Regional Airports

Regional Airports

WITPRESS WIT Press publishes leading books in Science and Technology. Visit our website for the current list of titles. www.witpress.com

WITeLibrary Home of the Transactions of the Wessex Institute, the WIT electronic-library provides the international scientific community with immediate and permanent access to individual papers presented at WIT conferences. Visit the WIT eLibrary at http://library.witpress.com

This page intentionally left blank

Regional Airports Editors: M. Nadia Postorino University of Reggio Clabria, Italy

M. Nadia Postorino University of Reggio Clabria, Italy

Published by WIT Press Ashurst Lodge, Ashurst, Southampton, SO40 7AA, UK Tel: 44 (0) 238 029 3223; Fax: 44 (0) 238 029 2853 E-Mail: [email protected] http://www.witpress.com For USA, Canada and Mexico WIT Press 25 Bridge Street, Billerica, MA 01821, USA Tel: 978 667 5841; Fax: 978 667 7582 E-Mail: [email protected] http://www.witpress.com British Library Cataloguing-in-Publication Data A Catalogue record for this book is available from the British Library ISBN: 978-1-84564-570-0 Library of Congress Catalog Card Number: 2011922776 The texts of the papers in this volume were set individually by the authors or under their supervision. No responsibility is assumed by the Publisher, the Editors and Authors for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. The Publisher does not necessarily endorse the ideas held, or views expressed by the Editors or Authors of the material contained in its publications. © WIT Press 2011. Printed in Great Britain by Quay Digital, Bristol The material contained herein is reprinted from a special editions of Sustainable Development and Planning, Vol.5, No.2, published by WIT Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the Publisher.

CONTENTS Preface ............................................................................................................................................. vii Editorial ............................................................................................................................................. 1 Exploring multi-criteria decision analysis method as a tool to choose regional airport hubs within Africa B. SSAMULA ................................................................................................................................... 3 Airport–airline relationships: opportunities for Italian regional airports S. CEPOLINA & G. PROFUMO .................................................................................................... 19 Potential demand for new high speed rail services in high dense air transport corridors C. ROMÁN & J.C. MARTÍN .......................................................................................................... 35 Regional airports and the accessibility of mountain areas: networks, importance and contribution to development X. BERNIER ................................................................................................................................... 51 Analysis of the regional air passenger transport system in Brazil: some aspects of its evolution and diagnosis S.C. RIBEIRO, C.C.L. FRAGA & M.P.S. SANTOS ...................................................................... 63 Regional airports’ environmental management key messages from the evaluation of ten European airports D.J. DIMITRIOU & A.J. VOSKAKI .............................................................................................. 73 Sustainable logistics platform in a regional Brazilian airport O.F. LIMA Jr, E.W. RUTKOWSKI, C.C. de CARVALHO & J.C.F. LIMA .................................. 87 Regional airport: study on economic and social profitability S. AMOROSO & L. CARUSO ....................................................................................................... 99 Assessment of air pollution from Tehran-Mehrabad airport, Iran G. BADALIANS GHOLIKANDI, M. LASHKARI, H.R. ORUMIEH, H.R. TASHAOUIE & S. HADDADI ............................................................................................................................ 109 Environmental effects of airport nodes: a methodological approach M.N. POSTORINO ....................................................................................................................... 117 Architectural design standards for Muslims prayer facilities in airports A. MOKHTAR .............................................................................................................................. 131

This page intentionally left blank

PREFACE Regional Airports have become increasingly important elements of the air network system, both as feeders of hub-and-spoke services and as origins or destination of point-to-point services. Congestion at the main hubs and increasing demand for air transportation - both for passengers and freight services – necessitates revaluation of the overall air systems, with regional airports taking an ever expanding role. Optimisation of air transportation systems within the framework of other forms of transport plays an important part in the present quest for sustainability. Congestion nowadays is not only associated with countries such as the USA and those in the EU, but also a variety of other countries with fast developing economies where there is a strong increase in air transportation demand. The revolution of the existing airport system, including regional airports requires the developing of new optimisation tools which can simulate the whole process and produce optimal solutions. These models are also essential to predict future demands and, in particular the role that regional airports will play. The siting of new airports involves taking into consideration a variety of environmental, ecological, social and economic factors which transcend the problem of transportation resources optimisation itself. Regional Airports can be a powerful driving force behind the development of an area and conversely can result in major problems if they are wrongly sited.

Carlos A. Brebbia Director Wessex Institute of Technology The New Forest, 2011

This page intentionally left blank

1

EDITORIAL The papers selected for this book are extended versions of those presented at an International Seminar on Regional Airports organized by the Wessex Institute of Technology (UK) and the University of Reggio Calabria (Italy). The idea of convening a meeting on Regional Airports was prompted by the need to review the specific policies set up to promote their development in different countries. What is the role of regional airports in the current air transport framework? The answer is not simple and requires looking at the extraordinary growth of air transport all over the world during the past few decades. Air t ransport is faster than other transport modes, capable of transporting passengers and goods to the most remote countries. It can also be an engine to promote the economic development of a region. Airline deregulation started first in the USA, followed by the EU and afterwards in other countries, has further facilitated the growth of air traffic. Air transport has become accessible to more people thanks to increased frequencies and lower fares partly resulting from competition among airlines. The next step, ie airport deregulation, is expected to produce still more benefits to all the main stakeholders, ie airlines, passengers and airports themselves. This remarkable growth has also had some negative effects over the years such as high congestion levels both at attractive airports and along routes that, in turn, mean poor air service performances; increased environmental impacts (noise and atmospheric pollution), mainly due to the larger numberof flights rather than to the technological characteristics of aircraft which have become much more efficient in terms of fuel consumption and pollutant emissions. Those environmental impacts contribute at world level to the overall greenhouse effect while at the local level they can reduce the quality of life for communities located in the airport neighborhood. Regional airports have an important role to play in the air network and can contribute to finding sustainable transport systems. They help, for instance, to spread air traffic thus avoiding high congestion levels along a small number of routes or at hub nodes. Past experiences can help the development of better regional airports by taking into account apriori the effects of air traffic on the environment and on the surrounding communities. Recent EU rules, for instance, place important constraints in terms of acoustic pollution at an airport. The EU also encourages studies to estimate the airport carbon footprint at local scale. This publication collects papers dealing with regional airport topics such as environmental impacts and their evaluation, airport economics and location, airport network configuration, travelerair choices and regulatory tendencies, but also social and cultural aspects not always discussed in the literature. Regional airports can improve the air transport system and if properly sited can be a powerful engine for the development of an area. M Nadia PostorinoUniversity of Reggio Calabria, Italy

This page intentionally left blank

Regional Airports

EXPLORING MULTI-CRITERIA DECISION ANALYSIS METHOD AS A TOOL TO CHOOSE REGIONAL AIRPORT HUBS WITHIN AFRICA B. SSAMULA Built Environment, CSIR, South Africa.

ABSTRACT The aviation industry in the African region, in order to compete with the global market, is exploring and actively pursing expansion either through strategic alliances or by adopting the hub and spoke (H&S) network development. Hubbing has the major benefits of consolidating passengers, thus increasing frequency of travel while increasing accessibility and improving the economies of scale to operate the service. The African region is vast and is characterised by sparse passenger demand; therefore, the decisions made in locating airports as hubs pose a challenge. This paper aims to explore the use of the multi-criteria decision analysis (MCDA) tools as a method of choosing hub airports. The reason why the MCDA tools are most appropriate is because choosing a hub airport is a complex decision which has to take into account various issues like: network costs, infrastructure costs, security, economic viability, safety, passenger travel time expenditure, etc. The various tools, processes and methodologies used in decision making theory are explored and applied in order to choose hub airports with the lowest transport costs in an efficient H&S network. The major findings in this study show that because Africa has a sparse network, with a few role players, the choice in hub location options relies greatly on the cost of routing passengers through the hub airport. MCDA is shown to be a useful tool whose only limitation is maintaining the uniformity of weighting the criteria for the whole region. Keywords: Africa aviation, airports, decision analysis, hub location, network design.

1 AFRICA AS A SPARSE REGION Africa is a large continent of 30 million km2, with dimensions three times the size of Europe and distances from the south to the north of about 8,000 km. Although the population of the continent is over 860 million, the average population density of Africa is 28 inhabitants per km2, which falls below the world’s average population density of 44 persons per km2, thus defining the continent as a sparsely populated region. In 2002, Africa’s population comprised 13% of the world’s total, with Africa’s air passenger traffic contributing only 4.1% to the world’s total air passenger traffic, making it the smallest region for air services worldwide [1]. The passenger data used in this study show that the annual number of air trips per inhabitant in Africa is equal to only 0.14. The low demand for air travel within the region is due to the fact that it is an expensive unaffordable means of travel. Furthermore, the load factor, which is the ratio of the revenue passenger kilometres (RPK) to the available seat kilometres (ASK), is one of the critical determinants of profitability in relation to the breakeven load factor. Fig. 1 shows that the African region has the lowest load factor at 62.56%, compared with other regions of the world. The Far East and Pacific regions have relatively high load factors, averaging 76.32%. The low load factors are a reflection of the scanty routes in the African region. The routes are scanty because of the much higher air fares compared with those in other regions of the world and because of a relatively poor population, hence the sparse travel demand on the continent. Even though Africa shows great potential with initiatives like the open skies initiative adopted in the Yamoussoukro Decision, under the auspices of the African Union and New Partnership for Africa’s Development (NEPAD), the challenge will be to make air travel affordable and accessible, so as to improve trade and tourism through the economic impact of air transport on countries.

4

Regional Airports

1.1 Hub and spoke network design Button et al. [2] state that in order to minimise costs and keep airfares down, airlines need to keep aircraft in the air for the longest possible time to achieve the highest possible load factor, and to coordinate their aircraft, crew and maintenance schedules. To achieve this, many airlines operate hub and spoke (H&S) networks which entail consolidating traffic from a diverse range of origins, destined for a diverse range of final destinations at hub airports. H&S networks involve the collection of passengers from their origin (i), transferring them through hubs nodes (k and l) and then distribution of passengers to their destination node j as illustrated in Fig. 2. The advantages of hubbing for routes with low passenger demand are very apparent. A traditional airline would not serve these routes because the operating costs needed to meet the low demand make them unprofitable. Accessibility within the continent would actually increase with hubbing 90% 80%

73%

75.60%

76.32%

70%

65.23%

63.43%

62.56%

Asia & Middle East

Latin America

Africa

60% 50% 40% 30% 20% 10% 0% USA

Europe

Far East & Pacific

Figure 1: Load factors for world regions. Source: Chingosho [1].

Figure 2: Schematic representation of the hub and spoke network design. Source: CMISRO [3].

Regional Airports

5

due to the fact that the flight frequency of the airlines would increase, which is an advantage to users of the service because they have more options. The hub network allows flexibility of planning and operations for the service provider, with adequate utilisation of aircraft on routes with reasonably high load factors, yielding profitability in a market of scarce passenger demand [4]. The methodology involved in the design of H&S network involves, first, locating the hub airports through which all the flow will pass. Once the hubs have been located, the nodes are allocated to the hubs using single assignment to the closest hub. Thereafter, the pattern of flow for the network is established and the passenger numbers along each link are calculated. The network is then costed by calculating the cost of transferring all the passengers from their origins to their destinations through the hub links. The crucial step in H&S network design is the hub location problem. Boland et al. [5] define the hub location problem as one concerned with creating H&S networks, which involves locating hubs and assigning non-hub nodes to hubs with the objective of minimising transportation costs across the network. This determines the cost effectiveness of the H&S network as compared to point-to-point travel. Operational researchers have carried out various studies to try and solve the hub location problem in the most practical, meaningful and realistic way for airlines. For the purpose of this study, the location of the hub airports is the decision that needs to be made using the multi-criteria decision analysis (MCDA) process. With Africa’s air network characteristic of low passenger demand and the vastness of the continent, designing a hub network will be a challenge. This paper focuses on simplifying one of the steps entailed in designing an H&S network using the MCDA method for the African region as a way to foster the development of air transport. This paper will define the H&S network design elements that are crucial for sparse networks. Thereafter, the MCDA process and methodology will be introduced and applied to the hub location procedure. An evaluation of the criteria will be carried out to highlight how the MCDA method can simplify a complex decision making process in the H&S network design methodology. 1.1.1 Importance of hub location Hubs are defined as collection points that serve the purpose of consolidating traffic flow. The concentration or consolidation of flow can reduce movement costs (i.e. transportation or transmission) through economies of scale, even though the distance travelled may increase as stated by Campbell [6]. Hubs are usually found within air networks, mail delivery systems and in telecommunications. Schnell and Huschelrath [7] suggest that hubs can be defined in two general ways: (1) denoting whether an airport represents a hub within a carrier-independent system of air transport (i.e. airport level) and (2) denoting its role within a carrier-specific network (i.e. airline level). In the analysis of hubbing, the definition of what constitutes a hub becomes crucial. For the purposes of this study, a hub airport will be defined at airport level, based on route structure, i.e. its function as a distribution point for air travel to and from its surrounding catchment area, with connecting services, irrespective of the number of originating passengers and airlines serving it. 1.1.2 Limitations of the study 1.

