Documente Academic
Documente Profesional
Documente Cultură
Edited by
Matthias Finger
Professor, Management of Network Industries, Ecole
Polytechnique Fédérale Lausanne, Switzerland
Kenneth Button
University Professor, Schar School of Policy and Government,
George Mason University, USA
Published by
Edward Elgar Publishing Limited
The Lypiatts
15 Lansdown Road
Cheltenham
Glos GL50 2JA
UK
1 Introduction 1
Matthias Finger and Kenneth Button
379
Index
vii
This book offers a critical and wide-ranging assessment of the air trans-
port liberalization process during the past 40 years.
This process is first located within the broader deregulation movement
of the late 1970s and 1980s, which, in addition to specific national events,
was partly the product of economic globalization at that time, resulting in
economic integration and competitive pressure on nation states and their
protected industries. All network industries were concerned, to varying
degrees. In addition, air transport was only one sector subject to reform;
other affected networks include the liberalization of national telecommuni-
cations and postal sectors, of energy, of surface transport and of the water
and sanitation sectors. The Anglo-Saxon countries typically deregulated the
operations of their network sectors early, with continental Europe typically
following suit, often driven by the European Commission, although Eastern
and Central Europe countries were driven by larger shifts in their political
structures. While such regulatory reform – or, to use the American jargon,
“deregulation” – triggered fierce ideological debates, it gradually became
clear that greater liberalization of both operations and infrastructure sectors
allowed the network industries to better adapt to the efficiency and com-
petitiveness requirements of an increasingly integrated and global economy.
Of course, there were often good reasons why the different national
network infrastructures and their operators were originally protected and
regulated in order to, depending on their intrinsic natures, minimize the
distortions that can occur in excessively competitive or monopolist struc-
tures. This was particularly the case when networks were vital elements in
the economy, either in terms of conventional economic development or as
conduits for allowing greater social and political integration. In aviation in
the 1920s and 1930s, such reasons initially pertained to ensuring mail net-
works, to tying ex-colonies to their mother countries, to passenger services
to less populated areas and to social services more generally, to national
interests and corresponding control over flag-carriers, to the support of
the military, and of course to safety, along with broadly defined national
interests, as later enshrined in the 1944 Chicago Convention.
(Chapter 5), India (Chapter 6), and China (Chapter 7) – and regions, such
as Europe (Chapter 8), Latin America (Chapter 9), and Africa (Chapter
10).
After considering the main national and regional developments in the
aviation sector, the second section of the book discusses the main fears
or objections that have been leveled against deregulation of air transport;
namely, safety (Chapter 11), social services (Chapter 12), oligopoliza-
tion (Chapter 13), the ensuing persistence of the hub-and-spoke system
(Chapter 14), and market instability (Chapter 15). While most of these
fears have turned out to be unfounded, air transport deregulation, and
especially its success, has actually led to new, originally unanticipated
challenges. In the third section, we discuss two emerging challenges in the
aviation sector: mounting concerns over security (Chapter 16), and the
challenges in air traffic control in Europe (Chapter 17) and other parts
of the world (Chapter 18). Air traffic control is the remaining element of
global aviation that is still heavily regulated with continuing diversity in
national approaches. As well as national institutional rigidities, in terms of
technology, the global network is still based upon radar systems awaiting
a shift to more flexible and efficient satellite-based air traffic management.
There are inevitably other challenges that we have not addressed, includ-
ing reducing the adverse environmental imprint of aviation; developing
institutions that permit the maximum exploitation of economies of scale,
scope and density without the economic distortions that can accompany
monopoly power; and ensuring that less developed parts of the globe are
not excluded from the benefits of adequate air transportation. However,
such challenges will have to be left for further research and publications.
Here, we have simply gathered the most relevant and highest quality
inputs on all the various dimensions of air transport liberalization.
INTRODUCTION
Initial regulation of international air service was initiated with the 1919
Paris Convention. The overriding objective of the convention was to
establish regulations that were universally applicable to international avia-
tion transport. In addition to states’ rights to regulate entry and flights,
safety regulation was also a key component of the Paris Convention. The
general principles of the convention recognized that each contracting state
was entitled to maintain sovereign rights over its airspace, and therefore
had the right to regulate entry and flights into and through its airspace.
