This paper reports the first test evaluating whether stated choices as
part of a contingent valuation survey discriminate between significan
t and trivial causes. Using a random digit dialed sample for North Car
olina, two different plans were posed to respondents-one to expand a p
opular North Carolina highway flower planting program nationwide and a
second to facilitate the use of recycled tires in making asphalt for
highways. Random assignment, identical payment mechanisms, and overall
question designs were used The results indicate clear discrimination
in choices, choice functions, and the estimated willingness to pay for
the two plaits.