2.

Technicalities that exist in the airline industry as a business, which include regulatory components of service agreements, operational constraints such as degrees of freedom permitted, competition, available time slots, security and pollution, will not be included. The network cost results of this study are neither deemed to be an accurate representation of the transportation costs nor realistic for airlines in the region. This is purely an academic exercise and therefore the real potential of this exercise is to use the results to develop a methodology to identify

6

Regional Airports

cost-effective criterion to determine the various MCDA alternatives. This study can be applied to real airline operations data in assessing hub network optimisation strategies using MCDA. 3. The environmental costs of hub networks as explained by Morrell and Lu [8] will not be taken into consideration when calculating the environmental costs resulting from an H&S network design. 4. The airline service being considered is a traditional passenger airline which transports its passengers to their destinations at the minimum frequency needed to meet existing demand. 2 MULTI-CRITERIA DECISION ANALYSIS Tecle and Duckstein [9] define MCDA as a vast field of study which includes decision making in the presence of two or more conflicting objectives and/or decision analysis processes involving two or more attributes. The general objective of MCDA is to assist a decision maker or a group of decision makers to choose the best alternative from a range of alternatives in an environment of conflicting and competing criteria. The methodology of decision analysis provides a framework to combine traditional techniques of operations research, management science and systems analysis with professional judgments and values in a unified analysis to support decision making as stated by Pieterson [10]. Decision analysis focuses on aspects fundamental to all decision problems namely: 1. 2. 3. 4. 5.

A perceived need to accomplish some objectives. Several alternatives, one of which must be selected. The consequences associated with alternatives are different. Uncertainty usually about the consequences of each alternative. The possible consequences are not all equally valued.

In recent years, several methods have been proposed to deal with MCDA problems. Belton and Stewart [11] highlighted the value function method as one of the several methods most suited for complex problems. Furthermore, Belton and Stewart [11] state that value function methods are some of the more widely applied MCDA methods and have benefited from the long-standing interests of psychologists, engineers and management scientists who have been nurtured through a continuing awareness of behavioural and social issues as well as the underlying theory. The methods are able to deal with complex issues, can accommodate the involvement of multiple stakeholders and allow processes to be facilitative and transparent. Value function methods can assist in the problem formulation phase and in informing stakeholders about the decision processes [12]. The value function method synthesises assessment of the performance of alternatives against individual criteria, together with inter-criteria information reflecting the relative importance of the different criteria, to give an overall evaluation of each alternative indicative of the decision-makers’ preference. The reason why this method is the most appropriate is because choosing a hub airport is a complex decision which has to take into account various issues like: network costs, infrastructure costs, security, economic viability, safety, passenger travel time expenditure, etc., all of which have different and varied impacts and implications. 2.1 Methodology The MCDA process is used because this approach uses several criteria, in which the analyst aims at establishing comparisons on the basis of the evaluation of the alternatives according to several criteria. In an approach using a single criterion, the analyst seeks to build a unique criterion taking into account all the relevant aspects of the problem. The methodology used in this paper is classified into four steps. These steps are outlined below.

Regional Airports

1. 2.

3. 4.

7

Structuring the decision problem: This involves the generation of alternatives and the specification of objectives. Assessing possible impacts of each alternative: In this step, the analyst determines the impact of each alternative. If it were possible to precisely forecast impact, we could associate one consequence with each alternative. Then the evaluation of alternatives would boil down to a choice of the best consequence, preferably based on research. Determine preferences (values) of decision makers: This step, unique to decision analysis, involves the creation of a model of values/preferences to evaluate each of the alternatives. Evaluate and compare alternatives: Once a decision problem is structured, the magnitude and the associated likelihoods of consequences determined, and the preference structure established, the information must be synthesised in a logical manner to evaluate the alternatives. 3 APPLYING MCDA TO THE HUB AIRPORT LOCATION

3.1 Decision problem: hub location analysis Locational analysis is a procedure in operational research used to locate the hubs and route flow via the hubs in an H&S network system. The two systems defined by O’Kelly and Bryan [13] are:

• •

A delivery system, in which the decision-maker positions the facilities and determines the rules of allocation to the centres. A user-attracting system, where the facility is located by one agent but the allocation decisions are decentralised and the planner has to make some reasonable guesses as to how the public will make use of the facilities.

There are various factors, pointed out by Schnell and Huschelrath [7] that influence the likelihood of an airport becoming a hub. Some of these factors are:

• • • • • • • • •

climatological characteristics of the location, geographical location and topographical surroundings, market size of the airport, inhabitants’ income, level of development of business and leisure centres to increase the attractiveness of the airport, potential of the airport to increase its capacity when there is congestion, number of flights operated, number of destinations served, number of gates available at the airport.

Africa faces the dilemma of not having many international airports with the capacity, demand, market size and infrastructure for hubbing. This is due to the low levels of income in most countries and the expense associated with air travel and its infrastructure. As discussed in Section 1, the decision to be made is the evaluation of various criteria to determine and solve the hub airport location problem, with the aim of minimising network costs. 4 ASSESSING RESEARCH-BASED IMPACT OF DIFFERENT CRITERIA The criteria that are used to assess the various hub airport location have been summarised and categorised in Table 1.

8

Regional Airports

Table 1: Criteria for MCDA. Functional criteria

Operational criteria

Locational criteria

Airport infrastructural capacity Flow thresholds

Passenger demand N–H costs Network costs

Shortest paths Centrality of the hub

4.1 Functional criteria This criterion analyses the elements that influence the functioning of an airport as a hub. Most of the challenges of airports that act as hubs include the ability of the airport to serve the additional demand. 4.1.1 Airport infrastructural capacity CMISRO [3] defines airport capacity as the limit on the amount of flow being collected by non-hub nodes from hub nodes. Capacity is important at hub airports because of the congestion that can arise at such airports due to the limitations in facilities (in terms of gates, runways and hangars) that are realised when an airport becomes a hub. Schnell and Huschelrath [7] state that for airlines there is a restriction on expansion at congested hub airports due to lack of slots in which planes can land. As a result, there is reduced flexibility on scheduling, which increases susceptibility to delays in emergency situations. 4.1.2 Flow thresholds The CMISRO [3] defines flow thresholds as the minimum flow that is needed on some or all of the links. The flow thresholds for each of these hubs could be taken into consideration, so that the flow carried would correspond to the capacity of the airport to serve as a hub. Some of these elements could be defined by:

• •

Number of airlines operating: This reflects the airport’s operational capacity in terms of gates, slots, baggage-handling processes and aircraft turnaround time. Airport passenger capacity: This assesses the ability of an airport to accommodate and serve high passenger numbers currently, based on current hubbing functionality and role, either geographically or operationally, to ease the transition into becoming a hub. Airport capacity is also of concern when there is need to consider an alternative route or a direct flight between i and j if this will cause the capacity of the hub airport to be exceeded when flow is consolidated.

4.2 Locational criteria Locational criteria define the elements of the airport in terms of its location within the region and the H&S network, with specific focus to the impact on sector distances. 4.2.1 Shortest paths Generally, as distances increase, the costs per passenger increase as well, due to the increasing operating costs incurred with higher aircraft utilisation costs in terms of depreciation, fuel and labour. This means that in order to ensure low operating costs, the sector distances flown should be kept as short as possible. In this method, the allocation problem of collecting and distributing flow can be solved by finding the shortest path between each pair of nodes in the directed graph, allowing collection from any node

9

Regional Airports

to any hub, transfer between hubs and distribution from any hub to any node. Ssamula [14] proved that in route networks, the shortest path usually implies that the costs on the route are minimised, because of the ability to fly smaller aircraft which are cheap to operate on these routes. Fig. 3 defines the relationship between weekly costs per passenger and increasing sector distances for various aircraft types, with an annual passenger demand of 30,000. The relationship between distance and cost for a given aircraft is linear because depreciation, fuel and labour increase with flight time or distance, which is similar to the findings of Swan and Adler’s study [15]. The general advantage of flying short routes is that a small fleet size is needed to operate the O–D pairs with both high and low passenger demand. Even when the frequency of flights increases with increasing passenger demand in the hub network, the fleet size will remain small because the flights are shorter [4]. 4.2.2 Centrality of the hub A procedure was performed by Topcuoglu et al. [16] to test for the cost benefits of locating a hub as centrally as possible within a cluster. It involved finding the geographical location of the mid-point using the latitude and longitude of all nodes in the cluster and then choosing the node that was nearest to the mid-point to act as a hub. Klincewicz [17] used the clustering heuristics methodology as one of the methods for choosing hubs in the facility-location problem by dividing the area into clusters and the different airports were given indexes in terms of probabilities. It uses the principle that the airport in a cluster that is most suitable as a hub would be the airport with the shortest node–hub distances and the highest passenger demand cluster, making it a more effective way of optimising the movement of flow. Based on this methodology, Ssamula [4] applied the method and found that this method in which the emphasis is placed on the strategic location of the hub, leads to a reduction in both node–hub and hub–hub costs. 4.3 Operational criteria The operational criterion includes the factors that influence airports’ attractiveness as a hub based on elements that lower or optimise operational costs. These elements include passenger demand, node–hub cost analysis and network costs. 30 000 Annual Passengers F-50 737-200 737-400 A320-200 A340-200 737-800 767-200 747-200 767-300ER 747-400 Erj135-jet

3,000 2,500 2,000 1,500 1,000 500

00

00

98

00

92

00

86

00

80

00

74

00

68

00

62

00

56

00

50

00

44

00

38

00

32

00

26

20

0

00 14

80

0

0

20

Weekly costs per pass (US$)

3,500

Increasing Distance

Figure 3: The linear increase of weekly operating costs with sector distance.

10

Regional Airports

4.3.1 Passenger demand The general trend is that as passenger numbers increase, the benefits of economies of scale increase. This is because the costs per unit flow decrease exponentially as demand increases until they become constant. Doganis [18] states that, economies from route traffic density arise because the higher seat load factors lower the costs per passenger mile. Fig. 4 derived from an airline route costing study by Ssamula [4] confirms the exponential decrease of costs per passenger as the weekly passenger numbers increase on a 3,000-km route for the 11 different aircraft. The study also found that high passenger demand reflects the infrastructure capacity of the airport and economies-of-scale benefits on the node–hub links. Hubs chosen coincidentally have the highest passenger demand within the region. The high passenger demand lowers the node– hub costs because the aircraft fly at high load factors. 4.3.2 Node–hub cost analysis The node–hub links in any hub network contribute more to network costs than the hub–hub links as determined by O’Kelly and Bryan [13] and confirmed by Ssamula [4] that node–hub costs contributed an average of about 56% to network costs. This is because the hub–hub links benefit more from the economies of scale gained from consolidated flow. This implies that since the node–hub portion of the journey is more costly, a strategy aimed at minimising the costs on the node–hub link needs to be explored. If the distances on the node–hub link can be minimised, the operating costs will be lower and this will encourage the use of smaller, cheaper short-range aircraft, which will minimise costs. In the study done by Ssamula [4] the node–hub cost analysis was used as a hub location methodology in trying to design an H&S network. The method and the findings of that methodology are outlined below. Fig. 5 represents the cost of transporting flow (Ci−n) ‘from’ each node (Oi). The total cost from node i to each of the n nodes in the network is summed in eqn (1). The nodes that have the cheapest costs of transporting flow from them are used as hub location options.

Costs per Passenger (US$/Pass)

10000

8000

6000

4000

2000

0 0

400

800

1200

1600

2000

Passengers Erj 135 JET Cost/pass

737-200 Cost/pass

A320-200 Cost/pass

A340 200 Cost/pass

767-200 Cost/pass

767-300ER Cost/pass

Figure 4: Exponential decrease of costs with increasing number of passengers.

Regional Airports

11

Oi = ∑ Ci − n = Ci −1 + Ci − 2 + L + Ci − n + Ci − 50 .

