Ecosystem-Service Choice Experiment

Contingent valuation typically asks for willingness to pay for a single, bundled environmental change, often via a referendum-style question. A choice experiment instead describes options as bundles of attributes and asks respondents to choose among them repeatedly, so it recovers an implicit price for each attribute rather than one global value. Hanley, Wright and Adamowicz promoted choice experiments precisely because this attribute-based structure values multiple marginal changes at once and can sidestep some criticisms of contingent valuation, though both are stated-preference methods sharing vulnerability to hypothetical bias.

The cost or price attribute is what makes monetary valuation possible. Because utility is modeled as linear in the attributes and in cost, the marginal willingness to pay for any environmental attribute is the negative ratio of that attribute's coefficient to the cost coefficient. Without a cost attribute there is no common monetary metric to convert estimated preferences into willingness to pay. The cost coefficient should come out negative, confirming respondents dislike paying more, and because every welfare estimate divides by it, its plausibility and precision are critical.

Hypothetical bias is the tendency for stated willingness to pay in a hypothetical survey to exceed what respondents would actually pay in a real transaction. Because choices in a choice experiment carry no real financial consequence, this bias can inflate value estimates. Common mitigations include cheap-talk scripts that warn respondents about the tendency, consequentiality reminders that stress the survey may influence policy, certainty follow-up questions, and, where feasible, calibration against revealed-preference or actual-payment benchmarks. Careful, realistic scenario design is the first line of defense.