A Distance Metric Approach to Attributes-based Consumer Choices: Applications to (i) Brand Choice in Ready-to-Eat Cereals and (ii) Store Choice in a Single Geographic Market

Abstract

To investigate two consumer-choice applications, this research uses the Pinkse, Slade, and Brett distance-metric (DM) method, which frames cross-product substitution as a function of relative distance in physical or virtual attribute space. The first application investigates the demand for health-related product attributes in ready-to-eat breakfast cereals, and finds that hypothetical policies that convert cereals to whole grains increase total welfare but policies, that decrease sugar content decrease welfare. The second application investigates food store choice in a major U.S. market and finds that distance does play an important role: distance affects both price-increase induced store switching and the total expenditure behaviors.