Accepted Manuscript

Title: Traffic Demand Forecasting and Flexible Planning in

Airport Capacity Expansions: Lessons from the
Madrid-Barajas New Terminal Area Master Plan

Author: Athina Sismanidou Joan Tarradellas

PII: S2213-624X(16)30035-9
DOI: http://dx.doi.org/doi:10.1016/j.cstp.2016.08.003
Reference: CSTP 110

To appear in:

Received date: 25-5-2014

Revised date: 10-2-2015
Accepted date: 19-8-2016

Please cite this article as: Sismanidou, Athina, Tarradellas, Joan, Traffic Demand
Forecasting and Flexible Planning in Airport Capacity Expansions: Lessons from the
Madrid-Barajas New Terminal Area Master Plan.Case Studies on Transport Policy
http://dx.doi.org/10.1016/j.cstp.2016.08.003

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Traffic Demand Forecasting and Flexible Planning in Airport Capacity
Expansions: Lessons from the Madrid-Barajas New Terminal Area Master Plan

Athina Sismanidoua, Joan Tarradellasb,

a,b EADA Business School, c/ Aragó 204, 08011, Barcelona, Spain

Corresponding author:
Athina Sismanidou
Tel: (0034) 616 94 06 53
Fax: (0034) 934 01 60 54
E-mail: asismanidou@eada.edu or atids@yahoo.com

Highlights

> Infrastructure projects can gain from robust and neutral forecasting
methods.
> Capacity expansions can benefit from a flexible planning approach.
> The Madrid-Barajas master plan lacked neutrality and flexibility.
> Airport infrastructure projects can benefit from a real options methodology.

1
Abstract: Liberalization and constant but irregular traffic growth has made the airport sector hard
to recognize compared to some decades ago. The new paradigm for airport management calls for
flexible capacity planning models. However, airport infrastructure investments are cumbersome
and almost never efficient in the short run. It is difficult to assess the right time for and right size
of such an investment. Traffic demand forecasts become of critical importance in the planning
phase. This case study performs an ex post evaluation of the demand forecasts included in the
Madrid-Barajas master plan, in view of the airport's last capacity expansion in 2006. We deduct
that a more robust (and neutral) forecasting methodology, coupled with a more flexible approach
to planning in general, could have led to a more efficient and economically sound infrastructure
project.

Keywords: Flexible planning in transport infrastructure; Air traffic demand forecasting; Real
Option based infrastructure appraisal

1. Introduction

There is a longstanding belief that government investments in public transportation infrastructure

generate employment opportunities, increase productivity, and benefit countries and their people,
directly or indirectly (Phang, 2003; Pantazis and Liefner, 2006; Percoco, 2010). Some researchers
argue that transport investment above necessary levels works to the detriment of other economic
sectors, (Costas-Centivany, 1999) and demonstrate, with their analysis, the absence of a causal
relationship between air capacity infrastructure projects and economic well-fare (Banister and
Berechman, 2001). However, most analysts confirm the existence of the positive impact the
increase in airport capacity has on GDP and investment growth (Hong et al., 2001). In line with this
view that investment in airport infrastructure supports economic growth, policy makers across the
globe consider it vital to protect and develop the aviation sector through occasional airport capacity
expansion projects.

Air transport, at the moment, is one of the strongest growth sectors across the world. Despite the
cyclical nature of the aviation sector, both demand and investment in air transportation continue to
increase. Forecasts through 2032 predict that yearly air passenger traffic growth rates will reach
five to six percent worldwide and four percent in Europe (Airbus, 2013; Boeing, 2013).

Growth in air passenger traffic has resulted in many major worldwide airports confronting capacity
constraints (Cohen and Morrison, 2003). The problem has been aggravated by the hub-and-spoke
strategies of these airports’ central carriers. In order to be competitive, airlines with hub-and-spoke
strategies schedule most feeder flights during certain peak day times that correspond to traveler
preferred choices (Fageda and Flores-Fillol, 2012). Blending feeder traffic over a short period of
time results in congestion (Hamzawi, 1992), a problem shared to a greater or lesser extent by many
international airports. In Europe, more than 25 large airports are close to reaching saturation levels

2
and demand at six of Europe’s largest airports exceeds capacity most hours of the day (Madras and
Zografos, 2010).

To deal with capacity constraints, increasing airport capacity potential has been the objective of
extensive investment projects in Europe. Over the last decades, all leading European international
hubs—with the exception of London-Heathrow, where there still is an ongoing discussion among
public and private stakeholders—have undergone major investment in capacity expansion. Paris-
Charles-de-Gaule has developed two pairs of independent runways and has completed several
terminal expansion projects to increase capacity. Amsterdam-Shiphol has introduced two additional
runways to form the only European triple parallel landings management system. Frankfurt-am-
Main inaugurated a third runway in 2011, and Flughafen Münich introduced a second terminal,
currently being expanded.

Airport infrastructure investment is inherently cumbersome. The size of such investment is usually
large and the time span for implementation typically long (Miller and Clarke, 2007). Thus, the
investment may not be fully capitalized for many years (Pels et al., 2001). While investment in
additional capacity is critical to fight congestion and facilitate air transport growth (Phang, 2003), a
“build it and they will come” approach to demand forecasting can lead to an overestimation of the
true potential of an airport, and to the construction of underutilized infrastructure (Flyvbjerg, 2005;
Flyvbjerg et al., 2005). Although long-run capacity constraints need to be addressed by
infrastructure investment, the critical challenge lies in deciding the time and the appropriate level
of the investment (Forsyth, 2007).

Considering the economic risk associated with airport capacity expansion projects, any lessons
from similar past ventures can prove valuable. In this study, we examine the 2006 Madrid-Barajas
airport expansion to draw conclusions about two relevant topics in relation to airport and other
transport infrastructure expansions: 1. demand forecasting methodologies, and 2. flexible planning.
In the analysis that follows, we show how the master plan drafted for the Madrid Barajas terminal
expansion could have profited from an improved demand forecast methodology and a more flexible
approach to planning.

The remainder of this paper is organized as follows. Section 2 provides a literature review. In
section 3, we describe the demand forecast approach in the Madrid-Barajas master plan and its
limitations in view of subsequent events in the Spanish economy and aviation sector. In section 4,
we discuss how the master plan could have been improved by incorporating a flexible approach to
planning. In section 5, we present our conclusions on the infrastructure policy lessons to be learned
from the Madrid-Barajas case study.

3
2. Supporting Literature

A number of researchers (Szyliowicz and Goetz, 1997; Burghouwt, 2007; Xiao et al., 2013)
conclude traditional master planning and the rigidness of the “rational” model must be replaced by
more flexible approaches. A principal area in the planning process where flexibility can be
introduced is demand forecasting. As large-scale projects find their justification in passenger traffic
matching the dimensions of the expansion, demand forecasting has become an integral part of the
planning process (Solak et al., 2009).
The most important difficulty with forecasting lies in the unpredictability of the future that renders
any forecasting exercise inaccurate in nature (Xiao et al., 2013). Different areas of methodological
improvement have been identified in literature to deal with the limitations of forecasting.

First, in order to deal with “uncertainty,” analysts suggest using sensitivity analysis and different
market scenarios. The longer the demand outlook, the more scenarios will be involved and
measured (Burghouwt, 2007).

Another area of improvement that has been suggested consists of including advanced “judgmental”
approaches, e.g., taking into account the opinion of all stakeholders (Armstrong, 2005; Burghouwt,
2007). Different stakeholder participation in the forecasts is also a safeguard against the lack of
neutrality associated with infrastructure project planning. As Szyliowicz and Goetz (1997) explain,
one cannot ignore the political dimensions involved in the planning process, and, according to the
same authors, it is a well-known fact that forecast analysts, on many occasions, end up revising
their figures until they meet their superiors’ expectations.

