Abstract
During the COVID-19 pandemic, passenger demand for air transportation declined drastically. In the Unites States (U.S.), the Coronavirus Aid, Relief, and Economic Security (CARES) Act provided financial assistance. In return, commercial passenger airlines were given minimum service obligations, which allowed airlines to remove markets (flights between origin and destination airport pairs) from their networks as long as they continued operating in all cities that they serviced pre-pandemic. A binary logit methodology is used to model airline-market level decisions to continue operating in a market or to exit it. Two time periods are modeled: during normal operating conditions (before the pandemic) and after a major shock event (after the beginning of the pandemic). Results show that after the pandemic, 8.4 times more airline markets are exited as compared to before. Interestingly, the probability of exit is found to vary widely across markets, airports, and airlines. Some market characteristics have a high probability of exit both before and after the pandemic, including low passenger revenue per available seat mile, low flight frequencies, and flights to/from multi-airport cities. In contrast, other market characteristics impact airlines' market exit decisions in only one time period rather than both. For example, during normal operating conditions, airport size does not impact market exit. However, after the pandemic, the probability of exit is 1.8 to 2.2 times higher for the larger hub airports as compared to the smallest airports (non-hubs), a result that is explained within the context of the CARES Act minimum service obligations.