Effects of Airline Industry Changes on Small- and Non-Hub Airports (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

15 C H A P T E R 2 2.1 Introduction The U.S. aviation industry has undergone major changes in the past 15 years. First, the events of September 11, 2001, led to a drop-off in flying, and the imposition of new security procedures have affected the costs of providing air service and lengthened passenger travel times. In addi- tion, recessions in 2001 and 2007–2009 reduced demand for air travel, while the price of jet fuel increased substantially. These factors caused billions of dollars in losses to U.S. airlines, leading to bankruptcies, liquidations, mergers, and acquisitions. The industry responded by reducing capacity, retiring inefficient aircraft types, and raising revenue through the imposition of new and expanded ancillary fees. These responses allowed the industry to return to profitability in the past few years. In addition, new “ultra LCCs” (ULCCs) have emerged that focus on point-to-point air service instead of hub-and-spoke network structures. However, the changes in the industry have not had a uniform effect on all airports. Airports serving smaller communities have been particularly affected by the changes, resulting in reduced service levels, less airline competition, and poorer service quality. Reduced air service is a concern for smaller communities—several applied and academic studies have found positive, significant relationships between increased access to commercial air service and factors related to economic development in local communities (Goetz 1992; Brueckner 2003; Green 2007). Officials at airports serving smaller communities need to understand how the recent history and performance of the U.S. aviation industry has affected air service in their com- munities, so that they can develop air service to promote local economic development. 2.2 Risk and Uncertainty The recent history and performance of the U.S. aviation industry provides important lessons on the risks and uncertainties of air service at small- and non-hub airports. Smaller communities have had very limited ability to respond to major trends affecting the aviation industry in recent years, including economic recessions and the price of jet fuel. Smaller communities also have had little influence on airlines’ responses to these trends, which have included reducing service levels and retiring inefficient aircraft types. Although some smaller communities have tried to reduce the risk of losing air service by providing incentives or subsidies for air service, these efforts have involved much uncertainty. In some cases, carriers exited markets when incentives or subsidies ended. Although smaller communities may try to reduce risk, there will still be uncertainty because of factors outside the Literature Review of Airline Industry Trends

16 Effects of Airline Industry Changes on Small- and Non-Hub Airports control of the community. A review of the recent history of the U.S. aviation industry provides lessons in the risks and uncertainty of air service at small- and non-hub airports. 2.3 Recent History and Performance of the U.S. Aviation Industry The recent history and performance of the U.S. aviation industry can be seen in the context of external shocks, the industry’s response to those shocks, and the results of the industry’s response. The main external shocks to the U.S. aviation industry in recent years were economic reces- sions, which reduced the demand for air travel, and the increased price of jet fuel, which increased airline costs, further reducing the overall level of air travel demand. The aviation industry responded to these external shocks by reducing capacity, consolidating, retiring inefficient air- craft types, and developing new sources of revenue. 2.3.1 External Shocks The U.S. economy has experienced two recessions since 2000, which have negatively affected the demand for air travel. According to the National Bureau of Economic Research, the two recessions in the past 15 years occurred from March 2001 to November 2001 and from Decem- ber 2007 to June 2009 (NBER 2014). In 2012, the U.S. DOT’s Office of Inspector General (OIG) issued a wide-ranging report on the performance of the aviation industry between 2008 and 2011 (USDOT OIG 2012). The report found significant evidence of the 2008–09 recession’s negative effects on the demand for air travel. Exhibit 2-1 shows the trend line and actual amount of per capita disposable income for the United States from 2000 to 2011. This reduction in purchasing power reduced the demand for leisure air travel; the slowdown in the economy also reduced business travel demand. Per capita disposable income declined during the 2008–2009 economic recession and has grown very slowly since the recession, relative to historical growth rates. On the cost side, the price of jet fuel rose dramatically between 2004 and 2008, then temporarily crashed, and returned to historically high levels by 2011. These increases have been a major source Source: Bureau of Economic Analysis (Census Bureau) Exhibit 2-1. Inflation-adjusted per capita disposable income, 2000–2011 (USDOT OIG 2012, p 4).