Improvements in forecasting methodology also call for better judgment on what should be the
required level of sophistication. The degree of methodological complexity is dictated by the
requirements of the infrastructure under consideration, as well as the data, skills, and facilities
available to the planner (Ashley et al., 1995). Traditionally, the forecasting methods employed by
air transportation researchers and industry practitioners vary in sophistication from basic, trend
extrapolation methodologies to econometric-based models, the latter ranging from linear
regressions to more complicated disaggregate mode choice models. Disaggregate mode choice
models have now become prevalent options for transportation researchers, but linear regression
methodologies are being employed heavily as well (Pels et al., 2000; Clewlow et al., 2014; Ishii et
al., 2009). Disaggregate mode choice models are particularly well suited to study alternatives in
transportation, i.e. competition among airports or between air and rail travel. The foundation of the
behavioral theory behind these models is that individuals maximize the “utility” they attribute to

4
travel services and that utilities can be added up. In general, different advantages and disadvantages
have been associated with econometric modeling and frequently different forecasting techniques
and models are used in a complementary fashion, depending also on the demand level examined,
e.g., by airport or regionally. Although, sophistication may not necessarily guarantee accuracy, a
careful look at the different variables affecting demand allows planners to identify risk areas and
treat them adequately (Carson et al, 2011). Selecting the right variables is usually a byproduct of
regression analyses of historic data that establish the relationship between traffic drivers and the
traffic segment under analysis (Wei and Hansen, 2006; Pels et al., 2009).

Another area of investigation gaining increased attention by researchers is the application of the
real option methodology in airport infrastructure valuation (Martins et al., 2014). Real options
allow planners to financially evaluate an expansion plan depending on the outcome of different
scenarios based on a staged approach to the build-out. Unlike the net present value (NPV) valuation
approach, the real option methodology allows planners to evaluate options such as “abandoning” a
project, “switching” the purpose of the infrastructure, or considering a “modular implementation,”
and compare such options against a one-time, upfront engagement. The ability to adapt usually has
an increased cost compared to a single implementation. By correctly evaluating the value of such
options, conclusions can be drawn as to the financial consequences of a flexible plan versus one
that cannot be adapted to changes in the environment (Ingersoll, 1992; Trigeorgis, 1996). In
contrast to traditional master planning, the real option approach does not rely on a single, most-
likely forecast. Instead, it employs techniques that seek to maintain the flexibility to adapt
regardless of any future scenario. The prevalence of real options methodology, as an instrument of
value in infrastructure projects, further supports the replacement of rigid strategies with more
flexible ones.

3. The Adolfo Suarez Madrid-Barajas Expansion

3.1 The New Terminal Area T4 and T4S

Adolfo Suarez Madrid-Barajas airport, with 39.7 million passengers in 2013, is the largest airport
in Spain and the fifth largest airport in passenger volume in Europe. The Spanish airport system
comprises 47 airports and is publicly managed by the Spanish National Airport Authority, AENA.
Madrid-Barajas was renamed in April 2014 in memory of Adolfo Suarez, a key political figure
during the Spanish transition to democracy. Since its first commercial flights in 1933, the airport
had grown steadily. In 1997, air traffic had risen to almost 24 million passengers a year from one
million back in 1970, positioning Madrid-Barajas as one of the world’s major airports.

5
In response to continued passenger traffic growth and large capacity constraints in the Spanish
aviation market, AENA invested in a third runway, put in place in 1998, and pushed peak hour
capacity to 74 operations per hour, handling a demand of 28 million passengers.

In 1997, one year before implementation of the third runway, a new major expansion of the airport
was deemed necessary to increase passenger capacity to 70 million and peak hour capacity to 120
operations. Similar to most policy makers, who regard air transportation as a key strategic asset for
national progress (Miller and Clarke, 2007), Spain’s government was eager to invest in the Madrid-
Barajas airport expansion to maintain and boost the Spanish aviation sector’s competitiveness. The
objective was to reinforce Madrid as a major international hub airport, which would act as the
principle gateway between Latin America and Europe, competing with new generation airports,
both in terms of efficiency and customer service.

The new Madrid-Barajas, with one million square meters distributed among five terminals, was, at
its inauguration in early 2006, the largest airport in the world in terms of terminal area. The cost 6.2
billion Euros (El Pais, 4/2/2006), about double the initial budget, and would allow some 35 million
additional passengers a year, almost double the Madrid-Barajas capacity before implementation.

Passenger traffic in Madrid continued to grow for one more year following the inauguration,
reaching 52 million passengers in 2007, and immediately thereafter began a steady drop through
2013. During the 2007–2013 period, passenger flow at Madrid-Barajas decreased 23.6%, from 52
to 39.7 million.

3.2 The Master plan’s demand forecast versus actual demand

The New Terminal Area project was described and justified in the Madrid-Barajas Master plan
(1999), ratified by the Ministry of Public Works and Transport.

Justifying the expansion project, the strategic rationale of establishing Madrid-Barajas as the
“principal” gateway for EU-Latin America traffic and consolidating its position as an international
hub was supported by a passenger demand forecast over a 25-year period. Future demand was
forecasted as a linear, causal relationship with GDP forecasts for three different segments:
domestic, intra-EU, and international non-EU traffic. In order to support the use of GDP as the
main, and, in this case, almost sole parameter in forecasting demand at the airport level, an
interview with DeAnne Julius, British Airways’ chief economist, was cited. Finally, in the forecast
methodology description, it was clarified that fares were assumed to remain stable over the
projected period.

6
The first yearly series for the domestic passenger segment was based on a linear regression using
previous year data, from 1986 to 1997, where the Spanish GDP was the variable explaining 95% of
the historic traffic. The historic results of the regression were then applied to third party forecasts
of the Spanish GDP (IMF, Bank of America, and the Wharton Econometric Forecasting
Associates).

The second yearly series for the intra-European traffic demand was based on one independent
variable, the weighted average of the GDP forecast for Spain and the European Union on a 50/50
basis. The combined domestic and intra-European passenger forecasts explained 93% of the
demand growth at the Barajas-Airport during the period 1986 to 1997.

The third yearly series, international non-European traffic demand, was based on the assumption
that long-haul routes from Madrid would mainly have Latin American destinations. In this case, the
forecast used two independent variables, the Latin American average GDP and the Spanish GDP.
This assumption was based on actual data from 1997, showing that Latin American destinations
accounted for 10 percent of the airport’s total volume compared to five percent for US destinations.

The forecasts were presented in three scenarios, a baseline, plus a pessimistic and an optimistic
scenario; the latter two were created using minus and plus 10 percent variations of the baseline
scenario GDP forecasts (Table 1).

The GDP baseline traffic demand forecast was then fine-tuned with some additional elements.
First, the domestic segment demand was adjusted to account for the introduction of the High Speed
Rail (HSR) service in the Barcelona-Madrid route. Second, the impact of future improvements in
hub connectivity and a higher ratio of connecting traffic at the Barajas Airport were included. The
impact of this improved connectivity increased baseline scenario volumes by 0.7 million
passengers in 10 years.

[Table 1 goes here]

When looking at the data, we can observe that during the first years of the forecast, the analyzed
traffic was actually understated rather than overstated. However, as of 2014, the demand forecasts
are eight years behind schedule.

After the 2008 world financial and economic recession, passenger traffic volumes started to decline
in all major European airports (Fig. 1). However, while major European airports, as well as many

7
other domestic airports in Spain such as El Prat-Barcelona (Fig. 1 and 2), resumed growth in 2009,
the Madrid-Barajas airport passenger traffic continued to decline, dropping to 39.4 million
passengers in 2013, i.e. the pre-expansion traffic levels and a 24 percent deviation from the
forecast. Furthermore, when we explore the relationship between the Spanish GDP and, in theory,
the related domestic traffic segment evolution for the 2008-2013 period, we find that the two
parameters are unrelated (R2 = 0.03808).

