TILEC Work In Progress: Gyula Seres
10:45-11:45, M 1003
Strategic Anchoring: An Experimental Test in Auctions
Abstract: Anchoring is a widely observed cognitive bias triggered by an obviously irrelevant number that influences judgement. Following this line line of research, we focus on anchoring effects on strategies. In some markets including bargaining and auctions, agents are exposed to potential anchors which limit the strategy space, but do not convey information. We test for the presence of anchoring in games with an experiment in private-value first-price and silent Dutch auctions. Subjects are exposed to two anchors. For one, some treatments include a very high upper limit of bids. For two, subjects may receive a randomly allotted number before the auctions. We find evidence of anchoring bias of both types. Moreover, the interaction effects are negative, suggesting that different type of anchors weaken each other. We also find that the sign of the effect of the upper limit is context dependent as it is negative and significant in silent Dutch auctions.