(1)

50

1

The cheapest nodes from which to fly the passengers as hub options, in this method the cheapest hub nodes chosen coincidentally, had the highest passenger demand within the region. Fig. 6 illustrates the calculation of the cost of transporting flow (Cn−i) ‘to’ each node (Di). The total costs for all the nodes in the network were summed. The nodes that have the lowest costs to fly as destinations are used as hub location options. 50

Di = ∑ Cn − i = C1− i + C2 − i + L + Cn − i + C50 − i .

(2)

1

In this method, centrality of the hub nodes with the shortest total distances was found to be a commonality with the cheapest hub network design. The strategic location of the hubs was found to outweigh the economies of scale achieved through high traffic volumes. 4.3.3 Network costs By definition, network costs mean the total costs of transporting passengers from their origin to their destination through the hubs. The costs are calculated as a product of the costs per unit flow and the 1 2

3 Node i 4

5

Figure 5: The cheapest hub to fly ‘from’. 1 2 3

4 Node 5

Figure 6: The cheapest hub to fly ‘to’.

12

Regional Airports

flow along all the routes. An analysis of total network costs is crucial to find out which hub locations achieve lower costs on a larger scale. Ssamula [4] summarised some of the important design elements for optimum sparse H&S network designing sparse networks as 1. 2. 3. 4.

The transmission flow costs which were found to be cheapest for hub location options which have high passenger demand. The sector distance was also found to be crucial in lowering operating costs, in sparse markets, as smaller more efficient short-range aircraft can be operated. Since sector distances are crucial in lowering costs the optimum number of hubs/clusters in sparse markets is determined by the distance threshold for the efficient aircraft. Nodes are assigned more efficiently to the closest hub in order to lower node–hub costs by minimising N–H links.

Even though many network cost equations exist in literature, for the purposes of this study, the equation that is used is developed from the uncapacitated single allocation r hub median problem, using the quadratic optimisation problem with linear constraints of the hub location problem developed by O’Kelly and Bryan [13] and rewritten by Klincewicz [17]. This eqn (3) is used to calculate network costs for networks in which the hub capacity has no threshold and each node is allocated to a single hub to limit complexity. The equation used to calculate the network costs, is given as f(x) = ∑i∑k XikCik(Oi + Di) + ∑i∑k Xik ∑k ∑m XkmCkmWkm.

(3)

The first term in eqn (3) involves the calculation of the collection and distribution costs (node–hub movement); this part of the equation includes:

• • • • • • •

Oi and Di represent the total amount of flow originating and terminating at node i, since all those passengers have to undergo that leg of the journey regardless of their final origin or destination node. The factor Xik is the constraint that addresses the fact that all nodes go through at least one hub. It is represented as 1 if that node–hub movement occurs and as 0 otherwise. Cik represents the cost per passenger from node i to the nearest hub, k. The second term in eqn (3) calculates the cost of moving the people who are travelling through the hubs k and m. The factor Xik is represented as 1 if that node–hub movement occurs and as 0 otherwise. The factor Xkm is represented as 1 if the hub–hub movement occurs for a given O–D path and as 0 otherwise. This means that only the passenger flow Wkm that is determined by the N–H–H–N, H–H–N or H–H movement is included in this part. Ckm represents the cost per passenger on the H–H links from hub k to hub m.

4.4 Applying criteria to the evaluate the various alternatives For purposes of this study, each of 50 countries is represented by one major international airport that is used as a node within the African network. All the international airports used in the database of the cost model were analysed. Special consideration was given to airports that are currently being used as hubs in these regions. The continent was divided into four geographical regional clusters as shown in Fig. 7, based on the optimum number of hubs derived by Ssamula [4]. This process entailed generically assessing the optimal number of clusters to produce an efficient network. In order to find the optimum number of hubs for the network, virtual hubs are found at the centre of each cluster for a two-, three-, four- and five-cluster network. This systematic method is then used to analyse the effect of increasing the number of hubs on network costs illustrated in Fig. 8. Based on the total network costs, it appears that the optimum network for the continent would be either a four-hub network because it had the lowest costs.

13

Regional Airports

North

West

East

South

Figure 7: Four-cluster network. 3,500,000,000

Costs (US$)

3,000,000,000 2,500,000,000 2,000,000,000 1,500,000,000 1,000,000,000 500,000,000 0 Clusters 2

Clusters 3

Clusters 4

Clusters 5

Network types N-H Costs

H-H Costs

Total network costs

Figure 8: Cost variations as clusters increase in a network. The region was then divided into four clusters which are also aligned with the existing Regional Economic Communities (RECs). The RECs listed in their specific regions illustrated in Fig. 7 are: 1. 2. 3. 4.

United Maghreb Union (UMA) in the north, East African Community (EAC) or Common Market for Eastern Southern Africa (COMESA) in the east, Economic Community for West African States (ECOWAS) in the west, Southern African Development Community (SADC) in the south.

The various cost elements (cost per passenger) and the various alternatives used in MCDA process will be derived from the cost model developed by Ssamula [19]. This cost model calculates the operating costs incurred by flying along a specified route, and the database for this model contains Africa-specific data. The costs used are calculated by selecting the aircraft (chosen from 11 different aircraft types of varying capacity) most commonly used in Africa that produces the lowest operating costs for the route.

14

Regional Airports

4.5 Determining preference values for criteria The parameters used to justify potential hub airport options for each criterion will be based on the hub airport locations that provide the criteria that have potential for lowering total network costs and reducing passenger travel time. The preference value, which determines how the criteria achieves its aim for each of the criteria, and the data source for the information are outlined in Table 2. 4.6 Evaluation criteria The process of determining which nodes are chosen as the most suitable hubs is elaborated below. 1. 2. 3. 4.

An index of 1 is awarded to the airports within that cluster that meets the criteria, i.e. for highest passenger number = 1. The indexes for the rest of the airports within the row are calculated in proportion ratios of 1 to the index of the airport that met the criteria. Indexes are calculated for each criterion and for each airport. The airport with the highest index within the cluster becomes the most probable hub airport.

Table 2: Preference values and data sources for the criteria. Criteria

Preference value

Data source

Airport infrastructural The presence of adequate infrastructure in terms World Bank Data Query, capacity of runways, gates and aprons to accommodate aircraft departures per a high frequency of flights is vital. This would year, for the year 2001 mean that minimum additional investment would be needed when converting airports to hubs Flow thresholds

Highest number of airlines operating currently, this capacity can handle the extra flow

Passenger demand

The presence of high passenger demand at an airport implies that the airport is already a popular destination. The economies of scale enjoyed on routes to and from these busy airports would mean lower transportation costs on the node–hub links Lowest total sector distance for the cluster 50-by-50 distance matrix from an on-line airport mileage calculator Shortest distance from the geographical 50-by-50 distance matrix centre of the hub. The hub airport should be from an on-line airport conveniently located geographically, so that mileage calculator it is well connected as a hub and does not inconvenience passengers Cheapest node–hub costs Eqns (1) and (2) Lowest total network costs Eqn (3) [4]

Shortest paths

Centrality of the hub

N–H costs Network costs

Individual airport information: world airport data Furness’ method of a double-constrained gravity model based on world bank data

15

Regional Airports

The MCDA method allows for decision makers to weigh each of these alternatives based on the overall desired effect. In this study, each of the alternatives is weighted equally, based on the lack of factual evidence that one criterion has a higher effect on the suitability of a hub location option over another hub. Therefore, the methodology that will be adapted in the four steps is summarised below. 1. 2.

3.

4.

Decision problem: To locate the four cheapest hub location option within the African continent. Assessing possible impacts of each alternative: The possible impact of each of these alternatives is in gauging the suitability and cost effectiveness of the hub location option for the H&S network. These are differentiated into the functional, locational and operational criteria. Determine preferences (values) of decision makers: Based on the various findings in literature the preference values are also linked to designing a cost effective H&S network, summarised in Table 2. Evaluate and compare alternatives: The alternative hub airports will be assigned individually for each criterion. The airport with the highest total index criterion value will be chosen as the hub airport.

5 EVALUATE AND COMPARE ALTERNATIVES All the international airports used in the database of the cost model were analysed using the criterion set out in Table 3. The indexes are calculated and summarised for the airports within each of the clusters that have the highest indexes. Table 3: Hub location evaluation. Region Airport code Country Infrastructural capacity Airport capacity Airlines served Passenger numbers Centrality/ shortest path Node–hub (costs per pass US$) Network costs (US$) Total index

Northern cluster FEZ

ALG

CAI

Morocco Algeria Egypt

Southern cluster

Eastern cluster

Western cluster

JNB

NBO

LOS

HRE

ADD

DKR

South Africa Zimbabwe Kenya Ethiopia Nigeria Senegal

0.250

0.500

1.000

1.000

0.333

1.000

0.500

1.000

1.000

0.023 0.327 1.000

0.273 0.286 1.000

1.000 1.000 0.935

1.000 1.000 1.000

0.294 0.214 0.072

1.000 1.000 0.879

0.778 0.422 1.000

0.500 1.000 1.000

1.000 0.778 0.077

0.942

1.000

0.616

0.975

1.000

1.000

0.924

1.000

0.728

0.509

1.000

0.759

1.000

0.266

0.953

1.000

1.000

0.712

0.509

1.000

0.267

1.000

0.857

0.838

0.850

1.000

0.750

3.559 5.058 5.576 6.975 3.036 6.671 5.474 6.500 5.045 Cairo International O R Tambo Airport Jomo Kenyatta Lagos Airport (CAI) (JNB) Airport (NBO) International Airport (LOS)

16

Regional Airports

Based on the evaluation criteria, for each cluster the airport with the highest total index is the most suitable hub option in the cluster. In the north, south, east and west, all the airports with highest airport capacity and passenger numbers are chosen as the suitable hubs. The usefulness of this study is that it is able to prove that MCDA is a simple useful tool that can be used to further explore the ways in which an optimally efficient hub network design, specific to sparse markets. MCDA provides a solution to the hub-location problem that can incorporate elements that range from operational to functional while preserving the methodology of each individual criterion, and furthermore this criterion can be ranked in order of importance, so as to come to an appropriate decision. The major findings in this study shows that because Africa has a sparse network, with a few role players, the choice in hub location options relies greatly on the cost of routing passengers through the hub airport. The hub airport should be near the economic heart of the region so that it is able to nurture economic growth through employment, infrastructure and development. MCDA is shown to be a useful tool whose only limitation is the assumed uniform weighting of the different criteria for the region based on lack of suitable data. REFERENCES [1] Chingosho, E., African Airlines in the Era of Liberalization; Surviving the Competitive Jungle. E-Book, ISBN 9966-05-011-6, 2005. [2] Button, K., Lall, S., Stough, R. & Trice, M., Debunking some common myths about airport hubs. Journal of Air Transport Management, 8, pp. 177–188, 2002. [3] CMISRO (Center for Mathematical and Information Sciences), Operations Research – Hub Location Report. Australia, 2003. http://www.cmis.csiro.au/or/hubLocation/ accessed 3/02/04. [4] Ssamula, B., Strategies to Design a Cost-effective Hub Network for Sparse Air Travel Demand in Africa. PhD Thesis, University of Pretoria, South Africa, 2008. [5] Boland, N., Krishnamoorthy, M., Ernst, A.T. & Ebery, J., Preprocessing and cutting for multiple allocation hub location problems. European Journal of Operational Research, 155(3), pp. 638–653, 2004. [6] Campbell, J.F., Hub location and the ρ-hub median problem. Operations Research, 44, pp. 923–935, 1996. [7] Schnell, M.C.A. & Huschelrath, K., Existing and new evidence on the effects of airline hubs. International Journal of Transport Economics, XXXI(1), pp. 99–121, 2004. [8] Morrell, P. & Lu, C., The environmental cost implication of hub–hub versus hub by-pass flight networks. Transportation Research Part D, 12, pp. 143–157, 2007. [9] Tecle, A. & Duckstein, L., Concepts of multicriterion decision making. Multicriteria Analysis in Water Resources Management, eds J.J. Bogardi & H.P. Nachtnebel, UNESCO: Paris, pp. 33–62, 1994. [10] Pieterson, K., Multiple criteria decision analysis (MCDA); a tool to support sustainable management of groundwater resources in South Africa. Water SA, 32(2), pp. 119–128, 2006. ISSN 0378-4738. [11] Belton, V. & Stewart, T.J., Multiple Criteria Decision Analysis – An Integrated Approach, Kluwer Academic Publishers: Boston/Dordrecht/London, 2002. [12] Keeney, R.L., Decision analysis: an overview. Operations Research, 30(5), pp. 803–838, 1982. [13] O’Kelly, M.E. & Bryan, D., Hub location with flow economies of scale. Transportation Research Part B, 32(8), pp. 605–616, 1998.

