17.10.2013 Views

The Airport Choice of Freighter Operators in Europe

The Airport Choice of Freighter Operators in Europe

The Airport Choice of Freighter Operators in Europe

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>The</strong> <strong>Airport</strong> <strong>Choice</strong> <strong>of</strong> <strong>Freighter</strong> <strong>Operators</strong> <strong>in</strong><br />

<strong>Europe</strong>:<br />

Insights from a Discrete <strong>Choice</strong> Analysis<br />

Franziska Kupfer<br />

Department <strong>of</strong> Transport and Regional Economics<br />

University <strong>of</strong> Antwerp<br />

Work <strong>in</strong> cooperation with:<br />

P. Goos, R. Kessels, E. Van de Voorde, A. Verhetsel<br />

GARS Air Cargo Conference<br />

– Wildau, 20 October 2011


1. Introduction<br />

2. Previous Research<br />

3. <strong>The</strong> Discrete <strong>Choice</strong> Approach<br />

4. Conclusions<br />

Overview<br />

1


<strong>Europe</strong>‘s Top 20 cargo airports 2010<br />

Introduction<br />

Own composition based on ACI data<br />

3


<strong>The</strong> <strong>Airport</strong> <strong>Choice</strong> <strong>of</strong> <strong>Freighter</strong><br />

<strong>Operators</strong> <strong>in</strong> <strong>Europe</strong><br />

All cargo airl<strong>in</strong>es, cargo<br />

subsidiaries, cargo divisions<br />

with freighter operations<br />

to/from <strong>Europe</strong>, with the<br />

exeption <strong>of</strong> <strong>in</strong>tegrators<br />

<strong>Airport</strong> choice concerns<br />

choice for scheduled<br />

freighter operations <strong>in</strong><br />

<strong>Europe</strong><br />

Introduction<br />

Source: Lufthansa Cargo<br />

4


1. Introduction<br />

2. Previous research<br />

3. <strong>The</strong> Discrete <strong>Choice</strong> Approach<br />

4. Conclusions<br />

Overview<br />

5


Previous Research<br />

Studies with <strong>in</strong>formation about airport choice<br />

ma<strong>in</strong>ly deal with the airport choices <strong>of</strong> passenger<br />

airl<strong>in</strong>es (see for example Huston, Butler, 1991 and Berechman, Wit, 1994) or<br />

the hub choice (Watanabe et al, 2009, Ohashi et al., 2005, Martín, Román,<br />

2004)<br />

Methodologies used <strong>in</strong> literature: Weber Least<br />

Cost Model (Wantanabe et al, 2009; Dennis, 1994) or general<br />

surveys (see e.g. Gard<strong>in</strong>er, Ison, 2008)<br />

Series <strong>of</strong> studies look<strong>in</strong>g at airl<strong>in</strong>e choice <strong>of</strong><br />

passengers (Blackstone, Buck, Hakim, 2006; Pels, Nijkamp, Rietfeld, 2000;<br />

Hess, Adler, Polak, 2007)<br />

6


<strong>Airport</strong> choice factors<br />

Restrictions<br />

Market factors<br />

Cost factors<br />

Time factors<br />

Strategic factors<br />

(Perception) <strong>of</strong> airport quality<br />

Other factors<br />

Previous Research<br />

Source: Cargolux<br />

7


1. Introduction<br />

2. Previous research<br />

3. <strong>The</strong> Discrete <strong>Choice</strong> Approach<br />

3.1. Experimental Setup and Data<br />

3.2. Results<br />

4. Conclusions<br />

Overview<br />

8


Prelim<strong>in</strong>ary Interviews:<br />

Experimental Setup and Data<br />

5 airl<strong>in</strong>es and 2 airports<br />

Aim: narrow down<br />

factors/attributes identified <strong>in</strong><br />

literature review<br />

Rat<strong>in</strong>g <strong>of</strong> choice attributes<br />

used as prior <strong>in</strong>formation for<br />

choice design<br />

Def<strong>in</strong>ition <strong>of</strong> factor levels<br />

Source: Korean Air Cargo<br />

9


Experimental Setup and Data<br />

Orig<strong>in</strong>-Dest<strong>in</strong>ation demand (potential revenue)<br />

