A multitude of natural resources in Nevada (such as public lands and wild horses) are allocated via auctions whereby the individual who places the highest bid obtains ownership rights and pays a price determined by the associated auction rules. Despite the widespread use of auctions to allocate Nevada’s valuable natural resources, surprisingly little is known about the historical performance and design of Bureau of Land Management (BLM) sales in the state. This project develops a research proposal designed to address this deficiency and determine the optimal mechanism/s for selling Nevada’s scarce natural resources.

Beginning with the work of Vickrey (1961), an extensive literature has developed that examines the design and application of auctions to allocate property rights and assign ownership to goods and services. This literature has largely assumed that the auctioneer sells a single, indivisible object and employed a simplifying assumption of no resale opportunities. Goods sold at an auction were ultimately consumed by the winning bidder. With a competitive market and homogeneous agents, bidding strategies are monotonically increasing in bidder types and a symmetric separating equilibrium exists where goods are allocated to the bidder with the highest expected use value. Further, revenue equivalence theorems suggest that for auctions with an infinite number of participants, performance measures are independent of the auction format—i.e., first-price, second-price, Dutch, or English auction.

Real auction markets, however, often present bidders with the opportunity to bid on multiple items with possible ex post exchange opportunities which may be influenced by specific rules established by the auctioneer. For example, when bidders anticipate resale opportunities, both the initial auction format and secondary market structure directly influence the efficiency and performance of the primary market. For instance, the Perfect Bayesian Nash bidding strategy in an English auction with resale is the same as that for an English auction without resale when the private types of bidders are publicly known on the secondary market (Haile, 1999). However, the Perfect Bayesian Nash bidding strategy for a first-price auction with resale differs from that of a first-price auction without resale even when private types are common knowledge on the resale market (Haile, 2003).

We argue that existing BLM auction design may not account for important factors such as multi-unit sales and subsequent resale opportunities that influence equilibrium bidding strategies. This is not surprising given the difficulties in executing clean empirical tests to identify the effects of such factors on bidder behavior. As such, it is difficult to determine how alternate auction formats may attenuate the tendency for bidders to strategically manipulate their behavior for personal gain. For example, the impacts of resale on bidder behavior is linked to the informational structure and pricing mechanism assumed in the secondary market. Without an ability to examine data from a “natural experiment” with a controlled secondary market, it is difficult to determine how changes in the primary auction format would effect market outcomes. Experimental markets and laboratory studies alleviate many of these concerns and provide an attractive basis for analyzing bidder behavior. The proposed research will use insights garnered from historical BLM auction data to calibrate lab experiments designed to evaluate the optimal auction format for selling natural resources in Nevada.