« WITTMAN AND SWELBAR: DESPITE SMALLER INDUSTRY, NETWORK CONNECTIVITY REMAINS »

This post contains excerpts from Modeling Changes in Connectivity at U.S. Airports [it is available online here: http://hdl.handle.net/1721.1/79091] the second paper in MIT’s Small Community Air Service White Paper Series. The aim of the paper series is to examine and analyze the past, current, and anticipated future trends of small community air service in the United States. The authors of this paper series hope that these reports will serve to inform the policy debate with relevant and accurate statistical analysis, such that those responsible for deciding the future of small community air service will do so armed with factual basis for their actions.

The authors of the MIT Small Community Air Service White Paper series are members of the Massachusetts Institute of Technology’s International Center for Air Transportation, one of the nation’s premier centers for aviation, airline, and airport research. Financial support for study authors has been provided in part by the MIT Airline Industry Consortium, an interdisciplinary group of airlines, airport councils, manufacturers, suppliers, policy makers, and advocacy groups dedicated to improving the state of the practice of air transportation research in the United States. However, any views or analyses presented in this and all future reports are the sole opinions of the authors and do not reflect the positions of MIT Airline Industry Consortium members or MIT.

The first report in the series, Trends and Market Forces Shaping Small Community Air Service in the United States, reviewed changes in capacity at U.S. airports from 2007-2012. It is available online here: http://dspace.mit.edu/handle/1721.1/78844.

Acknowledgements

The authors wish to thank Peter Belobaba and the members of the MIT Airline Industry Consortium for their helpful comments and suggestions during the completion of this study.

Executive Summary

As described in the first paper in the MIT Small Community White Paper Series (Wittman and Swelbar 2013), the U.S. air transportation system has undergone a series of changes in response to the financial crisis of 2007-2009, high fuel prices, and a new wave of profitability-focused “capacity discipline” airline management strategies. More than 14.3% of yearly scheduled domestic flights were cut from the U.S. air transportation network from 2007-2012, mostly due to actions of the network carriers. Smaller airports were disproportionally affected by the cuts in service, losing 21.3% of their scheduled domestic flights as compared to an 8.8% decline at the 29 largest U.S. airports.

However, simply examining gains or losses in flight volumes does not provide a complete picture of the strength of commercial air service at an airport. For instance, many smaller airports lost service from network carriers from 2007-2012 but saw new service from ultra-low cost carriers (ULCCs) like Allegiant Air or Spirit Airlines. These ULCCs typically serve vacation destinations and offer limited connecting service to other U.S. airports or the global air transportation network. The small airports that lost network carrier service only to receive replacement service from ULCCs may not have seen significant decreases in flight volumes, but their connectivity was likely adversely affected.

As the pace of globalization has increased in recent years, commercial air service that provides connections to the global air transportation network has become increasingly important for economic, social, and demographic reasons. While air connectivity is important for communities of all sizes, research has suggested that small communities can obtain significant economic benefits from well-connected commercial air service. However, recent work has shown that small- and mid-sized airports have been disproportionally affected by cuts in commercial air service in the U.S. over the past six years.

An airport’s connectivity to the global air transportation network is challenging to measure because it cannot be observed directly through published statistics. There is currently no industry-standard metric to assess an airport’s connection to the global air transportation system. This creates challenges for airport managers and policy-makers in interpreting the effects of gains or losses in flights or seats on an airport’s connectivity. On a regional level, it is also valuable to analyze which airports have seen increases or decreases in connectivity over a given period. Previous attempts at defining connectivity metrics have often not taken into account the quality of connecting destinations, been too complex for non-technical audiences to understand and adopt, and have often included no analysis of connectivity at smaller airports. The discussion paper introduces a new, relatively easy-to-compute metric that can be used to assess these changes in connectivity to global air transportation service at U.S. airports.

The Airport Connectivity Quality Index (ACQI) introduced in the paper computes airport connectivity as a function of the frequency of available scheduled flights, the quantity and quality of destinations served, and the quantity and quality of connecting destinations. Unlike other connectivity models, the ACQI model considers connecting opportunities from a given airport as well as the quality of destinations served, such that an additional flight to a large city or a major connecting hub is more valuable than an additional flight to a smaller community with limited connecting options. The analysis also pays particular attention to connectivity at smaller airports, which have been largely ignored in previous work. The report computes ACQI connectivity scores for 462 U.S. airports for each year from 2007-2012; the ACQI scores for these airports are available in several appendices.

