Is the literature on the WTP-WTA disparity biased?

Abstract

Abstract Empirical studies demonstrate the difference between the willingness-to-pay for a good and the willingness-to-accept giving it away, known as the WTP-WTA disparity. Understanding this effect helps intangible goods to be properly valued or the plausibility of suggested explanations to be adequately judged. We verified the existence of a publication bias in the literature (are studies finding larger disparities more likely to be published?) using 102 studies published over the period 1980–2018. We analysed how the estimate of disparity is associated with the estimation uncertainty using rank correlation and linear/quantile/panel/meta regression; this multiplicity of approaches was used to check the robustness of the results against confounding, skewness, hierarchical data, and possible changes over time. With few exceptions, a bias-suggesting positive association is present (yet decreases over time). Our analysis suggests that the mean disparity estimates may be overestimated, by approximately 12% (3% in more recent studies). Disparity is greater for non-market goods (confirming previous findings) and when the same respondents valued both WTP and WTA. Authors and editors should not withhold publication simply because the results are not regarded as sufficiently controversial (and it seems this problem has been largely overcome). When performing meta-analysis, the latest studies should also be included.