Information Cost and Consumer Choices of Healthy Foods

This paper explores how lowering consumer search costs based on labeling formats affects the probability of choosing healthy foods. We propose a theoretical model of the links between information costs and consumer choices of healthy foods and empirically test the ensuing propositions with scanner data from ready-to-eat breakfast cereals (RTEC). Based on a natural experiment with changes in labeling for otherwise identical products, we apply an alternative-specific conditional logit model to approximately 1.13 million observations derived from Nielsen Homescan weekly purchases data matched to advertising and product nutritional and labeling data. Empirical results confirm the theoretical propositions that more convenient labeling significantly increases the probability of a healthier product being chosen. We also find that consumers with a higher volume of RTEC purchases are more sensitive to information cost reductions. Overall, the use of front-of-package labels is effective in inducing consumers to select healthier RTEC products.