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. While individual bid amounts and the price paid by the winning bidder will vary according to auction format, a standard result from the auction literature is that the winning bidder will be the individual with the greatest value for the good being sold. However, this efficiency result only holds when the auctioneer is selling a single, indivisible good for personal consumption to bidders with independent and private values for the commodity.

In alternate environments, i.e., auctions where bidders compete for multiple items or anticipate the opportunity to re-sell the item on a secondary market, it need not be the case that goods are sold to the individual who values them most. In such markets, individuals may strategically manipulate their bid on any given item to affect subsequent outcomes. Since traditional models do not account for such strategic behavior, a number of auction formats such as the first-price, sealed bid tender or ascending open outcry are susceptible to manipulations that can generate large welfare losses. However, there are alternate auction formats that theoretically prevent such strategic manipulations and induce truthful bidding. Given that many of the auction markets run by the BLM are characterized by multi-unit sales and subsequent resale opportunities, it is important they employ auctions that limit the possibilities for bidders to engage in strategic behavior that negatively impacts the welfare of state residents.

Despite the widespread use of auctions to allocate Nevada’s valuable natural resources, surprisingly little is known about the historical performance and design of BLM sales in the state. This project outlines a research proposal designed to fill these gaps by gathering historical data on the structure of BLM auctions and the factors that influence the allocation (and sale price) of natural resources in Nevada. The significance of the proposed work is to combine insights drawn from an empirical analysis of behavior in naturally occurring markets and data gathered from experimental auctions to determine whether the design of existing BLM auctions ensures efficient allocations. In particular, the proposed research will use historical auction data to calibrate experimental markets designed to provide policy makers with empirical evidence that can aid in the design and implementation of optimal mechanisms for the sale of Nevada’s natural resources.

From a normative perspective, such experimental studies provide an important guide for the design and implementation of auction markets in Nevada. In experimental markets, the researcher can alter key model parameters and institutions to compare actual outcomes under a variety of proposed alternatives. Such comparisons provide a basis for selecting amongst an array of policy alternatives that are integral to the design and implementation of actual institutions.

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.