A Meta-Analysis on the Endowment Effect in Experiments

While the discussion on the phenomenon of endowment effect has been on-going, it is surprising that few meta-analyses have examined this phenomenon exclusively. Three previous meta-analyses have investigated on the disparity that exists between Willingness to-Pay (WTP) and Willingness-to-Accept (WTA). Experimental and survey studies often find that WTA is much greater than WTP; however, this disparity may not come exclusively from endowment of an object but could also be attributable to seller/buyer perspective, income effects, transaction costs (Randall and Stoll, 1980), commitment costs (Zhao and Kling, 2004), or other psychological variables such as framing (Thaler, 1980). The most recent meta-analysis on the WTP/WTA gap, by Tuncel and Hammit (2014), is based on a previous meta-analysis by Horitz and McConnel (2002), which reviewed 45 studies and examined the effects of type of good and experimental conditions. Horitz and McConnel’s main finding showed that the WTP/WTA disparity was greater for public goods or none-market goods compared to ordinary private goods. Tuncel and Hammit expanded the previous analysis by including new studies up until early 2012 with a total of 76 studies. Tuncel and Hammit replicate prior finding that there exists systematic differences in the WTP/WTA disparity by type of good. However, they showed that this disparity is related to participants’ experience (real world experience and experimental trial experience) in valuing the good. Furthermore, the magnitude of this parity has decreased overtime. The third meta-analysis by Sayman and Onculer (2005) examined 39 studies and found that incentive-compatible designs (such as the goods are tangible or intangible) decreased the WTP/WTA disparity.