The Willingness to Pay-Willingness to Accept Gap, the "Endowment Effect," Subject Misconceptions, and Experimental Procedures for Eliciting Valuations

TLDR: Isoni et al. as discussed by the authors used mug valuations to test endowment effect theory and showed that the gap for commodities can be turned on and off by implementing procedures designed to control for subject misconceptions about the value elicitation procedures.

Abstract: The purpose of Plott and Zeiler (2005)—henceforth, PZ—was to investigate whether previously published experiments using consumption goods such as mugs and candy bars to measure gaps between willingness to pay (WTP) and willingnessto-accept (WTA) support endowment effect theory (EET). Our results demonstrate that the gap for commodities can be turned on and off by implementing procedures designed to control for subject misconceptions about the value elicitation procedures. Following experiments traditionally used to demonstrate the endowment effect, we used mug valuations to test EET. We used lottery rounds only to provide our subjects with paid practice using the value elicitation device prior to employing that device to elicit subjects’ mug valuations. In a footnote, we report evidence of contamination in the lottery data that rendered it inappropriate for our purposes (PZ 2005, fn. 15). Our footnote summarized the details of misconceptions reported in a data supplement provided to all who request our lottery data. The supplement was not referenced in our paper, so we make it available as an online Appendix to our Reply. 1 Andrea Isoni, Graham Loomes, and Robert Sugden (2011)—henceforth, ILS—use experimental procedures similar to ours and observe, just as we did, no gap in mug valuations. ILS claim that our 2005 paper is misleading and has misled researchers. They are concerned that our paper produces, and has been interpreted as producing, a set of procedures sufficient to remove all gaps, including gaps in lotteries. To justify their concern they focus on the wording of our abstract and overlook the context of paragraphs from which they quote sentences to support their thesis. The result is what we consider to be a misleading picture of the content of our paper and the facts that we report, namely that mug gaps disappear after we implement controls for misconceptions and that none of our data provides support for EET. We want to emphasize that our focus was on EET and not on more general theories of prefer ence formation, reference effects and decision processes that have emerged in the literature more recently and might explain our results. 2 In Section I, we demonstrate