Literature Review of Airline Industry Trends 17 of financial difficulty for airlines (USDOT OIG 2012, p. 4). Exhibit 2-2 shows the average monthly spot price per barrel of jet fuel from January 2000 to December 2013. The price nearly quadrupled during the period, from $33 per barrel in January 2000 to $124 per barrel in December 2013. ACRP Report 48 found that, as a result of rising jet fuel prices, fuel comprised approximately 35% to 40% of total airline operating expenses in 2009, relative to the 15% of operating expenses in 2001 as shown in Exhibit 2-3. The report also found this was accompanied by a decline in overall domestic seat capacity offered by commercial carriers. (Though not shown in the exhibits above, the recent sharp decline in oil prices in the second half of 2014 has clearly benefitted the major air carriers, but there is not yet any indication that they plan to increase capacity as a result.) 2.3.2 Industry Response The airline industry has responded to the financial strain caused by high fuel prices and reduced demand for air travel in several ways, including consolidation, capacity reduction, retirement of inefficient aircraft types, fare increases, and growing reliance on ancillary revenue. Sources: Bureau of Transportation Statistics (recent months are preliminary and subject to restatement); Energy Information Administration Exhibit 2-2. Average jet fuel spot price ($ per barrel), 2000–2013 (Airlines for America 2014).

18 Effects of Airline Industry Changes on Small- and Non-Hub Airports The effects of the fuel cost increase on the air transportation system can be explained through supply-side and demand-side effects. According to a recent study, “Supply-side effects include increases in direct operating costs of airlines, resulting in changes to networks and fleet assignments. Demand-side effects are due to reductions in overall economic activ- ity, as well as passenger and freight sensitivity to fare increases” (Morrison, Bonnefoy, and Hansman 2010, p. 7). During the fuel cost increase, airlines responded by reducing use of fuel-intensive aircraft and by increasing fares to pass fuel costs on to passengers, both of which are supply-side effects. Airlines passed some of the increase in fuel costs on to passengers through fare increases, fuel surcharges, and unbundling of services by introducing ancillary fees for services (e.g., checked baggage and onboard meals) (Morrison, Bonnefoy, and Hansman 2010, p. 9). The financial strain on airlines resulted in the bankruptcies of 49 U.S. passenger and cargo airlines between 2001 and 2013, of which 13 occurred in 2008. Most bankruptcies did not result in a carrier ceasing operations, because the U.S. Bankruptcy Code allows companies to reorga- nize under Chapter 11. Exhibit 2-4 shows U.S. airline bankruptcies from 2001 to 2013 by type: Chapter 7 (liquidation) or Chapter 11 (reorganization). In some cases, bankruptcy reorganizations made it easier for carriers to merge with one another. A series of airline mergers during the 2000s resulted in substantial consolidation of seat capacity as Exhibit 2-3. Fuel prices and domestic scheduled seats, 2003–2009 (Spitz and Berardino 2011). Source: Airlines for America (A4A) data, U.S. Bankruptcies and Services Cessations Exhibit 2-4. U.S. airline bankruptcies, 2001–2013.

Literature Review of Airline Industry Trends 19 shown in Exhibit 2-5. In 2001, 90% of domestic seats were split among 10 major carriers (Trans World Airlines is not shown because it merged with American Airlines during 2001). After the American Airlines–US Airways and the Southwest Airlines–AirTran Airways mergers are com- pleted, four carriers (i.e., American, Delta, Southwest, and United) will control approximately 85% of domestic passenger capacity. As a result of industry consolidation and bankruptcies, airlines closed some under-performing hubs (USDOT OIG 2012). Exhibit 2-6 shows the change in the number of flights at five airline hubs that experienced particularly drastic service decreases in recent years. The number of scheduled passenger flights at Cincinnati, Cleveland, Memphis, Pittsburgh, and St. Louis declined by nearly 40% from 2007 to 2012. In 2012, the U.S. DOT OIG noted that “the rising and volatile price of fuel now has a much greater influence on whether the airlines add or cut a flight and how frequently fares need to be adjusted” (USDOT OIG 2012, p. 4). The influence of the price of fuel and the reduced demand for air travel due to the economic slowdown and airline consolidation can be seen in the dra- matic decline in the number of domestic flights operated by U.S. passenger carriers from 2003 Source: Official Airline Guide (OAG), October data, 2001-2013; consolidations shown based on effective date of mergers. Exhibit 2-5. Seat shares of top U.S. carriers, 2001–2013. Hub Airport Hub Airline Scheduled Passenger Flights June 2007 Scheduled Passenger Flights June 2012 Percent Change Cincinnati Delta 12,781 4,710 -63.1% Cleveland Continental 9,070 6,684 -26.3% Memphis Northwest 8,227 5,308 -35.5% Pittsburgh US Airways 7,462 4,470 -40.1% St. Louis American 9,503 7,127 -25.0% Total 5 Hubs 47,043 28,299 -39.8% Source: Federal Aviation Administration Exhibit 2-6. Flight changes at selected hub airports (USDOT OIG 2012, p. 12).