Regional Airports

17

[14] Ssamula, B., Analysing aircraft selection for route operating cost for short-haul routes. Journal of the South African Institute for Civil Engineers (SAICE), 48(2), pp. 2–9, 2006. [15] Swan, W.M. & Adler, N., Aircraft trip cost parameters: a function of stage length and seat capacity. Transportation Research Part E, 42, pp. 105–115, 2006. [16] Topcuoglu, H., Corut, F., Ermis, M. & Yilmaz, G., Solving the uncapacitated hub location problem using genetic algorithms. Computers & Operations Research, 32(4), pp. 967–984, 2005. [17] Klincewicz, J.G., A dual algorithm for the uncapacitated hub location problem. Location Science, 4, pp. 173–184, 1996. [18] Doganis, R., The Airline Industry in the 21st Century, 1st edn, Routledge: London, 2001. [19] Ssamula, B., Developing a Cost Model for Running an Airline Service. MEng Dissertation, University of Pretoria, South Africa, 2004.

This page intentionally left blank

Regional Airports

AIRPORT–AIRLINE RELATIONSHIPS: OPPORTUNITIES FOR ITALIAN REGIONAL AIRPORTS S. CEPOLINA1 & G. PROFUMO2 Centre of Excellence on Integrated Logistics (CIELI), University of Genoa, Italy. 2Department of Business Studies, University of Naples “Parthenope”, Italy.

1Italian

ABSTRACT In recent years, the aviation industry is facing fast evolution patterns and strong competition in its main sectors. On the one hand, the increasing competition between airlines is lowering industry profitability. On the other hand, airports are facing new managerial challenges due to privatisation processes. Not only international airports but regional airports also are affected by these phenomena. This turbulent environment is pushing companies in the aviation industry to follow new strategic paths in order to face the new competitive arena; one of the paths that is gaining increasing attention is cooperative strategy, created in order to reduce uncertainty, sharing risks and costs. The literature has primarily focused its attention on horizontal alliances, while less attention has been given to vertical integration strategies between airports and airlines. The aim of the present work is, therefore, to cover this literature gap, analysing the airport–airline relationships from a strategic management perspective. After a brief classification of all the possible forms of relationship, the potential benefits for the two actors, airlines and airports, will be analysed, trying to underline the possible cost and revenue synergies. Particular attention will be given to regional airports, some of which are now experiencing fast development, due to the interaction with airline companies. This is the case of the two Italian regional airports investigated in the second part of the paper. Keywords: airport–airline interaction, regional airports, vertical integration strategy.

1 INTRODUCTION In the last few years, the aviation industry has been affected by external events and developments, such as globalisation and liberalisation/deregulation processes, that have challenged all the actors belonging to the value system, with special reference to airports and airlines. On the one hand, the increasing competition between airlines, boosted by the entry of low cost carriers (LCCs) in the market, is squeezing companies to a price war, diminishing industry profitability. On the other hand, airports are experiencing, because of the privatisation processes, managerial challenges which, in some cases, still have to be matched with political objectives. This is the case of regional airports, a large part of which is still public owned. As the environment is becoming more and more turbulent, companies operating in the aviation industry are seeking strategic paths in order to survive in the new competitive arena. Growth strategies, in particular, may enable firms to face high competition levels, maintaining a sufficient profitability, reaching economies of scale and scope and holding higher bargaining power, but to be effective, in an uncertain and very dynamic framework, the implementation process has to be fast. Internal expansion, known also as organic growth, may not be suitable, because it is usually a slow process that requires large financial resources. So, the resources and competences needed for the expansion should be found in other firms, through different forms of relationship that range from simple agreements to mergers and acquisitions (external growth strategies). In recent years, the literature has extensively focused on horizontal alliances and mergers and acquisitions (M&A) processes inside many industries of the air transport value system, while less attention has been given to vertical integration strategies between airports and airlines in particular. Aiming at covering this literature gap, the present work explores airport–airline relationships from a strategic management perspective. Airports and airlines are very different subjects with a separate regulation regime, diverse competitive arenas, not similar strategic behaviours, but, at the same time, they are linked by a customer–supplier

20

Regional Airports

relationship and they serve the same final customers, passengers [1]. These aspects, together with recent airport–airline interaction experiences, like Lufthansa’s cooperation with German airports of Frankfurt and Munich, require a deeper analysis of airport–airline forms of interaction. Moving from these considerations, the present paper focuses on airport–airline relationships in the Italian aviation industry. The Italian case is rather peculiar, at least for two reasons. Firstly, the network of national regional airports is highly spread, with many small airports characterised by different growth rates and there is a lack of a strong flag carrier. Secondly, the Italian regulatory framework, with its particular ‘concession regime’, still has a very strong influence on airport operations and the development of any relationship with airlines. Moreover, the Italian aviation market is the fifth largest European market in terms of passengers, with 135 million passengers in 2007, and it has recorded an average annual growth rate of 8.2% in the last 5 years, consistently higher than other more developed markets [2]. The paper is structured as follows. Section 2 shows managerial literature on the different forms of relationships between firms using corporate control as the lens of analysis and focuses attention on the literature contributions related to airport–airline interactions. This part of literature is less developed and there are still comprehension and behavioural gaps to overlap. Section 3 is aimed at identifying the drivers that may push airlines and airports to engage vertical integration strategies through different external implementation processes. Section 4 presents an up to date portrait of the Italian aviation industry, focused on the principal characteristics of the airport system and the airlines operating in the national area. Section 5 tries to verify the drivers of vertical integration strategies in two selected case studies of Italian regional airports. Finally, Section 6 concludes with a brief discussion of results and possible future fields of research. Although the paper is the outcome of a collective work, Paragraphs 1; 2.2; 4 and 5.2 can be attributed to S. Cepolina, while Paragraph 2.1; 3; 5.1 and 6 can be attributed to G. Profumo. 2 AIRPORT–AIRLINE VERTICAL FORMS OF INTERACTION The managerial literature on firms’ relationships is really deep and rich, following different lens of analysis. In this section, the approach based on corporate control will be used in order to evaluate the specificities of the airport–airline vertical forms of interactions. In the aviation industry, literature has extensively focused on firms’ relationships inside each sector (e.g. airline’s alliances and M&A, aircraft constructors’ concentration processes), while less attention has been given to the study of vertical forms of interaction between companies belonging to different stages of the value system. 2.1 Managerial literature on firms’ relationships Following their corporate strategies, firms may choose to grow by enhancing the resources and competences that are internally generated (organic growth) or by utilising and developing resources, knowledge and capabilities available in other companies. This second implementation process may take several forms, depending on the business relationship developed between the two firms. The existing managerial literature has elaborated different taxonomies of the phenomena. For the purpose of our paper, it is useful to classify business relationships according to the type of corporate control; following this classification, they can range in a spectrum that goes from a simple, shortterm transactional relationship to a full acquisition or merger, in which a company takes the entire ownership of another [3], as indicated in Fig. 1. Moving towards M&A implies, on the one hand, a continuous increase in the commitment of the companies involved in the relationship to achieve the foreseen objectives and a greater steadiness of the interaction; on the other hand, a higher business risk, due to the greater investment and the difficulty to exit from the relationship.

21

Regional Airports Transaction Contract

Alliance Non equity alliance

Equity alliance

Purchase Joint Venture

Acquisition

Merger

Risk and control

Limited

Shared

Total

Duration

Short Term

Medium to Long Term

Permanent

Legal status

No new legal entity formed

New legal entity formed

Legal status of old entities changed

Figure 1: The spectrum of business relationships. Source: Adapted from Ref. [3]. In the case of a transactional relationship, as for example the interaction between a customer and its supplier, one company may, to some degree, influence the other firm, but the control is limited in scope (only to what is written in the contract) and duration. As the partners usually do not invest too many efforts, in terms of resources and competences, in this short-term relationship, their business risk is also limited. The different forms of alliance fall in the middle of the spectrum. They can be defined as inter-firm relationships in which two or more companies jointly invest in a common activity over a number of years, sharing risks and returns, but remaining legally independent. Only in the case of a joint venture there is the creation of a new legal entity. The term ‘strategic alliance’ includes a wide range of relationships that vary from long-term purchasing agreements to marketing and research and development (R&D) collaborations, to joint ventures [4, 5]. Despite the differences, all the alliance forms present at least a few common features [4]: the link between the alliance’s scope and the strategic intent of each partner [6], the sharing of resources and knowledge among partners [7] and the creation of opportunities for organisational learning [8]. Alliances are more complex to manage than transactional relationships and usually have a longer lifetime, even if they have a clear endpoint. They are very useful in uncertain and risky market conditions, as they limit the resources a company must commit to the new venture [3]; they are, indeed, often viewed as a mechanism to cope with uncertainty. In some forms of alliance, the companies reciprocally purchase minority equity stakes in order to maximise the commitment to the joint project, this is the case of equity alliances. Because partners in an alliance remain independent, a single partner is not able to control the others completely and there is the multiplication of decision-making centres, which implies longer and more complex decisions on controversial issues, such as eliminating redundant assets, rationalising product lines or specialising facilities. Alliances are also transient in nature; they can be closed without too many difficulties. For these reasons, alliances are less effective when there is economic value to be gained through rationalisation, which implies cost cutting: ‘horizontal acquisition will always outperform scale alliances’ [9]. On the other side of the spectrum, we can find M&A, in which the control on the other company is permanent and complete. In a merger, the level of integration between firms is maximum, as the companies become one new expanded legal entity, such an instrument is very complex to manage [10], as it implies the full blend of managers, staffs, competences and values. In an acquisition, a firm takes an ownership stake in another company, sufficient to exercise the control; in this case, the integration process may be absent or focused on specific functions, such as information and technology (IT) or R&D. The full ownership control presents some specificities: on the one hand, a company has to invest resources, knowledge and to assume the responsibility for the acquired assets, increasing its business

22

Regional Airports

risk. So, if the market conditions are already very risky and uncertain, a company should choose other forms of relationships. On the other hand, the full corporate control allows the company to better achieve tough decisions much more rapidly than in case of alliances, in which the decision process may be blocked by the intervention of many actors [9]. 2.2 Airport–airline interactions literature Airport–airline interactions are frequent only in few parts of the world, e.g. Asia, Australia and Arab nations, where airport operators and airlines are part of the same group and share the mission to support each other for the sake of the country’s competitiveness and economic development [11]. This consideration supports the limited amount of research and only recent developments in this field. Over the last decade, however, airport–airline interactions have gained increasing attention in US and Europe also, where they are seen as a strategic answer to the straightening competition in the aviation industry and to the privatisation process that characterises the entire aviation industry. On the one hand, airlines are squeezed among their neighbours in the aviation value system that leverages local monopolies (such as airports) or oligopolies (such as aircraft equipment manufacturers) [12]. Despite ongoing liberalisation, the regulatory framework still has not reached a common European level, to be able to push consolidation processes in the aviation industry. Airport price regulation forms are particularly relevant and there are strong differences in their stage of adoption among European countries [13, 14]. On the other hand, airports across the world are modifying their business model, focusing more and more on non-aeronautical revenues (retailing, advertisement, ground transport and property development) to generate financial resources. In the case of hub airports, resources are designated to increase capacity to meet infrastructure demands; in the case of secondary airports, resources are designated to increase airport attractiveness and to gain air traffic [15]. Secondary airports management often has an additional critical point to face because of the local public ownership, which seeks to balance economic aims (profit maximisation) with political and social aims (occupation, local well-being). The top management needs to reach remarkable levels of air traffic with a limited bargain power. Under these conditions, the strategy of secondary airports is to look more and more to LCCs as partners, because of their traffic generating ability. Since these new competitive matters are gaining more and more relevance, researchers have started to highlight them and literature has been developed. The emerging literature on airport–airline relationships may be classified into two big categories. The first category is based on the different aims of the agreement between airlines and airports. These types of classifications are numerous and are generally articulated in air service agreements and land service agreements. The first tends to develop new traffic and is devoted to airports with overcapacity status, like numerous European secondary airports. The second, however, tends to improve efficiency to better utilise existing capacity. They are devoted to highly congested airports, like European hub airports. In this category, some authors [16, 17] identify three agreement levels: marketing oriented (called land services), capacity oriented (called air side services) and security/technology oriented. The first two levels are strongly strategic, aiming at growth of airports and airlines; security/ technology oriented collaborations are instead more operational, aiming at increasing performance. They usually tend to improve airport security level as well as process efficiency (baggage processing technology) and they do not need a long-term relationship. The second literature category is based on the characteristics of the relationship, like time coverage, steadiness and actors’ commitment. This category originates from general managerial relationships

