Multi-Airport Systems in Era of Low-Cost Carriers Dr. Richard de Neufville

Professor of Engineering Systems and of Civil and Environmental Engineering Massachusetts Institute of Technology Airport Systems Planning RdN

Theme

  

“Low cost” airlines are developing a “parallel network” of travel “network choice” (rather than “airport choice”) may determine traffic in multi-airport systems Competition between “low cost” and “legacy” airlines leading to struggle between “low cost” and “legacy” hubs

 Boston/Logan vs. Boston/Providence, etc., etc. Airport Systems Planning RdN

What is a Multi-Airport System?



the significant airports serving transport in a metropolitan region, without regard to ownership or political control



Ex: Boston, Providence, Manchester



Ex: Copenhagen (Denmark), Malmo (Sweden)



Discussion



This is reality for travelers



Contrasts with focus on ownership (as used by ACI – Airports Council International, and others)

Airport Systems Planning RdN

Discussion of Tables 1 – 4 derived from de Neufville database

  

Table 1: Except for very crowded situations, secondary airports are a small fraction of total at airport with most passengers.

 This reflects concentration factor …

Table 2: Catalog of Major Secondary Airports Table 3: MAS ranked by Estimated Number of Originating Passengers, eliminating transfers

  This focuses on a Major Driver of MAS

Table 4: Potential MAS, mostly driven by Low-Cost airlines.

 New ones constantly emerging… Airport Systems Planning RdN

Planning Issue



Many ‘mistakes’ in planning multi airport systems

 Washington/Dulles – planned as major DC airport, but had only ~ 3 MAP for 20 years  London/Stansted – similar story – only developed with Ryanair hub around 2002  Osaka/Kansai – Osaka/Itami did not close  Montreal/Mirabel – huge airfield, now “closed” to passenger traffic  Et cetera… Airport Systems Planning RdN

Why mistakes happened



Failure to appreciate traffic concentration at primary airports



… Because planners/forecasters using wrong mental model

Airport Systems Planning RdN

What drives traffic allocation in Multi-Airport System?



Airline competition has been primary



S-shaped market share/frequency share Market Share Frequency Share



Drives airlines to

 Match flights => Allocate flights to major markets  Concentrate Traffic at primary airports Airport Systems Planning RdN

Right model: “Concentration” not “Catchment Areas”



Concentration is standard urban phenomenon

 e.g.: financial, jewelry, etc. districts 

Driven by what suppliers offer



Customers choose which location (airport) depending on where they find what they need -- not just most convenient facility

Airport Systems Planning RdN

“Concentration” persists - until high level of local traffic



When local originating traffic high…



More flights add little at major airports



Airlines place flights at second airports



There appears to be a ‘threshold”…



Currently ~ 16 million originations/year



Note: higher as “average” aircraft carries more passengers, larger aircraft or higher load factor

Airport Systems Planning RdN

New Reality: No-frill airlines setting up “parallel network”



Low cost carriers “parallel” majors



Major fare distinctions



Ticket distribution separate

 Internet direct to users, ‘no’ travel agents 

Parallel service between cities

 Providence/Baltimore not Boston/Washington 

‘No’ interlining of bags, tickets



‘Not’ in Reservation systems

Airport Systems Planning RdN

Metropolitan areas with significant multi-airport systems Metropolitan Region London New York Tokyo Los Angeles Paris Chicago Hong Kong Miami San Francisco Washington/Baltimore Shanghai Seoul Moscow Osaka Traffic in Millions For Region Originating 137 106 99 85 84 100 67 59 57 61 43 42 37 38 52 43 42 37 32 30 27 25 24 21 19 19 18 17 de Neufville data base for 2007 Multi-Airport System Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

Airport Systems Planning RdN

Metropolitan areas with significant multi-airport systems Metropolitan Region Bangkok Beijing Boston Sao Paulo Milan Manchester (UK) Las Vegas Taipei Barcelona Amsterdam Frankfurt Mexico City Atlanta Jakarta Traffic in Millions For Region Originating 42 49 36 35 37 35 45 33 35 48 57 36 84 32 17 17 17 17 16 16 16 15 15 15 15 15 15 15 de Neufville data base for 2007 Multi-Airport System Yes Yes Yes Yes Yes, low-cost Yes Yes, low-cost Yes, low-cost Yes, low-cost Yes

Airport Systems Planning RdN

Major exceptions to rule: technical or political



Until recently, major exceptions to concentration rule were:



Technical -- runways too short

 Belfast, Belo Horizonte, Buenos Aires, Rio de Janeiro, Taipei 

Political -- or military...