20 Effects of Airline Industry Changes on Small- and Non-Hub Airports to 2013. The number of flights declined by more than 12% during the period, from 9.5 million in 2003 to 8.3 million in 2013 as shown in Exhibit 2-7. The decline in flights in recent years was not uniform across all U.S. carriers. While all of the net- work carriers reduced flights by more than 10% between 2007 and 2012, most LCCs and ultra-low- cost carriers (ULCCs) increased flights during the same time period (USDOT OIG 2012, p. 28). Exhibit 2-8 shows the change in scheduled flights by carrier between June 2007 and June 2012. Much of the decline in flights by network carriers occurred in the short-haul sector. The U.S. DOT OIG investigated the change in flights from 2007 to 2012 by length of haul. The number of available short-haul flights of less than 250 miles in 2012 was 24% lower than in 2007. Exhibit 2-9 shows the percent change in scheduled flights by flight distance from June 2007 to June 2012. The change in flights by length of haul can be explained by passenger demand elasticity, relying on the well-established observations that short-haul air travel is more price elastic than long-haul air travel, domestic air travel is more price elastic than international air travel, and leisure travel is more price elastic than business travel. Higher price elasticity indicates that consumers are more sensitive to changes in price. In response to these factors, a recent study suggests that airlines made strategic Source: Bureau of Transportation Statistics, T-100 Segment Data Exhibit 2-7. U.S. passenger carrier domestic flights, 2003–2013. Source: Federal Aviation Administration Exhibit 2-8. Change in scheduled flights by carrier, 2007 versus 2012 (USDOT OIG 2012, p. 28).

Literature Review of Airline Industry Trends 21 decisions to reduce service in markets that were more price elastic and to maintain or increase service in markets that were more price inelastic (Morrison, Bonnefoy, and Hansman 2010). Despite the substantial reduction in the number of flights since 2007, the number of available seat miles has been largely unchanged. Airlines targeted short-haul flights and flights with smaller aircraft for most of the flight reductions, which means that the overall number of available seats was not impacted as significantly as the overall number of flights. Exhibit 2-10 shows domes- tic capacity by carrier type (network airline or low-cost airline) from 2007 to 2012. According to an OIG report, the recent changes in domestic capacity “demonstrate that the airlines have adapted and developed new means for managing excess capacity” (USDOT OIG 2012, p. 8). When adjusted for stage length and seat size, smaller aircraft (such as regional jets) are gener- ally more fuel intensive than larger aircraft, so airlines reduced the miles flown of fuel-inefficient smaller aircraft while increasing the miles flown of fuel-efficient larger aircraft (Morrison, Bonnefoy, and Hansman, 2010). The number of domestic flights by regional jet aircraft with 30 to 70 seats decreased by 20% from June 2007 to June 2012, as shown in Exhibit 2-11 (USDOT OIG 2012, p. 31). Source: Federal Aviation Administration Exhibit 2-9. Change in flights by length of haul, 2007 versus 2012 (USDOT OIG 2012, p. 29). Source: GAO analysis of DOT data. Exhibit 2-10. Domestic capacity changes by carrier type, 2007–2012 (GAO 2014a, p. 15).

22 Effects of Airline Industry Changes on Small- and Non-Hub Airports The trend of fewer flights by aircraft with fewer seats is expected to continue as virtually the entire small regional jet fleet in the United States will be retired over the coming years, as shown in Exhibit 2-12. Despite the recent decrease in use of small regional jets, network carriers increased their reli- ance on regional carriers in the past decade. The share of flights marketed by American, Delta, United, and US Airways and operated by regional partners increased from 40% in 2000 to 61% in 2011 (USDOT OIG 2012, p. 8). Source: Federal Aviation Administration Exhibit 2-11. Scheduled domestic flights by aircraft size, 2007 versus 2012 (USDOT OIG 2012, p. 31). Source: Greenslet/ESG Aviation Services, United States Airlines Fleet Forecast data, 2011-2032 Exhibit 2-12. Historical and projected domestic small regional jet fleet, 2000–2032.