[Figures 1 and 2 go here]

While the global financial crisis was an unpredictable event, there existed at least two, in our
opinion, predictable elements that made demand drop more in Madrid-Barajas than it did in other
Spanish airports. One was liberalization of Europe’s air traffic market, encouraging the rise of low-
cost carriers (LCCs) and point-to-point flights from Spain’s secondary airports to domestic and
other European destinations, avoiding the hub airport of Madrid-Barajas. The other was the
introduction of four, high-speed rail corridors, instead of only the Madrid-Barcelona route
envisaged in the master plan forecast. The impact of these two elements is discussed in more detail
in the sections that follow.

3.3 Impact of liberalization and LCCs on Madrid-Barajas passenger traffic

Liberalization in Europe came in a phased manner, in the form of three legislative “packages” in
1987, 1990, and 1992. The reform liberalized, among other things, access to routes and airport
slots. By 1997, the right to operate traffic in a foreign territory, known as “cabotage,” was
safeguarded by community legislation. The new legal framework introduced the end to monopolies
and incentivized the entry and growth of low cost carriers (Barbot, 2009; Barbot et al., 2013).

At the time the Madrid-Barajas master plan was being drafted, liberalization in the US market had
already resulted in the re-organization of domestic air networks (Graham, 1999; Reynolds-Feighan
and McLay, 2006; Graham and Shaw, 2008). By 2002, LCCs had reached eight percent in Europe
and were presenting an aggressive growth trend (Franke, 2004). Moreover, first-mover European
LCCs, like Ryanair and easyJet, had announced their aircraft purchases and leasing pipelines. Thus,
it would be reasonable to expect that at the time any demand forecasting exercise for the European
region would include comprehensive projections on the impact of these new market players.
Therefore, it is remarkable that LCCs did not form part of the demand equation in the Madrid-
Barajas master plan.

8
Low cost carriers not only capitalized on the spare capacity available in secondary and third-tier
airports, but also took advantage of the capacity expansion projects in major airports (Dobruszkes
and Mondou, 2013). When the Madrid-Barajas airport inaugurated its new terminal, designed as a
hub for Iberia and the Oneworld alliance partners, its former terminal opened up to low cost
carriers including Ryanair. Similarly, as soon as Barcelona-El Prat was expanded with a new
terminal in 2009, the old terminal was dedicated to the low cost airlines. In the case of the latter,
further capacity had been opened up earlier in 2005, when Iberia moved away from a two-hub
scheme to a single hub based in Madrid (Suau-Sanchez and Burghouwt, 2011, 2012). Fig. 3
illustrates the change in the market dynamics at Barcelona-El Prat.

In the Spanish airport system, spare capacity resulting either from Iberia withdrawing from the
airport or because the airports were not congested (e.g., Valencia and Bilbao), allowed new
competitors to service not only destinations with new demand but also destinations where True
Origin Destination (TOD) traffic justified the route. The impact on the Madrid-Barajas airport of
new direct services linking Spanish cities can be seen in Fig. 4 and 5. The figures illustrate how
other airports in Spain have grown in domestic traffic market share at the detriment of connecting
traffic via Madrid-Barajas with hub carrier, Iberia. A similar decrease in the connecting traffic via
Madrid-Barajas is also seen at the intra-European traffic level (Fig. 6). While EU traffic has been
relatively resilient to the crisis, connecting traffic for the segment at Madrid-Barajas, expressed as a
percentage of total traffic, has decreased considerably.

[Figures 3, 4, 5, and 6 go here]

With regard to long-haul traffic, the objective described in the Madrid-Barajas master plan was to
reinforce Madrid-Barajas as an international hub. The aim was to position Madrid-Barajas
alongside Paris-Charles-de-Gaulle, London-Heathrow, and Frankfurt-am-Main for the highly
lucrative long-haul traffic. Fig. 7 and 8 show that while overall EU/non-EU traffic has been
growing, connecting traffic for this market segment has been decreasing, thereby losing market
share. This is especially true for major European airports including Madrid-Barajas.

[Figures 7 and 8 go here]

The diminishing size of connecting international traffic in Europe can be explained by different
reasons such as the emergence of new economic poles, supporting the establishment of new direct
routes, the rise of Istanbul and Dubai as major hubs for international traffic, and airport
privatization, with new management seeking growth. Furthermore, long-haul traffic has been
influenced by a continuously growing number of liberalization initiatives at a cross-continent level,

9
including the EU-US Open Skies agreement. As expected (Burghouwt and de Wit, 2005), the EU-
US Open Skies agreement has contributed to more flexibility, vis-à-vis the airport points served
before its signature, and offers new potential for transatlantic routes from secondary airports (Bel
and Fageda, 2007, 2010). The successful inauguration, over the last three to four years, of
numerous point-to-point long-haul routes from non-hub airports in Europe, such as Manchester and
Barcelona, lead us to believe that the potential for further transatlantic routes from secondary
airports is a real trend that will further shape the aviation market.

In the Europe-Latin America traffic market (the focus of Madrid-Barajas), we observe that the
number of direct passengers has also been growing more aggressively than connecting traffic (Fig.
9 and 10).

[Figures 9 and 10 go here]

3.4 The High Speed Rail (HSR) impact

Spain experienced the strong effect of the HSR on airline traffic soon after the opening of its first
AVE corridor, joining Madrid to Seville in 1992. Within one year the train’s market share
increased from 21 percent to 82 percent, further affecting the market share of hub-and-spoke
network carriers in the long-haul market segment (Fu et al., 2012; Jiménez and Betancor, 2012).
Similarly, in the London-Paris route, rail captured about 80 percent of the traffic since being
introduced in 1994.

There is a wide array of literature on the social and economic impact of the HSR as well as its
impact on air travel demand (Janic, 2003; González-Savignat, 2004; Givoni and Banister, 2006).
Most studies examining the substitution of air transportation with HSR transport conclude that it is
hard for airlines to compete with rail between two cities with a short-haul distance and that the
improvement of travel times by the HSR reduces demand for short-haul and domestic air
transportation (Phang, 2003; Andrés López-Pita, 2005; Dobruszkes, 2010; Sánchez-Borràs et al.,
2012).

As expected in the Madrid-Barajas master plan, domestic air traffic from/to Madrid-Barajas was
affected by the introduction of additional HSR lines in the Spanish geography. Oddly, in the master
plan, only the impact of the Madrid-Barcelona line was anticipated, even though at the time there
were plans for three more HSR lines: Madrid-Malaga in 2007, Madrid-Valencia in 2010, and more
recently, Madrid-Alicante in 2013.

10
Table 2 shows the traffic performance of the new HSR services in Spain between 2007 and 2013.
Although the new HSR services have probably attracted additional traffic at each route (Jiménez
and Betancor, 2012), the numbers suggest that there has been a more serious disruption in air
demand than that foreseen in the Madrid-Barajas master plan. It remains to be seen if the expected
arrival of HSR to the Madrid-Barajas New Terminal Area will remedy the disruption by bringing
additional passengers or if it will lead to further over-capacity at Madrid-Barajas.

[Table 2 goes here]

4. The Madrid-Barajas Master Plan Limitations

The Madrid-Barajas master plan could have benefitted from several improvements in terms of
demand forecasting methodology and a more flexible planning approach.

To start with, several risk areas were either omitted or understated in the demand forecasting
analysis. First, the launching of LCC competitors was not even mentioned in the plan, although the
trend was clearly a reality in many European and non-European markets at the time the investment
was being assessed. Specifically, the impact LCC carriers would have on domestic traffic at the
Madrid-Barajas airport was understated. A possible explanation for excluding the LCCs impact
from the master plan could be that before the low-cost phenomenon spread in Europe, a stable level
of competition between airports and airlines for the intra-EU and international traffic created a
deceptive sense of certainty about future air traffic demand (Jimenez at al., 2014). The entry
barriers imposed by the grandfathering of slots (Dobruszkes, 2006; Barbot et al., 2013) made
forecasters believe that the national Spanish carrier Iberia was safeguarded against competition.
However, in a liberalized airline market, expanding capacity at congested airports removes such
barriers and encourages competition (Fageda and Fernández-Villadangos, 2009).

