Benefit estimates for landscape improvements: sequential Bayesian design and respondents' rationality in a choice experiment

Abstract

A multi-attribute, stated-preference approach is used to value low and high impact actions on four major landscape components addressed by the Rural Environment Protection Scheme in Ireland. Several methodological issues are addressed: the use of prior beliefs on the relative magnitudes of parameters, standardized description of different levels of landscape improvements via image manipulation software, adoption of efficiency-increasing sequential experimental design, and sensitivity of benefit estimates to inclusion of responses from ‘‘irrational’’ respondents. Results suggest that Bayesian design updating delivers significant efficiency gains without loss in respondent efficiency, and estimates are upward-biased when irrational respondents are included. (JEL Q24, Q51)