Does Purchase Without Search Explain Counter-Cyclic Pricing?

Basic economic theory tells us to expect that an increase in demand should lead to an increase in price. However, previous studies have found the opposite trend in the prices of seasonal goods, such as canned soup. I propose an explanation of this phenomenon: consumers are more likely to purchase without search in low demand periods, reducing the gains of temporary price reductions, and decreasing estimated price sensitivity. Purchase without search is consistent with consumers using shopping lists to make their purchase decisions before observing prices. I test this explanation using a novel dynamic, structural inventory model where consumers make decisions on whether to search, which reveals price promotions, and which products to purchase given their search decision. Estimating this model using previous methods is a computational challenge because of the expansion of the state space required to model seasonal preferences. To overcome this challenge, I develop a cyclic-successive approximation algorithm, which removes the computational burden of adding cyclic variables to the state space of a dynamic model. I find that consumers purchase without search 61% of the time in winter, and 82% of the time in summer. This causes price elasticities that are twice as large in winter as they are in the summer. I find that the dominant cause of seasonal search is seasonal consumption preferences, rather than seasonal price variation. Basic economic theory tells us to expect that an increase in demand should lead to an increase in price. However, previous studies have found the opposite trend in the prices of seasonal goods, such as canned soup, tuna, and ice cream, where prices have been observed to decrease in periods of high demand. This phenomenon has been termed " counter-cyclic pricing ". In this paper, I propose that counter-cyclic pricing is partially a reaction to seasonal changes in consumers' propensity to make purchase decisions without searching the category. For example, shoppers who use a shopping list will choose both the variety and quantity of a good to purchase before observing prices at the store. An increase in purchases without search will increase average prices because non-searching consumers will not react to discounts, and so holding a sale will simply decrease the price that these consumers pay. I posit that consumers make a higher proportion of their purchases without search in low demand periods for two reasons. First, the average purchase size is smaller, and so the expected …