Landing spot allocation and market incentives
When your airplane takes off and lands, what do you think about? Do you think about the ground below getting larger and larger? The lights of the cities and the airport growing ever brighter as the plane makes its final approach? Or the engines whine as the increasingly thick air speeds past the blurred turbine blades?
You probably don’t think about how much it costs to land or take off at an airport. Not just the costs of the fuel or maintenance or the pilot’s salary, but the permission to take off and land an airport. Most airports do not have that many runways. Even John F. Kennedy Airport in New York City, one of the largest in the world, has only four runways for planes to take off and land on. Considering that JFK handles over 422,000 planes a year (an average of well over a thousand a day), and has to contend with weather conditions, different priority flights, and the scheduling constraints of almost 100 different airlines from 50 different countries, scheduling can get very difficult.
Instead of having a strictly centralized assignment system, which would get very complicated very quickly, JFK (as well as most other airports of its class) use landing slots, which are the right to take off or land at corresponding airports at certain times of day. Since these landing slots are integral to an airline’s ability to operate, they are highly prized and are valued in the millions of dollars. The highest price ever paid for one was a $75 million landing slot for an early morning arrival.
Landing slots are able to be traded, which naturally sets up an market for them. For example, United Airlines and Delta recently swapped landing slots at different airports. These slots can be used to consolidate an airline’s route at a certain airport (letting them run more flights out of the same airport) or let an airline expand into another airport or region.
In this paper, the authors discuss the incentives in situations such as inclement weather when it is advantageous to reschedule delayed flights into earlier landing slots. However, this rescheduling is dependent on airplanes self-reporting flight data to a centralized system, where there might be incentives not to cancel. Additionally, this is not a free market where airlines can use their landing slots without oversight. For example, landing slots are usually forfeited if airline usage of that landing slot drops below 80%, leading airlines to run “ghost flights” with few or no passengers on them. Also, in conditions of inclement weather, air traffic control can force airlines to swap slots temporarily to address demand changes and delays or cancellations.
The authors found that under current regulations, it is often advantageous for airlines to not report delayed flights or cancellations. This creates inefficiencies in the market in terms of later flights being delayed or cancelled in order for the original airline to benefit. For example, an airline may be able to “sacrifice” one of their landing slots to gain valuable landing slots later. The authors considered three criteria – lights’ feasible arrival times, cancellations, and relative delay costs – in order to design better incentives for airlines. Additionally, the authors found that distinguishing between reporting delays and reporting cancellations – two things that were combined in previous models – allows the authors to prove that their rules are better and less prone to abuse by airlines.
While the actual mathematics used in the paper is beyond the scope of this course, it is similar to a combination of matching markets, auctions, and Nash equilibria. One important difference from the material we learned in class is that the Nash equilibria is affected not only by multiple players (the airlines) but also by an external force (the weather). Additionally, instead of a free matching market and auction system, the outcome of the Nash equilibria can affect the market (delays and cancellations of landing slots can change prices and even allocations of landing slots subject to the airport’s discretion). Thus, the model that the authors examined combines many of the concepts we learned in class into a larger model dependent on much more variables.
The author’s findings may not seem very important. In the larger context of government regulations and flights out of airports across the world, this has far-reaching implications. Flight delays cost the airline industry $8 billion a year in maintenance, crew salaries, and more, while passengers lose an estimated $15 billion a year in cancellation fees, missed flights, and economic loss due to delayed and cancelled flights. If this new model gives the airlines a dominant strategy in the context of the landing slot market by being forthright and prompt about delays and cancellations, everyone stands to gain a lot of benefit.