"Costs and Concentration: Airport Hubs Are Changing"

Sunday, November 14, 2004

Costs and Concentration: Airport Hubs Are Changing  
By Michael Mecham  
Aviation Week & Space Technology


NATURAL SELECTION 

The art of the airport hub is changing. The basics--feeder routes funneling
passengers from small to larger airports--remain similar to the model that
emerged with U.S. deregulation in 1978. But the concept has gone global and
the competition has tightened. 

Everything is spreading worldwide: deregulated markets, discount airlines,
the mortality of network carriers and the growing importance of airline
alliances. That airlines need to reduce their costs to survive is old news;
that airports need to help their best client airlines survive and attract
new ones is less well known. 

This year's report on international air transport looks at the global
evolution of hubs--from the congestion and constraints they face in the U.S.
to the consolidation and alternative business models underway in Europe.
Asia's hubs have the blessings of continued growth and the economic pull of
the Chinese market, but face the uncertainties of a new phenomenon--discount
airlines. 

In a recent study, the Boston Consulting Group (BCG) sees nine "super hubs"
emerging in coming years, three on each major airline continent. Their
candidate cities are London, Frankfurt and Paris in Europe; Chicago,
Dallas-Fort Worth and Atlanta in the U.S.; and Tokyo, Singapore and Hong
Kong in Asia. 

The BCG's groupings are weighted to the home base of key members of the
three major alliances--Oneworld, SkyTeam and the Star Alliance. During the
terrorism-fed downturn of 2001-02, they accounted for 55% of global
passenger volumes and also racked up $20.4 billion in combined losses. 

"To reduce costs, each of the three main airline alliances is likely to
concentrate future long-distance passenger growth into one mega-hub in each
continent," the study's authors write. "These airports will have three key
characteristics: a central geographic location, a large and affluent
catchment area and a resident carrier that is both financially sound and a
major player in its respective alliance." 

That analysis ties the fortunes of airports directly to their airline
clients. More than three-fourths of the world's top 50 airports rely on a
single carrier for 40% or more of their traffic. The dependence rises to 80%
in some cases. 

"During the sharp downturn in volumes between 2000 and 2002, the smaller
'bottom-quartile' hubs in the U.S. lost 12.8% of their traffic, while the
larger top-quartile hubs suffered only a 6.3% fall," the BCG says. In fact,
the rankings of the largest airport hubs have changed little over the past
two decades. 

Airport consolidation is only increasing the importance of "fortress hubs"
for the major carriers--Delta Air Lines in Atlanta Hartsfield, having pulled
back from Dallas-Fort Worth, and British Airways in London Heathrow, having
pulled back from London Gatwick, are examples. 

"The legacy carriers are focusing on lucrative profits out of fortress
cities," says Fitch Ratings Director Douglas Kilcommons. "Legacy carriers
aren't abandoning hubs. They're concentrating on markets where they have the
best presence." 

WORRIED ABOUT THE EROSION of its passenger base to such low-fare competitors
as Southwest Airlines, United Airlines considered whether it was time to
adopt a point-to-point strategy, but rejected the idea as unworkable for its
passengers. As an example, United's chief operating officer, Pete McDonald,
points to Chicago O'Hare International Airport, where the carrier accounts
for nearly 50% of all flight activity. O'Hare has all four of United's
products--its mainline operations, international service, low-cost carrier
Ted for leisure markets, and the United Express commuter/regional service.
O'Hare becomes the essential way for United "to connect traffic to those
four products, which generates significant revenue for us," he says. "Let me
leave you in no doubt, the hub-and-spoke model is right for us and we will
continually optimize the network model. We believe in it and we plan to
stick with it." 

McDonald acknowledges United's need to cut costs, but doesn't believe the
hub-and-spoke system itself is the source of the carrier's problem. "We are
addressing the cost structure at United. We are well on our way to finishing
our work to get where it will let us be competitive with our hub-and-spoke
network." 

