Abstract
We compare hypothetical and real payments in a Vickrey auction. The use of hypothetical payments overestimates the real willingness to pay. The results cast doubt on the validity of hypothetical willingness to pay questions for the typical respondent.