Regulation following the 1919 Convention imposed increasingly stringent
regulations on US air transport services. For instance, foreign investment
in US air carriers was significantly restricted following passage of the Air
Commerce Act of 1926. This Act required US citizens to own at least 51
per cent of any air carrier registered in the United States, and at least two-
thirds of the board of directors were required to be US citizens; US citi-
zenship was also required of the president of the company to operate in the
US (Furlan, 2006; Hardaway, 2010; Westra, 2007). Succeeding legislation
further increased the US citizenship threshold for air carrier ownership, as
the 1938 Civil Aeronautic Act (CAA) also established the Civil Aeronautics
Board (CAB) as a regulatory body with authority over the setting of rates
and the restriction of air carrier market entry. This regulatory oversight
applied to carriers serving national and international routes departing and
arriving at US airports. In sum, early legislation of international aviation
operations in the US created a highly regulated business environment with
routes increased by 134 per cent for alliance members for the 1992–99
sample period, compared to a 50 per cent increase for other carriers.
While mean fare and traffic findings reported by the US DOT support
the hypothesis that Open Skies Agreements create a business environment
that promotes relatively low fares and higher flight capacity, these find-
ings may be misleading, since they do not control for other determinants
such as quality of service. Nonetheless, findings of enhanced consumer
welfare and flight capacity are reported when adjusting for differences
in fare determinants such as flight distance and number of competi-
tors in the market. Using US DOT Databank 1b (D1B1) data and T100
International Segment data for the 1993 to 2008 sample observation,
Cristea et al. (2014) reported that passengers on US–international flights
covered by OSAs pay lower fares and have a choice of more direct flights.
These findings are consistent with fare and flight frequency findings
by Winston and Yan (2015). Using international airline data collected
by the International Air Transport Association (IATA) for the 2005–09
observation sample, they reported almost a 15 per cent reduction in fares
associated with operating flights on routes covered by OSAs. Winston and
Yan used the findings to simulate welfare gains that could arise if the US
were able to negotiate OSAs with other countries experiencing a signifi-
cant amount of international traffic. Their simulation results suggest that
negotiating OSAs with remaining passengers would generate a gain of $4
billion for passengers.
A significant number of countries have signed bilateral OSAs with the
US following the 2009 observation period used in past research (Winston
and Yan, 2015). Notable among these signatories are countries with large
volumes of passenger traffic with the US, such as Japan, Colombia and
Brazil. Pricing pressure on routes without OSA coverage is potentially
high, assuming OSAs facilitated low fares charged on routes serving these
new signatory countries. Indeed, over 109 countries had signed OSAs with
the US by 2012, depicting a large global network of international destina-
tions with the ability to charge low fares and the capacity to serve a non-
trivial number of passengers. Maintaining high fares on routes that are not
covered by OSAs has the potential to suppress passenger demand in those
markets. In addition, as the US DOT observes, carriers can take advantage
of their hubs in Open Skies countries to provide lower-cost services to non-
signatory countries. Hence, estimation of fares along US–international
routes using more current data is warranted to examine the potential gains
from negotiating OSAs with remaining non-signatories.
Data
for this chapter’s estimation, each of these fares would have the same
distance, the same load factor, the same equipment size, the same flight
frequency, and so on. By averaging the fares, there is one observation
for each carrier/route combination. Findings are partitioned into three
categories: (1) O–D pairs that are covered by Open Skies Agreements;
(2) O–D pairs that are only provisionally approved by Open Skies
Agreements; and (3) O–D pairs that are not covered by a ratified Open
Skies Agreement or a provisionally accepted agreement. These mean
findings suggest flights on routes covered by Open Skies Agreements
actually charge passengers higher fares per mile. Fares along these routes
have a mean value of 39 cents per mile compared to 24 and 34.5 cents per
mile for routes covered by provisional agreements or without any agree-
ment, respectively. Analysis of other information on US-international
flights suggests that differences in route and carrier characteristics are
likely to contribute to an inflated mean fare for Open Skies routes. For
instance, compared to US carriers serving routes with provisional Open
Skies Agreements, US carriers serving international routes covered by
Open Skies Agreements fly shorter distances with lower load factors,
greater number of seats per aircraft and greater frequency of service.
