Demand shifting across flights and airports in a spatial competition model

This paper investigates the nature of day-to-day competition between flights using a unique panel data set on prices and inventories. We use instrumental variables methods and several spatial autoregressive models (SAR) to estimate price reaction functions. The primary source of product differentiation is departure time. After controlling for flight-specific characteristics and various sources of price dispersion, we find important evidence of demand shifting between competing flights. Most of the shift is being captured by flights scheduled to depart within a 3-hour window. We find no evidence of demand shifting between airports.