Reminder Nudge, Attribute Nonattendance, and Willingness to Pay in a Discrete Choice Experiment

Abstract

Attribute non-attendance (ANA) is one of the choice simplification strategies respondents employ in choosing among alternatives in stated preference elicitation methods. Studies have shown the importance of accounting for ANA in estimating demand functions for non-marketable quality differentiated goods and services. This study addresses the question whether reminding respondents on the need to attend to all attributes in discrete choice experiments (DCE) affects ANA and improves quality of model fitting. We compare ANA patterns and willingness to pay (WTP) values ‘before’ and ‘after’ a reminder for attention to all attributes. We report on a study using DCE data elicited from 960 respondents generating 11520 observations. Bayesian D efficiency criterion was used to design the experiments and attribute nonattendance was inferred using constrained latent class econometric models. We employed mixed logit model in WTP space to estimate WTP values for the different livestock market services in Ethiopia. We find that a nudge in the form of reminding full attention improves data fitting quality, reduces ANA and distributes WTP more evenly across the different services considered. Our results imply that researchers studying behaviors of rural communities in developing countries might be able to estimate the implicit prices of attributes more precisely if they employ reminders while conducting DCE.