Testing the initial endowment effect in experimental auctions

Abstract

In experimental auctions, participants are usually given an initial endowment or cash compensation that will cover the costs associated with their bids in the experiment as well as their participation time. This study analyses participants' bids in a randomly binding second-price auction to test the effect of the initial endowment on participants' willingness-to-pay estimates. Three different endowments: $2, $4, and $6 are used to compensate participants in an experimental auction. It is found that participants receiving $4 or $6 as an initial endowment, bid statistically higher than those receiving only $2. Thus, willingness-to-pay estimates are sensitive to the initial endowment. Based on the results obtained in this experiment, it is concluded that endowments that are closer to the value of the auctioned good may be a more appropriate way to compensate auction participants in order to reveal their true willingness to pay for a private good, and to reduce overbidding. Additionally, the findings illustrate that an initial compensation closer to the participants' value of time may inflate their bids, and consequently may not be a correct compensation mechanism to elicit true willingness-to-pay for private goods. New ways of compensating participants in experimental auctions should be investigated.