WTA estimates using the method of paired comparison: Tests of robustness

The method of paired comparison is modified to allow choices between two al ternative gains so as to estimate willingness to accept (NTA) without loss aversion. The robustness of WTA values for two public goods is tested with respect to sensitivity of the WTA measure to the context of the bundle of g oods used in the paired comparison exercise and to the scope (scale) of the public good project. There was no statistical difference in WTA measures f or open space based on independent treatments with different contexts. One treatment involved valuing open space within a set of goods with similar va lue. The other treatment involved valuing open space within a choice set of goods which had lower dollar values than open space. There was a statistic al difference in WTA between a permanent expansion in the bus system and a temporary expansion in the bus system. We conclude the method of paired com parison appears to be a very promising approach to elicit measures of WTA.