Putting a spin on hubs
By Dave Demerjian
Airways: A Global Review of Commercial Flight, December 2006

Some travelers love them; others will do anything to avoid them. But few can deny that hub airports have become a hallmark of air transport in the USA. Proponents of hubs say they are a logical response to the 1978 deregulation of the airline industry, and point to the explosive growth in passenger traffic over the past two decades as evidence that the hub-and-spoke system works. Critics call hubs congested, inefficient, and expensive, arguing that they stifle competition and keep ticket prices high on many routes.

US airlines, reeling from high oil prices, unrelenting competition, and heightened security concerns, have in recent years closed or reduced the size of some hubs, while significantly expanding others. In a market where regional jets are directly linking more cities than ever before, and airlines are engaged in a never-ending battle to reduce costs, do big hub airports still have a place?

A regulated world
Before 1978, all routes, prices, and schedules of US airlines required approval by the Civil Aeronautics Board (CAB), which had begun regulating the industry 40 years earlier. While the agency's mission was to ensure that airlines provided safe, reliable service on both low- and high-volume routes across the United States, some argued that by limiting competition and access to key routes, the CAB provided airlines little impetus to run their businesses efficiently. The CAB itself became known as a bloated bureaucracy, and notorious for its lengthy delays in approving new routes.

According to former Air Transport Association Chief Economist David Swierenga, the move toward hub-and-spoke networks began in the Seventies, and was accelerated by the energy crisis later that decade. The airlines believed that a hub-based system was the most efficient way for them to operate large fleets while allowing them to extend the size of their networks.
"The airlines realized that they needed to cut capacity in order to conserve fuel," Swierenga says. "The hub-and-spoke system is quite common in other modes of transportation, and airlines saw this as their best option."

R. John Hansman, the Director of the International Center for Air Transportation at the Massachusetts Institute of Technology, explains that providing nonstop service between 50 markets involves a total of 2,450 flights. Add a central hub where passengers change airplanes to reach their destinations, and the entire network can be connected with a total of only 98 flights. The CAB, however, wasn't convinced the hub system could be successfully applied to commercial aviation, and was reluctant to allow carriers to reshape their networks in a way that reduced non-stop service, even if they did ultimately serve more destinations. CAB regulation also increased the risk to airlines.

Complicating matters was the arrival of wide-body jetliners. During the Seventies, many trunk airlines took delivery of 350-seat Boeing 747s and 200/250-seat McDonnell Douglas DC-10 and Lockheed L-1011 tri-jets, but few operated routes demanding that many seats. Airlines figured that flying passengers into a hub, where they could transfer to other flights, as a good way to utilize this additional capacity.

At the same time, the idea of deregulation took hold. It was an issue embraced by White House hopeful Edward Kennedy, and later by President Jimmy Carter, who saw deregulation legislation as an easy way to demonstrate his administration's commitment to reforming government. The government was also eager to avoid a repeat of its expensive bailout of the US rail industry in 1970, and while the move was strongly opposed by some airline chiefs, deregulation of the US airline industry became law in 1978.

New rules
The 'script' that the airlines had used for decades suddenly changed dramatically. Carriers were now free to compete on any domestic route, setting schedules, frequencies, and – eventually – fares based on market demand. But regulation had left most US airlines with routes concentrated in certain regions, and few could boast a true national route network. To serve the cities now up for grabs, airlines needed to either buy hundreds of additional airplanes – a cost-prohibitive proposition during the deep recession at the end of the Seventies – or quickly begin to expand hub-based networks.

One of the first to try it was Braniff Airlines, which had launched scores of new city pairs immediately following deregulation. Forced by the government's 'use it or lose it' rule to begin operating these routes immediately, Braniff significantly increased its point-to-point flying, but also set up a hub at Dallas/Ft. Worth (DFW), Texas, funneling passengers onto its lone 747-100, which offered nonstop service between Dallas and Honolulu.

Both Delta Air Lines and Eastern Air Lines used Atlanta (Georgia) as a hub before deregulation, and American was quick to adopt and refine the idea. Each time a new spoke was added to a hub, the number of passengers using that hub increased exponentially without a significant increase in fixed costs. For example, adding ten new destinations to and from a hub might increase the number of passengers using it by 20%, but would increase the hub's fixed costs by only a fraction of that amount. Scheduling waves of arrivals and departures close to one another – a process known as 'banking' flights – increased connection opportunities for passengers, further enhancing a hub's appeal.

Redrawing the map
In the brutally competitive post-deregulation environment, building route networks around a series of hubs became an essential strategy for US airlines. American was arguably the most aggressive, continuing to expand at DFW, opening hubs in Nashville (Tennesee), Raleigh/Durham (North Carolina), San Jose (California), Miami (Florida), and San Juan (Puerto Rico), and spending nearly $100 million to build a hub at Chicago-O'Hare that could compete with United's giant operation there. In addition to developing Chicago, United slugged it out with Continental and Frontier in Denver (Colorado), and later opened hubs at San Francisco, Los Angeles (California), and Washington Dulles. Most major US airlines worked to establish dominant positions at key airports.