Standard airline economic theory suggests shorter one-segment flight
distances can contribute to lower fares, since airline costs are a function
of the number of take-offs and landings, which vary less than propor-
tionally with distance flown.8 Average load factor, or the average number
of passengers per seat, is expected to have a negative influence on fares,
as many airline costs vary less than proportionally with the number of
passengers. For example, flight-crew costs, maintenance costs, fuel costs,
gate fees and terminal rental do not vary proportionally with passengers.
Similarly, as shown by Graham et al. (1983), there are cost savings asso-
ciated with operating larger aircraft. Thus, we expect that average plane
size will have a negative effect on fares, a priori. As reported in Bitzan et
al. (2014), it is important to note that each of the variables hypothesized
to influence the costs of a particular movement may also potentially
influence demand. For example, higher load factors may mean less
comfort due to crowding, decreasing demand. As another example, as
Borenstein (1989) suggested, larger aircraft are often more comfortable
and perceived to be safer than smaller aircraft. This may increase the
demand for travel on larger aircraft. However, previous researchers have
shown that the cost effects of these variables dominate any demand-side
effects that might occur (e.g. Bitzan and Chi, 2006 or Borenstein, 1989).
The final cost variable depicts frequency of service and is likely to influ-
ence fares through its impacts on demand. Higher flight frequency is
likely to mean more convenient schedules for passengers, increasing the
Notes:
a Low-cost US carriers serving US–Mexico routes are JetBlue Airways, Airtran Airways,
Spirit Airlines, Frontier Airlines, Sun Country Airlines and Virgin America Airlines.
b Legacy US carriers serving US–Mexico routes are United Airlines, American Airlines,
Alaska Airlines, US Airways and Delta Air Lines.
c Other (regional/charter) carriers serving US–Mexico routes are American Eagle, Mesa
Airlines, ExpressJet Airlines, SkyWest Airlines and Horizon Air.
While some characteristics of carriers serving Open Skies O–D pairs are
associated with high fares, evidence in Table 2.1 also indicates character-
istics associated with low fares. These routes face relatively greater actual
competition, as the Herfindahl index is lower than that for routes without
an in-force Open Skies agreement. The promotion of market competition
is consistent with the objective of OSAs and, as stated earlier, mean values
actually understate the true level of competition, since the data source used
in this chapter does not include non-US carriers serving US–international
routes. The mean number of slot-controlled airports at the originating or
destination airport is smaller for Open Skies in-force routes. Because take-
off and landing slots are limited at the slot-controlled airports, the ability
of an existing or new carrier to meet the demand for more flights is limited.
In this way, the slot control can serve as an inhibitor to competition.
The variation in characteristics across the three groups of US–
international routes highlights the need to employ multivariable estima-
tion approaches to provide greater insight on the relative fares of flights
serving Open Skies in-force routes, Open Skies provisional routes and
routes without such agreements.
Empirical Approach
where FPMij is the mean airfare per passenger-mile for the carrier between
origin i and destination j. This variable is often referred to in other studies
as yield. The variable Distanceij is the mean one-way distance between
origin i and destination j, the variable Load is the mean load factor for the
carrier between origin i and destination j, which is computed by taking the
ratio of passenger-miles/seat-miles weighted by the number of departures
performed, and the variable Equipij is the mean aircraft size for the carrier
between origin i and destination j, which is computed by weighting the
number of seats per aircraft by the number of departures performed. These
three dependent variables are included to account for differences in the cost
that carriers incur on different routes. It is predicted that factors contribut-
ing to higher unit cost would be passed on to customers in the form of higher
fares. Hence, negative values are expected for the coefficients on these three
cost variables. The variable Mexicoij is also included for sample populations
comparing fares on Open Skies routes to fares on flights without Open Skies
Agreements to account for the cost differences, since a substantial share
of US–international flights that depart or arrive in airports in Mexico are
served by low-cost carriers and low-fare regional and charter carriers.
In addition to the inclusion of variables influencing operating costs or
identifying countries with a large prevalence of low-cost carriers serving
US–international routes, this specification also includes factors influenc-
ing the strength of air passenger demand on a particular route and factors
influencing the effectiveness of competition and potential competition.
