The endowment effect is the finding that possession of an item adds to its value. We introduce a new procedure for testing this effect: participants are divided into two groups. Possession group participants inspect a numbered lottery ticket and know it is theirs, while inspection group participants only inspect a lottery ticket without being endowed with it. Subsequently participants choose between playing the lottery with this (possessed or inspected) ticket, or exchanging it for another one. Our procedure tests for the effect of endowment while controlling for the influence of transaction costs as well as for inspection effects and the influence of bargaining roles (buyer vs. seller), which often afflict experimentation with the endowment effect. In a real setting, tickets in possession were valued significantly higher than inspected tickets. Contrary to some findings in the literature participants also correctly predicted these valuation differences in a hypothetical situation, both for themselves as well as for others. Furthermore, our results suggest that regret rather than loss aversion may be the source of the endowment effect in an experimental setting using lottery tickets. Applying our procedure to a setting employing riskless objects in form of mugs revealed rather ambiguous results, thus emphasizing that the role of regret might be less prominent in non-lottery settings.
Keywords: Endowment effect; Ownership; Prediction; Regret; Simulation.