Different arbitration mechanisms have been proposed to improve upon Conventional Arbitration. While some of them have the theoretical prediction that risk-neutral and risk-averse people will reach an agreement, laboratory experiments show that most of the proposed mechanisms have low agreement rates. This paper proposes an as yet unstudied factor to explain disagreements between disputants. Using a utility function proposed by Fehr and Schmidt (1999) that includes inequality aversion, the model predicts that two risk-neutral disputants might not have the economic incentives to reach an agreement.