This study provides empirical evidence documenting how COVID-19 affects intertemporal pricing and price dispersion in the U.S. domestic airline market. Studying a unique panel of 43 million fares collected before and after the outbreak of the pandemic, we find that airlines discounted fares by an average of 57% in the first five months of the pandemic relative to the five months that immediately preceded the pandemic. We also find that flight-level prices increased at a lower rate, particularly in the last week to departure. As a consequence, flight-level price dispersion decreased. These findings are consistent with the theoretical predictions arising from models of stochastic peak-load pricing and intertemporal price discrimination.

When faced with capacity constraints, firms may moderate demand by increasing prices when demand is known to be high ex-ante (i.e., systematic peak-load pricing). In this article, we examine the extent and duration of systematic peak-load pricing in the days surrounding public holidays in the U.S. airline industry. Applying two-stage least squares techniques to a unique panel of over 18 million fares, we estimate travel premiums ranging from 4.3% to 83.1% in the days surrounding national holidays and from 2.7% to 34.7% in the days surrounding federal holidays. We also find that the duration of the peak-travel period is longer for national holidays and shorter for federal holidays. Examining heterogeneity in holiday peak-load pricing, we find some evidence that travel premiums during national holidays are larger on longer-distance routes, on routes to or from slot-controlled airports, on routes to leisure destinations, and on ultra-low-cost carriers.

Discounts during Thanksgiving and Christmas are common in a variety of retail markets. In this article, we examine whether holiday discounts extend to the airline industry. In contrast to many retail markets where purchased goods are meant for immediate consumption (e.g., groceries), goods in airline markets are often consumed in the future due to advance purchases. Exploiting a unique panel of almost 22 million fares, we find that fares purchased on a holiday for flights in the 60-day period following the holiday are 1.9% cheaper, supporting the conjecture that airlines price discriminate when demand is lower than average or when the mix of purchasing passengers makes demand more elastic. These holiday discounts also do not vary with the level of competition, indicating that market structure has no impact on the magnitude of the holiday purchase discount.

We develop a theory of optimal aircraft size, where the cost of the flight crew is the primary factor driving the use of larger aircraft, while passenger utility is primary factor driving the use of smaller aircraft.After fitting our model to U.S. data, we perform a counterfactual experiment where the minimum crew size requirement is relaxed from two pilots to one, a policy currently being discussed by aviation experts.Implications are derived for the number of aircraft demanded and its size distribution, demand for pilots, passenger traffic, flight frequency, and where new nonstop service may be introduced.

We offer a comprehensive empirical study on hidden-city ticketing (HCT), a pricing phenomenon in the airline industry that occurs when the fare for a nonstop trip from A to B is more expensive than a connecting trip from A to B and B to C. Exploiting a unique panel of over 473 thousand fares for flights departing between October 1st, 2019 and December 31st, 2019, we find that HCT depends on route competition (both on A–B and A–C routes), largely occurs in the last week to departure, is less likely when airport C is a hub, and primarily occurs on carriers that operate large hub-and-spoke networks (e.g., American, Delta, and United).

Discounts during Thanksgiving and Christmas are common in a variety of retail markets. Although classical economic theory predicts that prices should increase when aggregate demand is high, one possibility is that consumers are more price elastic during seasonal demand peaks. In this article, we examine holiday pricing in the airline industry. Exploiting a unique panel of almost 22 million fares, we find that fares purchased on a holiday for flights in the sixty-day period following the holiday are 1.8% cheaper, supporting the conjecture that airlines price discriminate when the mix of purchasing passengers makes demand more elastic. An increase in competition is associated with larger holiday purchase discounts, with the largest discounts reserved for flights within one-week of departure.

Maximum employee work-hour restrictions are implemented to reduce accidents. However, because they decrease the stock of work-hours available to employers, they may also have detrimental effects. A quasi-experiment suggests that pilot hours-of-service reforms, which decreased the number of flights and hours a pilot may work, reduced consumer choice and increased fares in the airline industry. We find that regional and low-cost carriers reduced scheduled flight frequency, while less constrained legacy carriers (and potentially their wholly owned subsidiaries) were unaffected. Further, we find evidence that market concentration increased on many routes, implying that fare increases may be due to a decrease in competition. These findings illustrate a situation where a policy implemented to correct one market failure, airlines not internalizing the full social costs of accidents by allowing dangerously fatigued pilots to fly, exacerbated another market failure by decreasing competition.