For instance, the variable Freqij, which depicts the mean number of
flights per quarter for the carrier between origin i and destination j, is an
indicator of passenger demand. The variables Herfij, PCarrij and Slotij
account for route-level competition and potential competition. Herfi is
the Herfindahl–Hirschman Index for flights between origin i and desti-
nation j, and is computed by taking the summation of carrier shares of
passengers squared for US carriers. This variable is normalized to range
from zero to one; thus, a value of one is indicative of a monopolistic route
and a value of zero is indicative of a perfectly competitive route. While
standard microeconomics suggests lower prices in competitive markets,
past estimation of US–international fares provides evidence supporting
the theory of imperfect contestability, such that the presence of potential
competitors at the originating or destination airport contributes to lower
fares (Bitzan et al., 2014). The specification of Equation 2.1 accounts for
the possible fare effect of potential competition by including the variable
PCarri,j. This variable depicts the number of carriers serving the origin
airport and/or the destination airport, but not serving the route. The
remaining variable Slotij that accounts for competition along O–D routes
is a dummy variable that indicates whether landing and departure slots
are airport-controlled at origin i or destination j. Since slot control is
viewed as an inhibitor of competition, it is expected that its coefficient
estimate will be negative.
The remaining fare determinant, specified as Open Skiesij in Equation
2.1, is included to examine whether Open Skies Agreements facilitate car-
riers charging customers low relative fares, all else equal. The Open Skies
variable is a dummy equalling one if the route includes a non-US destina-
tion or originating location that is covered by an in-force or provisional US
Open Skies agreement. The estimated coefficient on this dummy variable
depicts the log-fare differential for flights on Open Skies Agreement routes
compared to flights without Open Skies Agreements. This coefficient is
converted into a percentage differential by using the formula (expb − 1)
× 100; however, interpreting findings derived from making this conver-
sion might be problematic if heteroscedasticity is associated with the log-
transformed specification (Ai and Norton, 2000; Baser, 2007). Distorted
estimated fare differentials arise due to the introduction of enhanced bias
that is associated with the anti-log transformation when heteroscedastic-
ity is present. A Breusch–Pagan test for the log-fare estimate for each of
the three population samples rejects the hypothesis of homoscedasticity
as the chi-squared value varies from a low of 56.27 to a high of 181.66,
which is statistically significant at the 0.01 level. Given the presence of
heteroscedasticity, a gamma-based generalized linear method (GLM) is
used to estimate the log-wage equation to compute consistent estimates
(Manning and Mullahy, 2001).9 Estimation bias could still persist even
Fare findings derived when estimating Equation 2.1 for three sample
populations is presented in columns 1–3 of Table 2.3. Column 1 contains
results when estimating Equation 2.1 for the sample population of routes
serving in-force Open Skies routes and those serving routes without Open
Skies Agreements. Column 2 uses the sample population of routes covered
by in-force or provisional Open Skies routes, and column 3 uses the sample
population of routes covered by provisional Open Skies Agreements and
routes without an Open Skies Agreement.
Findings on the control variables are consistent with standard trans-
portation economic theory. Lower fares are associated with the vari-
ables accounting for cost variation by routes. For instance, the estimated
coefficients on route distance, load factors and seats per aircraft are
negative and statistically significant for all but the sample population
for routes covered by provisional Open Skies and routes without Open
Skies Agreements (column 3). Even for that sample, only the estimated
coefficient on the load factor parameter lacks significance. The estimated
coefficient on the Mexico destination–originating route dummy is also
negative and is consistent with the mean findings, indicating a large share
of US routes arriving to or departing from Mexican airports are served
by low-fare airlines. The remaining parameter estimates on the variables
accounting for factors influencing the strength of air passenger demand
on a particular route, and factors influencing the effectiveness of com-
Notes:
a T-statistic presented in parentheses.
three population samples. This fare finding for actual competition does
not support the hypothesis of perfect contestability, but does support
the rationale for negotiating Open Skies Agreements that enhance direct
competition. The estimated coefficient on the airport slot control is also
positive but lacks statistical significance for the sample population of
routes with provisional Open Skies Agreements and routes without Open
Skies Agreements. Findings depicting potential competition at the route
level suggest fares are lower on routes with a larger number of carriers
that could compete for service on the observed route. This fare finding on
potential competition is consistent with past research showing support for
the theory of imperfect contestability on US–international flights (Bitzan
et al., 2014).
