Can airfares tell? An alternative empirical strategy for airport congestion internalization

In this paper, we propose an alternative empirical strategy to test whether airlines internalize airport congestion corresponding to their share of traffic at the airport. In particular, we construct a hypothesis from theoretical derivation that if airlines do internalize airport congestion the airfare prices would be positively correlated with the interactive term of the airline's passenger number at the origin airport and the congestion delay level of this airport. We test this hypothesis with the Airline Origin and Destination Survey (DB1B) database and the Airline On-Time Performance database published by the US Bureau of Transportation Statistics (BTS). We find that the hypothesis is supported by the data, suggesting that airlines’ behaviour is in line with the internalization theory. We further implement subsample analysis and find that airline type also plays a role in this matter. In particular, while full service carriers internalize airport congestion, low cost carriers do not.