Files
Abstract
Payment cards, which identify upper and lower bounds on individual willingness to pay, are
one approach to improve efficiency over the popular dichotomous choice approach to
contingent valuation. This paper reports a split sample test of the impacts on benefit estimates
and efficiency arising from differences in the numbers of divisions on payment cards. Prior
expectations were for increased cell numbers to improve efficiency, but that efficiency gains
would eventually be offset because of increased response variance as cell numbers increased.
Contrary to prior expectations, parameter estimates, standard errors and benefit measures
were invariant to cell numbers.