The estimated coefficient of the parameter on the key variable of
interest is the Open Skies dummy. Findings on this estimated coefficient
reported in column 1 of Table 2.3 suggest statistically significantly lower
fares of 5.5 per cent on routes covered by in-force Open Skies Agreements
compared to routes without Open Skies Agreement coverage. This finding
is consistent with past research on liberalization of US–international
airline operations (Winston and Yan, 2015, and Cristea et al., 2014). In
contrast, findings reported in column 2 reveal that airfares on routes with
provisional Open Skies Agreements closely resemble fares on routes with
in-force Open Skies agreements. However, fares are measurably lower on
routes covered by provisional Open Skies Agreements compared to routes
without Open Skies Agreements. Indeed, findings in column 3 reveal a 4.5
per cent provisional Open Skies discount, which is close to the discount
observed in column 1 for in-force Open Skies Agreements. However, the
provisional Open Skies fare results lack statistical significance, most likely
due to the relatively small sample of routes covered by this type of Open
Skies Agreement.
CONCLUDING REMARKS
Since 1992, 120 countries have negotiated Open Skies Agreements with
the US. A review of the history of US aviation service agreements shows
that over 70 years have elapsed from initial regulation of international
service to the development of the current state of international aviation
service agreements. On the one hand, economic theory suggests that the
easing of fare, entry and capacity restrictions derived from negotiating
these agreements should enhance consumer welfare and present pas-
sengers with the opportunity to enjoy greater frequency of air transport
services. On the other hand, international alliances formed following
NOTES
REFERENCES
Ai, C. and E. Norton (2000), ‘Standard Errors for the Retransformation Problem
with Heteroscedasticity’, Journal of Health Economics, 19(5), 697–718.
Alford, E. and R. Champley (2007), ‘The Impact of the 2007 US–EU Open
Skies Air Transport Agreements’, International Trade Administration Occasional
Paper, no. 07–001.
Baser, O. (2007), ‘Modeling Transformed Health Care Cost with Unknown
Heteroskedasticity’, Applied Economics Research Bulletin, 01, 1–6.
Bitzan, J. and J. Chi (2006), ‘Higher Airfares to Small and Medium Sized
Communities: Costly Service or Market Power?’, Journal of Transport Economics
and Policy, September, 40, 473–501.
Bitzan, J., A. Kones and J. Peoples (2014), ‘Airfares and Competition on
International Routes’, in J. Peoples (ed.), The Economics of International Airline
Transport, Bingley: Emerald Press.
Borenstein, S. (1989), ‘Hubs and High Fares: Dominance and Market Power in
the U.S. Airline Industry’, RAND Journal of Economics, Autumn, 20, pp. 344–65.
Button, K. (2009), ‘The Impact of US–EU “Opens Skies” Agreement on Airline
INTRODUCTION
27
was treated like a public utility. As a public utility, the airline provided a
public service and the policy was to preserve service and employment in a
financially sustainable way; there was no consideration of the consumer’s
interests. This narrow view of the public interest defined as ‘to preserve
service by a national airline in a financially sustainable way’ has driven
most – if not all – subsequent Canadian air policy decisions; Transport
Canada has chosen Air Canada to be ‘the winner’, and structures and
manages air policy accordingly.
EVOLUTION TO DEREGULATION
The shift from regulation to deregulation came in the early 1980s, with
both internal and external pressures. Domestic deregulation in the US in
1978 offered a window on the impact of such changes. Increased com-
petition, increased entry, and lower fares were appealing to consumers
and politicians alike. The leakage of large numbers of Canadian air pas-
sengers to nearby US points at that time (as it is now in 2016) was also a
stimulus. The Crown-carrier Air Canada (AC) was regulated under a new
Act passed in 1977; this Act placed the carrier on an equal basis before
the regulator, the Canadian Transport Commission – at least in principle.
Under the new Act, AC’s shares were held by the Minister of Transport on
behalf of the government. The carrier was ordered to have a more com-
mercial orientation and to be profit motivated. At the time, however, the
incentives, rewards or penalties for AC were not changed, thus pursuing
profit was not like the challenge at other carriers. There was little or no
downside risk, as the government would not have let AC fail. This is most
likely true today as well, even with privatization. To be fair, AC was still
charged explicitly, as well as implicitly in the eyes of Canadians, with pro-
viding service that under strict commercial rules would be unlikely to be
offered. The internal pressures came from the treatment of carriers other
than AC. The regional carriers were restricted under the Regional Air
Carrier programme, and Canadian Pacific (CP) was limited in a number
of ways from competing with AC including limitations on the capacity it
could offer on transcontinental and some overseas routes. These restric-
tions had disappeared by 1980.
