Reducing Attribute Non-Attendance Risk in Choice Experiment: An application of Design Attribute Relative Importance Index for Waste Management Service Provision in Segmented Markets

Abstract

Attribute non-attendance is an estimation problem that biases welfare estimates in the economic valuation of environmental goods when choice experiment (CE) is used. The potential for observing the problem is increased when segmented markets with dissimilar preferences exist for environmental goods but not captured in the experimental design for CE. The current procedure for avoiding this problem relies on qualitative techniques, which limits applications to large samples and thus prone to bias. To reduce such bias, this study used an alternative technique based on the design-attribute relative importance index (DARII) to determine respondents’ prioritization of attributes within and across markets. Our results in solid waste management services demonstrate that all selected attributes were highly ranked across markets. While such finding conforms to expectations based on the interview, we found a statistically significant difference in attributes prioritization across markets. We thus conclude that assuming equal priority for attributes in different market segments may be improper but possibly tangled when the qualitative method is used. As such, the study recommends CE studies to use DARII before the experimental design stage.