Author(s): Luttmann, Alexander Phillip | Advisor(s): Brueckner, Jan K | Abstract: This dissertation contains three chapters and focuses on price discrimination and airport congestion in the U.S. airline industry.The first chapter explores possible determinants that may affect an airline's decision to charge passengers different roundtrip fares depending on trip origin, a case of directional price discrimination. Such fare differences cannot be the result of differences in cost, as the cost of flying a roundtrip passenger does not significantly differ depending on direction. It is argued that directional fare differences result from airlines recognizing that passenger price elasticities differ between route endpoints. A price discriminating airline will then charge a higher roundtrip fare at the endpoint where the passenger price elasticity of demand is comparatively lower. Evidence is found suggesting that airlines do use differences in income to price discriminate when setting roundtrip fares. Fares are found to be $0.18-$0.43 higher on average for each $1000 difference in average per capita income between origin and destination metro areas. This finding is sensible assuming that higher incomes reduce the price elasticity of demand for air travel, with richer passengers being less sensitive to the cost of travel. The second chapter investigates the trade-off between providing convenient flight connections for passengers and reducing airport congestion. From the passenger perspective, layovers are detrimental since the addition to total travel time relative to a nonstop itinerary is a cost incurred by the passenger. An airline is able to reduce a passenger's layover time by narrowing the gap between flights at the connecting airport. However, narrowing this flight gap has the adverse effect of increasing airport congestion. Taking these perspectives into account, it is clear that layover time influences a prospective passenger's purchasing decision and an airline's flight scheduling decision. Using published fare and itinerary data from Google Flights, this chapter provides insight into both decisions by providing empirical estimates on the value of layover time in the U.S. airline industry. Passengers are found to be compensated with a fare that is $42.74-$47.60 cheaper per hour of layover time.The last chapter evaluates the effectiveness of slot controls (restrictions on the number of departing and arriving flights) as a congestion management policy. Utilizing the introduction of slot controls at John F. Kennedy (JFK) and Newark (EWR) airports in 2008 as a quasi-experiment, no evidence is found of a reduction in flight delays at both airports. In the months after slot controls were introduced, the average arrival delay at EWR actually increased by 7 minutes. Further, the length of Delta's departure banks (high-volume periods of departing flights) decreased by about 2 minutes at JFK while the scheduled time of EWR flights decreased by 1.5-2.2 minutes. These findings are consistent with Ater (2012), who suggested that policies aimed at reducing congestion at highly concentrated airports will only have a limited impact because dominant airlines already internalize congestion. The results highlight the need for policymakers to carefully consider how the allocation of airport slots will impact flight scheduling decisions when implementing similar policies in the future.

American, Delta, and United have organized their operations into extensive hub-and-spoke networks that typically require passengers originating and concluding travel in non-hub cities to board a connecting flight at a hub en route to the final destination. From the passenger perspective, layovers are detrimental since the addition to total travel time relative to a nonstop itinerary is a cost incurred by the passenger. An airline is able to reduce a passenger's layover time by narrowing the gap between flights at the connecting airport. However, narrowing this flight gap has the adverse effect of increasing airport congestion. Taking these perspectives into account, it is clear that layover time influences a prospective passenger's purchasing decision and an airline's flight scheduling decision. Using published fare and itinerary data from Google Flights, this paper provides insight into both decisions by providing empirical estimates on the value of layover time in the U.S. airline industry. This paper finds that passengers are compensated with a fare that is $42.74-$47.60 cheaper per hour of layover time. Of the three dominant legacy carriers, United passengers are found to be compensated at an even higher rate of $61.89 per hour.

Slot controls (restrictions on the number of departing and arriving flights) are implemented at over 200 capacity constrained airports worldwide to mitigate persistent congestion and delays. However, evidence is limited regarding the effectiveness of these policies. This paper uses a quasi-experiment to evaluate whether the introduction of slot controls at John F. Kennedy (JFK) and Newark (EWR) in 2008 reduced the incidence and severity of flight delays. No evidence is found of a reduction in delays at both airports. In the months after slot controls were introduced, the average arrival delay at EWR actually increased by 7 minutes. Further, the length of Delta's departure banks (high-volume periods of departing flights) decreased by about 2 minutes at JFK while the scheduled time of EWR flights decreased by 1.52-2.24 minutes. These findings are consistent with Ater (2012), who suggested that policies aimed at reducing congestion at highly concentrated airports will only have a limited impact because dominant airlines already internalize congestion. The results highlight the need for policymakers to carefully consider how the allocation of airport slots will impact flight scheduling decisions when implementing similar policies in the future.

This paper explores possible determinants that may affect an airline's decision to charge passengers different roundtrip fares depending on trip origin, a case of directional price discrimination. Such fare differences cannot be the result of differences in cost, as the cost of flying a roundtrip passenger does not significantly differ depending on direction. It is argued that directional fare differences result from airlines recognizing that passenger price elasticities differ between route endpoints. A price discriminating airline will then charge a higher roundtrip fare at the endpoint where the passenger price elasticity of demand is comparatively lower. Evidence is found suggesting that airlines do use differences in income to price discriminate when setting roundtrip fares. Fares are found to be $0.18-$0.43 higher on average for each $1,000 difference in average per capita income between origin and destination metro areas. This finding is sensible assuming that higher incomes reduce the price elasticity of demand for air travel, with richer passengers being less sensitive to the cost of travel.