A model of the anchoring effect in dichotomous choice valuation with follow-up.

This paper focuses on modelling and estimating the starting point bias in closed-ended follow-up questions, where several bids are presented successively, depending on previous answers. Although the contingent valuation literature took off in the last decade, there is only one study modelling the starting point bias. We propose a new modelling of this anchoring effect based on the assumption the first proposed bid has a direct influence on the individual's willingness-to-pay, i.e respondents modify their willingness-to-pay when presented with the first bid just before they answer the first question. Monte Carlo results support the specification of our model. An application is provided based on data from a contingent valuation study conducted concerning air quality in Strasbourg.