Literature Review of Airline Industry Trends 23 By reducing capacity, carriers were able to increase load factors on flights. With fuller planes, carriers were able to reduce the number of discounted fares offered (USDOT OIG 2012, p. 8). With reduced capacity and increased fares, network airlines were able to increase revenue per available seat mile (RASM) by 23% from 2007 to 2012, while low-cost airlines increased RASM by 27% during the same period (GAO 2014a, p. 15–16). Exhibit 2-13 shows RASM by carrier type (network airline or low-cost airline) from 2007 to 2012. Exhibit 2-14 shows average fares by flight length from 2000 to 2011. Carriers increased fares most sharply in short-haul markets that experienced the greatest capacity reduction. Source: GAO analysis of DOT data. Note: Unit revenues, or revenue per available seat mile, are calculated as operating revenues excluding transport revenues divided by total available seat miles. Exhibit 2-13. Unit revenue by carrier type, 2007–2012 (GAO 2014a, p. 16). Source: Bureau of Transportation Statistics Exhibit 2-14. Average fares by flight length, 2000–2011 (USDOT OIG 2012, p. 9).

24 Effects of Airline Industry Changes on Small- and Non-Hub Airports Source: Bureau of Transportation Statistics, Form 41, Schedule P-1.2 Exhibit 2-15. Ancillary revenue collected by U.S. passenger airlines, 2000–2013. According to DOT, short-haul fares increased from $116 in 2000 to $164 in 2011 (USDOT OIG 2012, p. 9). In addition to increasing passenger fares, airlines increased revenue by adopting ancillary fees to recapture some revenue lost because of reduced demand for air travel and to offset increased fuel costs. The recent growth in ancillary fees results from new fees and expanded existing fees for services that were previously included in base airfares to generate additional ancillary rev- enue. Examples of services include checked baggage, reservation changes and cancellations, seat selection, food, and other miscellaneous items. Between 2000 and 2010, average round-trip pas- senger fees increased from $3 to $22 (USDOT OIG 2012, p. 9). Exhibit 2-15 shows the amount of baggage fees, reservation and cancellation fees, and mis- cellaneous revenue collected by scheduled U.S. passenger airlines from 2000 to 2013. Ancillary revenue increased from approximately $1.7 billion in 2000 to approximately $9.8 billion in 2013. 2.3.3 Results High fuel costs and reduced demand for air travel resulted in increased financial stress for airlines in the past 15 years. Legacy carriers accumulated more than $62.8 billion in financial losses between 2000 and 2009, while LCCs accumulated limited profits of $2.1 billion during the same period (USDOT OIG 2012, p. 5). The same overall pattern of profit and loss is evident if looking at net income (which includes the effects of non-operating revenues and expenses, taxes, and interest) rather than operating income. Exhibit 2-16 shows overall operating income of U.S. passenger airlines from 2000 to 2011 (USDOT OIG 2012, p. 10). Since 2009, industry consolidation, capacity reduction, fare increases, and increased ancillary revenue have resulted in improved financial performance. Major U.S.

Literature Review of Airline Industry Trends 25 airlines broke even in 2009 and generated operating profits in 2010 and 2011, in sharp contrast to the operating losses in 2001 through 2005 and 2008. The improved financial performance of U.S. airlines is partially a result of reduced competi- tion in the industry. According to OIG, “Since 2007, overall competition as measured by the number of airlines serving an airport has declined substantially” (USDOT OIG 2012, p. 31). But in addition, the U.S. airline industry has benefitted greatly from the significant economic recovery that has taken place since 2010 and has been further buoyed by the continued growth of new and enhanced ancillary (non-fare) revenues. In recent years, airline service quality has improved markedly (as measured by the percent- age of on-time and completed flights). Exhibit 2-17 shows the number of flights by outcome (on-time or delayed/cancelled) from 2000 to 2011. The peak year for delays and cancellations was 2007 when 29% of flights were delayed or cancelled. By 2011, the rate had declined to 23% (USDOT OIG 2012, p. 13). Part of this improvement likely resulted from the recent trend of Source: Bureau of Transportation Statistics Exhibit 2-16. U.S. passenger airlines operating income, 2000–2011 (USDOT OIG 2012, p. 10). Source: Federal Aviation Administration Exhibit 2-17. Total flight operations by outcome, 2000–2011 (USDOT OIG 2012, p. 13).

