This paper studies empirically three different, but not mutually exclusive, hypotheses on entry deterrence that have been proposed to explain the observed patterns of airline entry-exit decisions in city-pair markets. The first hypothesis establishes that the dominant status of an airline in an airport can help to deter entry. The second hypothesis is based on product proliferation: an incumbent airline may want to provide more differentiated products (schedules) in order to lower the potential profitability of the competitors. The third hypothesis establishes that a hub-and-spoke network can be an effective mechanism to deter entry in spoke markets. I propose an approach to empirically distinguish the contribution of these hypotheses to explain airlines' entry and exit decision in city-pair markets. Our results are based on new measures that result from the merger of two databases from the US Department of Transportation: DB1B and T100 databases. I estimate a structural game of entry with incomplete information, and use the estimated model to separate the contributions of these three hypotheses. I find that the strategic behavior of incumbents benefit them by reducing the entry probability of potential entrants. I found empirical evidence for all of the three hypotheses. |