Regional Airports

23

literature, modified in order to fit aviation industry specificities. Different interaction forms are identified on the basis of the classification highlighted in Fig. 1: cooperation agreements, alliances, joint ventures and M&A [18, 19]. Many recent aviation industry phenomena can be observed through this lens of analysis; long-term usage contracts between airlines and airports as cooperation agreements, signatory airlines as a form of alliance, joint ventures and acquisitions. We also include in this second category the studies which analyse airport–airline interactions on the basis of the different kinds of subjects involved, like the copious contributions focused on LCC and secondary airport interactions [1, 20]. 3 AIRPORT–AIRLINE INTERACTION DRIVERS The existing literature on the determinants of the different forms of business relationships is wide and rich, even if fragmented. There are several studies focused on each type of relationship, in particular, on the motivations of alliances and M&A, but less attention has been given to create a common framework from which arise the drivers that may push firms to expand using the resources and competences present in other companies. Related research has shown that the determinants of M&A and alliances, associated with the maximising of a firm’s value, are quite similar and may be broadly categorised into [21–23] efficiency or operational drivers and market power drivers. The former synergies are related to economies of scale and scope and all the other cost economies that may be achieved by larger firms; the latter synergies, however, emerge from the possibility to access or create new markets or ‘strategic windows’, to develop knowledge and capabilities that are not present in the firm, and from the ability of the partners to control the price, the quantity or the nature of the products sold, thereby generating extra-normal profits (collusive synergies) [22]. The process of value creation may also involve taking strategic actions related to financial and risk diversification, in particular, in case of M&A [24, 25]. Alliances are the only form of external expansion that can be used in case of legislative barriers of entry (in the airline business, for example, as the regulatory framework strictly prohibits cross-border mergers, alliances represent the only instrument that allows airlines to serve the global market). The business relationship drivers have been extensively studied inside each stage of the aviation value system, in particular, in the case of alliances and M&A among airlines [26–30], less attention has been given to understand the motivations of vertical integration strategies. This literature is, in fact, more recent and it is mainly focused on case study analysis [17, 31–34]. The few studies related to the vertical forms of interaction between airports and airlines highlight that, although the relationship’s final attempt is to jointly serve customers and cope with traffic demand in a profitable, efficient and sustainable way [17], there are specific drivers for airports and airlines that push them to interact. The relationship is goal oriented; both parties enter into it for their own benefit [11]. The airport–airline interaction has also changed in nature in the last few years, passing from a pure transactional ‘supplier–customer’ relationship to a more strategic agreement. Our attempt is to identify the drivers that may push airlines and airports to engage in vertical integration strategies through different external implementation processes and, then, to verify them in two case studies belonging to the Italian aviation industry. Following the drivers emerging from the general literature, airport–airline interaction drivers may be distinguished into efficiency driven and market power driven. Inside each category, it is then possible to identify specific drivers for airports and airlines, as summarised in Table 1. The efficiency drivers are in both cases related to the cost economies that may be achieved through the interaction. As regards airports, the increasing traffic emerging from the relationship with an airline helps them to enhance operational capacity and consequently reduce unit costs; studies have

24

Regional Airports

Table 1: Airport–airline interaction drivers. Interaction category

Airlines drivers

Efficiency/operational drivers Cost economies, risk sharing, higher service quality Market power/strategic drivers Create a strong hub, create a feeding traffic to the hub, control of airport slots, develop traffic in the basin of the airport, better image

Airports drivers Cost economies, risk sharing, higher service quality Improved connectivity, increasing bargaining power, longer term planning, regional social benefits, better image, increasing non-aviation revenues

Source: Our elaboration. in fact demonstrated that unit costs decrease significantly as traffic increases up to 1.5 million work load units (WLU – defined as one passenger or 100 kg of freight) per annum and continue to fall until traffic reaches 3.0 million WLUs per annum [13]. As regards airlines, cost economies may arise, besides from traffic development, from a large set of activities undertaken by airports, called air service development [16]. Many airports, in fact, especially regional airports with low bargaining power, are now taking a share of the costs (and of the risks) of developing airline networks, providing services traditionally under the responsibility of airlines, such as analysis of the potential demand for a particular route, marketing activities for the development of the route traffic, financial incentives and handling cost reductions. The operational drivers are also related to the increase in service quality, due to more customised services and due to the development of operational airport capacity by joint management of on time performance [16]. However, one of the primary drivers for the formation of relationships between airports and airlines is the reduction of risk and uncertainty for both parties [17], obtained through the sharing of investment costs. The market power drivers are related to the acquisition of ‘revenue’ synergies, linked to the development of traffic and the preference of consumers. From an airline’s point of view, a strong interaction with an airport may be driven by the will to create a feeding traffic towards its primary hub, to set up a strong hub and to control airport slots (in particular in already congested airports), in order to offer customers seamless connections and gaining their preference. However, in case of M&A, some anticompetitive effects have been highlighted [32], such as decreasing quality for rival airlines, discrimination in the access to ground handling services and predatory practices towards competing airlines using cross-subsidies. From an airport’s point of view, the market power/strategic drivers are, instead, related to: the increase in its accessibility and connectivity; the possibility to plan with a longer term; the development of traffic in the airport basin, with social benefits for the region; and the increase in non-aeronautical revenues. In both cases, there could also be an improvement in the image and reputation of the partners that in some cases may profit from the standing of each other. 4 ITALIAN AVIATION INDUSTRY: LEGISLATIVE FRAMEWORK AND SPECIFICITIES The Italian airline sector is one of the most attractive markets in the European panorama. Its weight in terms of passenger traffic is relatively small (11.88% of the EU total pax traffic based on OAG data), but it is destined to increase consistently in the next years.

25

Regional Airports

The evolution trend for past traffic, with strong growth rates, in particular in the domestic sector, highlights the relative newness of the Italian market where the propensity to fly is still lower than the European average. As a young market, Italy is also a very dynamic market, characterised by legislative openness and a consequent high level of competition. Italian market competitiveness is also affected by the low market share hold by the Italian national flag carrier, Alitalia, which in 2006 was 25%, compared to the 45% hold by Air France in France. Moreover, in 2006 the Alitalia crisis emerged, resulting in a period of decline that ended only in 2008. In that year, a new company, Cai-Alitalia, was constituted with the purpose of integrating Alitalia and AirOne, the two most important Italian flying carriers, by acquiring their main assets. This event had strong traffic implications: comparing the data for the last 2 years (2007–2008), Alitalia has seen a fall of 26% in passengers carried (from 24.4 to 18 million); the nearly 6.5 million passengers who have abandoned Alitalia have not just moved to AirOne, whose clients have increased only from 7.1 to 7.4 million, but went elsewhere. So, by combining the two companies, there is a fall from 31.5 million passengers in 2007 to 25.5 in 2008. A last consideration refers to LCCs, which account for 29.9% of national supply (available seat kilometres) in 2008, with an annual growth rate of 48.5% in the period 2001–2008 (based on OAG source). In this business model there is no competitive national player (the first Italian LCC, Myair, with a market share of just 6%, is now close to bankruptcy) and the great majority of the growth is imputable to foreign LCCs (Ryanair and Easyjet, in particular, have a market share, respectively, of 32.6% and 17.5%). As regards the Italian airport sector, the market is composed of 101 airports. Among them, 45 are classified by the Italian regulatory authority ENAC as international (they can schedule international flights) while the remaining 56 are labelled as domestic (they can schedule only domestic flights) [35]. The framework highlights a widespread dissemination of Italian airports, which is confirmed by air traffic statistics. Table 2 shows a low level of air traffic concentration, which has decreased from 2000 to 2007. Concentration levels are calculated for different airports classes, including airports with the highest passenger traffic. The first two classes (top 5 and top 10 Italian airports) have lost much more traffic, in the time gap considered, than the other classes (top 20 and top 30 Italian airports). This trend may be explained by the more intense growth of Italian secondary airports supported by the advent of LCCs and by the increasing weakness of the national flag carrier. The development of secondary airports may be better appreciated by looking at Table 3, which shows total growth rate between 2000 and 2007 for different classes of Italian airports. In this case, airport classes are identified on the basis of the passenger traffic related to year 2000. The total growth rates are indirectly related to the airports’ dimensions in terms of passengers: smaller airports have grown more than their bigger counterparts. Table 2: Italian air traffic concentration. Airports

Pax 2000

First 5 First 10 First 20 First 30 Total

60,471,235 76,261,828 87,947,712 91,056,923 91,434,374

Concentration % 2000 66.14 83.41 96.19 99.59

Source: Our elaboration on ENAC data.

Pax 2007 79,200,150 104,769,202 128,360,982 134,406,226 135,315,674

Concentration % 2007 58.53 77.43 94.86 99.33

26

Regional Airports

Table 3: Total growth rate in different Italian airports’ classes (2000–2007). Airports 5,000 pax Total

Pax 2000

Pax 2007

Δ07–00

Δ07–00%

5,710,300 7,425,291 7,202,153 18,699,558 52,397,072 91,434,374

11,242,147 15,594,506 12,448,595 29,908,252 66,121,388 135,315,674

5,531,847 8,169,215 5,246,442 11,208,694 13,724,316 43,881,300

96.87 110.02 72.85 59.94 26.19 47.99

Source: Our elaboration on ENAC data.

A second Italian aviation industry specificity relates to airport governance. The great majority of Italian airports are managed by independent companies through a license delivered by the Italian regulatory authority (ENAC) (only 2 airports of the 45 international airports are public managed directly by ENAC). Licenses can be distinguished into total licenses and partial licenses. With the former, managing companies are responsible for the airport infrastructures and get all the airport’s charges; in the case of the latter, managing companies are responsible only for passenger and freight infrastructures (like terminals) and get only related charges. There is a third license category called precarious, similar to partial license with reference to infrastructures provision: in this case, managing companies obtain revenues only from terminal commercial activities (no collection of charges is foreseen) [36]. We have analysed airport governance with reference to airports with more than 1 million pax in 2007 (see Table 4) by classifying shareholders into two main categories: public and private subjects, of which the latter is split into airport management companies, airlines and a residual category. The resulting sample, which includes 34 airports, shows an advanced stadium of the privatisation process, formally started in 1993 (with Italian law n. 537/1993 and Italian law n. 351/1995), although public influence is still very strong. In fact, most of the airports (21) are controlled by public local authorities, like regional administrations and municipalities. Two airports are directly managed by ENAC and the remaining airports (11) are owned by private subjects. There are, however, very few cases in which airlines participate in airports’ capital. From these observations, it is clear that airlines prefer soft forms of relationships, like transactional relationships and alliances. In order to complete this scenario of the Italian airports sector we introduce some information on airport charge national regulation. In 2005, the Italian Government approved a new law (n. 248/2008) changing the framework for setting airport’s charges. The new policy set up a price cap hybrid till (allowing costs netted of 50% of commercial margin), abolishing any pre-existing alignment of Italian airport charges with the rest of Europe. As a consequence, the charges for Italian airports are now lower than the European average, in a range between 19% and 49%, depending on different variables [37]. The Italian association of airports (Assaeroporti), as well as ACI Europe, at the European level [38], states that the present average return on capital of airports is very low and might not be acceptable for the private sector, in order to sustain the necessary investments to maintain, upgrade or expand the long-term tangible assets of airports, such as terminals, runways, access roads and car parks, as well as to expand the capacity in order to follow a growing demand.

27

Regional Airports

Table 4: Italian airports classified by shareholders’ categories. Shareholders category

Airport 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34

Alghero Ancona Bari Bergamo Bologna Brescia Brindisi Cagliari Catania Crotone Firenze Forlì Genova Lamezia Terme Lampedusa Napoli Olbia Milano Malpensa Milano Linate Palermo Pantelleria Parma Pescara Pisa Reggio Calabria Rimini Roma Ciampino Roma Fiumicino Torino Trapani Treviso Trieste Venezia Verona

Private

Public

Airport management Other private companies Airlines subjects

Local Authorities

30% 30% 100% 90% >67% 100%

65%

5% 5% >5%

25% >20% >95% >95% 100% 100%

10% >5% 5% >4% 80%

>95% >95% >44% 10%

>20% >95% >50% 100% >80% 33% >90%

Source: Our elaboration on AIDA data and company websites. To meet these requirements the European Commission has recently prepared a directive (2009/12/ CE of March 11, 2009), which establishes a general framework, setting common principles for the levying. Sharing the aim of the community legislator, we stress the narrowness of its application, limited to airports with more than 5 million passengers. In Italy, the directive could be applicable in only eight airports, which account for 70% of the total national traffic.