26 Effects of Airline Industry Changes on Small- and Non-Hub Airports providing fewer flights but with larger aircraft and from rationalizing scheduled block times so that more flights arrive “on time.” 2.4 Effects on Airports Serving Small Communities The airline industry’s response to external shocks in the past 15 years has had a particu- larly severe effect on service at airports serving small communities as identified in a number of studies (USDOT OIG 2012; GAO 2014a and b, 2011a, 2003a and b, 2005; Morrison, Bonnefoy and Hansman 2010; Wittman 2014; Wittman and Swelbar 2013a and b). Airlines often look to reduce service at airports serving smaller communities first because the opportunity for profit at these airports is smaller because the population base of potential travelers is smaller. In addition, if a small community is within a reasonable driving distance of a larger airport, residents may choose to drive to the larger airport that has better service options (further reducing the demand for air service at the airport serving the smaller community) (GAO 2003a). 2.4.1 Impact on Seats Exhibit 2-18 shows the annual change in available seats relative to October 2001 by airport hub group and the number of available seats for October of each year. The overall number of available seats was lower in 2013 than in 2001 for each hub group, although the decline was much greater at medium-hub airports than at other hub groups. Airport hub groups are deter- mined annually by FAA based on the number of enplanements. Thus, part of the medium-hub Source: OAG, October data, 2001-2013; includes domestic service only. Exhibit 2-18. Available seats by airport hub group, 2001–2013.

Literature Review of Airline Industry Trends 27 airport group’s performance can be explained by the reassignment of some medium-hub air- ports to the small-hub airport group during the study period. Exhibit 2-19 provides snapshots of the U.S. passenger airline industry by airport hub group in October 2001 and October 2013. The number of large- and medium-hub airports decreased while the number of small-hub airports increased. This reflects the reclassification of some of the larger airports as small-hub airports because of fewer enplaned passengers at the larger airports. There were also 16 fewer non-hub airports in 2013 relative to 2001, reflecting reduced service at the nation’s smallest airports with scheduled service. The increasing consolidation of passenger service at larger airports is evident when examin- ing the share of seats and number of airports in each airport group. The 67 large- and medium- hub airports in 2001 accounted for 87.9% of total seats. Although the total share of seats in 2013 was essentially identical (88.0%), there were only 63 large- and medium-hub airports in 2013, which means that the number of seats at large- and medium-hub airports increased, on average, from 2001 to 2013. 2.4.2 Impact on Flights Exhibit 2-20 shows the annual change in flights relative to October 2001 by airport hub group and the number of flights for October of each year. The number of overall flights in 2013 was lower than in 2001 for each of the airport hub groups, although the decline at large-hub airports was less substantial than at airports in the small-hub groups. Source: OAG, October data, 2001-2013; includes domestic service only. Exhibit 2-19. Changes in the share of seats and number of airports by airport hub group from 2001 to 2013.

28 Effects of Airline Industry Changes on Small- and Non-Hub Airports Part of the change in flights to smaller airports can be explained by the growth of LCCs relative to network carriers. Network carriers traditionally provided service to smaller airports to feed hub operations at larger airports. LCCs typically only serve larger airports with point-to-point service. The low fares offered by LCCs often induce residents of smaller communities to drive to larger air- ports, which reduces the demand for air service at smaller communities. According to a survey con- ducted by GAO, “Eighty-one percent of [small community airport officials] attributed the leakage to the availability of lower fares from a major airline at the alternative airport” (GAO 2003b, p. 5). Airlines reduced service, sometimes dramatically, at some under-performing hubs as a result of industry consolidation and bankruptcies. Small- and non-hub airports were significantly affected by these changes in the airline hub structure. Exhibit 2-21 shows the change in flights from small- and non-hub airports to airline hub airports. The size and color of the circle reflects the absolute change from 2001 to 2013. Pittsburgh International Airport (PIT) experienced the largest negative change; in 2013 there were no flights from small- and non-hub airports to PIT in contrast to the 7,240 flights in 2001. John F. Kennedy International Airport (JFK) in New York experienced the largest positive change; in 2013, there were 41% more flights from small- and non-hub airports to JFK than in 2001. For ease of discussion, this section refers to airports by their FAA location identifiers. Appendix A contains a table with descriptive information for each airport in the data set used for this report. Changes were not uniform across all airports. Those airports that served as hubs for network carriers (e.g., American Airlines [STL], Delta Air Lines [CVG], and US Airways [PIT]) often Source: OAG, October data, 2001-2013; includes domestic service only. Exhibit 2-20. Flights by airport hub group, 2001–2013.

