OFSTRA · Optimal Fairness and Safety for Task and Resource Assignment

Lines connect the coordinator with its partners.

Project objective

OFSTRA will explore the deep connection between the existence of optimally fair allocation rules and the strategic safety of their implementations. An allocation rule is ""fair"" if it passes all the permutation tests deemed relevant by policy makers. Thus, if policy deems that two participants in the allocation problem can consider trading places, and if one of them desires to do so, then the permutation test is failed and the allocation not fair. Fair allocations exist in many applications of game theory, and seemingly miraculously, there is often an allocation that all participants agree is the best of these. More miraculous still is the following: in the process of calculating this best allocation for the problem at hand, there is often no need for participants to strategize about what they should reveal. Fairness of course depends on the preferences of the participants, and there are game theoretic circumstances in which hiding or even falsifying one's preferences can be advantageous. In such circumstances, participants must be strategically sophisticated---about what to hide and what to reveal---to ensure that their welfare is respected. When a process does not require such strategic sophisitication, we call it ""safe"". It is not yet fully understood why optimal fairness and safety so often coincide, or even the full extent of the circumstances in which they do. This project will delineate the boundaries on which they coincide and understand the structure that generates these boundaries.""

Original text from CORDIS.

Participants

• TILBURG UNIVERSITY- UNIVERSITEIT VAN TILBURG · TilburgCoordinatorNetherlands

Links

• View on CORDIS
• DOI: 10.3030/101277157