Multi-agent, multi-site resource allocation under quotas with a Stackelberg leader and network externalities

Abstract

We model multi-agent resource allocations consisting of many agents and sites, analyze serial dictatorship and simultaneous allocation approaches, and then apply them to an actual case of mussel harvesting. For tractability, the model is developed in the context of a multi-fisher, multi-site fishery where one fisher may emerge as a Stackelberg leader while the rest are followers, but the generalized model can be used to study many other resource allocation problems. Network externalities are introduced in the form of congestion costs, and management systems based on individual and collective quotas are analyzed. The presence of a Stackelberg leader is a novel approach that extends results obtained in resource allocation and indicates the possibility of full elimination of followers, with impacts on the sustainability of fishing communities. We compare allocation outcomes under different assumptions and discuss policies that could prevent the followers from exiting the market.