Low Information Games - Experimental Evidence on Learning in Ultimatum Bargaining

This paper reports experimental evidence on behaviour in an Ultimatum Game where responders have low structural information and feedback so that they have to learn the nature of the game during repeated play. The results lend support to the view that certain learning conditions are less favourable in terms of individual outcomes than others as suggested by the contingent learning approach (Slembeck, 1998). Furthermore, there is evidence that proposers behave "less fair" when responders lack structural information, which contrasts with common notions of fairness or "manners" in ultimatum bargaining (Camerer and Thaler 1995).