Literature Review of Airline Industry Trends 29 suffered a substantial decline in flights. Conversely, those airports that served as hubs for LCCs (e.g., JetBlue Airways [JFK] and Southwest Airlines [DEN and LAS]) often attracted more flights. 2.4.3 Impact on Connectivity A study by Morrison, Bonnefoy, and Hansman (2010) provides more evidence of the con- nection between fuel prices and observed carrier network and fleet changes. The study found that air transportation networks did not change uniformly in response to fuel cost increases and attempts by airlines to improve fleet fuel efficiency; non-hub airports lost 12% of connections to the air transportation network, compared to an average loss of 2.8%. Analyses were performed over medium-term (July 2004 to July 2008) and short-term (July 2007 to July 2008) periods. These periods were selected because the price of fuel changed sub- stantially, while the potential effect of exogenous events (e.g., changes in GDP, airline competi- tion, and accidents and security incidents) was low. In the short term, service was reduced for small and remote communities. For airports that lost all service, the average driving distance to the next nearest airport with service was 57 miles. The share of the continental U.S. population living within 40 miles of an airport with regular service declined to 88.9% in July 2008 from 90.3% in July 2007. Wittman and Swelbar (2013b) examined recent commercial airline domestic scheduling trends, with particular emphasis on trends at smaller airports (defined as non-, small-, and medium-hub airports). Most airports have seen reductions in scheduled domestic flights, stem- ming from economic conditions and the airlines’ response of capacity discipline, but some airports have seen larger reductions than others. The major network airlines (i.e., American, Delta, United, and US Airways) and South- west Airlines responded to recent difficult economic conditions by shifting from a strategy of Source: OAG, October data, 2001-2013; includes domestic service only. Exhibit 2-21. Change in flights from small- and non-hub airports to airline hub airports, 2001 vs. 2013.

30 Effects of Airline Industry Changes on Small- and Non-Hub Airports prioritizing market share to a strategy of prioritizing profitability. The new strategy entailed rationalizing capacity to reduce operating losses and increase yields by increasing load factors. The strategy resulted in an overall small number of domestic flights and a concentration of the remaining flights at large-hub airports. In some instances, the network carriers relied more on their regional partners flying small regional jets to serve smaller markets; in other cases, LCCs such as JetBlue, ULCCs such as Spirit Airlines, and/or ultra-regional carriers such as Cape Air and Great Lakes Airlines have filled the void in whole or in part left by network airlines at smaller airports. Flying at small- and non-hub airports changed during recent years in many cases. The number of flights by 37 to 50 seat regional jets declined substantially at small-hub airports as network air- lines reduced frequency to their connecting hubs and/or replaced multiple flights with 37–50 seat regional jets with fewer flights by 50–76 seat regional jets. In other cases, network airline service from smaller airports to connecting hub airports was terminated entirely and LCCs or ULCCs began service from the smaller airports to vacation destinations or other smaller airports. By analyzing the smaller airports that experienced reduced domestic flights in recent years, Wittman and Swelbar developed a taxonomy that could be used to identify airports which may be at risk of future service loss. They identified lack of local demand, proximity to a nearby hub, and presence of ULCCs (given ULCCs history of filling voids left by network airlines) as the main risk factors for future service losses. The authors also identified several other exogenous factors that could result in future service loss at smaller airports, including potential pilot short- ages at regional airlines resulting from changed federal regulations, the effect of federal seques- tration on air traffic control service at smaller airports, and the future budget of the Essential Air Service Program. In a subsequent paper, Wittman and Swelbar (2013a) developed a metric they call the Airport Connectivity Quality Index (ACQI) to assess an airport’s connection to the global air transporta- tion system and to assess changes over time. The ACQI uses the frequency of available scheduled flights, the quantity and quality of destinations served, and the quantity and quality of connect- ing destinations to produce a relatively easy-to-compute metric. The ACQI includes two parameters that influence the model outcome. One parameter reflects the relative quality of a destination airport, by weighting FAA hub airport types based on aver- age enplanements such that service to airport types with more enplanements is weighted more heavily than service to airport types with fewer enplanements. The other parameter reflects the relative values of non-stop and one-stop service, by weighting service based on values derived from the literature showing that passengers value a non-stop itinerary up to 8 times more than a connecting itinerary. The report places special emphasis on connectivity at smaller airports, which had not received much attention in previous work on airport connectivity. ACQI connectivity scores were com- puted for 462 U.S. airports for each year from 2007 to 2012. Smaller airports experienced greater declines in ACQI connectivity scores (15.6% at medium- hub airports and 11.0% at small-hub airports) than large-hub airports (3.9%). Non-hub and EAS airports experienced a smaller decline (8.2%) than medium-and small-hub airports, but a greater decline than large-hub airports. However, there was great variation in the ACQI con- nectivity scores for individual non-hub and EAS airports because some airports gained new service by network airlines (causing large increases in scores), lost all service (causing scores to drop to zero), or had service by LCCs or ULCCs replace service previously provided by network airlines (causing lower scores). Percent change declines in capacity (flights and available seats) were greater than percentage declines in ACQI connectivity scores, which suggests that airlines’ recent strategy of rationalizing