Similar to the capacity reductions in flights and seats discussed in the first white paper, medium-hub and small-hub airports suffered the greatest losses in connectivity over the last six years. ACQI connectivity scores at medium-hub airports fell by 15.6% on average between 2007 and 2012, compared to a 11.0% decline in connectivity at small-hub airports and a 3.9% decline at large-hub airports. The decline in connectivity can be attributed to airlines cutting capacity and destinations as a result of challenging macroeconomic events and more restrictive capacity management strategies.

It should be noted, however, that percentage changes in connectivity at most airports were less than percentage changes in flights or available seats. This suggests that some of the service cuts as a result of recent “capacity discipline” strategy did not directly harm connectivity, but instead removed redundant flying to secondary hubs. At many small airports, removing a flight to a secondary hub would not result in a substantial loss in connectivity as long as flights to other, larger connecting hubs remain. However, connectivity at the secondary hubs themselves (which are often medium-hub airports) was adversely affected over the last six years. Future changes in connectivity in the United States will largely depend on whether the capacity discipline equilibrium remains in place or if a new capacity management paradigm evolves in response to ongoing changes to the structure of the U.S. airline industry.

Computing ACQI Scores for U.S. Airports

Data sources

ACQI scores for each of 462 U.S. airports from 2007-2012 were computed using schedule data from the Diio Mi Market Intelligence Portal. The Diio Mi data is sourced from Innovata SRS, which provides up-to-date schedule data for 99% of airlines worldwide.

Data was collected for all airlines, domestic and international, with scheduled flights from the United States. Code-share connecting destinations were included by grouping appropriate airlines into each of the three major alliances: Star Alliance, Skyteam, and Oneworld.

ACQI Score Overview

In 2012, large-hub airports were roughly three times more connected than medium-hubs, six times more connected than small-hubs, and about fifteen times more connected than non-hubs. Each of the top 25 most connected airports in the U.S. in 2012 was a large hub.

The average ACQI score fell for each airport hub type during the study period, suggesting that airport connectivity as a whole in the United States has declined during the events of 2007-2012. However, just as capacity discipline strategies were not applied evenly across all airport types, all U.S. airports did not feel the reduction in connectivity equally.

Connectivity at medium-hub airports fell the most between 2007 and 2012, with these airports’ ACQI scores declining by 15.6% over those years. On the other hand, large hub airports did relatively well, only losing 3.9% of their connectivity over the same period. In all, connectivity declined by 8.3% across all airports in the United States between 2007 and 2012, compared to a 12.8% decline in connectivity at smaller airports alone during those years.

The pattern of connectivity of medium-hub airports is different than that of large-hub airports. While both large-hubs and medium-hubs lost connectivity during the economic slowdown of 2007-2009, medium-hub airports did not undergo the same recovery in 2010 and 2011 that large-hub airports did. Instead, connectivity at medium-hubs continued to fall after 2010 as a result of capacity discipline strategies that targeted these airports as a primary focus for service reductions.

Small-hubs have also been hit hard by airline capacity discipline. As with medium-hub airports, small-hubs are retaining their scheduled domestic service to large-hubs (albeit at reduced frequencies) while losing direct service to other smaller destinations. Much of this previous point-to-point service between nearby smaller airports has started to disappear as the network carriers and Southwest continue to consolidate service at their connecting points. However, the number of destinations reachable with a one-stop connecting itinerary has increased over the last six years at many airports.

In the aggregate, fluctuations in connectivity at non-hub and Essential Air Service airports have been relatively minor compared to the significant decreases in ACQI at medium- and small-hubs. The average ACQI score at non-hub and EAS airports decreased by 8.2% from 2007-2012, compared to a 15.6% decline at medium-hubs and an 11.0% reduction at small hubs. This is likely due to federally mandated levels of air service at Essential Air Service airports; these airports have so far avoided the wide-spread capacity cutting that occurred at slightly larger airports.

