Government‐industrial‐research cooperation in virtual water strategy: A multi‐agent evolutionary game analysis

The virtual water strategy (VWS) is an effective tool to balance regional water resource endowments and guarantee water supply security. However, because of self‐interested games around VWS (human decision bias), there is a need for methods to maintain reliable cooperation between governments, virtual water (VW) enterprises and research institutions. This study builds a multi‐agent evolutionary game model to analyse the relationship of players and their impacts on VWS through changing decision mechanisms and the paths to enhance their confidence in cooperation. Considering differences in initial willingness to cooperate and changing factors affecting payoffs, an evolutionary game can produce changing stable equilibriums or stable cooperations, even if some players are reluctant to cooperate. Therefore, to promote the development of VWS, a multistep support mechanism can be built for the VW industry, which fosters model enterprises and optimizes the cooperation framework to stimulate research innovations at scientific institutions. Highlights Traditional virtual water strategy (VWS) studies ignored the specific implementation and stakeholder game issues. A multi‐agent evolutionary game model is built to analyse the stable states and decision mechanisms for VWS. The payoffs of each game player will affect its strategy choice and the speed of model evolution. It is clearly possible for the stakeholders to form a stable equilibrium of long‐term cooperation around the VWS. Changing the factors of one game player may indirectly lead to changes in the strategies of other players.