An analysis of dynamic price discrimination in airlines

Prices for the same flight change substantially depending on the time of purchase. This paper uses a unique dataset with round-the-clock posted fares to document significant within-day price variation. Labeling time-variation as discriminatory is difficult because the cost of an unsold airline seat changes with inventory, days before departure and aggregate demand expectations. After controlling for these factors and aggregating hourly fares to have a framework with two consumer types, we are able to identify a component that is largely consistent with dynamic price discrimination. We find higher prices during office hours (when business travelers are likely to buy) and lower prices in the evening (when leisure travelers are more likely to purchase). As the proportion of business travelers increases closer to departure, both price dispersion and price discrimination become larger. We provide an alternative explanation for the observed within-day price differentials which is related to Edgeworth price cycles.