Across the border, United's fellow Star Alliance member, Air Canada, has had
its own legacy cost issues to overcome. It recently emerged from bankruptcy
protection as a reinvented low-cost carrier with a concentration on two
hubs, its main base at Toronto Pearson International Airport (which handles
40% of all Canadian flights) and Vancouver International Airport for Asian
connections. 

CEO Robert A. Milton expects smaller hubs to vanish as the industry
consolidates. But he doesn't expect hubs to rule air travel. Customer
preferences for point-to-point services will be the dominant flight pattern,
he says. To that end, Air Canada plans high-frequency services with regional
jets, which will double as feeder jets to Vancouver and Toronto to support
Air Canada's international flights. 

STRONG HUBS ARE ESSENTIAL for global alliances, but they're not the first
criterion for membership, says Rahsaan Johnson, a spokesman for SkyTeam
member Continental Airlines. "You think first of the region, then which
carriers to link with, then how to connect with them and where," he says. 

Continental's route to SkyTeam membership is traceable to long-standing
domestic alliances with Northwest Airlines, which had its own relations with
KLM Royal Dutch Airlines, which had been code-sharing with SkyTeam founder
Air France for more than a decade. "Continental and KLM elected to join
SkyTeam because of those connections," says Johnson. "Our European partners
had a lot of services to Continental and Northwest hubs." Examples include
Air France flights to Continental's hubs in Houston and Newark. 

For their customers, the international traveler, a carrier's membership in a
global alliance is a marketing tool. Common ticketing, common check-in
kiosks, common lounges, common gate areas and the ability to apply
frequent-flier awards earned on one carrier to another--all are designed to
take the complexity and inconvenience out of global travel, says Stephan
Nagel, director of Star Alliance's route network operations. And all require
well-functioning hub airports. 

But sometimes you have to make do. Japan is the world's second largest
economy, so it has obvious appeal for airlines. But its main gateway,
Tokyo's Narita International Airport, doesn't serve domestic routes.
Instead, they're connected by Haneda airport, which is a bus ride across
town. That's obviously not optimal for All Nippon Airways, Star Alliance's
Japanese partner. To optimize the situation, new domestic feeder flights
were added and Star members adjusted their international schedules. A
three-bank connecting system was initiated and dedicated Star Alliance
transfer buses begun. The result: the number of transfer passengers at
Narita from Haneda has doubled. 

One of the airline industry's biggest experiments concerns "de-peaking" or
"unbanking" hub flight schedules. 

"IF YOU THINK OF HUBS the way we traditionally thought of them a few years
ago," says Bernie DeSena, American Airlines vice president for O'Hare, "it
may make less sense. We think of them differently today." 

American, credited by many as leading the way among major carriers to
de-peak hub-and-spoke operations, started on that path two years ago (AW&ST
Sept. 2, 2002, p. 26). Depeaking attacks several issues: greater use of
personnel by spreading their workload throughout the day and "hopefully, a
more attractive schedule for the local customer." But there's a tradeoff.
"Your connection times are slightly elongated from where they were before,"
DeSena acknowledges. 

Nagel readily acknowledges the squeeze of coordinating arrival and departing
passengers in tight 2-hr. morning and afternoon peak periods. High staff
levels are needed to handle the rush of passengers and aircraft. And
connection times must be just long enough for passengers to make their next
flight but not so long as to keep aircraft waiting. 

He says de-peaked operations, sometimes called "rolling hubs," are best
applied where there's a large flight frequency. Airlines risk losing
passengers if their connection times are too long. "There's no clear answer
as to whether the rolling hub is a good hub," says Nagel. 

The market is showing that passengers aren't willing to pay the premiums
that once supported multiple opportunities for convenient connections, says
Bill Swelbar, president of Eclat Consulting. That's forcing the kind of hub
scheduling changes that American has applied at Dallas-Fort Worth and Delta
has recently begun implementing at Atlanta Hartsfield. Consumers are willing
to give up time on their schedule for the lower priced tickets that network
carriers can offer if they don't have to maximize scheduling connections. 

