LAUSR

This paper examines the impact of health information in different media outlets on bottled water consumption through consumer learning. We develop a random coefficient discrete choice model with Bayesian learning process to capture consumers’ learning of health information and changes in their beverage choices over time. Consumers are assumed to have initial prior beliefs about the health effect of different beverages and to update their beliefs using health information received from different media types. Empirical results show that consumers’ perceived quality of bottled water kept increasing during our sample period, and this learning process accounted for 24.44% of the industry’s revenue, which is about 4.8 billion dollars per year. Comparing the effectiveness of different media outlets, we find that the sales impact of traditional media (TV and radio) is greater than online sources. Our findings highlight the contribution of health information to the bottled water industry and provide policy makers with a new direction to reduce high-calorie food consumption and improve public health.