Recent theoretical developments in multi-unit demand (one sided) auctions make it clear that uniform price auctions create an incentive for demand reductions (bid shaving) resulting in inefficient outcomes. In contrast, extensions of the Vickrey auction rule to multi-unit demand auctions promise, in theory, to restore economic efficiency and to raise expected revenue relative to uniform price auctions under a wide variety of circumstances. These predictions rest critically on bidders recognizing and responding appropriately to the strategic opportunities for demand reductions in uniform price auctions and following a "sincere" bidding strategy in the Vickrey auctions. The purpose of this research is to explore behavior in multi-unit demand auctions through both theory and experiments. The experiments will focus on the extent to which bidders reduce demand, and the resulting inefficiencies and revenue impact, in uniform price auctions when bidders have equal and unequal demands and with more or less intense competition. Vickrey auction experiments will focus on the effects of different implementation rules for inducing sincere bidding using a relatively transparent, open English clock auction procedure and a sealed bid version of the clock procedure that may prove more tractable in field settings. Behavior will be studied in both interactive and non-interactive (games against computerized bidders) environments with the latter designed to eliminate the "noise" associated with other players' actions on learning to behave strategically. Theoretical work will provide benchmark models against which to evaluate the experimental outcomes and to identify conditions where there are tradeoffs in theory, between efficiency and revenue under the different auction institutions. Work in multi-unit demand auctions is important because (i) most auctions involve multi-unit demands and (ii) a number of insights obtained from unit demand auctions fail to hold with multi-un it demands. Many of these issues have come to the forefront-with the design and implementation of the recent spectrum auctions.