The de-peaking issue raises another question: how many hubs are needed? The
BCG study says lower tier hubs are under threat. One way hubs are adjusting
is to specialize: Iberia's home in Madrid has become its key linkage for
flights into South America rather than high-priced Miami. Austrian Airlines
saw an opportunity to connect the emerging countries of Eastern Europe
through its Vienna hub. 

But some think the world has too many hubs. "Why does a
Greensville-Spartanburg need 25 choices on how to get to LAX?" asks Swelbar,
thinking of the variety of connecting hubs he could use to fly from the
South Carolina city to Los Angeles International. "The market is not that
big." 

Aircraft technology is as much a part of the hub story as runways and
terminals. Boeing's emphasis on long-range, midsize jets has led it to
develop the 7E7 as a continent-hopping point-to-point specialist. Seeing
congestion and airport concentration, Airbus is focusing on development of
the industry's largest long-range transport, the A380. 

"There's no question that there are a very significant number of markets
that will always be served by some sort of aggregate approach," says Boeing
Vice President A. Kent Fisher, a market strategist. "Portland, Ore., and
Charlotte, N.C., are not likely to have a nonstop between them anytime
soon." 

What has changed over the past few years is the introduction of single-aisle
aircraft with transcontinental ranges. That means airlines no longer have to
fly a half-full large aircraft between major hubs simply because it has the
necessary range. Longer range aircraft also lengthen the spokes that may fan
out from a given hub, raking away any semblance of geographic order. So
American Airlines connects Orange County, Calif., out of DFW. The passenger
flying from Portland to Charlotte may choose to make a hub connection in
Philadelphia or DFW, all based on his individual schedule. 

Meanwhile, point-to-point connections among secondary cities will continue
to grow, particularly for international operations. Aircraft choices such as
the 7E7 will expand such possibilities, says Fisher. 

Airbus Vice President Laurent Rouaud, who heads its market forecasting, says
the shrinkage of hubs is the clue to the need for large aircraft to serve
concentrated markets. "The bulk of air travel is generated by a
concentration of population that is leading to concentrated networks," he
notes. 

He points to a Solomon Smith Barney study that concludes U.S. international
non-hub traffic is shrinking an average 4.9% a year, while hub traffic is
increasing 6.7% a year. 

AIRBUS SEES CONTINUITY where Boeing sees fragmentation in intercontinental
services. Currently, 20 top city pairs on the North Atlantic account for 40%
of all seat capacity, the same as 20 years ago. The transpacific routes are
dominated by 11 primary cities, as always. 

"Congestion is a reality today," says Rouaud, that requires big aircraft
feeding hubs. "In 2022, more than half of 1,170 very large transports will
be operated on flights serving the top 10 airports." 

So the question becomes, how many large aircraft will global carriers
require at these very large hubs? And how many smaller long-range aircraft
will they need to serve secondary cities in new regions or to provide
off-peak services for their main hubs? The two manufacturers' sales depend
on the answers; so will airlines' hub strategies. 

Christopher Fotos and Frances Fiorino in Washington, Pierre Sparaco in Paris
and Neelam Mathews in New Delhi contributed to this report. 

