A multi-agent cellular automata model to explore water trading potential under information transaction costs

The economic literature has extensively researched water markets, finding that this instrument can deliver superior resource allocations under growing scarcity. However, these assessments typically ignore the information, bargaining, and enforcement costs that occur in trades, known as transaction costs, which can be particularly relevant in emerging markets such as those for water. This paper presents a multi-agent cellular automata model that couples a positive multi-attribute utility programming (PMAUP) model with an agent-based model (ABM) to elicit the information transaction costs of water trading under alternative market and climate scenarios. A first experiment evaluates an ideal water market with no transaction costs, which is then compared to alternative decentralized spot market setups representing different degrees of information availability that are modeled using the coupled PMAUP-ABM. The difference between the economic surplus under the ideal market and that achieved under alternative market setups with information constraints is used as a proxy value of information transaction costs. Methods are illustrated with an application to the Douro River Basin in central Spain. Results show that information transaction costs can reduce the economic surplus between 0.6% and 45% depending on the scenario.