* 20% less orig<strong>in</strong>-dest<strong>in</strong>ation demand<br />

* 10% less orig<strong>in</strong>-dest<strong>in</strong>ation demand<br />

* Equal orig<strong>in</strong>-dest<strong>in</strong>ation demand<br />

* 10% more orig<strong>in</strong>-dest<strong>in</strong>ation demand<br />

* 20% more orig<strong>in</strong>-dest<strong>in</strong>ation demand<br />

Night time restrictions<br />

* No night time restrictions<br />

* Limited or very expensive night time slots<br />

* Night time flight prohibitions<br />

Presence <strong>of</strong> forwarders<br />

* No forwarders present<br />

* Only major forwarders present<br />

* Broad range <strong>of</strong> forwarders present<br />

<strong>Airport</strong> charges (<strong>in</strong>clud<strong>in</strong>g handl<strong>in</strong>g)<br />

* 20% lower airport charges<br />

* 10% lower airport charges<br />

* Equal airport charges<br />

* 10% higher airport charges<br />

* 20% higher airport charges<br />

Presence <strong>of</strong> passenger airl<strong>in</strong>es<br />

* No passenger airl<strong>in</strong>e operations at airport<br />

* Only passenger operations <strong>of</strong> own<br />

airl<strong>in</strong>e/group or <strong>of</strong> ma<strong>in</strong> passenger airl<strong>in</strong>e partner<br />

at airport<br />

* Different passenger airl<strong>in</strong>e operations from<br />

own airl<strong>in</strong>e/group as well as other airl<strong>in</strong>es<br />

<strong>Airport</strong> experience with cargo<br />

* No experience with cargo<br />

* Limited experience with cargo<br />

* Extended experience with cargo<br />

10


Experimental Setup and Data<br />

Partial pr<strong>of</strong>ile design:<br />

To deal with a large number <strong>of</strong> attributes<br />

To reduce the complexity <strong>of</strong> the choice task so<br />

as to overcome respondent fatigue and<br />

<strong>in</strong>consistent choices<br />

To rule out lexicographic choice behavior, that<br />

is, people who make choices based on the<br />

levels <strong>of</strong> just one attribute or a small subset <strong>of</strong><br />

the attributes<br />

11


Experimental Setup and Data<br />

2-stage design algorithm:<br />

Stage 1: Determ<strong>in</strong>ation <strong>of</strong> the<br />

constant attributes <strong>in</strong> each<br />

choice set<br />

Stage 2: Determ<strong>in</strong>ation <strong>of</strong> the<br />

levels <strong>of</strong> the non-constant<br />

attributes<br />

Source: Avient<br />

12


<strong>Choice</strong> Design:<br />

Experimental Setup and Data<br />

As orig<strong>in</strong>-dest<strong>in</strong>ation demand was predicted to<br />