Second, the master plan authors could have contemplated the risk of unutilized capacity being
detrimental to Iberia, as well as incorporated in more detail Iberia’s strategic plans for the future.
Instead, the Madrid-Barajas master plan only dedicated a few paragraphs to Iberia’s future strategy.
Different researchers have emphasized the significance of density economies and network
externalities present in the hub-and-spoke systems, and have highlighted the importance of the
relationship between hub carriers and their airports (Starkie, 2002; Oum et al., 2008, Barbot et al.,
2013). The Iberia, Madrid Barajas relationship provides an example of this, in addition to other
notable cases in Europe such as British Airways and London-Heathrow, Alitalia and Rome-
Fiumicino, and Air France and Paris-Charles De Gaulle. If the Madrid-Barajas airport was to
compete directly with Europe’s mega hubs, the master plan demand calculations should have been

11
aligned with the partner carrier’s strategy in regards to its whole network of spokes. Such an
approach, would have allowed better insight into Iberia’s feeder operations network plans and an
early identification of unreasonable assumptions or risks. Instead, the risks associated with Iberia as
the hub airline, including increasing competition from other airlines flying from secondary hub
airports in Europe and the Middle East, did not form part of the future demand equation.

Third, although four Spanish government HSR routes were underway at the time the Master plan
was drafted, only the impact of the Barcelona corridor was included, in an incomplete manner, as
the estimated share of route cannibalization was not in line with evidence from other HSR route
impact already seen in Spain and in Europe.

Since all these risk areas were not totally unknown at the time the master plan was drafted, we are
compelled to argue that the planning most probably suffered from a lack of neutrality and a flawed
judgmental approach. The participation in the analysis process of a wider number of stakeholders
and aviation experts, including carriers other than Iberia, would most likely have resulted in the
inclusion of the above mentioned risk areas as well as the utilization of a more sophisticated
methodology.

While academic studies offer the basis for relatively sophisticated forecasting methodologies, the
degree of forecast sophistication varies among industry practitioners. Notwithstanding the
uncertainties inherent in the future, major aviation industry opinion leaders, such as Boeing,
Airbus, IATA, and Eurocontrol, perform econometric analyses based on hundreds of regression
models to develop demand elasticity estimates for different levels of aggregations, e.g., route level,
airline level, national level, and supra-national level. Had the Madrid-Barajas analysts followed a
more sophisticated methodology, they would have investigated separately the different traffic
segments and their risks, employing the most appropriate case methodology for each. The GDP
linear correlation, followed by the Madrid-Barajas analysts, was a very simplistic approach that
should have been fine-tuned using additional methodologies. As an example, competition from the
HSR could have been examined using a disaggregate mode choice model.

A more sophisticated approach, taking into account wider parameters, could also have formed the
basis for more comprehensive future scenario building. The concept of flexible planning assumes
that the identification and measurement of different scenarios are revisited on a continuous basis in
order to adapt infrastructure-related decisions to a changing reality (Burghouwt, 2007). At the same
time, the establishment of different scenarios, with their associated risks, would have allowed a
proper evaluation of alternative options available to the planners.

12
Several alternative options were available and could have been considered by the Madrid-Barajas
planners. To begin, an air congestion problem can be solved, reducing the ratio of demand to
capacity, either by decreasing demand or by increasing capacity, or a combination of the two. Thus,
increasing capacity by building new facilities would be one of the different alternatives
hypothetically available. Many airports close to saturation, like London-Heathrow, Frankfurt, and
Tokyo-Haneda, have taken a variety of such alternative measures to enhance capacity capabilities
and can serve as examples for other airport systems with similar requirements and structures.
Alternatives to a full-fledged infrastructure expansion consist of implementing improvements in
terminal design and planning, terminal processing, and operational practices to optimize existing
infrastructure capacity. Technological and operational innovations allow more and more efficient
solutions in these areas. Furthermore, airport pricing is another powerful tool for controlling the
demand part of the equation. Discriminative airport pricing can be employed to motivate larger
aircrafts with higher load factors, or disperse flight operations throughout the day. These “soft”
alternatives might not have sufficed to meet the Madrid-Barajas airport’s full needs and potential,
but could have been envisaged in combination with a lower scale infrastructure project.

Another alternative that could have been considered by AENA in dealing with airport capacity
congestion was diverting some long-haul routes to other secondary hubs (Forsyth, 2007). The
multi-hub strategy may be appropriate when the airports used individually generate enough traffic
based on high catchment areas and considerable flight operations in service (Sismanidou et al.,
2013). Especially when it comes to large long-haul markets, Burghouwt (2014) illustrates how
these are generally served from both a primary and secondary hub. The dual-hub strategy has been
tested, among others, by some airports in the US as well as in Germany, Switzerland, Brazil, China,
and to some extent, in Italy. In the case of the Spanish airport system, the lower scale expansion at
Barcelona-El Prat could have served as an alternative to provide capacity relief for Madrid-Barajas
and its hub airline Iberia. Instead, the approach, following the Madrid-Barajas expansion, was for
national carrier Iberia to use Barcelona-El Prat as a feeder airport until 2006, and thereafter almost
totally withdraw from that airport, giving space to the entry of new competitors there for key routes
previously served by Iberia, including long-haul destinations such as New York and Miami. One
could argue that the multi-hub alternative depends entirely on the carrier and not airport
management. However, the interdependence in the relationships between hub carriers and the hub,
make it difficult to believe that, at the time of the new airport planning, AENA and Iberia were not
in agreement on focusing connecting traffic in Spain via Madrid-Barajas, or that either side was
“forced” to reject a dual-hub policy.

Since the alternatives to capacity expansion discussed here may not always be sufficient to fight
congestion, new facilities may be deemed necessary. Nonetheless, Miller and Clarke (2007)

13
demonstrate that there may be an optimal congestion threshold for triggering capacity expansions.
The underlying logic is that the flexibility provided by an early investment must correspond to a
financial compensation from the ability to quickly meet market demand if traffic grows as
expected. If the expectations for demand growth are met, the value of being able to react in time
may be significant. However, the authors conclude that a strategy of delivering capacity in small
increments can be more beneficial compared to strategies developing large projects and long
response times. When air traffic growth projections are not met, alternative or lower scale
investments to increase capacity yield more benefits. The use of incremental, phased or “modular”
expansions, as alternatives to building colossal infrastructures right away, are being increasingly
documented (Smit, 2003). In 2004, even IATA advised governments to look at existing runway and
terminal capacities and then define mid- and long-term objectives for the “phased expansion or
development of new or existing airports.” By deferring significant contractual and design decisions
until they become absolutely necessary, the negative impacts of unforeseen changes can be
minimized.

Two additional, smaller infrastructure scale alternatives were briefly envisaged in the Madrid-
Barajas master plan, but immediately discarded on the basis that expected future capacity
requirements would soon lead to air congestion and make such alternatives obsolete. Without
further analysis, the Madrid-Barajas planners opted for a full-size solution, and, through one
project, increased capacity to satisfy demand levels estimated for a twenty-year time frame.

A final remark on the Madrid-Barajas master plan limitations is that the different infrastructure
growth alternatives could have been subjected to real option valuation. A valuation of each
alternative based on the real option methodology would have allowed a more accurate evaluation of
the risks and uncertainties involved and justified, on more accurate grounds, one strategy over
another.

5. Summary and Infrastructure Policy Lessons to be Learned

There is little doubt that in order to compete effectively for transfer traffic from other airports in
Europe, Madrid-Barajas required an infrastructure that would enable a smooth transfer process for
Iberia’s connecting passengers, adequate peak capacity, convenient connecting times, and attractive
leisure/shopping facilities. At the time the master plan was drafted, the already congested Madrid-
Barajas airport was experiencing continuous growth, along with the Spanish economy. Thus, the
expansion of Madrid-Barajas airport facilities seemed reasonable as well as imperative.

