The Economics of Sales

Abstract

Temporary price reductions or “sales” account for a small fraction of price quotes and a large fraction of items sold. Nevertheless, several recent macroeconomic studies have concluded that sales are not an important source of price flexibility. Using a large set of scanner data covering 31 product categories in 50 markets, I investigate the aggregate features of sales over time and find that sales have a large impact on short to medium term fluctuations in average unit price. These fluctuations are not captured by the CPI and appear to be relevant to macroeconomists. I provide evidence that sales are a relevant margin used by firms to adjust to fluctuations in demand.

I argue that sales are best thought of as price dispersion, as they are in the Industrial Organization literature. I propose a relatively simple model of price dispersion with a clean interpretation of when a price is a sale or not. The primary effect of sales is to discriminate between groups of shoppers characterized by their ability to become informed about the realization of prices. If the share of informed shoppers is correlated with the business cycle, then sales may be an important source of price flexibility.

The final chapter is joint work with Mattew Jaremski. In this paper, we present a model of repeated price competition to illustrate how entry causes incumbents to alternate between high and low prices. Using a six year panel of weekly observations from a grocery chain, we find that individual stores employ more sales as the distance to Wal-Mart falls. Moreover, the increase in the frequency of sales was concentrated on the most popular products.