Sealed‐bid Versus Open Auctions: The Evidence

Data from U.S. Forest Service timber auctions are used to test Vickrey's original proposition that sealed‐bid and open auctions yield equal revenue. Models for the high‐bid in an auction are specified and estimated using three different procedures: ordinary least squares, two stage least squares, and maximum likelihood. The second two procedures include a second equation which accounts for the Forest Service's auction choice and therefore control for selection bias. These two procedures, which unlike ordinary least squares yield consistent estimators, imply a statistically and economically insignificant difference in high‐bids between sealed‐bid and open auctions.