Literature Review of Airline Industry Trends 31 capacity resulted in reducing frequency or eliminating service to secondary connecting hubs, rather than eliminating all service to both secondary and primary connecting hubs. However, ACQI connectivity scores did indicate that the secondary connecting hubs themselves were adversely affected by recent changes in airline service. 2.4.4 Impact on Airline Competition In response to concerns about competition in the U.S. passenger airline industry given recent airline mergers, GAO analyzed passenger itinerary data from 2007 through 2012 (GAO 2014a). GAO found that the average number of competitors in most markets and market concentration has not changed substantially in recent years. However, when markets were categorized in quin- tiles based on the number of passengers served, the smallest (5th) quintile showed a decrease in competitors from 3.3 in 2007 to 3.0 in 2012, as shown in Exhibit 2-22. Similarly, market concen- tration in the smallest quintile has also increased since 2007. 2.4.5 Impact on Service Quality Service quality is a concern for smaller airports, because there are fewer opportunities for pas- sengers to recover from delayed, cancelled, or diverted flights at small airports with few sched- uled flights relative to larger airports. Furthermore, smaller airports may be more likely to suffer from airline service disruptions because the aircraft serving smaller airports generally have less capacity, resulting in fewer affected passengers if service is disrupted. Recent studies have found that airports in smaller communities experience higher rates of service disruptions than larger communities (GAO 2011a). Source: GAO analysis of DOT data. Note: Each quintile contains approximately the same number of passengers, but the number of city-pair markets differs. For example, in 2012 the first, second, third, fourth, and fifth quintiles contained 37, 99, 237, 682, and 9,379 city-pair markets, respectively, each with 20 percent of the 411 million passengers in our sample. Exhibit 2-22. U.S. passenger airline competition by city-pair market size, 2007–2012 (GAO 2014a, p. 24).

32 Effects of Airline Industry Changes on Small- and Non-Hub Airports Exhibit 2-23 shows the percentage of cancelled or diverted departures by community size in 2010. Rural communities suffered substantially greater rates of cancellations and diversions than larger communities. Although the percentage of delayed arrivals was also higher at rural com- munities than at larger communities, the difference was not as substantial. Service quality is related to the level of airline competition at an airport. A recent OIG study found that, “when airline markets become less competitive both the average length of flight delays and percentage of late flights increased.” When the number of airlines serving a market decreased from three to two, the length of delays increased by more than 25% and the rate of cancellations increased by 7% (USDOT OIG 2012, p. 2). These statistics are particularly relevant to airports serving smaller communities because competition has decreased in recent years. 2.5 Links Between Local Economic Development and Air Service In 2011, the FAA estimated that commercial aviation activity supported 10.5 million jobs and contributed over $650 billion (5.0% of GDP) to the U.S. economy (FAA 2011). Commercial air service to cities and regions across the United States and the world is viewed as a critical resource by local economic development officials. Specifically, economic development officials use a region’s access to reliable and affordable air service as a selling point to potential businesses looking to relocate to an area. In an increasingly interconnected world, access to a large net- work of destinations has become a significant factor for large corporations in deciding where to expand business operations. Access to commercial air service helps connect existing businesses and community members with economic opportunities around the globe through enterprises such as tourism and the just-in-time delivery industry. Several applied and academic studies have found positive and significant relationships between increased access to commercial air service and factors related to economic development in local Source: GAO analysis of FlightStats data. Note: Community sizes are as follows: large metropolitan (greater than or equal to 1 million), midsized metropolitan (250,000 to 999,999), small metropolitan (50,000 to 249,999), and rural (fewer than 50,000). These data reflect the vast majority of scheduled passenger flights. In 2010, for example, the data included performance information for about 98 percent of U.S. scheduled passenger flights from primary airports excluding airports in U.S. territories. Exhibit 2-23. 2010 cancellations and diversions by community size (GAO 2011a, p 14).