28

Regional Airports

5 AIRPORT–AIRLINE INTERACTION DRIVERS: LESSONS FROM ITALIAN EXPERIENCES In order to verify the airport–airline interaction drivers identified in Section 3, we have selected two case studies, both regarding regional airports, emerging from the Italian airport sector. The first case, Olbia Airport in Sardinia Island, with 1,741,120 passengers in 2007, is the only Italian case of an airport management company (Geasar S.p.A.) controlled by an airline, Meridiana, a regional carrier which holds 79.79% of Geasar shares. The second case, Genoa Airport, with 1,105,802 passengers in 2007, is a regional airport facing very high competition. In the same catchment area, in fact, at a distance lower than 200 km, there are five international airports: Nice (France), Turin, Milan Linate, Milan Malpensa and Pisa. In order to improve its competitive position, Genoa Airport has developed relationships with many carriers of different nature. 5.1 Case study A: Olbia Airport Olbia Airport is a regional airport located on the northeastern coast of Sardinia Island, with a prevalence of tourist traffic. The airport management company Geasar S.p.A. represents a rather peculiar case in the Italian airport sector, as at the moment it is the only one controlled by an airline company. Meridiana Airlines, in fact, holds the majority of its shares (79.79%), whereas few public regional authorities are minority shareholders: Sassari Chamber of Commerce (with a share of 10%), Nuoro Chamber of Commerce (with a share of 8%), Sardinia Region Administration (with a share of 2%) and Emerald Coast Consortium (with a share of 0.2%). Olbia Airport started its operations in 1974, by substituting the former airport of Venafiorita, and was managed (with a partial license) by Geasar S.p.A since March 1989. In 2005, the company obtained the total license to manage the airport for 40 years. Meridiana Airlines, the principal shareholder of the airport management company, is a regional carrier established in March 1963, with the name Alisarda, by Prince Aga Khan. The carrier started its operations as an air taxi and charter operator, holding the objective of favouring the development of the tourist industry in the Emerald Coast, which until then was accessible only by sea. In 1966, the airline started to serve Rome and Milan from Olbia; in the following 2 years, it opened other national routes. In 1991, the carrier changed its name to Meridiana, following the shareholders’ agreement on the fact that the airline’s future was the pan-European market and it entered the European market with new international routes (Barcelona, Paris, London and Frankfurt). In December 2006, Meridiana acquired a 29.95% stake in the growing leisure carrier Eurofly, through the acquisition of 4 million shares from Spinnaker. In 2008, the carrier increased its stake in Eurofly to 46.1% of the shares. The ownership of Meridiana Airlines is held directly and indirectly by Aga Khan, with a majority stake (79.29%of shares), while the company’s employees and a bank foundation (Fondazione Cariplo) represent the other principal minority shareholders. The acquisition process of the Olbia Airport may be explained by the will of the airline to contribute to the development of tourism (and of the traffic) in the area, increasing regional social benefits. Prince Aga Khan is, in fact, the founder of the Emerald Coast Consortium, whose primary objective is the creation of an integrated tourist value system. The diversification strategy that the carrier has followed in the last few years may be read in the same way: in June 2006, in fact, Meridiana launched Wokita.com, the company’s on-line tour operator. Wokita’s ambition is to offer tourist booking services (hotels, car rentals, flights, holiday deals) on all the destinations served by Meridiana. Then, in September 2006, the carrier acquired 15% of the shares of ADF, the management company of Florence Airport.

Regional Airports

29

The other airline’s drivers of the acquisition process may be related to the customisation of services, cost sharing and the possibility to create a base in the airport, also with maintenance infrastructures. The airport’s drivers are firstly related to the development of passenger traffic and the improvement of connectivity. The expansion of the airport has been strictly associated with the development of Meridiana Airlines which, obviously, is the dominant carrier, with 56.9% of the traffic held in 2007. The rest of the traffic is generated by LCCs, which account for 33.4% of the total passenger traffic [39] and other charter operators; traditional carriers are almost absent. The airline business models serving Olbia Airport are perfectly coherent with the characteristics of passenger traffic, i.e. seasonal and primarily related to tourism (as the prevalent north versus south traffic highlighted by the route map in Fig. 2 shows). Starting from 1989, the year in which Geasar S.p.A. began its activities, the airport traffic has continuously grown at an average rate of 5% per annum, reaching 1.5 million passengers in 2003 (in the same year the annual growth rate was 12%). This passenger traffic increase, emerging mostly from the relationship with Meridiana Airlines, has helped the airport to enhance its operational capacity and consequently reduce unit costs. The sharing of risk and investment costs has been another important motivation for the airport. In 2004, for example, the process of renovation and expansion of the airport infrastructures was completed, with an investment of 46 million Euros. Part of this investment was also related to the creation of a commercial area for developing non-aeronautical revenues, according to the new airport’s commercial business model. All the drivers mentioned before are summarised in Table 5.

Figure 2: Olbia Airport route map. Source: www.rati.com.

30

Regional Airports

Table 5: Olbia Airport –airline interaction drivers. Interaction category

Airlines drivers

Airports drivers

Efficiency/ Cost and risk sharing (maintenance operational drivers infrastructures), customised services Market power/ strategic drivers

Risk and cost sharing (airport expansion investment; declining unit costs), higher service quality (increased operational capacity) Create a strong hub (Olbia is Improved connectivity Meridiana hub), control of airport (development pax traffic), slots (Meridiana is the main shareholder regional social benefits (tourism of the airport management company), development), increasing regional social benefits non-aviation revenues (new commercial area)

Source: Our elaboration. 5.2 Case study B: Genoa Airport Genoa is a regional airport located on the northwest coast of Italy, which is public held (Genoa Port Authority 60%, Genoa Chamber of Commerce 25% and Roma Airport Management Company 15%). Genoa Airport is facing strong competition from five other airports which serve the same catchment area (the number of people living within the area in which Genoa Airport is reachable in approximately 2 hours of transport, i.e. less than 2 million people). This specificity, together with the absence of a focalisation strategy, explains Genoa Airport’s traffic trend, which shows a slower growth rate than its counterparts. In order to improve its competitive position, the management company has started, in the last few years, to invest in a growth strategy based on different types of interaction with airlines. The final goal is to increase passenger traffic through higher connectivity (in fact Fig. 3 shows a limited number of routes) and market diversification. In this way, the airport has intensified its relational network with different types of subjects, which can be classified into traditional carrier, charter carrier and LCC. Interaction drivers between Genoa Airport and different airline’s categories are next investigated and summarised in Table 6. The five traditional carriers that operate at the moment at Genoa Airport (Cai-Alitalia, Air France, Lufthansa, Iberia and British Airways) represent the great majority of airport traffic. Cai-Alitalia, in particular, accounts for almost 50% of passenger traffic, operating as the dominant carrier. The airport strategy towards this typology of carriers is focused on strengthening the company’s bargaining power by trying to decrease the dependence on the dominant carrier. In order to achieve this objective, Genoa Airport is building relationships with other traditional carriers. These types of relationship, mainly transactional based, are guided by the following airport’s drivers: promoting the region as a tourist destination as well as a connection airport, sharing market risk on a wider customer portfolio. Traditional carriers are interested in less congested airports, where it is easier to find slot availability and to acquire new traffic to redirect towards their hub. This is the kind of relationship developed between Genoa Airport and Lufthansa Airlines, through its controlled regional carrier Air Dolomiti. Genoa Airport is connected by Air Dolomiti to Munich, Lufthansa’s second hub, using small size vehicles with a high frequency (four connections per day). The two charter carriers that operate at Genoa Airport (Air Italy and TUI Fly) account for only 3% of the total passenger traffic. They began to use Genoa Airport only recently (in the last 2–3 years)

31

Regional Airports

Figure 3: Genoa Airport route map. Source: www.rati.com. Table 6: Genoa Airport–airline interaction drivers. Interaction category Efficiency/ operational drivers Market power/ strategic drivers

Airlines drivers

Airports drivers

Cost economies (better load factor), Risk sharing (wider customers’ market risk sharing (airport financial portfolio), higher service quality support), higher service quality (higher (customised services like free flight frequency, customised services) parking for pax and lounge room) Create a feeding traffic to the hub Face neighbour airports competition (Dolomiti airlines carries pax to Munich (higher pax traffic and improved Malpensa hub), higher slots availability, connectivity), increase bargaining develop traffic in the basin of the airport power (decreasing the dependence (co-terminalisation strategy) from Alitalia), regional social benefits (tourism development)

Source: Our elaboration.

32

Regional Airports

and their traffic has strongly improved. With reference to these partners, totally dedicated to the leisure industry, the airport’s non-economic objectives, like regional touristic development, acquire more relevance. In order to increase its attractiveness to outgoing charter networks, the airport management company has developed a co-terminalisation strategy with other Italian airports (e.g. Milan Malpensa) and has offered special conditions to passengers (e.g. free parking). The co-terminalisation strategy allows carriers to carry passengers originating from multiple points (in this case Genoa and Milan) to the same final destination on the same plane. In this case, the airline’s drivers seem to be the increased flight frequency and the achievement of a better load factor, while the airport’s drivers are cost and risk sharing. Regarding incoming charter traffic, Genoa Airport may not rely on the strong touristic attractiveness of the region, so it has opted to exploit some niche markets, like the cruise industry, which has in Genoa’s port the second Italian home port, and emerging markets like Russia and Scandinavia. Genoa Airport offers marketing incentives and facilities (dedicated lounge room) to tour operators to stimulate development of new routes. Genoa Airport is at the moment working on a plan for the establishment of a Ryanair base. Ryanair requirements could not be matched by the only airport company: it needs the support and the involvement of the local public administration. The success of new LCC routes depends on a deep and widespread marketing plan involving public entities of destination and origin airports. Bureaucratic and legislative obstacles in addition to time lags in organising and defying public intervention hardly fit in with the strategy of LCCs, which tends to catch strategic windows and to exploit market opportunities for brief periods of time. The four LCCs that operate at Genoa Airport (Ryanair, Belleair, Transavia and Blu-Express) account for 11.53% of the total passenger traffic. This type of interaction is particularly crucial for Genoa Airport, which has failed till now to establish a stable relationship with any LCC and has seen a dynamic opening and closing of air routes in the last 6–7 years. This trend has had an impact in terms of passenger rate volatility: for example, the LCC passenger rate has fluctuated in the last 2 years ranging from 12.52% in 2007 to 11.53% in 2008. The main interaction problems are related to low bargaining power of the airport, imbalance in the incoming and outgoing passenger traffic, scarce number of repeating passengers, high load factor requirements and limited contractual obligations for LCCs. 6 CONCLUSIONS In this paper, we have tried to explore vertical integration strategies between airports and airlines, with the aim of thoroughly understanding the interaction drivers. We have proposed a set of determinants distinguished by subject which can be broadly categorised into efficiency driven and market power driven (see Table 1) and then we have applied our assumptions in two case studies related to the Italian aviation industry. The investigated cases, which involve different types of subjects (regional airports, traditional airlines as well as charter carriers and LCCs), have shown the validity of the proposed drivers (as Tables 5 and 6 show). In particular, they seem to be not dependent on the different types of interaction, whereas they appear to be much more related to the characteristics of the partners involved in the relationship. In the case of regional airports, one of the most critical drivers is the creation of social benefits, whereas for LCCs efficiency drivers are prevailing (cost economies and risk sharing). These preliminary conclusions require a deeper and more well-documented analysis in the future. In order to achieve stronger support, the research needs more empirical investigation through a wider number of case studies.

