Ascending price clock auctions with drop-out information typically yield outcomes closer to equilibrium predictions than do comparable sealed-bid auctions. However clock auctions require congregating all bidders for a fixed time interval, which has limited field applicability and introduces inefficiencies of its own due to the cost of congregating bidders. This paper studies whether such inefficiencies can be removed by implementing a survival auction mechanism. Survival auctions are multi-round sealed-bid auctions with an information revelation component, in which bidders are successively eliminated from the auction from one round to the next. Such auctions have been shown to be strategically equivalent to ascending-bid clock auctions with drop-out information (see Fujishima et. al. 1999). We conduct an experimental study of a multi-unit demand Vickrey survival auction in which all but the lowest bidder make it to the next round and compare the outcomes to a single round, static-Vickrey auction and to Ausubel's dynamic version of the Vickrey auction with drop-out information provided. The paper investigates whether subjects behavior supports the theoretical isomorphism between the Ausubel auction and the survival version of the auction, and therefore whether such auctions can be successfully implemented in practice.