be very dom<strong>in</strong>ant: 40 choice sets without<br />

orig<strong>in</strong>-dest<strong>in</strong>ation demand<br />

2 x 40 choice sets (four blocks <strong>of</strong> 10)<br />

Each respondent had to make 20 (+1) choices<br />

13


Experimental Setup and Data<br />

14


Sample:<br />

Experimental Setup and Data<br />

26 airl<strong>in</strong>es (32 completed<br />

surveys)<br />

More than 50% <strong>of</strong> population<br />

From which:<br />

11 <strong>in</strong>dependent cargo carriers<br />

15 subsidiaries or cargo<br />

divisions <strong>of</strong> comb<strong>in</strong>ation<br />

carriers<br />

Source: Air France/KLM Cargo<br />

15


MNL MNL noPax MNL noPax<br />

Cl<strong>in</strong> DL<strong>in</strong><br />

parameter t-test parameter t-test parameter t-test<br />

estimate<br />

estimate<br />

estimate<br />

Night time<br />

no restrictions 0.27563 0.2772 0.2777<br />

limited 0.00937 0.1 0.0178 0.18 0.0163 0.17<br />

prohibitions<br />

Experience<br />

-0.285 -2.17** -0.295 -2.24** -0.294 -2.25**<br />

extended 0.6 0.587 0.592<br />

limited -0.174 -1.82 -0.164 -1.74 -0.169 -1.8<br />

no<br />

Forwarders<br />

-0.426 -3.22* -0.423 -3.21* -0.423 -3.21*<br />

broad range 0.754 0.758 0.759<br />

major 0.426 4.28* 0.422 4.25* 0.421 4.24*<br />

no<br />

Passenger<br />

-1.18 -8.59* -1.18 -8.63* -1.18 -8.61*<br />

different 0.1979<br />

no -0.11 -1.1<br />

sibl<strong>in</strong>g -0.0879 -0.9<br />

Demand 0.0325 3.63<br />

20% 0.69528 0.68948<br />

10% 0.283 1.38 0.294 1.45<br />

equal 0.00672 0.03 0.00152 0.01<br />

-10% -0.387 -2.05** -0.383 -2.04**<br />

-20% -0.598 -2.78* -0.602 -2.8*<br />

Charges -0.0294 -5.11*<br />

-20% 0.5373 0.53673<br />

-10% 0.384 2.85* 0.384 0.384*<br />

equal 0.0117 0.09 0.00527 0.04<br />

10% -0.379 -2.72* -0.362 -2.63*<br />

20% -0.554 -3.83* -0.564 -3.89*<br />

Results<br />

* significant at 1%<br />

** significant at 5%<br />

17


Results<br />

18


Forwarders<br />

Charges<br />

Experience<br />

Demand<br />

Night time<br />

0 20 40 60 80 100<br />

L-R ChiSquare<br />

Results<br />

MML MML noPax Cl<strong>in</strong>Dl<strong>in</strong><br />

DF ChiSquare Prob>ChiSq ChiSquare Prob>ChiSq<br />

Night time 2 6.291 0.043** 6.62 0.037**<br />

Experience 2 27.625 < 0.0001* 27.454 < 0.0001*<br />

Forwarders 2 91.854 < 0.0001* 92.45 < 0.0001*<br />

Passenger 2 3.761 0.1525<br />

Demand 4 13.396 0.0095* 13.316 0.0098*<br />

Charges 4 28.913 < 0.0001* 28.106 < 0.0001*<br />

* significant at 1%<br />

** significant at 5%<br />

Effects likelihood ratio tests<br />

19


Results<br />

20


Conclusions<br />

Studies concern<strong>in</strong>g airport choice ma<strong>in</strong>ly look at hub<br />

choice and choices <strong>of</strong> passenger airl<strong>in</strong>es<br />

Analysis <strong>of</strong> factors br<strong>in</strong>gs more <strong>in</strong>formation on what is<br />

important for airl<strong>in</strong>es<br />

Discrete choice approach br<strong>in</strong>gs new <strong>in</strong>sights to problem<br />

First results show that the presence <strong>of</strong> forwarders is the<br />

most important attributes and passenger operations are<br />

not significant<br />

Further research:<br />

* more advanced models (MMNL)<br />

* <strong>in</strong>tegrate socio-economic factors<br />

* calculate trade-<strong>of</strong>fs<br />

21


Franziska Kupfer<br />

Universiteit Antwerpen<br />

Pr<strong>in</strong>sstraat 13 – SB 408<br />

2000 Antwerp<br />

Franziska.Kupfer@ua.ac.be<br />

+32 3 265 41 87<br />

Contact<br />

Source: www.kenlawrence.eu<br />

22

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!