14
Nevertheless, when we evaluate the Madrid-Barajas master plan we encounter several limitations
in the plan from which we can draw conclusions that can be helpful for future airport planners. The
first lessons to be learned relate to the demand forecasting methodology chosen by the Madrid-
Barajas planners. Capacity expansion master plans, especially for large, multi-billion dollar scale
projects, can gain from the latest methodological advances in demand forecasting, proposed by
researchers and industry practitioners. Although complex demand forecast models might be
unrealistic for air traffic demand planners, a too simplistic estimating approach can lead to
unnecessary deviation from real demand and leave many large infrastructure projects unjustified. In
the case of the Madrid Barajas expansion plan, the use of GDP as the almost sole variable to
predict traffic was, in our opinion, insufficient. First, the traffic demand- GDP relationship is not
always linear. According to an analysis by major transport-consulting firm, Steer Davies and
Gleave, the relationship between passengers and GDP growth rates during the period 2004-2012 in
Europe can be best explained with a non-linear function. Analysts agree that the GDP-traffic
relationship is subject to considerable variations depending on the economic cycle, the traffic
segment analyzed, and the specific factors influencing traffic such as quality of service, fares and
competition from other modes (Oum et al., 1992; Calderon J., 1997; Brons et al., 2002).
Notwithstanding the often, but not always, linear relationship between traffic demand and GDP,
passenger traffic can be best estimated if fine-tuned with additional analysis of the driving factors
of the demand. In particular, for airport level demand analysis, disaggregate mode choice models
are particularly adequate for studying airport competition as well as competition with rail
transportation. Furthermore, today’s improved database mining capabilities make it possible to
access a growing amount of historical data from real traffic figures and passenger surveys, allowing
for better identification of demand drivers and follow up at the hub airport level.

Demand forecasting, in terms of airport capacity expansions, can also be improved by engaging as
many stakeholders as possible in the planning process. In our opinion, the new terminal planning
could have improved in judgment and neutrality with the participation in the discussion of a wider
selection of aviation and transport industry experts. At the time the Madrid-Barajas project was
being evaluated, there were some dominant market trends that shaped the market over the
following decade. These trends were omitted or underestimated as a result of, what we consider,
avoidable judgmental flaws. A parallel today, would be omitting from an airport expansion master
plan the trend of the growing number of long haul destination point-to-point flights, a side effect of
growing liberalization in the aviation market and new rising economies.

The other lessons to be learned from this case relate to the “monolithic” approach in the master
plan. This approach consisted of estimating demand for a single baseline scenario and reaching the
conclusion that the only solution to the expected capacity constraints was a major, new terminal

15
expansion. In our opinion, when dealing with airport capacity expansion, many different capacity
enabling alternatives must be considered and, to the extent possible, measured, in particular those
that can adapt better to future uncertainties. Such an alternative would be, for example, a project
with phased facility expansions, allowing planners to adapt construction to real traffic demand. The
real options methodology can aid the evaluation process by allowing planners to compare the risks
and benefits associated with each alternative.

Other capacity enabling alternatives, to be evaluated alone or in combination with a modular

approach to infrastructure expansion, include measures to optimize existing capacity. Existing
capacity can be optimized by increasing load factors or airplane capacity of the operations, or by
optimizing connecting traffic, scheduling, traffic control, and runway use to manage traffic peaks.
In this context, congestion-based pricing schemes, as well as market driven slot allocation
mechanisms, can be useful traffic demand management tools (Madas and Zografos, 2008).

The Madrid-Barajas master plan is not the only case, nor the last one, with rigid planning. As Khan
(1989) explains, most types of transport capacity investments can be made sequentially, but “there
exists a tendency on the part of many transportation planners to prepare inflexible master plans
which offer little opportunity to change the course of system development should circumstances
necessitate such as change." The current developments in the air transport sector, such as the traffic
volatilities brought by the rise of LCCs and the free-market regime, generate intensifying
uncertainty in traffic and passenger requirements. Thus, the ability to counter uncertainty through
flexible planning solutions is increasingly important. Had the Madrid-Barajas policy planners
allowed for uncertainty in the master plan demand projections, instead of basing their decision on a
single growth estimate, it is probable that they would also have opted for one of the phased runway
and terminal design alternatives, only vaguely described in the master plan.

Acknowledgements
We would like to thank our reviewers for their valuable remarks and Aeroports de Catalunya for
their help with True Origin Destination data.

16
References

Adler, N., Nash, C., Pels, E., 2008. High-Speed Rail & Air Transport Competition. Tinbergen Inst. Discuss.
Pap.

Adler, N., Pels, E., Nash, C., 2010. High-speed rail and air transport competition: Game engineering as tool
for cost-benefit analysis. Transp. Res. Part B Methodol. 44, 812–833. doi:10.1016/j.trb.2010.01.001

Alderighi, M., Cento, A., Nijkamp, P., Rietveld, P., 2005. Network competition - The coexistence of hub-
and-spoke and point-to-point systems. J. Air Transp. Manag. 11, 328–334.
doi:10.1016/j.jairtraman.2005.07.006

Andrés López-Pita, F.R., 2005. Impact of High-Speed Lines in Relation to Very High Frequency Air
Services, pdf, J. Public Transp. 8, 17–35.

Armstrong, J.S., Green, K.C., 2005. Demand Forecasting: Evidence-based Methods, Methodology.

Ashley, D.J., Hanson, P., Veldhuis, J., 1995. A policy-sensitive traffic forecasting model for Schiphol
Airport. J. Air Transp. Manag. doi:10.1016/0969-6997(95)00034-8

Banister, D., Berechman, Y., 2001. Transport investment and the promotion of economic growth. J. Transp.
Geogr. 9, 209–218. doi:10.1016/S0966-6923(01)00013-8

Barbot, C., 2009. Airport and airlines competition: Incentives for vertical collusion. Transp. Res. Part B
Methodol. 43, 952–965. doi:10.1016/j.trb.2009.04.001

Barbot, C., D’Alfonso, T., Malighetti, P., Redondi, R., 2013. Vertical collusion between airports and airlines:
An empirical test for the European case. Transp. Res. Part E Logist. Transp. Rev. 57, 3–15.
doi:10.1016/j.tre.2013.01.002

Bel, G., Fageda, X., 2007. Airport Management and Airline Competition in OECD Markets, in: Essays on
Transport Economics. pp. 159–181. doi:10.1007/978-3-7908-1765-2_12

Bel, G., Fageda, X., 2010. Privatization, regulation and airport pricing: An empirical analysis for Europe. J.
Regul. Econ. 37, 142–161. doi:10.1007/s11149-009-9110-7

Brons, M., Pels, E., Nijkamp, P., Rietveld, P., 2002. Price elasticities of demand for passenger air travel: A
meta-analysis. J. Air Transp. Manag. 8, 165–175. doi:10.1016/S0969-6997(01)00050-3

Burghouwt, G., 2007. Airlines Network Development in Europe and its Implication for Airport Planning.
Ashgate, Aldershot.

Burghouwt, G., 2014. Long-haul specialization patterns in European multihub airline networks - An
exploratory analysis. J. Air Transp. Manag. 34, 30–41. doi:10.1016/j.jairtraman.2013.07.008

Burghouwt, G., de Wit, J., 2005. Temporal configurations of European airline networks. J. Air Transp.
Manag. 11, 185–198. doi:10.1016/j.jairtraman.2004.08.003

Burghouwt, G., Hakfoort, J., 2001. The evolution of the European aviation network, 1990-1998. J. Air
Transp. Manag. 7, 311–318. doi:10.1016/S0969-6997(01)00024-2

Carson, R.T., Cenesizoglu, T., Parker, R., 2011. Forecasting (aggregate) demand for US commercial air
travel. Int. J. Forecast. 27, 923–941. doi:10.1016/j.ijforecast.2010.02.010

Clewlow, R.R., Sussman, J.M., Balakrishnan, H., 2014. The impact of high-speed rail and low-cost carriers
on European air passenger traffic. Transp. Policy 33, 136–143. doi:10.1016/j.tranpol.2014.01.015

17
Cohen, J.P., Morrison Paul, C.J., 2003. Airport infrastructure spillovers in a network system. J. Urban Econ.
54, 459–473. doi:10.1016/j.jue.2003.06.002

Costas-Centivany, C.M., 1999. Spain’s airport infrastructure: Adaptations to liberalization and privatization.
J. Transp. Geogr. 7, 215–223. doi:10.1016/S0966-6923(99)00007-1

Dobruszkes, F., 2006. An analysis of European low-cost airlines and their networks. J. Transp. Geogr. 14,
249–264. doi:10.1016/j.jtrangeo.2005.08.005

Dobruszkes, F., 2010. Airline and high-speed rail competition in Europe: Towards a comeback of air
transport.pdf, in: ERSA 50th Conference.

