Price Search and Consumption Inequality : Robust, Credible, and Valid Inference

This paper investigates whether heterogeneity in price search mitigates consumption inequality. I propose a model where the consumer can pay weakly lower prices by shopping more frequently. The extent by which shopping intensity decreases prices paid depends on the consumer shopping technology. I show that the model is generically not identified and develop a revealed preference methodology to set identify the shopping technology in a computationally tractable fashion. My approach allows for nonparametric concave preferences, rich heterogeneity, and measurement error in prices. Using a novel statistical framework, I show that data on shopping expenditures from the Nielsen Homescan Dataset are consistent with the model. I find that low-income consumers have slightly higher shopping intensities and a slightly better average shopping technology compared to high-income consumers