The acquisition of information is an important feature in most auctions where one’s exact private valuation is unknown ex-ante. We conducted the first experiment in testing a risk-neutral expected surplus maximization model with this feature. Varying the auction format and the cost of information acquisition we found bidders in most cases acquired too much information. Moreover, bidders who remained uninformed placed bids significantly below the optimal bid. The general prediction concerning revenue and efficiency remains valid, as a higher information cost was associated with lower revenues and efficiency rates. We explore different ex-post explanations for the observed behavior and show that regret avoidance can explain the data while risk aversion and ambiguity aversion cannot.

Gretschko, Vitali and Alexander Rajko (2015), Excess information acquisition in auctions, Experimental Economics 18, 335–355. Download


Gretschko, Vitali
Rajko, Alexander


Dynamic auctions, Information acquisition, Bidding behavior