PRICE ELASTICITY FOR SMARTPHONES IN THE UNITED STATES: RESULTS FROM THREE METHODOLOGICAL APPROACHES
The Berry, Levinsohn and Pakes (BLP) model is a widely accepted model when using aggregate data to examine an industry selling differentiated products. This dissertation contributes to existing literature of the New Empirical Industrial Organization (NEIO), focusing on the BLP models and provides insights into consumer demand for smartphones by estimating two BLP models which explore own-price and cross-price elasticity of demand. The first paper summarizes a BLP model estimated using aggregate data and compares the result to a consumer level choice model. Both models are estimated with data from the smartphone phone industry. Customer level survey data was aggregated to mimic data used when estimating a BLP model allowing for a general review for consistency between elasticity estimates. Final estimates show the BLP estimation methodology generally has higher own-price and cross-price elasticities compared with a mixed logit estimated using consumer level choice data. The second paper included a near perfect branded complement, network connectivity, by calculating shares using smartphone, network carrier combinations and adding the carrier price. Elasticities for each smartphone dropped relative to the original smartphone only BLP. As with many BLP models, the smartphone only and the branded complement models both suffered from poor price instrumental variables. Including the branded complement provided improved insights into the smartphone industry and allowed for more detailed cross-price elasticity calculations. BLP, in this case, can provide estimates of own-price and cross-price elasticities for consumer noticed branded complements. This has obvious practical use as many industry participants may not have access to consumer level data of an important branded complement.