Dobruszkes, F., Mondou, V., 2013. Aviation liberalization as a means to promote international tourism: The
EU-Morocco case. J. Air Transp. Manag. 29, 23–34. doi:10.1016/j.jairtraman.2013.02.001

Fageda, X., Fernández-Villadangos, L., 2009. Triggering competition in the Spanish airline market: The role
of airport capacity and low-cost carriers. J. Air Transp. Manag. 15, 36–40.
doi:10.1016/j.jairtraman.2008.07.002

Fageda, X., Flores-Fillol, R., 2012. On the optimal distribution of traffic of network airlines. Eur. Econ. Rev.
56, 1164–1179. doi:10.1016/j.euroecorev.2012.05.001

Flyvbjerg, B., 2005. Measuring inaccuracy in travel demand forecasting: Methodological considerations
regarding ramp up and sampling. Transp. Res. Part A Policy Pract. 39, 522–530.
doi:10.1016/j.tra.2005.02.003

Flyvbjerg, B., Skamris Holm, M.K., Buhl, S.L., 2005. How (In)accurate Are Demand Forecasts in Public
Works Projects?: The Case of Transportation. J. Am. Plan. Assoc. doi:10.1080/01944360508976688

Forsyth, P., 2007. The impacts of emerging aviation trends on airport infrastructure. J. Air Transp. Manag.
13, 45–52. doi:10.1016/j.jairtraman.2006.10.004

Franke, M., 2004. Competition between network carriers and low-cost carriers - Retreat battle or
breakthrough to a new level of efficiency? J. Air Transp. Manag. 10, 15–21.
doi:10.1016/j.jairtraman.2003.10.008

Fu, X., Zhang, A., Lei, Z., 2012. Will China’s airline industry survive the entry of high-speed rail? Res.
Transp. Econ. 35, 13–25. doi:10.1016/j.retrec.2011.11.006

Givoni, M., Banister, D., 2006. Airline and railway integration. Transp. Policy 13, 386–397.
doi:10.1016/j.tranpol.2006.02.001

González-Savignat, M., 2004. Competition in Air Transport - The Case of the High Speed Train. J. Transp.
Econ. Policy 38, 77–108.

Graham, B., 1999. Airport-specific traffic forecasts: A critical perspective, in: Journal of Transport
Geography. pp. 285–289. doi:10.1016/S0966-6923(99)00019-8

Graham, B., Shaw, J., 2008. Low-cost airlines in Europe: Reconciling liberalization and sustainability.
Geoforum 39, 1439–1451. doi:10.1016/j.geoforum.2007.12.006

Hamzawi, S.G., 1992. Lack of airport capacity: Exploration of alternative solutions. Transp. Res. Part A
Policy Pract. doi:10.1016/0965-8564(92)90044-8

Hong, J., Chu, Z., Wang, Q., 2011. Transport infrastructure and regional economic growth: evidence from
China. Transportation (Amst). doi:10.1007/s11116-011-9349-6

18
IATA, I.A.T.A., 2004. Airport development reference manual: 9th Edition.

Ingersoll, Jr., J.E., Ross, S.A., 1992. Waiting to Invest: Investment and Uncertainty. J. Bus.
doi:10.1086/296555

Ishii, J., Jun, S., Van Dender, K., 2009. Air travel choices in multi-airport markets. J. Urban Econ. 65, 216–
227. doi:10.1016/j.jue.2008.12.001

Janic, M., 2003. High-speed rail and air passenger transport: a comparison of the operational environmental
performance. Proc. Inst. Mech. Eng. Part F J. Rail Rapid Transit 217, 259–269.
doi:10.1243/095440903322712865

Jiménez, J.L., Betancor, O., 2012. When trains go faster than planes: The strategic reaction of airlines in
Spain. Transp. Policy 23, 34–41. doi:10.1016/j.tranpol.2012.06.003

Jimenez, E., Claro, J., de Sousa, J.P., 2014. The Airport Business in a Competitive Environment. Procedia -
Soc. Behav. Sci. 111, 947–954. doi:10.1016/j.sbspro.2014.01.129

Jorge-Calderón, J.D., 1997. A demand model for scheduled airline services on international European routes.
J. Air Transp. Manag. doi:10.1016/S0969-6997(97)82789-5

Khan, A., 1989. Realistic Planning for Transportation A Flexible Approach. Long Range Plann. 22, 128–136.

Madas, M.A., Zografos, K.G., 2008. Airport capacity vs. demand: Mismatch or mismanagement? Transp.
Res. Part A Policy Pract. 42, 203–226. doi:10.1016/j.tra.2007.08.002

Madas, M.A., Zografos, K.G., 2010. Airport slot allocation: A time for change? Transp. Policy 17, 274–285.
doi:10.1016/j.tranpol.2010.02.002

Martins, J., Marques, R.C., Cruz, C.O., 2014. Maximizing the value for money of PPP arrangements through
flexibility: An application to airports. J. Air Transp. Manag. 39, 72–80.
doi:10.1016/j.jairtraman.2014.04.003

Miller, B., Clarke, J.P., 2007. The hidden value of air transportation infrastructure. Technol. Forecast. Soc.
Change 74, 18–35. doi:10.1016/j.techfore.2004.03.011

Oum T.H., Waters II, W.G., Yong, J.-S., 1992. Concepts of Price Elasticities of Transport Demand and
Recent Empirical Estimates: An Interpretative Survey. J. Transp. Econ. Policy 26, 139–154.
doi:10.2307/20052976

Oum, T.H., Zhang, A., Fu, X., 2008. Air transport liberalization and its impacts on airline competition and air
passenger traffic. Transp. J. 49, 371–390. doi:10.1016

Pantazis, N., Liefner, I., 2006. The impact of low-cost carriers on catchment areas of established international
airports: The case of Hanover Airport, Germany. J. Transp. Geogr. 14, 265–272.
doi:10.1016/j.jtrangeo.2005.05.002

Pels, E., Nijkamp, P., Rietveld, P., 2000. Airport and Airline Competition for Passengers Departing from a
Large Metropolitan Area. J. Urban Econ. 48, 29–45. doi:10.1006/juec.1999.2156

Pels, E., Nijkamp, P., Rietveld, P., 2001. Relative efficiency of European airports. Transp. Policy 8, 183–192.
doi:10.1016/S0967-070X(01)00012-9

Pels, E., Njegovan, N., Behrens, C., 2009. Low-cost airlines and airport competition. Transp. Res. Part E
Logist. Transp. Rev. 45, 335–344. doi:10.1016/j.tre.2008.09.005

19
Percoco, M., 2010. Airport Activity and Local Development: Evidence from Italy. Urban Stud.
doi:10.1177/0042098009357966

Phang, S.Y., 2003. Strategic development of airport and rail infrastructure: The case of Singapore. Transp.
Policy 10, 27–33. doi:10.1016/S0967-070X(02)00027-6

Reynolds-Feighan, A., McLay, P., 2006. Accessibility and attractiveness of European airports: A simple
small community perspective. J. Air Transp. Manag. 12, 313–323.
doi:10.1016/j.jairtraman.2006.07.009

Sánchez-Borràs, M., Robusté, F., Criado, O., 2012. High-Speed Railways in Spain. Transp. Res. Rec. J.
Transp. Res. Board. doi:10.3141/2261-05

Sismanidou, A., Tarradellas, J., Bel, G., Fageda, X., 2013. Estimating potential long-haul air passenger traffic
in national networks containing two or more dominant cities. J. Transp. Geogr. 26, 108–116.