The first vestiges of competition began with the government allowing
discount fares by the two transcontinental carriers, AC and CP. All carriers
introduced a series of ad hoc lower fares in various forms. These included
discounts to students, seniors and families travelling together, to name a
few. The discounts were not across the board and were not actually the
outcome of competition, but they did serve to respond in some degree to
earnings. Deficits were much less frequent after 1950, but between 1937
and 1962 Air Canada received ten subsidies amounting to $23.9 million.5
It has been claimed that since 1964 the airline has paid dividends
amounting to $55 million. However, it should be noted that the govern-
ment forgave portions of the Crown airline’s debt. In 1977, for example,
the federal government converted a major portion of Air Canada’s debt
to equity, thus reducing its burden of interest payments. Air Canada states
that since 1978 it has repaid $120 million in government loans and that it
has paid interest on these loans totalling $148 million.
When Air Canada was privatized in 1989, there were two important
aspects to the privatization that affected the ability of others to compete
fairly. The first was limitations on ownership and the second was the way
that the privatization funds were distributed (discussed in detail below).
Both actions resulted in a strong equity position for AC and showed that
the competition between carriers was biased, as the private sector carriers
were faced with private sector market conditions. On the other hand, AC
has continued to receive favourable treatment by government as the gov-
ernment adjusted the requirements for AC to meet pension obligations,
and twice has threatened back-to-work legislation when groups within AC
threatened to go on strike.
The path to the new air transportation policy began in September 1983,
when the then Minister of Transport, Lloyd Axworthy, sent a letter to the
acting president of the Canadian Transport Commission (CTC) stating
that recent events relating to ‘discount fares highlighted the need for a
comprehensive discussion of the entire domestic air travel pricing policy’.
He activated a section of the National Transportation Act and asked
the Minister of Justice to appoint an advocate for the public interest. In
December 1983, Mr Axworthy formed an interdepartmental taskforce
on airline regulation. It was composed of a ‘steering group’ consisting of
deputy ministers and a working group of middle-level public servants.
In early 1984, the CTC began hearings on the regulation of domestic
airfares. Very quickly, a wide range of interveners pushed the task force
hearings toward the broader issue of deregulation. A group of airline
unions strongly protested against the idea of deregulation in Canada. Air
Canada, however, changed its position in midstream and ceased to be a
vociferous opponent of deregulation.
In May 1984 the government announced its ‘New Canadian Air Policy’.
It recommended some relaxation of regulation, ‘but not deregulation’. The
under liberalization the carriers had more time and a somewhat more
sheltered environment (for example, entry to the industry was virtually
impossible, while entry to specific routes was easier but not free) in which
to get ready for deregulation. Third, the policy was followed, in only 14
months, by the federal Conservative government’s Freedom to Move policy
paper, which proposed outright deregulation throughout Canada. This
obviously reinforced the move toward deregulation and put more pressure
on all existing carriers to get their costs down, engage in consolidations to
achieve the benefits of a large network and reconfigure their route struc-
tures so as to survive under open competition.
It should be noted, however, that the May 1984 policy statement made
no reference to the future of Air Canada and whether it should be pri-
vatized. However, the Crown carrier was instructed to operate on sound
business principles in contemplation of profit, and to refrain from overly
aggressive competitive pricing and scheduling practices.
DEREGULATION IN 1988
consequences for the structure of the industry in the longer term. Many
anticipated that the results observed in the United States after deregulation
in 1977 would be similar for Canada. Initially, some changes in industry
structure were seen, and there were some changes in the level of fares,
as well as the introduction of a greater variety of fares. In his study for
the Royal Commission on National Passenger Transportation, Morrison
(1993) said that deregulation in Canada had no effect on the average level
of fares, but had simply decreased long-haul fares and increased short-
haul fares. This is interesting, as the government – through the CTC – had
held the industry with such tight constraints by confining certain airlines
to certain geographic markets, that the potential efficiency gains from
deregulation were less than those in evidence in the US.