Literature Review of Airline Industry Trends 33 communities. Exploring commercial air service and urban development, Goetz (1992) found a positive and significant relationship between per capita commercial passenger flows and future growth in urban areas. Brueckner (2003) found that a 1% increase in enplanements increased employment in service-related industries by 0.1% in the Chicago metropolitan area. The results of Brueckner’s study indicate that expanding the number of enplanements in a metropolitan area may expand employment opportunities for the most vulnerable members of a community. A study by Green (2007) found a positive and significant relationship between enplanements and a local community’s population and employment growth. In exploring what drives per capita incomes in counties in Arkansas, Fullerton, Licerio, and Wangmo (2010) found that per capita incomes were significantly higher in counties with air- ports with commercial air service. A recent study by Tittle, McCarthy, and Xiao (2013) found a positive and significant relationship between runway length at large- and medium-hub airports and a region’s gross metropolitan product (GMP). Finally, a GAO study found a positive cor- relation between the number of enplanements and the number of air carriers serving a market and a host of economic indicators including per capita income and manufacturing employment (GAO 2011b). The strong evidence of the link between commercial air service availability and local eco- nomic development helps explain the sense of urgency among many local officials in trying to retain or attract new service to their communities. However, economic development offi- cials in smaller communities (e.g., those home to small- and non-hub airports) face several demand-related challenges in trying to attract air service to their communities. First, demand for air service is often positively correlated to a region’s population, which is often quite small in communities with small- and non-hub airports. Operating a flight to a commu- nity is in essence a series of fixed costs (e.g., labor, fuel, and maintenance) that remain the same, regardless of the number of passengers on a plane. Therefore, before an air carrier will risk starting new service, they must be convinced that a community’s population base will provide enough demand or revenue for the airline to operate the flight profitably. Second, demand for air service is also strongly correlated with per capita income and employment status of a region. The relationship between per capita income and employment and air service is often viewed as a “chicken and the egg” dilemma, with many officials believing that air service is the key ingre- dient to improving the economic condition of a region, but not being able to attract air service due to the same set of economic indicators. The air service cycle is built on the assumption that increased service will lead to more service as the economic indicators of a region improve, lead- ing to a larger market for travel and therefore, more air carriers (Hazel 2011). Finally, because smaller communities are often served by smaller regional aircraft and have fewer air service choices, the fares for air service at smaller communities are often higher than at larger airports. If a smaller airport is within driving distance of a larger airport with more service and lower fares, the smaller airport may experience demand “leakage,” where members of the local community will drive to the larger airport to access greater choice and lower fares. Leakage occurs for several reasons including proximity to a legacy-carrier network hub (with more direct flight options), proximity to an airport with LCCs (often with lower fares), or due to a fragmented catchment area where competition is high (Martin et al. 2009). A GAO study reported that of the 98 small- or non-hub airports surveyed, 83 indicated leakage was a sig- nificant problem in trying to retain or attract new air service (GAO 2011b). A study by Zhang and Xie (2005) studied the Golden Triangle Regional Airport in Mississippi which is less than 3 hours’ drive time from larger airports in Jackson, Birmingham, and Memphis. The authors found that fares, flight schedules, and distance to the airport were the most significant factors affecting airport choice behavior. An econometric study by Bhadra (2004) found that air travel

34 Effects of Airline Industry Changes on Small- and Non-Hub Airports from smaller communities without an airport is structurally different than those with small hubs—travel from communities with small hubs is relatively elastic, suggesting that travelers are willing to travel to other airports. Therefore, to attract service to their communities, local economic development officials must overcome two problems: unattractive market economics and leakage of passengers to other air- ports. However, local officials are often quite limited in the tools and resources that can be used to solve the public policy problems of growing their market or stopping leakage to other airports. Most evident, a community’s population and its geographic location (in relation to other com- munities) are fixed in the short run. Ultimately, however, some communities may not have the sheer size or level of economic activity or be able to compete with the lower fares and/or better service of a nearby airport, to maintain the necessary demand for air service. Thus, for certain smaller communities, sustainable service, without some form of government intervention, may be unachievable in the long term.