WORLD'S LEADING AIRPORTS--PASSENGERS/2003

Rank City (Airport) Total           %/Year 
                    Passengers      Change 
                    (2002 to 2003) 
1 ATLANTA (ATL)     79,086,792       2.9 
2 CHICAGO (ORD)     69,508,672       4.4 
3 LONDON (LHR)      63,487,136       0.2 
4 TOKYO (HND)       62,876,269       2.9 
5 LOS ANGELES (LAX) 54,982,838      (2.2) 
6 DALLAS/FORT WORTH AIRPORT (DFW) 
                    53,253,607       0.8 
7 FRANKFURT/MAIN (FRA) 
                    48,351,664      (0.2) 
8 PARIS (CDG)       48,220,436      (0.3) 
9 AMSTERDAM (AMS)   39,960,400      (1.9) 
10 DENVER (DEN)     37,505,138       5.2 
11 PHOENIX (PHX)    37,412,165       5.2 
12 LAS VEGAS (LAS)  36,285,932       3.6 
13 MADRID (MAD)     35,854,293       5.7 
14 HOUSTON (IAH)    34,154,574       0.7 
15 MINNEAPOLIS/ST. PAUL (MSP) 
                    33,201,860       1.8 
16 DETROIT (DTW)    32,664,620       0.6 
17 NEW YORK (JFK)   31,732,371       6.0 
18 BANGKOK (BKK)    30,175,379      (6.2) 
19 LONDON (LGW)     30,007,021       1.3 
20 MIAMI (MIA)      29,595,618      (1.5) 
21 NEWARK (EWR)     29,431,061       0.7 
22 SAN FRANCISCO (SFO) 
                    29,313,271      (6.8) 
23 ORLANDO (MCO)    27,319,223       2.5 
24 HONG KONG (HKG)  27,092,290     (20.0) 
25 SEATTLE (SEA)    26,755,888       0.2 
26 TOKYO (NRT)      26,537,406      (8.1) 
27 ROME (FCO)       26,284,478       3.7 
28 SYDNEY (SYD)     25,333,508       4.0 
29 TORONTO (YYZ)    24,739,312      (4.6) 
30 PHILADELPHIA (PHL) 
                    24,671,075      (0.5) 

Note: Total passengers arriving, departing and direct transit passengers
counted once.
Source: Airports Council International, total passengers

WORLD'S LEADING AIRPORTS--MOVEMENTS/2003

Rank City (Airport)  Total         %/Year  
                     Movements     Change 
                     (2002 to 2003) 
1 CHICAGO (ORD)      928,691        0.6 
2 ATLANTA (ATL)      911,723        2.4 
3 DALLAS/FORT WORTH AIRPORT (DFW) 
                     765,296        0.0 
4 LOS ANGELES (LAX)  622,378       (3.6) 
5 PHOENIX (PHX)      541,771       (0.7) 
6 PARIS (CDG)        515,025        1.0 
7 MINNEAPOLIS/ST. PAUL (MSP) 
                     510,382        0.5 
8 CINCINNATI HEBRON (CVG) 
                     505,557        3.9 
9 LAS VEGAS (LAS)    501,029        0.8 
10 DENVER (DEN)      499,794        1.0 
11 DETROIT (DTW)     491,073        0.0 
12 HOUSTON (IAH)     474,913        4.0 
13 LONDON (LHR)      463,650       (0.6) 
14 VAN NUYS (VNY)    460,734       (7.6) 
15 FRANKFURT/MAIN (FRA) 
                     458,865        0.1 
16 PHILADELPHIA (PHL) 
                     446,529       (3.6) 
17 CHARLOTTE (CLT)   443,394       (2.7) 
18 MIAMI (MIA)       417,423       (6.5) 
19 AMSTERDAM (AMS)   408,300       (2.1) 
20 NEWARK (EWR)      405,808       (0.0) 
21 MEMPHIS (MEM)     402,258        0.9 
22 SALT LAKE CITY (SLC) 
                     400,452       (1.6) 
23 PHOENIX (DVT)     389,307       (0.1) 
24 SANFORD (SFB)     385,303        3.2 
25 MADRID (MAD)      383,804        4.3 
26 ST. LOUIS (STL)   379,772      (13.1) 
27 NEW YORK (LGA)    374,952        3.5 
28 BOSTON (BOS)      373,304       (4.8) 
29 TORONTO (YYZ)     370,996       (3.2) 
30 PITTSBURGH (PIT)  361,329      (15.0) 

Note: Movements are landings and takeoffs of aircraft.
Source: Airports Council International, aircraft movements


 Do you have an opinion about this story?
Share it with other readers in our CAA Discussion Forums

http://www.californiaaviation.org/dcfp/dcboard.php


*****************************************