GROUP DECISION SUPPORT METHODS TO FACILITATE PARTICIPATIVE WATER RESOURCES MANAGEMENT
In previous WRC-funded projects, the decision analysis group in the Department of Statistical Sciences at the University of Cape Town has developed the concept of Scenario Based Policy Planning (SBPP) for use in the evaluation of strategic alternatives within any public sector planning context but in particular for water resources planning. The key features of SBPP are the following:
- The systematic generation of a relatively small number of scenarios (not in the sense of 'external' scenarios such as different population growths, but in the sense of hypothetical alternatives) to present the range of available strategic options. These scenarios are defined to a requisite level of detail to allow different stakeholders groups to identify a clear preference ordering amongst the alternatives, but no more detail than is necessary for this purpose.
- The use of tools from Multiple Criteria Decision Analysis (MCDA) to assist stakeholder or interest groups to formulate their preferences: The first step is to assist the group in structuring their evaluation in terms of a "value tree", i.e. the criteria against which alternatives should be assessed. The options are then evaluated initially against each criterion individually, after which the evaluations are aggregated into an overall preference ordering. In practice, we have recommended the use of value measurement theory for the evaluation step, primarily because it provides a common currency for comparing the preferences of different groups. In brief, value measurement theory involves a process of scoring options initially against the most basic criteria within which there is little conflict or ambiguity, and then gradually aggregating these scores across more-and- more divergent concerns and interests.
- Comparison of the MCDA outputs from each group in order to identify potential consensus solutions: These outputs provide a mechanism for communication of value judgements and preferences between different groups, and between the groups and policy makers. The resulting information can be used to identify (a) alternatives which are clearly not viable; (b) alternatives which are potentially good compromises between conflicting goals, but in need of specification to greater levels of detail before a final choice can be made; or (c) new alternatives consisting of combinations of features from the alternatives which have been evaluated. This process may be carried out in group discussion, possibly facilitated by other techniques of MCDA.
- Iterative process: SBPP is intrinsically iterative. After one pass through the above process, the surviving policy options may need to be refined and/or supplemented by- additional options, after which the process repeats until there is acceptance that the best consensus has been achieved.
The concepts of SBPP and MCDA have been detailed in previous WRC reports (WRC 296/1/93 and WRC 512/1/97), and are summarised together with some updated concepts in Chapter 2 and Chapter 3 of the current report.
Links to National Water and other Acts
The original development of the SBPP/MCDA procedures were motivated by a realization that strategic decision-making must involve all stakeholders effectively from an early stage of planning. This was recognized as good practice, even though there was at that time little in the way of legislative requirement for involvement of stakeholders in policy formulation and decision processes. More recently, however, the promulgation of the National Water Act of 1998 has both recognized the existing of many potentially conflicting criteria in water resources planning, and mandated the effective involvement of different stakeholders in the process (especially through catchment management agencies or CMAs). The development of the SBPP and related MCDA concepts has thus anticipated the requirements of the new act, and provides a mechanism whereby the intentions of the act can be realized. A detailed comparison is made in the report of the links between SBPP/MCDA and not only the new National Water Act, but also related legislation such as the National Environmental Management Act, the Environment Conservation Act and the National Forests Act. Within the context of the National Water Act, it is argued that the SBPPIMCDA process is directly relevant to the determination of management classes, the determination of reserve for basic human and ecosystem needs, the determination of resource quality objectives and the formulation of catchment management strategies and water allocation.
Conclusions reached from this evaluation of the requirements of the various acts in the light of the SBPP/MCDA process are the following:
- SBPP/MCDA offers a theoretically sound and broadly accessible framework for developing and evaluating alternatives as required by the acts;
- MCDA offers tools with which to define criteria contributing to the overall objectives against which the alternatives can be evaluated;
- MCDA provides the opportunity to include a wide range of inputs (qualitative and quantitative), from different stakeholders, helping to ensure the holistic and transparent assessment which appears to be the intention of the acts;
- MCDA offers a means for developing coherent and justifiable scoring systems for indices, to be used in determining priorities.
Case Studies and Action Research
Much of the research documented in the present report can be classified as "action research". The research team became intimately involved in a number of case studies, in many cases taking the initiative in organizing the group forums and discussions, and coordinating the data collection where necessary. Case studies reported in the main report are the following:
- Sand River project: This was run as a pilot project, commissioned by DWAF, to investigate approaches to catchment planning within an integrated catchment management framework. Our involvement related to the implementation of MCDA in evaluating land-use alternatives (scenarios in our terminology) and their associated water-use implications. Four workshops were conducted, during which impacts on ecological, social and economic issues were evaluated using the "thermometer scale" techniques of value measurement. Our group also needed to take responsibility for the development of a data base to support the evaluations. The overall recommendations are being carried through to a phase II of the project.
- Land-use and forestry in the Maclear district: Some of the initial work on this project was reported in the report WRC 512/1/97. Following on from a WWF-funded project on classifying ecosystems in the region, a series of four workshops were held with representatives of a variety of interests, including conservation, the forestry industry, and local political leadership. The aim was to establish some consensus concerning appropriate levels of afforestation in the region. In many senses, our group needed to take responsibility for arranging and facilitating the workshops and assembling the relevant data. Clear recommendations did emerge from the final workshop, and have been conveyed to the Forestry Review Panel. Final decisions have not been made, and further environmental impact assessments have been commissioned.
