Assessing the Stability of Social Planner Solutions in Multi-Participant Water Conflicts

The use of optimization techniques to identify socially optimal solutions to allocation problems is widely accepted in the water resources literature. For cases with multiple criteria and multiple decision makers (MC-MDM), techniques traditionally lump the decision makers (DM) into a single user and evaluate the criteria, thus not capturing the influence of power dynamics that occur among DMs. Through application of stability methods such as the power index, an alternative method for solving allocation problems, one can assess whether a decision rule is likely to be accepted based on the parties' relative power. This work presents a method for applying the power index to MCMDM problems to determine the most stable decision rule. Because stability is concerned with minimizing the dissatisfaction of the most powerful party, stability can produce different outcomes from the conventional system-level optimization. This work presents this comparative analysis through a case study of the Caspian Sea resource allocation negotiation between five littoral states (Iran, Russia, Kazakhstan, Turkmenistan, and Azerbaijan). Goal programming, compromise programming, and power index methods are employed to evaluate the best scheme for allocation of the Caspian Sea resources. Results show that optimal solutions, analyzed by social planner methods, may not be stable in practice, as certain parties may be too dissatisfied with their allocation to enter into an agreement. This finding has important policy implications as it justifies why stakeholders may not accept the socially optimal solution in practice and underlies the necessity of considering solution stability in social planning and decision making.