REFERENDUM CONTINGENT VALUATION, ANCHORING, AND WILLINGNESS-TO-PAY FOR PUBLIC-GOODS

This study reports on experiments that examine anchoring in single ref erendum questions in contingent valuation surveys on willingness to pa y for public goods, and on objective estimation. Strong anchoring effe cts are found that lead to systematically higher estimated mean respon ses from Yes/No referendum responses than from open-ended responses. T his response pattern is similar for contingent valuation questions and for objective estimation questions. The paper concludes that psychome tric anchoring effects, rather than incentive effects, are the likely cause of results commonly found in contingent valuation studies, and t hat the currently popular single referendum elicitation format is high ly vulnerable to anchoring. (C) 1998 Elsevier Science B.V.