- Baviaanskloof Wilderness Area: This work arose from a proposal by Eastern Cape Nature Conservation to expand the Baviaanskloof Wilderness Area. The SBPP/MCDA approach was used to provide inputs into the evaluation of different scenarios which might result from the proposal, in terms of direct, indirect and non-use values. Once again, our inputs involved the use of SBPP/MCDA within a workshop setting, as well as collating information into the required data bas~. At this stage, a preliminary ranking of alternatives has been developed and provided to the main participants. Further evaluation of conservation impacts is being undertaken as part of a WWF-funded project.
- Classification of Estuaries: This exercise differed from the previous three, in that the MCDA procedures were used not to assess policy scenarios directly, but to develop indices for classifying estuaries into management classes, taking into consideration a number of divergent criteria. Part of this exercise involved the use of questionnaires rather than workshops. Contributions emerging from this study are included in the DWAF Resource Directed Measures initiative.
The primary purpose of involvement in the above case studies was to develop an understanding of the dynamics of implementation of SBPP/MCDA in practice. For this reason, it is important to focus on the lessons which can be extracted from the experiences (see next section), and which provide guidelines for the wider implementation of the processes. A brief survey of key participants in some of the case studies indicated that almost all participants found the process itself useful, especially in terms of the holistic integration of the different views provided, while the majority found the basic tools easy to understand. There were some who found certain of the more intricate tools (such as the sensitivity analyses) less easy to understand, and this clearly needs to be addressed in the introductory courses on SBPP/MCDA which are planned for presentation in the next months.
Principles arising from case studies
As indicated in the previous paragraph, an important part of the research was to document the key lessons for implementation of SBPP/MCDA for water resources planning in South Africa that can be extracted from the case studies. These are as follows.
- Role of the facilitator / decision analyst: The decision analysts cannot simply be neutral advisors or meeting facilitators, but need to become an integral part of the project team. The experience from the case studies was that the decision analysis team had an important role to play in interpreting user inputs, in identifying information needs (see next point), and in coordinating data collection and collation.
- Discipline of the MCDA process in identifying critical information needs: The systematic process of evaluation of alternatives in terms of identified criteria, coupled to sensitivity analyses reveals clearly what additional information or quantitative data is or is not important to reaching a justifiable and robust solution.
- Consistency checks and feedback to participants: The MCDA process requires participants to express many value judgements in sometimes quite qualitative and intuitive terms. It is important that the implications of these judgements be fed back to the groups, in terms, for example, of implied trade-offs (such as implied monetary equivalents of social and environmental goals). This is easily incorporated in to the process, and provides participants with a global sense of whether the results are consistent and justifiable. The key point is that while the theoretical foundations of MCDA in general, and value measurement in particular, provide justification for the procedural rationality of the process, it is these consistency checks which provide the basis for claiming substantive rationality, i.e. that the conclusions themselves have validity.
- Allowance of adequate time for the process: By definition, we are dealing here with complex strategic decisions. Although the SBPP/MCDA process can facilitate the process, making it both effective and efficient in communicating values between interest groups, there is still time needed to allow all participants to develop and to share insights, and to establish relevant information needs. The experience from the case studies suggests that for non-trivial problems it would be expected that four or more workshop sessions are typically needed, separated by periods of data gathering and reflection.
- Use of appropriate technology: The process is best supported by some form of decision analysis software (see next section), particularly to allow for rapid feedback of sensitivity analyses and consistency checks. On the other hand, not all participants may be comfortable with direct use of computer tools, and there may be advantages in using "pencil-and-paper" or flip-chart processes in the workshop, with an analyst present to capture the results electronically. One possibility to be investigated within the follow-up project is the extent to which internet-based systems may be advantageous, allowing users to experiment with inputs in their own environment which may be less threatening than in an open workshop.
One of the objectives of this research project was to evaluate and to develop where necessary the appropriate decision support software to implement the SBPP/MCDA process.
The general experience has been that the commercially available V.I.S.A software provides almost all of the support needs for use in workshops and for extensive analysis between workshops. Some possible extensions may be improved links to spreadsheet models, and to GIS systems.
The project leader collaborated with the Institute for Environmental Studies at the Free University of Amsterdam, on the development of a multi-criteria decision support system for use in environmental impact assessments. The intention of this system is to allow specialist groups representing different interests to carry out evaluations at their own time and place. This software is being released in The Netherlands under the name "DEFINITE for Windows", and may be useful for the same purpose here.
As previously indicated, there may be advantages in an internet-based support system. and a first experimental version of such software has been developed and is under testing. This development will continue in a follow-up project.
Recommendations for further research
The results of the research reported here have clearly demonstrated both the viability and the value of the SBPP/MCDA procedures for a variety of water resource planning problems. The following needs for additional research have nevertheless been identified:
- Effective means of integrating the SBPP/MCDA procedures into the regular operational activities of catchment management agencies and other groups concerned with assessing and recommending flow requirements and management plans.
- Full development and implementation testing of internet-based s9ftware support systems, as described in Chapter 10.
- The effective integration of spreadsheet, GIS and other data management systems into the MCDA software.