Those airlines unable to develop strong hubs found themselves at a competitive disadvantage. While Pan Am boasted a huge fleet of wide-body jets and an enviable international route network, it lacked a strong domestic operation to feed traffic onto its passenger-hungry 747s. Its attempt to build domestic feed by purchasing National Airlines in 1980 was unsuccessful. Eastern, long plagued by labor and operational problems, was also hurt by a lack of strong hubs. It was a weak number two to Delta in Atlanta and came under attack by American in its stronghold of Miami. Eastern's attempt to expand westward by establishing a hub at Kansas City also failed.

In the consolidation that followed deregulation, many carriers picked up hubs through mergers and acquisitions. US Air (now US Airways) added hubs in Dayton (Ohio), Charlotte (North Carolina), and Syracuse (New York) when it purchased Piedmont in 1989 (all but Charlotte were later shuttered); TWA became dominant in St. Louis, Missouri, with the buyout of Ozark in 1986; and Northwest strengthened its position in Minneapolis (Minnesota) and Detroit (Michigan) after purchasing Republic Airlines.

By the end of the Eighties, almost all major US airlines had opened hubs across the United States. While a handful of airports served as hubs for two or more airlines (American and Delta at DFW; Eastern and Delta at Atlanta; United, Continental, and Frontier at Denver; and American and United at O'Hare), this was the exception rather than the rule. For the most part, each airline focused on becoming the dominant carrier at a handful of airports, and allowed its competitors to do the same.

A delicate equilibrium
By the mid-Nineties, US airlines had entered a period of relative stability, with the industry reporting record profits beginning in 1996. Some of the nations' best-known airlines – as well as most of the post-deregulation startups – were gone, victims of a decade of intense competition and the first Persian Gulf War. Eastern, Pan Am, Braniff, and PeoplExpress had been forced out of business, while Piedmont, Republic, Ozark, and Western had dissapeared into other carriers. Those that remained moved away from the financially ruinous fare wars that had characterized early deregulation, and shifted their focus toward building hub networks.

As the decade progressed, critics began crying foul, accusing some airlines of building impenetrable fortress hubs where one carrier controlled the vast majority of traffic, and engaging in predatory pricing practices that made it impossible for others to compete. In 1999, the US Justice Department sued American Airlines, claiming the airline used predatory pricing practices at DFW against three low cost carriers: Vanguard, Western Pacific, and SunJet International. According to the complaint, "American Airlines repeatedly sought to drive small, start-up airlines out of [DFW] by saturating their routes with additional flights and cutting fares. After it drove out a new entrant, American re-established high fares and reduced service." The suit was dismissed in 2001, and while American remains the dominant carrier at DFW, all three of the upstarts have ceased operations.

In 2000, low-cost carrier Spirit Airlines filed suit against Northwest Airlines, claiming it had engaged in predatory pricing on several routes out of Detroit. That case is still pending.

Swierenga believes the predatory pricing arguments are for the most part unfounded. "It's painful for the guys trying to gain access, but that's how the marketplace works," he says of airlines dropping prices when new carriers begin service from their hubs. "When a competitor enters your marketplace, you drop fares. That simulates demand, and to meet that demand, you add capacity."
           
New realities 
Hub and spoke networks – and the airlines operating them – came under increased pressure as the Nineties progressed. One reason was the continued growth of low-fare, point-to-point carriers. Rather than setting up large hubs, these airlines built their route networks around high frequency, non-stop service between key markets. Southwest, the largest low-cost carrier – which began in 1971 with flights between three Texan cities: Dallas, Houston, and San Antonio – expanded rapidly during the Nineties, and served 50 cities by 1997. Another dozen destinations have been added since then.

As Southwest and others continued expanding, so did the number of US cities connected by nonstop flights, and the number of opportunities to 'overfly' congested hub airports. Passengers long-accustomed to connecting through a hub to reach small- and medium-sized markets suddenly found themselves able to travel nonstop between many destinations, often at lower prices than those offered by the network airlines. 

And while point-to-point carriers traditionally avoided competing directly with dominant airlines at the major hubs, this also began to change. Southwest's arrival at Baltimore/Washington International (BWI) – a US Airways stronghold – in 1993, was a particularly illustrative example of the damage that low cost carriers could wreck on the hubs of their supposedly entrenched network competitors.