Regional Airports

33

Further research in this field should also be focused on the future environmental developments that could affect the risk rate and the competition level in the aviation industry, such as the 2008 financial crisis. IATA’s latest data, in fact, outlines a strong passenger traffic drop (8% in the last 12 months). These environmental changes could, in fact, make vertical forms of interaction a strategic survival option for airports and airlines in order to maintain their competitive advantage. REFERENCES [1] Francis, G., Fidato, A. & Humphreys, I., Airport–airline interaction: the impact of low-cost carriers on two European airports. Journal of Air Transport Management, 9(4), pp. 267–273, 2003. [2] International Center for Competitiveness Studies in the Aviation Industry (ICCSAI), Fact Book 2008. La Competitività del Trasporto Aereo in Europa, ICCSAI Editors: Bergamo, 2008. [3] Cools, K. & Roos, A., The Role of Alliances in Corporate Strategy, The Boston Consulting Group Report, 2005. [4] Spekman, R.E., Forbes, T.M., Isabella, L.A. & MacAvoy, T.C., Alliance management: a view from the past and a look to the future. Journal of Management Studies, 35(6), pp. 747–772, 1998. [5] Lorange, P., Roos, J. & Broon, P.S., Building successful strategic alliances. Long Range Planning, 25(6), pp. 10–17, 1992. [6] Hamel, G., Competition for competence and inter-partner learning within international strategic alliances. Strategic Management Journal, 12, pp. 83–103, 1991. [7] Borys, B. & Jemison, D.B., Hybrid arrangements as strategic alliances: theoretical issues and organizational combinations. Academy of Management Review, 14(2), pp. 234–249, 1989. [8] Lei, D. & Slocum, J.W., Global strategy, competence building and strategic alliances. California Management Review, 35(1), pp. 81–97, 1992. [9] Rock, M.L. (ed), Fusioni e Acquisizioni. Aspetti Strategici, Finanziari e Organizzativi, McGraw-Hill: Milano, 1990. [10] Garette, B. & Dussauge, P., Alliances versus acquisitions: choosing the right option. European Management Journal, 18(1), pp. 63–69, 2000. [11] Goetsch, B. & Albers, S., Towards a model of airport–airline interaction. German Aviation Research Society, www.garsonline.de, 2007. [12] Franke, M., Innovation: the winning formula to regain profitability in aviation? Journal of Air Transport Management, 13(1), pp. 23–30, 2007. [13] Graham, A., Managing Airports: An International Perspective, Butterworth Heinemann: Oxford, 2001. [14] Klenk, M., New approaches in airline/airport relations: the charges framework of Frankfurt airport (Chapter 9). The Economic Regulation of Airports. Recent development in Australasia, North America and Europe, ed. P. Forsyth, Ashgate Editors: Berlin, pp. 125–139, 2004. [15] Meersman, H., Van de Voorde, E. & Vanelslander, T., The air transport sector after 2010: a modified market and ownership structure. European Journal of Transport and Infrastructure Research, 8(2), pp. 71–90, 2008. [16] Auerbach, S. & Koch, B., Cooperative approaches to managing air traffic efficiently – the airline perspective. Journal of Air Transport Management, 13(1), pp. 37–44, 2007. [17] Albers, S., Koch, B. & Ruff, C., Strategic alliances between airlines and airports – theoretical assessment and practical evidence. Journal of Air Transport Management, 11(2), pp. 48–58, 2005. [18] Koch, B., Opportunities and limitations to vertical alliance partnerships between airports and airlines. Proceedings of the 6th Conference on Applied Infrastructure Research, 2007.

34

Regional Airports

[19] Oum, T.H. & Fu, X., Impacts of airports on airlines competition: focus on airport performance and airport–airline vertical integration. Joint Research Transport Centre OECD, discussion paper 17, pp. 3–37, 2008. [20] Humphreys, I., Ison, S. & Francis, G., A review of the airport-low cost airline relationship. Review of Network Economics, 5(4), pp. 413–420, 2006. [21] Seth, A., Sources of value creation in acquisitions: an empirical investigation. Strategic Management Journal, 11(6), pp. 431–446, 1990. [22] Seth, A., Value creation in acquisitions: a re-examination of performance issues. Strategic Management Journal, 11(2), pp. 99–115, 1990. [23] Williamson, O.E., Economies as an antitrust defence: the welfare tradeoffs. American Economic Review, 58(1), pp. 18–36, 1968. [24] Lewellen, W.G., A pure financial rationale for the conglomerate merger. Journal of Finance, 26(2), pp. 521–537, 1971. [25] Vicari, S., Nuove Dimensioni della Concorrenza. Strategie nei Mercati Senza Confini, Egea: Milano, 1989. [26] Oum, T.H. & Park, J.H., Airline alliances: current status, policy issues, and future directions. Journal of Air Transport Management, 3(3), pp. 133–144, 1997. [27] Iatrou, K., Oretti, M., Airline Choices for the Future: From Alliances to Mergers, Ashgate Publishing Ltd.: Hampshire, 2007. [28] Clougherty, J.A., US domestic airline mergers: the neglected international determinants. International Journal of Industrial Organization, 20(4), pp. 557–576, 2002. [29] Doganis, R., The Airline Business, 2nd edn, Routledge: London, 2006. [30] Kim, E.H. & Singal, V., Mergers and market power: evidence from the airline industry. American Economic Review, 83(3), pp. 549–569, 1993. [31] Carney, M. & Mew, K., Airport governance reform: a strategic management perspective. Journal of Air Transport Management, 9(4), pp. 221–232, 2003. [32] Serebrisky, T., Market power: airports, vertical integration between airports and airlines. Public Policy Journal, Note Number 259, 2003. [33] Fuhr, J. & Beckers, T., Vertical governance between airlines and airports – a transaction cost analysis. Review of Network Economics, 5(4), pp. 386–412, 2006. [34] Kuchinke, B.A. & Sickmann, J., The joint venture terminal 2 at Munich airport and its consequences: an analysis of competition economics. Proceedings of the 4th Conference on Applied Infrastructure Research, eds F. Fichert, J. Haucap & K. Rommel, INFER Research Perspectives: Berlin, pp. 107–133, 2007. [35] Malighetti, P., Martini, G. & Paleari, S., An empirical investigation on the efficiency, capacity and ownership of Italian airports. Rivista di Politica Economica, 47(I–II), pp. 157–188, 2007. [36] Sciandra, L., Aeroporti e Infrastrutture: Prospettive e Criticità del Quadro Regolatorio, Rapporto ISAE, Priorità nazionali. Infrastrutture materiali e immateriali, 2008. [37] Assaeroporti, Analisi della Sostenibilità Economica e Proposte per lo Sviluppo la Mobilità, www. assaeroporti.it, 2006. [38] SH&E International Air Transport Consultancy, Capital Needs and Regulatory Oversight Arrangements. A Survey of European Airports. Aci Europe, 2006. [39] Rosato, P., Proprietà e Governo dell’Impresa di Gestione Aeroportuale, Cacucci Editore: Bari, 2008.

Regional Airports

POTENTIAL DEMAND FOR NEW HIGH SPEED RAIL SERVICES IN HIGH DENSE AIR TRANSPORT CORRIDORS C. ROMÁN & J.C. MARTÍN Departamento de Análisis Económico Aplicado, Universidad de Las Palmas de Gran Canaria, 35017 Las Palmas G.C., Spain.

ABSTRACT Demand analysis is a key element in the evaluation of public policies. The ex ante evaluation of large scale projects involving for example new high speed rail (HSR) services requires the estimation of reliable choice models to predict ridership shares of the new alternatives and to identify the main sources for traffic diversion and traffic generation. This paper analyses and forecasts potential demand for HSR services in the high dense air transport route: the line Madrid–Barcelona. The model aims to explain changes in the demand for interurban rail and air transport as a result of substantial improvements in the level of service due to the introduction of the HSR. Results highlight that the expected volume of demand for the HSR in the corridor is not enough to guarantee a positive social benefit of this project. Keywords: discrete choice modeling, intermodal competition stated preference, mixed RP/SP data.

1 INTRODUCTION Congestion at roads and airports terminals, road accidents and greenhouse gas emissions, represent nowadays the main externalities of the transport systems. These negative effects have raised serious concerns about the impact of infrastructures on regional development, the competitiveness of the transport systems and the environmental quality. To improve intercity mobility, attention has been focused on evaluating alternative transportation services which provide an efficient response to incremental demand in the near future. These include, among others, upgrading conventional rail services to new high speed services using advanced technologies. The impacts caused by investments in high speed rails (HSRs) have been analyzed in the literature in many different ways. Thus, we can classify the studies into the following groups: (i) general assessments [1–7]; (ii) evaluations of the economic profitability of particular corridors or areas, de Rus and Inglada [8, 9] for the HST Madrid–Sevilla, Levinson et al. [10] for Los Angeles–San Francisco, de Rus and Román [11] for the HST Madrid–Zaragoza–Barcelona, Steer Davies Gleave [12] and Atkins [13] for the case of the UK, de Rus and Nombela [14] for the European Union, and Martín and Nombela [15] for the case of Spain; (iii) assessments of the regional effects [16–19], studies of the impacts on accessibility [20–24] and, finally, regarding (iv) intermodal competition, Combes and Linnemer [25] studied the impacts of the creation of a new infrastructure connecting two points that coexists with old network infrastructure (like roads) using a game theoretic approach. Demand analysis is a key element in the evaluation of public policies. Investment decisions in transport infrastructures cannot be made independently of the volume of potential demand in the influence area of the new infrastructure, because it determines to a large extent the social benefits of the project. The ex ante evaluation of large scale projects involving for example new HSR services requires the estimation of reliable choice models to predict ridership shares of the new alternatives and to identify the main sources for traffic diversion and traffic generation. The analysis of the perceptions and preferences of passengers on interurban transport is not new in the literature. Discrete choice modeling is usually claimed as a proper methodology to assess and compare the preferences of passengers in the context of modal competition. The behavioral nature

36

Regional Airports

of disaggregate discrete choice models has led to a widespread use of this tool in the field of travel demand modeling. Some recent applications in the context of intercity travel mode choice can be found in Refs. [26–34] among others. The objective of this paper is to analyze and forecast potential demand for new HSR services, the Madrid–Barcelona route, entering in a market characterized by a high density air transport services, and then to compare the estimate and actual measured demand after the entrance of the HSR in the market. The analysis is based on the estimation of a disaggregate demand model that uses information about travelers preferences in the existing modes in this corridor. Preferences for the new HSR alternative are obtained from a stated choice experiment facing air plane users with this new option. The rest of the paper is organized as follows. Section 2 presents some relevant information about the market background. The main characteristics of the datasets used in the analysis are presented in Section 3. Section 4 provides the demand model as well as estimation results. An analysis of potential demand and potential competition of the HSR with air transport is presented in Section 5. Finally, our main conclusions are presented in Section 6. 2 MARKET BACKGROUND Rail sector has received considerable attention within the European transport policy during the past decades, focusing more recently on the development of the HSR networks. Spain has adopted this policy and will account by 2010 with one of the densest HSR network in the world, with more kilometers than Japan and France, countries that have been pioneers in the development of their HSR networks (Table 1). In fact, the Spanish Infrastructure Master Plan has considered an expenditure of nearly 250 billion Euros in the development of the HSR until the year 2020, and by this time, the Spanish network is expected to reach 10,000 km. Figure 1 shows the Spanish HSR network. Some lines are in operation (in green), some under construction (in yellow), some are planned (in red) and some other are in study (in pink). The picture gives an idea of the density of the network in the future. The new HSR link between Madrid and Barcelona is a good example of this type of policy. Madrid and Barcelona are the most important metropolitan areas of Spain with five and three million of inhabitants, respectively; they represent important centers of economic activity and there are more than seven million trips per year. Regarding this line, the main difference with respect to the policy followed within the European Union is that Spain accounts with a very good level of other transport facilities linking these two cities. Until the inauguration of the HSR line in February 2008, RENFE provided rail services using conventional trains (Talgo) that connected the two cities in 5 h and 30 min offering a service frequency of eight departures per day in each direction. There is a 625-km motorway connecting these two cities in about 6 h and 20 min (being the link Zaragoza–Barcelona a toll road). Madrid and Barcelona are also connected by air shuttle and regular services and this route has recently become the most important domestic market in the world, with near five million passengers per year. This route is comparable, in terms of traffic, with other important domestic markets, e.g. Sao Paulo–Rio de Janeiro, Melbourne–Sydney, and Cape Town–Johannesbourg (Table 2). Table 1: Length of the HSR networks. Country Spain (year 2010) Japan (Shinkansen) France (TGV)

HSR network (km) 2,230 2,090 1,893

37

Regional Airports

Figure 1: High speed railway network in Spain. Year 2008. Table 2: High dense air domestic markets. Route Madrid–Barcelona Sao Paulo–Rio de Janeiro Melbourne–Sydney Cape Town–Johannesbourg

Flights per week 971 894 851 831

Source: OAG BACK Aviation Solutions. The new high speed train (HST) replaced conventional train services and entered in the market with the objective of attracting new passengers and deviating traffic from the air transport (the principal mode in the route), offering an improved level of service. In Table 3 we compare the main service attributes of the HST with the rest of the modes. As we can see, the HST improves substantially the level of service of the conventional train in terms of travel time (50% reduction) and service frequency (more than 50% increment in departures per day). This new alternative represents a close substitute of the air transport mode, but while total travel time (in vehicle plus waiting time) is similar in both modes, the plane is still very competitive in service frequency. In fact, one of the air sector reactions to the HST entrance was to use planes of reduced size in order to maintain a good level of service frequency.