However, the relatively flat slope of the aggregate ACQI score decline for non-hubs and EAS airports masks some significant changes in connectivity at individual airports. Some of these smallest airports were successful in luring one or more network carriers to start service between 2007 and 2012, increasing their connectivity by many multiples. Other small airports lost all network carrier service over these years, causing a devastating drop in connectivity to an ACQI score of 0 in some years. Many of these airports have been able to win back service in recent years, often from an ultra-low cost carrier like Allegiant Air or Spirit Airlines. However, the resulting level of connectivity with ULCC service is often less than with network carrier service, since ULCCs generally provide point-to-point service to vacation destinations with few connecting itineraries available. The appendices of the report show how some airports gained or lost significant portions of their connectivity score throughout the last six years, highlighting the volatility that small airports face and the importance of each and every flight and destination in maintaining attractive levels of connectivity for potential passengers.

Capacity Discipline and Airport Connectivity

As a final question, the paper examines the extent to which capacity discipline in the form of reductions in scheduled domestic flights and available seats has directly impacted connectivity. If there is a direct correspondence between capacity discipline and connectivity, we should expect to see decreases in connectivity similar to the declines in flights and seats at these airports.

For each airport type, the percent change in connectivity was significantly less than the percent change in both domestic seats and domestic flights over the study period. This suggests that a significant portion of the airlines’ capacity discipline strategies did not directly decrease passengers’ access to the global air transportation network, and instead involved cutting redundant service. Service could be called redundant if the connecting options from one hub overlap nearly completely with connection options from another hub that is already served.  In the aggregate, these repeated cuts of redundant service at “duplicate hubs” explain the large decrease in connectivity at Memphis, Salt Lake City, Pittsburgh, and Cincinnati over the last six years.

Conclusions: Future Trends in Small Airport Connectivity in the U.S.

The Airport Connectivity Quality Index (ACQI) developed in the report provides a straightforward way to compare connectivity between multiple airports or at a single airport over a period of time. Airport managers and policy makers will likely be interested in examining the appendices of the report, which show how the connectivity scores and rankings of their local airports have changed over the past six years. In the aggregate, however, what trends can we extrapolate from the ACQI to anticipate changes in connectivity over the next five years?

Capacity discipline does indeed appear to be a dampening force on airport connectivity, particularly for smaller airports. On the whole, small community airports have struggled to gain back connectivity since airline capacity discipline started in earnest in 2011, as airlines kept domestic capacity deliberately restricted despite the start of macroeconomic recovery in the country and stability in fuel prices. Barring any significant positive or negative macroeconomic shock, the downward trend in connectivity at small- and medium-size airports will likely continue, but the pace will most likely slow as airlines have already removed most redundant flying from their networks. However, the American Airlines/US Airways merger could place further downward pressure on connectivity as schedule and route redundancies are removed from the combined airline’s new network.

This assumes that airlines will continue to practice capacity discipline strategies by adding little net nonstop service over the next five years. However, the question certainly remains whether capacity discipline is a stable competitive equilibrium. In a game theoretic context, capacity discipline could be examined in a classic prisoner’s dilemma construct. It would appear that individual airlines each have an incentive to deviate from the capacity discipline equilibrium and increase capacity in order to gain more market share and, ostensibly, increase profits. It is possible that in the near future, an airline will decide to bolster capacity in key markets, breaking with capacity discipline and perhaps causing other airlines to feel compelled to follow suit to avoid losing market share. However, if all airlines shift to a capacity expansion strategy, too much capacity will likely be introduced into the market, leading to lower profits across the industry.

In this outcome, connectivity at smaller airports would likely increase as airlines begin to compete once again on the sizes of their networks. However, this scenario currently appears unlikely. Airlines have been able to return to profitability as a result of capacity discipline, and as of early 2013 appear unlikely to break with the strategy in the near term in an effort to gain market share. Yet it only takes one airline making a move to add capacity to cause the entire equilibrium to destabilize.

Hence, we expect to see small community airport connectivity to continue to stagnate in the near future. Individual airports may, through clever packages of incentives, continue to induce airlines to provide new service, boosting connectivity on a case-by-case basis. However, only service that can prove itself to be profitable will remain a long-term part of the U.S. air transportation network. Airports that win new service should expect to see their connectivity continue to fluctuate as airlines evaluate the economic merits of the new flights and incentive packages.

Comments

[Robert Chapin]

This is an interesting topic that often comes up during media coverage of the "pilot shortage" and regional airline sustainability. Being able to quantify the robustness of a route network and how it changes over time must be a vital supply-side statistic.