What can I get for it? The relationship between the choice equivalent, willingness to accept and willingness to pay

We hypothesise and confirm a novel empirical result concerning the willingness to accept (WTA)-willingness to pay (WTP) disparity. Employing data from what has become the classic experimental design, we reveal systematic variation in the relative magnitudes of three valuations: WTA, WTP and choice equivalent (CE). Individuals with low CE relative to others set a proportionally higher WTA, while those with high CE set a proportionally lower WTP. The effect size is substantial in relation to the WTA-WTP disparity itself. These results are predicted by a model in which subjects behave as if maximising surplus over a sequence of exchange encounters, as typifies real exchanges outside the laboratory. They are at odds with models based on loss aversion, which either predict constant relativities between the three valuations or the opposite of the pattern observed.