Smit, H.T.J., 2003. Infrastructure Investment as a Real Options Game: The Case of European Airport
Expansion. Financ. Manag. 32, 27. doi:10.2307/3666135

Starkie, D., 2002. Airport regulation and competition. J. Air Transp. Manag. 8, 63–72. doi:10.1016/S0969-
6997(01)00015-1

Suau-Sanchez, P., Burghouwt, G., 2011. The geography of the Spanish airport system: Spatial concentration
and deconcentration patterns in seat capacity distribution, 2001-2008. J. Transp. Geogr. 19, 244–254.
doi:10.1016/j.jtrangeo.2010.03.019

Suau-Sanchez, P., Burghouwt, G., 2012. Connectivity levels and the competitive position of Spanish airports
and Iberia’s network rationalization strategy, 2001-2007. J. Air Transp. Manag. 18, 47–53.
doi:10.1016/j.jairtraman.2011.08.004

Szyliowicz, J.S., Goetz, A.R., 1997. Revisiting transportation planning and decision making theory: The case
of Denver International Airport. Transp. Res. Part A Policy Pract. 31, 263–280. doi:10.1016/S0965-
8564(96)00033-X

Trigeorgis, L., 1996. Real Options: Management Flexibility and Strategy in Resource Allocation. MIT Press,
Cambridge, MA.

Warnock-Smith, D., Potter, A., 2005. An exploratory study into airport choice factors for European low-cost
airlines. J. Air Transp. Manag. 11, 388–392. doi:10.1016/j.jairtraman.2005.05.003

Wei, W., Hansen, M., 2006. An aggregate demand model for air passenger traffic in the hub-and-spoke
network. Transp. Res. Part A Policy Pract. 40, 841–851. doi:10.1016/j.tra.2005.12.012

Xiao, Y., Fu, X., Zhang, A., 2013. Demand uncertainty and airport capacity choice. Transp. Res. Part B
Methodol. 57, 91–104. doi:10.1016/j.trb.2013.08.014

Web references

Airbus Global Market Forecast 2013-2032. Available on September 30, 2014 on

http://www.airbus.com/newsevents/news-events-single/detail/global-market-forecast-2013-2032-
continuing-the-trend-toward-larger-and-more-efficient-aircraft/

Boeing. Current Market Outlook 2014. Available on September 30, 2014 on

http://www.boeing.com/assets/pdf/commercial/cmo/pdf/CMO_2014_Presentation.pdf

20
Eurocontrol. Challenges of growth 2013. Task 4: European air traffic in 2035. Available on September 30,
2014 on http://www.eurocontrol.int/sites/default/files/article/content/documents/official-
documents/reports/201306-challenges-of-growth-2013-task-4.pdf

Eurostat statistics database. Available on September 30, 2014 on

http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/search_database

IATA. Air Travel Demand. IATA Economics Briefing Nº9, 2008. Available on September 30, 2014 on
http://www.iata.org/whatwedo/Documents/economics/air_travel_demand.pdf

Madrid-Barajas Master Plan (Plan Director Barajas), 1999. Available on September 30, 2014 on
http://www.fomento.gob.es/mfom/lang_castellano/direcciones_generales/aviacion_civil/politicas_aero
portuarias/ita/planes_directores/planes_directores/pd_madrid_barajas.htm

Steer Davies Gleave. Air passenger growth rates, by Cresswell P. Available on September 30, 2014 on
http://www.steerdaviesgleave.com/news-and-insights/air-passenger-growth-rates

21
Fig. 1. Total passenger traffic in major European airports 2005-2013. Source:
Eurostat.

80

70

60 LONDON/HEATHROW airport
Total passenger traffici n million

PARIS/CHARLES DE GAULLE airport

50 FRANKFURT/MAIN airport

MADRID/BARAJAS airport
40
MÜNCHEN airport

LONDON/GATWICK airport
30
BARCELONA airport

PARIS/ORLY airport
20
ZURICH airport

10

0
2005 2006 2007 2008 2009 2010 2011 2012 2013

22
Fig. 2. Total passenger traffic in major Spanish airports 2005-2013. Source: Eurostat.

60

50

40
MADRID/BARAJAS airport
Total passenger traffici n million

BARCELONA airport

PALMA DE MALLORCA airport

30 MALAGA airport

ALICANTE airport

IBIZA airport

VALENCIA airport

20 BILBAO airport

10

0
2005 2006 2007 2008 2009 2010 2011 2012 2013

23
Fig. 3. Passenger traffic by airline at the Barcelona-El Prat Airport (in million passengers). Source:
Eurostat and own elaboration

60

50

40 Other (mostly legacy)

Ryanair+Easyjet
30
Vueling/Clickair
20 Iberia

10

0
2000 2005 2009 2013

24
Fig. 4. Domestic traffic at Madrid-Barajas 2005-2013. Source: Eurostat for total domestic
passengers and MIDT for connecting domestic passengers (adapted to the Eurostat methodology: in MIDT
each connecting passenger is counted as one passenger whereas the same passenger is counted as two by Eurostat: one
arrival and one departing)

Connec ng Passengers Domes c Segment Domes c Passengers Madrid

25.00
22.64

20.69 20.55
19.85
20.00 18.86 18.76

17.02
in '000 000 passengers

14.41
15.00

as % of total domes c 11.92

traffic at Madrid-Barajas

10.00

26.0% 25.5% 24.6%

23.9%
5.58 23.1%
5.15 5.27 4.92 17.3%
4.36
5.00 15.5%
3.25 13.7% 13.6%
2.64
1.98 1.62

0.00
2005 2006 2007 2008 2009 2010 2011 2012 2013

25
Fig. 5. Share of Spanish airports in domestic traffic 2005-2013. Source: Eurostat.

2013 25.9% 22.1% 10.6% 5.0% 4.9% 31.5%

2012 26.7% 21.2% 10.3% 5.1% 5.2% 31.5%

2011 27.6% 20.5% 10.1% 4.5% 5.4% 31.8%

MADRID/BARAJAS airport
2010 30.0% 18.6% 9.9% 4.6% 4.6% 32.3%
BARCELONA airport

PALMA DE MALLORCA airport

BILBAO airport
2009 30.8% 18.0% 10.2% 4.4% 4.7% 31.8%
SEVILLA airport

Other Spanish airports

2008 31.0% 19.1% 9.5% 4.6% 4.7% 31.1%

2007 31.1% 20.7% 9.0% 4.4% 4.4% 30.4%

2006 31.0% 21.7% 9.5% 4.3% 4.3% 29.2%

2005 31.5% 21.6% 9.5% 4.5% 4.3% 28.6%

26
Fig. 6. Spain/EU traffic at Madrid-Barajas 2005-2013. Source: Eurostat for total Spain/EU
passengers and MIDT for connecting Spain/EU passengers (adapted to the Eurostat methodology: in MIDT
each connecting passenger is counted as one passenger whereas the same passenger is counted as two by Eurostat: one
arrival and one departing)

25.00
Connec ng Passengers Spain/EU Segment Spain/EU segment passengers Madrid-Barajas

19.46
20.00
18.40 18.12
18.03 17.67
17.45

15.80
Passengers in million

14.18
15.00
12.81

as % of total Spain/EU
10.00 traffic at Madrid-Barajas

24.0% 25.0%
24.8% 22.5% 18.4%
24.0% 4.20 4.51 14.6%
5.00 3.98
3.52 3.39 12.4% 11.3%
3.08 2.84
2.24
1.79

0.00
2005 2006 2007 2008 2009 2010 2011 2012 2013

27
Fig. 7. EU/non-EU traffic 2005-2013. Source: MIDT (One-way passengers only. Connecting passengers
counted as one passenger at the connecting airport)