Watching its BWI market share erode at the hands of Southwest, US Airways in 1998 launched MetroJet, a single class, low-fare division that competed directly with Southwest on several of its Baltimore routes. But MetroJet cannibalized passengers from other big US Airways markets, particularly Washington National and Philadelphia, and it flew older, inefficient Boeing 737-200s. And despite union concessions, MetroJet had significantly higher operating overhead than Southwest. In a 2001 interview, then CEO Dave Siegel revealed that MetroJet's cost per available seat mile was 8 cents, compared with 6 cents at Southwest, and 10 cents at US Airways mainline. After the 9/11 terrorist attacks, US Airways shuttered the division and abandoned Baltimore, an airport where it had enjoyed a 69% market share ten years earlier.

Retrenching
Even as passenger traffic began to recover in the months following 9/11, most hub-and-spoke airlines continued to hemorrhage money.  Buckling under high oil prices and competition from low-cost carriers, they laid off over 100,000 workers, parked airplanes, and slashed capacity, closing and reducing the importance of some hubs in the process.
In February 2003, America West shuttered its small hub at Columbus, Ohio, reducing its service there from 49 daily flights to just four, and eliminating service to nine destinations. In November 2003, American Airlines halved the size of its St. Louis hub, an operation it inherited after purchasing Trans World Airlines (TWA) in 2000, cutting service from 417 to 207 daily flights.

Even more dramatic was Delta's pullout from DFW in January 2004. The airline reduced its daily departures there from 256 to 21 and laid off thousands of workers. In November of the same year, bankrupt US Airways downgraded its operations at Pittsburgh from a hub to a 'focus city', completing a service pull-down it initiated several years earlier. The month the service reduction took effect US Airways enplaned 436,569 passengers at Pittsburgh, a 21.5% drop from the same period in 2003. Today US Airways operates 235 daily departures to 65 destinations from Pittsburgh, down from a peak of over 500 daily departures to 110 destinations in the late Nineties.

David Whitaker, a vice president at the Columbus Regional Airport Authority, says America West's pullout was a major challenge for the airport. "America West's hub represented nearly a quarter of our daily departures," he explains, saying the number of passengers using the airport dropped by over half a million the year the hub was closed, and the impact could be felt across the airport. "The food and retail businesses, the car rental companies, they absolutely all suffered," says Whitaker. "It was difficult on everyone." In Pittsburgh, the airport closed its 25-gate commuter terminal following the US Airways retreat.

Looking to the future
Do these closures and downsizings herald the beginning of the end for hubs, and a return to a network focused on point-to-point flying? Swierenga thinks it's unlikely. "My view is that the situation has stabilized, and that there aren't likely to be other closures in the near future," he says. "I think in ten years, the landscape will look pretty similar." He points to the Airbus 380-800 as evidence that hubs are not likely to disappear, saying that with up to 555 seats, this type of aircraft is far more likely to be used for funneling passengers into a central hub than for point-to-point flying.

In fact, some hubs have been aggressively expanded. Since closing down its DFW hub and downsizing its Cincinnati, Ohio operation by 26% in 2005, Delta Airlines has added over 20 new destinations from its giant hub at Atlanta, 30 from Salt Lake City, and has launched a major expansion at New York-JFK that involves both international and domestic service.

At many US hubs, international expansion has been a growth driver. The number five US carrier, Continental Airlines, served 138 international destinations as of May 2006, almost all from hubs at Newark (New Jersey) and Houston (Texas), and international growth is a major part of Delta's plan to emerge from bankruptcy. During the summer 2006 season, the airline served over 60 international destinations from Atlanta, and 25 from JFK, which it is attempting to build into a major hub operation.

Swierenga points out that today even those airlines that have long resisted building traditional hubs include some very hub-like characteristics in their network. "Southwest may not be scheduling their flights to maximize connections, and may not be calling them hubs," he says of that airline's big operations at airports like Dallas/Love Field, Phoenix (Arizona), Nashville, and Baltimore. "But its certainly possible to use these airports as connection points."

Adjusting the model
While hub closures have remained infrequent, many airlines are aggressively pursuing ways to make their hubs more efficient. 'De-peaking' operations to create 'rolling hubs' – where arrivals and departures are spread out more evenly rather than clustered into banks – is one method. While it means that passengers must often wait longer for connecting flights, it allows airlines to more efficiently deploy staff and resources at hubs, thus reducing costs. American, United, and Delta are some of those that have embraced de-peaking.

And those airports where hubs have been downsized or closed are adjusting to their new reality. At Pittsburgh, Airtran Airways, JetBlue, and USA3000 Airlines have moved in to fill some of the vacuum left by the departure of US Airways, and Whitaker says that Columbus is on track to increase its passenger numbers in 2006. "The day after America West made their announcement, Delta came to us and announced 11 new daily flights from the airport," he says, adding that other carriers have upgraded their equipment on routes to and from Columbus and JetBlue recently arrived at Columbus, serving JFK and Boston.

"It's not easy when the dominant carrier leaves," he says. "But there is life after a hub.