38

Regional Airports

3 DATA COLLECTION 3.1 The sample selection A survey of travelers in the Madrid–Barcelona corridor was conducted in order to obtain information about the principal modes of transport: car driver, car passenger, bus, conventional train, and airplane. Data were collected during the second term of the year 2004, avoiding vacation periods (Easter and local holidays). At this time, the HST was already operating between Madrid and Zaragoza, but rail services between Madrid and Barcelona were still provided by conventional trains. A specific revealed preference (RP) questionnaire was designed for each mode of transport. As the main purpose of the research was to study the potential demand for the new HST alternative to stated choice (SC) experiment was included in the questionnaire of plane users as this mode was thought to be a close substitute of the new service. A choice-based sample was selected in order to ensure that sufficient observations are obtained for each of the modes currently chosen by each traveler. Modal shares in the sample were determined trying to replicate modal shares in the population, given the available information at the time the surveys were carried out and considering a maximum error of 10% [35]. Table 4 presents the modal share in the sample. Air travel is by far the dominant mode in this corridor, transporting over 66% of travelers. Car travel is the next largest market (12%), followed by train (11%) and bus (8%). The survey was randomly administered to bus, train and plane travelers through personal interviews. Bus users were interviewed in the bus station, train users inside the train, while air transport users were approached at the corresponding boarding gates at the Barajas Airport (Madrid). For the latter, interviews were completed with the aid of personal computers that allowed us to implement a stated choice experiment (facing the actual alternative with the new HST) according to the current trip experience. Flights and scheduled trips by bus and train were sampled over 1 week and at various times of the day in order to capture both peak and off peak travelers. A self-administered Table 3: Comparison of the main service attributes. Attribute

HST

Travel time (in vehicle) Fare/(fuel + toll): Regular Shuttle service Frequency (departures/ day one way)

Plane Conventional train

2 h 38 min

1h

5 h 30 min

102€ 163€

96€ 199€

65€

18

138

8

Car

6 h 22 min 8–9 h 70.52€ (24.15€ toll) 28€–35€–41€ (46.37€ fuel) 26

Table 4: Modal share in the sample. Mode Car driver Car passenger Bus Train (conventional) Plane

Bus

Travelers

%

38 18 39 51 295

8.62 4.08 8.84 11.56 66.89

39

Regional Airports

questionnaire was randomly distributed to motorist traveling between Madrid and Barcelona. Individuals were located at petrol stations strategically placed in the national motorway A-II and were asked to complete the questionnaire and mail it back. 3.2 The revealed preference questionnaire In all cases, the RP questionnaire was divided into four sections of questions: identification data, current trip information, household information and personal information. Details of the current trip included trip origin and destination, travel cost, time spent on the main travel mode, access and egress mode and the time taken to access and egress the main mode, waiting time, trip purpose, trip frequency, habitual mode for a similar trip, and other details needed to measure the attributes of the non-chosen alternatives. Socioeconomic information was collected at household and individual level. Household questions included: members in the household, number of cars and household income. Individual level questions include: age, gender, education level, activity, number of working hours, job position and personal income. In our sample, total travel time by plane (3 h 10 min) is substantially less than in the rest of the modes (6 h 20 min in car, 7 h 8 min in train and 9 h 39 min in bus) but, in this mode the proportion of access and egress time reaches about 70% of the total trip duration. Nearly 56% of trips were mandatory (work and education) being this percentage over 60% for plane and train travelers. The highest proportion of non-mandatory trips is found among bus travelers (74%). Regarding gender, 54% of travelers were men. We also observed substantial differences in per capita weekly income, ranging from 167 € for car passengers to 351 € for plane users. Table 5 shows the classification of the habitual mode used for a similar trip in terms of the chosen mode for this trip. Figures show that most of plane users (85.76%), use regularly this mode for traveling between Madrid and Barcelona, and rarely use conventional train and bus. This is not the case for the rest of the modes where the fidelization rate does not exceed 52%. Also a substantial percentage of people in the sample who chose car declared that this was their first trip. This percentage is very low for plane users (7.12%) which demonstrate that this segment (plane users) corresponds to frequent travelers. 3.3 The stated choice experimental design Given that the HSR services were not currently available at the time this study, the ability to predict HSR market share using RP information about the existing modes is not possible. For this reason, a SC experiment was designed and included in the questionnaire devoted to plane travelers. The experiment creates hypothetical choice situations facing plane users with the new HSR alternative. These choice scenarios are created using the actual trip context to give the experiment more realism. Table 5: Chosen mode vs. habitual mode. Habitual mode (%)

Chosen mode

Plane

Car

Train

Bus

First trip

Plane Car Train Bus

85.76 10.71 13.73 0.00

5.08 41.07 13.73 10.26

1.69 8.93 47.06 0.00

0.34 5.36 0.00 51.28

7.12 33.93 25.49 38.46

40

Regional Airports

The attributes included in the experiment represent typical level-of-service variables like travel time, access time, travel cost and headway (time between two consecutive services). We also include the latent variables reliability and comfort. This set of variables helped us to define the global quality of the alternative in each choice situation. In order to simplify the exercise and reduce the respondent burden (avoiding unnecessary biases), the effect of other attributes, such as the waiting time, was measured only in the RP context. (Other reason for not including waiting time in the SP experiment is that a substantial amount of this time is imposed by the safety control regulations at airports and it is out of control of the managers of the transport system.) All the effects are posteriorly included in the hybrid utility constructed in the mixed RP/SP estimation method. To gain realism, the levels assigned to some attributes in the SC exercise were customized to each respondent experience pivoting the information provided by the RP questions around the reference alternative (the plane, in this case). Thus, the levels of travel cost and access time were defined in terms of the values experienced by the sample respondents and plausible percentage variations according to the available information about future fares and access time for the HSR were also considered. The attribute levels used in the SC experiment are summarized in Table 6. Table 6: Attributes and levels. Mode Attributes Travel cost (cv) Travel time (tv)

Access + egress time (ta)

Levels

Plane

HSR

0 1 2 0 1 2 0 1 2

cv × 1.10 cv cv × 0.90 1 h 20 min 1 h 10 min 1h ta × 1.20 ta ta × 0.80

cv cv × 0.90 cv × 0.80 2 h 45 min 2 h 30 min 2 h 15 min ta ta × 0.90 ta × 0.80

Departure Departure before 9:00 after 9:00 Every 30 min Every 60 min Every 15 min Every 30 min 30 min delay (Inside the plane) 15 min delay (in the boarding gate) Departure on time Low: Small leg room Narrow seats High: Ample leg room Wide seats

Departure Departure before 9:00 after 9:00 Every 60 min Every 90 min Every 30 min Every 60 min 10 min delay

Frequency (headway) (f)

Reliability (r)

0 1 0 1

Comfort (C)

2 0 1

cv = travel cost in plane. ta = access + egress time in plane.

5 min delay Departure on time

High: Ample leg room Wide seats

41

Regional Airports

An experimental design consisting in nine scenarios for each alternative was created using the program WINMINT (a standard software, developed by Rand Europe http://www.hpgholding.nl/ (the former Hague Consulting Group (HCG)), which is frequently used to conduct SC experiments). Table 7 presents the combination of attribute levels in the experimental design. The program automatically created nine different choice sets for each person, selecting at random one scenario in each alternative. Thus, for example, if the scenario 3 was selected for the plane and the scenario 8 for the HSR, the choice set in Table 8 would be presented to the traveler. Thus, every respondent (i.e. the 295 plane users) provided nine stated preference (SP) observations obtaining a total of 2,655 statistical observations. After removing 179 inconsistent responses (those where the individual chose the worse alternative), we obtained a mixed RP/SP database of 2,917 observations. 4 THE DEMAND MODEL Discrete choice models have been widely used to study consumers’ behavior. The main interest lies on their ability to predict decision maker’s choices and to analyze demand response to changes in the

Table 7: Attribute levels in the experimental design. Plane (attribute levels)

HSR (attribute levels)

Scenario.

cv

tv

ta

r

f

C

cv

tv

ta

r

f

C

1 2 3 4 5 6 7 8 9

0 0 0 1 1 1 2 2 2

0 1 2 0 1 2 0 1 2

0 1 2 1 2 0 2 0 1

0 2 1 1 0 2 2 1 0

0 0 0 1 1 1 0 0 0

0 1 0 0 1 0 0 1 0

0 0 0 1 1 1 2 2 2

0 1 2 0 1 2 0 1 2

0 1 2 1 2 0 2 0 1

0 2 1 1 0 2 2 1 0

0 0 0 1 1 1 0 0 0

1 1 1 1 1 1 1 1 1

Table 8: Example of choice situation.

Travel cost Travel time Access + egress time Reliability Service frequency Comfort

Plane

HSR

99 € 1h 36 min 15 min delay Every 30 min Low (small leg room)

72 € 2 h 30 min 45 min 5 min delay Every 60 min High (ample leg room)

Actual Travel cost: 90 €, Actual Access time: 45 min, Departure before 9:00 Which alternative do you prefer for a trip like this one?  Plane  HSR

42

Regional Airports

attributes of the alternatives. The classic model of rational choice is based on two fundamental properties: consistency and transitivity. The first implies that the same choice selection should be obtained under identical circumstances; the second provides a unique ordering of alternatives on a preference scale (see e.g. Ref. [36]). A choice among a set of alternatives requires the application of a decision rule. The utility maximization behavioral rule lies under the scheme of the rational choice and normally implies a compensatory decision process, i.e. individuals made trade-offs among attributes in determining the alternative with the highest utility. Since the analyst does not have full information about the utility of the decision maker q for the alternative j, it is modeled as the sum of two components: a deterministic or observable utility Vjq, expressed in terms of a vector of attributes (Xjq) of the alternative and a vector of socioeconomic characteristics of the individual (Sq); and a random term ejq, representing the portion of utility unknown to the analyst. Thus, the true utility to the decision maker is represented by the random variable Ujq = Vjq + ejq; and therefore, the analyst, under the assumption of utility maximization, is only able to model the choice probability of the different alternatives. Different assumptions about the distribution of the unobserved portion of utility ejq result in different representations of the choice model. Thus, the famous multinomial logit (MNL) and nested logit (NL) models are obtained when ejq are i.i.d. extreme value and a type of generalized extreme value, respectively (see Refs. [37, 35] for more details about the derivation of choice probabilities in random utility models). In this paper we use a mixed RP/SP data set. In order to obtain the same variance in the error terms for the RP (se2 ) and SP (sh2 ) utilities, a scale factor µ satisfying se2 = m2 sh2 needs to be estimated [38]. Bradley and Daly [39] proposed an estimation method based on the construction of an artificial NL structure, usually referred as the ‘Nested Logit trick’, where RP alternatives are placed just below the root and each SP alternative is placed in a single-alternative nest with a common nest parameter µ [35]. We specified modal utility in terms of the main level-of-service attributes, namely, travel time (tv), travel cost (cv), waiting time (te), access + egress time (ta) and service headway ( f ). We considered a linear-in-the-parameter (but not linear-in-the-attributes) specification that included transport costs divided by the expenditure rate (g, defined as per capita family income divided by available time, that is, total time per period (a week in this case) minus working hours) (following Jara-Díaz and Farah [40]). We also included cost squared terms divided by the expenditure rate and the income (I, defined as per capita family income per week), as we obtained a significant proportion of money spent in transport, for the different modes, as recommended by Jara-Díaz [41]. This specification indicates that the marginal utility of the travel cost (which coincides with minus the marginal utility of income) varies with g, I and cv yielding a different value for each individual. We also define two different interactions, namely T (trip motive) with travel time and access + egress time with a dummy variable Ta < 60 (1, if access + egress time is less than 60 min, 0 otherwise). The last dummy variable is used to capture the time intensity of this component on the total travel time. For the SP alternatives (the new HST and plane) the utility was defined in function of the attributes included in the choice experiment: travel time (tv), travel cost (cv), access + egress time (ta), service headway (f), reliability (r), expressed in terms of the delay time and comfort (C). The latter was specified interacting with travel time in order to obtain the perception of comfort in terms of the duration of the trip as well as the perception of travel time in terms of the level of comfort. Regarding model structure we tested different NL structures for the RP alternatives. The best fit was found in the case of correlation between the plane and the conventional train. Estimation results are shown in Table 9. All parameter estimates have the expected sign and are significant at a 95% confidence level, with the exception of the headway, the waiting time and the

43

Regional Airports

Table 9: Estimation results. Parameter

Estimates (t-test)

Car driver constant

Ccc

Car passenger constant

Cca

Bus constant

Cb

Train constant

Ct

Travel time (tv)

qtv

Travel cost/g (cv/g)

qcv/g

Headway (f)

qf

Travel cost2/gI (cv2/gI)

qcv2/gI

Access + egress time (ta)

qta

Waiting time (te)

qte

Travel time_Work + education (tv × T)

qtv_T

Access + egress time_T acc + egr < 60 (Ta