the steering group, are always in a systemic relationship where the process in which interaction is carried out has an essential impact on the outcome. The requirement for the co-operative consensus seeking process is, in fact, also stated in the Finnish Water Act. However, in the project much more attention was paid to stakeholder involvement and public participation than the Act would require.

Postal questionnaires, workshops, public hearings and working groups were among the means to learn the opinions and listen to the local people. There were more than 50 working group meetings and more than 100 different people participated in them. In order to collect opinions from the general public, a postal questionnaire was sent to over 2 000 property owners. The most important forum for the stakeholder involvement was the steering group which comprised eighteen representatives. Both public authorities and different interest groups were included (Table 2). The role of the steering group was to discuss and approve the annual working plans of the project. However, the most crucial task was to develop recommendations which would be approved by all the stakeholders involved.

In addition to the DAI method, the work of the steering group was supported by mathematical models used to simulate the lake and river hydrology (Hämäläinen and Mäntysaari 2001). The methods were applied complementarily and each of them had their own role in the planning process. In this article, we focus on the results and experiences of the DA interviews (Figure 1).

3 Phases of the process

The process was built on an environmental impact assessment (EIA) approach with two new features. First, there was continuous stakeholder involvement and public participation. This is different from the old EIA tradition where people are heard in the beginning and at the end of the process. Second, participation was tightly integrated into the planning process, and planning and participation were developed in parallel in a closely interconnected way. This reflects our vision that the recognition of the systemic nature of participation processes is of crucial importance. A process oriented SI approach helps to change the mental models of the stakeholders from conflict management to collaborative consensus seeking.

The process consisted of four main tasks: 1) Framing the problem, 2) Assessing the impacts of the old regulation, 3) Generation and comparison of the regulation options, and 4) Development of recommendations and follow-up measures. The first three tasks were run partly in parallel. Public involvement was important in each task.

3.1 Framing of the problem

The scope of the project and the needs for further data collection were determined in the working groups. We created preliminary regulation options and identified attributes which were used to assess the impacts. The attributes were organized hierarchically in a value tree. The project covered the entire water course affected by the regulation, including both LakePäijänne and the outflowing River Kymijoki. In addition to the revision of the lake regulation policy, the group also decided to consider different non-hydrological mitigation measures.

3.2 Assessing the impacts of the old regulation

The project comprised eighteen subprojects which generated an extensive amount of information on the ecological, social and economic attributes. Field studies, various ecological models and expert judgments were used in the impact assessment phase. Numerical estimates of the impacts were given whenever possible. Furthermore, the studies identified constraints for planning i.e. non-acceptable water levels and flow rates for various uses of the water course.

3.3Generation and comparison of options

An evolutionary three-step approach was developed in order to find feasible regulation options. First, each representative of the steering group was interviewed with the DAI method. In the interviews, the data acquired in the environmental impact analysis was combined with the subjective preferences of each participant to evaluate the overall subjective value of each option. The practical implementation and the results of the DAIs are described in sections 4 and 5.

In the second step, the consensus finding process was put to the test as the aim was to find generally acceptable objectives in the steering group for regulation policy under the different water scenarios. First, the results of the DA interviews were analysed and discussed in the steering group. These discussions improved the participants' understanding of the other stakeholders' opinions, and also helped to understand the diversity of the opinions. This phase also showed that without a comprehensive analysis which includes all relevant impacts and objectives it was not possible to find a commonly acceptable solution. The first target water levels and flows for LakePäijänne and the River Kymijoki over the year were determined based on these general objectives.

In the third step, a hydrological simulation model was run for the period 1971-1995 in order to study how well the target levels could be achieved during dry, normal and wet hydrological conditions and what kind of ecological, social and economic consequences there would be. The results of each simulation were carefully analysed and targets for water levels and flows were refined if the outcome was not acceptable. The main problem in the development of a feasible regulation policy was to find a balanced strategy which would not cause big losses in hydro power or large increases in flood risks.

3.4Recommendations and follow-up

The most important and challenging task was the composition of a commonly acceptable set of recommendations for the future regulation policy. The analysis of the impacts and the increased awareness of the hydrological dynamics related to the water course regulation provided a very good basis for that. It turned out that the opportunities to diminish the adverse impacts of regulation on the aquatic ecosystem and recreational use were fairly limited. Therefore, we looked for other mitigation measures that could be used. We introduced recommendations about fish stock management, habitat restorations of the shorelines of the lake and rapids in the River Kymijoki and informing of the public. The total number of final recommendations was 31, fifteen of which considered the regulation practice, five dealt with fish stock management, four the restoration of habitats, five the improvements of communication, and two the follow-up of the implementation of the water course regulation and the recommendations.

4 Decision analysis interviews

4.1 Objectives and their achievement

We had several objectives with the introduction of the DAIs. The most important one was to support the participatory consensus-seeking process by empowering the participants to personally compare regulation options with respect to both intangible and incommensurable impacts. This allowed the stakeholders to evaluate their own priorities and values in this context with the real data. The three main issues addressed were:

Which impacts are perceived to be the most important ones?
Which are the most preferred and the most disliked components in the regulation strategies?
What are the important differences in the opinions between the stakeholders?

Furthermore, we had the methodological objective of learning how to avoid behavioral biases (see e.g. Pöyhönen and Hämäläinen 2000, Hämäläinen and Alaja 2003) by a suitable structuring of the problem and description of the impact data.

The interviews were preceded by an extensive preparation and testing phase. During this the objectives and attributes were selected, the value tree was structured, the impacts of the regulation options were assessed, and the weighting technique was chosen. Once the preparations were completed the personal and interactive interviews were started. By working individually with each participant, the analyst could ensure that all the issues were clear and no misunderstandings remained (Marttunen and Hämäläinen 1995, Hämäläinen and others 2001). The total number of interviews was twenty and they lasted from three to six hours. The interactive value tree analyses were done with the HIPRE3+ software ( Hämäläinen and Lauri 1992), the predessor of the WEB-HIPRE software (Mustajoki and others 2004) used in some of the expert group meetings.