Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In a subjective claims problem agents have conflicting perceptions on what constitutes a fair division of a jointly produced cake. In a large-scale experimental study involving a three-agent subjective claims problem, we compare the performance of four mechanisms which use agents’ reports on fair shares as input and yield a division of the cake (or less) as output. The mechanisms differ with respect to the desirable properties they possess and they are compared in terms of efficiency and perceived allocative and procedural fairness. Successful in terms of both fairness and efficiency are two mechanisms that explicitly ask for an assessment of the partners’ fair shares and that do not induce agents to exaggerate their assessment of the own fair share. One of the two successful mechanisms does not ask for an assessment of the own fair share while the other punishes overly selfish own claims.