Total traffic EU/non-EU segment (one way only) Connec ng traffic EU/non-EU segment (one way only)

140

120 114.85
111.66

102.30
Passengers in million

98.81
100 95.19 94.67 95.30

85.75
80.07
80 73.34

64.58 64.56

60
46.67% 44.66% 42.50% 42.60%
49.50% 46.20% 46.13% 46.53%
52.63%
52.66%
52.66% 52.73% 48.92
40 45.58 45.69 47.46
42.44 43.98 44.18 44.34
42.14
38.62
34.01 34.04
20
Connec ng traffic as % of
total EU/non-EU traffic

0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

28
Fig. 8. EU/non-EU connecting traffic at major European hub airports expressed as a % of total
EU/non-EU traffic. Source: MIDT (One-way passengers only. Connecting passengers counted as one passenger
at the connecting airport)

25.0%

2.2% 2.4%
20.0% 2.5%
2.5%
2.6%

4.6% 4.5% 4.4% 2.5% 2.5%

2.4%
4.2% 2.4%
15.0% 4.3% 2.2% MAD share
3.9% 3.7% 3.9% 1.9% CDG share
3.9% 3.7% 3.4% 3.2% 1.7%
3.3% 3.3% AMS share
3.1% 2.8%
2.6% FRA share
10.0% 3.0% 3.3% 3.3% 3.0%
LHR share
2.9%
2.7%
6.0% 6.2% 6.3% 2.8%
5.8%
5.3%
4.9%
4.7% 4.6% 4.7%
4.3%
5.0% 3.9% 3.7%

4.7% 4.7% 4.7% 4.6% 4.3% 3.9% 3.6% 3.6% 3.2% 3.5% 3.0% 2.9%

0.0%
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

29
Fig. 9. EU/Latin America traffic 2005-2013. Source: MIDT (One-way passengers only. Connecting
passengers counted as one passenger at the connecting airport)

Total traffic EU/La n America segment (one way only) Connec ng traffic EU/La ng America segment (one way only)

12

10.20 10.29

10 9.28 9.28
8.98 9.04
In '000 000 passengers

8.42 8.46
8.11
8 7.48

6.52 6.66

6
56% 55% 52% 51%
59% 59% 57% 55% 59%
59%
61% 5.09 5.20 5.12 5.27 5.23
4.95 5.02 4.97
60% 4.78
4 4.44
3.92 4.06

2
Connec ng traffic as % of
total EU/La n America
traffic
0
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

30
Fig. 10. EU/Latin America connecting traffic at major European hub airports expressed as a % of
total EU/Latin America traffic. Source: MIDT (One-way passengers only. Connecting passengers counted as
one passenger at the connecting airport)

35.0%

30.0%

15.1%
25.0% 13.7% 15.5%
15.5%
16.6% 17.1%
16.6%
15.9% 15.8%
15.7% 13.6% 12.3%
MAD share
20.0%
CDG share

AMS share

15.0% FRA share

7.4% 7.7%
7.0% 7.4%
5.5% LHR share
7.5% 6.1% 5.7%
7.8% 7.3% 7.2% 6.2%
10.0%
4.0% 3.7% 3.1% 3.3%
3.0% 3.9%
3.3%
3.2% 2.7%
2.9% 2.9% 2.8%
5.0%
4.6% 4.5% 4.9% 4.9% 4.3% 3.9% 3.9%
3.4% 2.9% 3.5%
2.8% 3.1%
2.1% 2.3% 1.9% 1.8% 1.8% 1.6% 1.5% 2.1% 2.2% 2.4%
1.4% 1.2%
0.0%
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

31
Table 1
Madrid-Barajas master plan traffic forecast (base case scenario) and actual passengers (in million). Source: Spanish Ministry of Infrastructure and Transportation and Eurostat.

Interational non-
Domestic EU EU Total Deviation from GDP Growth Theoretical Traffic
Year Forecast Actual Forecast Actual Forecast Actual Forecast Actual Forecast (actual) based on real GDP

1999 13.4 13.4 8.0 8.3 6.4 5.8 27.8 27.5 -0.8% 4.7
2000 14.2 16.6 8.9 9.9 7.1 6.3 30.1 32.8 9.2% 5 29.2
2001 14.9 17.2 9.6 10.5 7.8 6.3 32.3 34.0 5.3% 3.7 31.2
2002 15.6 16.8 10.4 10.6 8.4 6.4 34.4 33.9 -1.5% 2.7 33.2
2003 16.3 17.6 11.1 11.2 9.1 7.1 36.5 35.9 -1.8% 3.1 35.5
2004 15.7 18.4 11.7 12.2 9.5 8.2 36.9 38.7 5.0% 3.3 37.7
2005 16.4 19.5 12.2 12.9 10.0 9.5 38.5 41.8 8.7% 3.6 38.2
2006 17.0 20.4 12.7 14.0 10.4 10.5 40.1 44.9 12.1% 4.1 40.1
2007 17.6 22.6 13.2 17.7 10.9 11.2 41.7 51.4 23.2% 3.5 41.5
2008 18.3 20.6 13.7 18.2 11.4 11.8 43.3 50.5 16.7% 0.9 42.1
2009 18.9 18.9 14.2 17.8 11.9 11.4 45.0 48.1 7.0% -3.8 41.7
2010 19.6 18.8 14.7 18.4 12.3 12.6 46.6 49.8 6.9% -0.2 44.9
2011 20.3 17.0 15.2 19.5 12.8 13.0 48.3 49.6 2.7% 0.1 46.7
2012 20.9 14.4 15.7 18.2 13.3 12.6 49.9 45.2 -9.5% -1.6 47.5
2013 21.6 11.9 16.2 15.8 13.8 11.9 51.5 39.7 -22.9% -1.2 49.3
2014 22.2 16.6 14.3 53.1
2015 22.7 17.1 14.8 54.6
2016 23.3 17.6 15.3 56.2
2017 23.9 18.0 15.8 57.8
2018 24.5 18.5 16.3 59.4
2019 25.1 19.0 16.9 61.0
2020 25.7 19.4 17.4 62.6
2021 26.3 19.9 18.0 64.2
2022 26.9 20.4 18.5 65.8
2023 27.5 20.9 19.1 67.5
2024 28.1 21.4 19.7 69.1
2025 28.6 21.8 20.3 70.8

Source: Master Plan of Barajas 1995, Spanish Ministry of Transportation and Eurostat

32
Table 2
Main HST Routes in Spain departing from Madrid Airport (millions of passengers)

2007 2008 2009 2010 2011 2012 2013

Route Traffic 5.40 5.60 5.70 5.70 5.60 5.20 5.30
Madrid-Barcelona Air Traffic 4.60 3.50 3.00 3.10 3.10 2.50 2.20
% Air Traffic 86% 62% 53% 54% 55% 49% 41%

Route Traffic 0.66 0.80

Madrid-Alicante Air Traffic 0.24 0.16
% Air Traffic 35% 20%

Route Traffic 1.80 1.80 2.30 2.00 2.10

Madrid-Valencia Air Traffic 1.00 1.00 0.44 0.26 0.25
% Air Traffic 57% 57% 19% 13% 12%

Route Traffic 2.10 2.60 2.20 2.00 2.00 1.80 1.80

Madrid-Málaga Air Traffic 1.50 1.10 0.73 0.61 0.53 0.40 0.28
% Air Traffic 71% 43% 33% 30% 27% 22% 15%

Route Traffic 7.50 8.20 9.70 9.50 9.90 9.66 10.00

Total Air Traffic 6.10 4.60 4.73 4.71 4.07 3.40 2.88
% Air Traffic 81% 56% 49% 50% 41% 35% 29%

Source. Spanish Government. https://www.fomento.gob.es/NR/rdonlyres/9C5511A6-9896-4167-BE79-96364E13EAE6/124707/Competencia_avionAVE2013.pdf

33