Multi-Agent, Multi-Site Harvesting Allocation Under Quotas With a Stackelberg Leader and Network Externalities

We study a harvesting allocation problem consisting of a given number of fishers and a given number of fishing sites. We develop a framework in which one fisher is designated as Stackelberg leader, while the rest are followers. Moreover we introduce network externalities by taking into account congestion costs among the fishers. The presence of a Stackelberg leader when a fishery operates under a maximum harvesting allowance system is a novel approach not only in the sense of extending outcomes obtained in resource allocation problems, but also in indicating the structure of markets that might emerge, which include the possibility of full elimination of followers. The associated policy implications relate to the sustainability of fishing communities. Harvesting allocation in the presence of a Stackelberg leader is compared to existing methods without a Stackelberg leader, specifically the sequential multi-agent resource allocation, the simultaneous cooperative, and the simultaneous competitive allocation in which each fisher tries to maximize their profit given the best response of the others. Results are derived both in terms of numerical examples and an actual case of mussel harvesting