Essays in Consumer Behavior
My doctoral dissertation consists of three essays on consumer behavior. The first chapter studies demand for experience goods. Consumers behave very differently when they do not have perfect information about all brands available on a shelf. This paper extends the benchmark discrete choice model of consumer demand to capture two distinct features of experience-goods markets: prior brand experience and shopping frequency. Although the current literature incorporates habit formation in consumer demand models, it has not considered a more fundamental question: how the first experience with a brand affects the consumer's choice. The model is estimated using data on purchases of ready-to-drink orange juice, which comes from a new consumer-level panel provided by a large supermarket chain in Brazil. The results show that for this product prior experience of a brand is more important for a consumer's choice than price. Furthermore, own- and cross-price elasticities change significantly when experience and shopping frequency are taken into account. The findings of this chapter have implications for both firms' strategies and for antitrust analysis related to experience-goods markets. The second chapter explores how umbrella branding can significantly decrease consumer's first-time experience cost. Multiproduct firms often market their products under the same brand name. When a firm launches a new product with the same brand name, consumers can pool their prior experience with the brand to infer a quality for the product. This strategy can be particularly useful when a firm decides to enter a market of experience goods, in which consumers face a cost for trying a new product. The main objective of this chapter is to study the process by which consumers' brand choices and first-time purchases for ready-to-drink orange juices are affected by their experience with the same brand in another category. The results are consistent with signaling theories of umbrella branding as they indicate that consumers' experience cost with a product decreases with experience of other products of the same brand. The third chapter is about a household's choice of retail formats. Thirty percent of households' food expenditure in the United States comes from clubstores, mass merchandisers, supercenters, drugstores and convenience stores. However, earlier work focused on consumers' shopping behavior mostly in grocery stores and has not examined consumers choice across different types of retail outlets. To address this gap a multinomial logit model is estimated on household-level scanner data for the United States to study how households' characteristics are related to their choice of retail outlets. The results show that income, household size and ethnicity significantly affect these choices. These findings are important for policies that target certain consumer groups.