Free Exchange, the blog at The Economist, has a nice find and write-up:
Un-Endowing the Endowment Effect, by Economist.com: Let's say you agree to participate in an economics experiment. You show up at the lab ... and are randomly assigned to a specific group. You are then given a coffee mug. Finally, you’re asked if you’d like to trade the coffee mug for a candy bar. If you’re like most ... participants..., you probably don’t trade, but stick with what you’ve got. And perhaps it really is an awfully nice coffee mug, so you've made the right decision. Yet something perplexes economists. When the experiment is repeated with the other group, where the candy bar is the endowed good, most of them keep the sweet instead of taking home the mug.
And that, according to the behavioral economists like Nobel laureate Daniel Kahneman and University of Chicago’s Richard Thaler, is a direct challenge to the deep premises of neoclassical economics. Since the goods were randomly distributed, neoclassical theory predicts that there should have been much more trading than there actually was. Thus the concept of the “endowment effect” was born. It seemed to explain a whole host of other exchange asymmetries, too, such as why people often require a higher price to sell a good than they would ... pay to buy it.
The theory is that everyone in the experiment was acting on ... “loss aversion”..., which causes us to worry more about losses than equivalently sized gains. Parting with an endowed good is perceived as a loss greater than the potential gain from acquiring another good of putatively equal value.
Now a new paper ... forthcoming ...[in] the American Economic Review argues that this asymmetry might not be as formidable as it seems. The paper is based on experiments conducted by Charles Plott of Cal-Tech ..., and Kathryn Zeiler of the Georgetown University Law Center. (A working paper version is available here.)
Plott and Zeiler thought that perhaps traditional signaling theory could help explain the results of those previous experiments. For instance, when the endowed good was handed to the experiment participant, they were usually told, “I’m giving you the mug. It is a gift. You own it. It is yours.” But what if that signaled a certain level of value to you as the recipient of the mug? You don’t know if that candy bar is any good, but the chap who handed you mug seemed really insistent that you should hold onto it.
So, Plott and Zeiler simply told the participants: “The mug is yours. You own it.” They also adjusted for other possible factors. ...[T]hey had students signal their decision to trade (or not) by anonymously marking a card, rather than raising their hands in the midst of a crowd. And the participants got to inspect the other good, without giving up the one they had, before they made their choice.
The result? The exchange asymmetries disappeared. ...