 Berlin, Dusseldorf/Bonn, Glasgow, Moscow Airport Systems Planning RdN

New Reality: No-frills choose different airports



Southwest, Westjet (Canada), Ryanair and Easyjet (UK) require:

 Cheap properties, no Taj Mahals (compare San Francisco/International and Oakland; London/Gatwick and Luton)  Low congestion and delays  Flexible work force 

They find this at aggressive, ‘hungry’ airports -- not in major facilities

Airport Systems Planning RdN

New Reality: US/Canada Network of Low-Cost Carrier Airports

Metropolitan Region Boston Dallas/Ft. Worth Houston Los Angeles Miami San Francisco Vancouver Secondary Airport Manchester Providence Love Field Hobby Long Beach Ft. Lauderdale Oakland Abbotsford Low-Cost Carrier Southwest Southwest Southwest Southwest Jet Blue Southwest Southwest Westjet

Airport Systems Planning RdN

New Reality: Europe Network of Low-Cost Carrier Airports

Metropolitan Region Brussels Copenhagen Düsseldorf Frankfurt Glasgow London Manchester Milan Oslo Paris Rome Stockholm Secondary Airport Charleroi Malmo Weeze Köln / Bonn Hahn Prestwick Luton Stansted Liverpool Orio al Serio Torp Beauvais Ciampino Skvasta Low-Cost Carrier Ryanair Ryanair Ryanair easyjet Ryanair Ryanair easyjet Ryanair easyjet Ryanair Ryanair Ryanair Ryanair + easyjet Ryanair

Airport Systems Planning RdN

Multi-Airport Systems in Brazil

Metropolitan Region Sao Paulo Rio de Janeiro Belo Horizonte International, Distant Name Traffic Millions Garulhos Galeao Confins 15.8 8.9 3.7 Airport Domestic, Close-in Name Traffic Millions Congonhas Santos Dumont Pampulha 18.5 3.6 0.8 Source: Infraero; de Neufville database 2007

Airport Systems Planning RdN

Importance of Parallel Network of close-in Brazilian airports

Airport Pair Passengers, 1000s Congonhas Congonhas Congonhas Congonhas Santos Dumont

Brasilia

Pampulha

Curitiba

Congonhas

Garulhos

Santos Dumont Santos Dumont

Porto Allegre Salvador Brasilia

Pampulha

Source: INFRAERO, 2002, Rabbani, 2002 1461 596 565 551 365 364 325 312

Rank

1 2 3 4 5 6 7 8 Airport Systems Planning RdN

Implications for modelling future of second airports



A new driver for second airports...

 Low cost carriers often ‘not’ competing at big airports  Frequency competition does not drive growth pattern of secondary airports 

Competition between networks may be primary…



… followed by catchment area model of airport choice

Airport Systems Planning RdN

Implications for future of second airports



No frills airlines are becoming ‘major’

 Southwest 2nd largest airline in world (pax)  Market Cap ~ 10 billion $ > any other pax airline  Ryanair Market Cap greater than British Airways 

Majors are shrinking (UAL, USAir, etc.)



Implies that Primary airports will lose significant traffic to second airports



This is already happening!!!

Airport Systems Planning RdN

Southwest entry in Boston

Since 1996, the Regional Airports H ave Captured M ore than 75% of the Region’s Air Passenger Growth to Regional Airports along with growth of Southwest at Providence and Manchester (NH) Distribution of New England Passenger Growth

1990 –1996 1996 –2000

Logan 77% (+2.3M) Logan 24% (+2.3M) Regional Airports 76% (+7.2M) Regional Airports 23% (+0.7M)

Regional airports include Providence, Manchester, W orcester, Ban gor, Burlington, Hartford, New Haven, and Portland.

Source: Airport Records and US DOT, Form 41 schedules.

Source: Louis Berger, New England Regional Aviation System Plan materials

Page 3

Airport Systems Planning RdN

Market Share of Boston/Logan is in decline

The Region is Less Reliant on Logan Airport Figure 2: The Boston/Logan traffic share dropped by a quarter over the past 20 years; half of this occurred with the Southwest growth in the late 1990s at Providence and Manchester (NH)

80%

78%

70% 60%

59% The 2004 Share is about 57% (SH&E, ’05)

50%

'80 '81 '82 '83 '84 '85 '86 '87 '88 '89 '90 '91 '92 '93 '94 '95 '96 '97 '98 '99 '00

Note: Includes enplaned passengers at Logan, Hartford/Bradley, T .F. Green/Providence, Manchester, Portland, Burlington, Bangor, Tweed New Haven, and Worcester.

Source: US DOT, Form 41 and Part 298/C. Airport records for Loga n and various regional airports.

Source: Louis Berger New England Regional Aviation System Plan

Airport Systems Planning RdN

Page 4

Summary



A new, parallel air transport network is emerging to compete with majors



This low-cost carrier network may become a major feature of industry



It implies growth and importance of low-cost second airports throughout North America, Europe -- and perhaps elsewhere

Airport Systems Planning RdN

Supplemental Comment



Meanwhile, a similar development is taking place in air cargo



Fedex and UPS are developing their own networks of cargo airports



Fedex: Memphis, Toronto/Hamilton, San Francisco/ Oakland, etc.



UPS: Louisville, Los Angeles/Ontario. Chicago/Rockford, etc.

Airport Systems Planning RdN