Choice experiments for ecosystems and wild birds: A review

Abstract

In this detailed review of about two dozen published choice experiments concerning ecosystems and wild birds, I highlight differences across these studies that affect their suitability for benefits-function transfer. Survey-based choice experiments are often used to reveal the types of trade-offs that people are willing to make among the attributes that describe some different alternatives they may face. Studies can be designed merely to illuminate the tradeoffs willingly made by the sampled respondents in their particular context. However, to maximize the usefulness of an expensive survey-based choice experiment, it is important to design the study in a way that maximizes its value for future exercises in ‘benefits-function transfer.’ Representativeness of the “study sample” is a crucial issue, as may be methods for systematic sample selection correction. Heterogeneous preferences should be accommodated, and not solely through random-parameters utility-function specifications. Estimated marginal utilities for specific attributes should be allowed to vary systematically with observable characteristics of respondents and/or their neighborhoods, so that benefits-function transfer can accommodate systematic differences in preferences to be expected when the mix of population characteristics differs between the study context and other contexts. Across the set of choice-experiment studies reviewed here, I inventory differences in their characterizations of the ecosystems services related to wild birds, the geographic area of focus, the species of birds studied, the survey design (mode, dates, sample sizes), the choice set structure and elicitation method, the formal choice-set design criteria, sample representativeness, estimation methods, use of continuous versus categorical attributes, interactions between attributes, approach to accommodating heterogeneous preferences, any quantified values for wild birds, caveats (both acknowledged by the authors and not), and recommendations for future research, both offered by the study authors and gleaned from this overall review. (This paper is currently being revised and updated. The new version will be co-authored with Sonja Kolstoe.)

Publication
In draft form
Date