An important gap in the 1987 National Transportation Act was the
lack of recognition of the continued existence of a government-owned air
carrier. There was deregulation, but at least initially there was no privatiza-
tion. The importance of this gap is that while the policy intended that there
be competition among air carriers, the initial conditions of such competi-
tion had not been established. In particular, the effectiveness of a competi-
tive deregulation policy – when several small private-sector air carriers had
to compete with a still apparently privileged, well-branded government-
owned air carrier with over 50 per cent of the total market – was not dealt
with; nor was the control that Air Canada had at Canada’s major airport
in Toronto. In the late 1980s, for example, Air Canada enjoyed a larger
market presence than all other carriers combined. It served over 60 points,
whereas the other national carrier served 35 points and the smaller regional
carriers served far fewer. The most important aspect of Air Canada’s pres-
ence and dominance in the market was that it attracted a disproportionate
share of the high-yield business traffic.
The next major policy change occurred in 1988, when Air Canada was pri-
vatized. Until this time, it had operated under the direction provided by the
Air Canada Act of 1977. This Act was ostensibly to place the airline on an
equal footing with other carriers, particularly in front of the regulator, the
CTC. Prior to 1977 Air Canada had listed only $5 million in paid equity
from its shareholder, the government of Canada. The airline operated with
significant amounts of government-guaranteed debt that was financed by
its then-owner Canadian National Railway. A significant amount of this
debt had been forgiven over the years, but with the new Air Canada Act
in 1977, $325 million in debt was converted to equity. As a result of this
carrier services, and giving preference where early start-up of air services
is planned; the size and maturity of the air transportation markets and
the potential for future growth; foreign government requests; Canada’s
international trade objectives; safety and security issues; foreign relations;
and bilateral irritants and disputes. Note that the consumer’s interest is not
considered.
The introduction of the Blue Sky policy did not indicate the changes
from the previous policy, nor the reasons for the changes. There was no
statement as to why, or in what way, the prior policy was deemed inad-
equate. It was not clear how the Blue Sky policy was going to fix something
that was not working. Given the way in which the ‘new’ international air
policy was designed, it was clear it was not open skies by any means –
hence the term ‘Blue Sky’.
One seeming failure of the international air service policy is the low
frequency of service allowed for the majority of foreign carriers. In many
cases, foreign carriers do not fly more than three flights per week. With
such infrequent service, there is little chance of serving high-yield business
traffic. Such low allowable service frequencies make it nearly impossible to
develop and grow a market. When examining the countries that Canada
has agreements with, many are small and in many cases fares are restric-
tive, with single disapproval conditions, and when designating intermediate
points, they are nearly all Star Alliance hubs. The US, ICAO and IATA all
define a bilateral agreement as ‘liberal’ if frequencies for third, fourth and
fifth freedom traffic are unlimited. Canada does not use this definition
of ‘liberal’. Canada uses a Canadian maximization model rather than a
global maximization model. If the market demands only three weekly fre-
quencies, the market will yield that outcome; therefore, permitting single
daily service will do no harm. Defining limited frequencies ex ante never
gives the market an opportunity to work. Limiting access by foreign carri-
ers is clearly designed to protect AC.
●● Phase I took effect in the first half of 2009. Airlines now have unlim-
ited freedom to operate direct services between any point in the
European Union and any point in Canada. There are no restrictions
on the number of airlines flying between the EU and Canada, or on
the number of services operated by any airline. Cargo airlines have
the right to fly onward to a third country.
●● Phase II of the Canada–EU Open Skies Agreement anticipates
that Canada will amend its legislation to enable European investors
to own up to 49 per cent of a Canadian carrier’s voting equity, an
increase from the current 25 per cent. At that time, further traffic
rights will be granted, including the right for cargo carriers to
operate services to a third country, for Canada or the EU, to a third
country, without connection to their point of origin.
●● In Phase III, both sides would allow investors to establish and
control new airlines in each other’s markets. At that point, passenger
airlines would be able to fly onward to a third country.
●● In Phase IV, EU and Canadian carriers would be granted full rights
to operate between, within and beyond both markets, including
between points in the territory of the other party (known as cabo-
tage). These rights would be granted after both sides complete the
necessary steps to allow each member full ownership and control of
their carriers.
TRANSBORDER MARKET
Prior to 1995, air services between Canada and the US were governed by
a relatively restrictive bilateral air services agreement. This agreement set
out conditions on fares, capacity, cities served, carrier designation and
airport access. As a result, there was limited service between the two coun-
tries. However, after the 1995 Open Skies Agreement was signed, within
three days 17 new city-pair routes were approved by the US Department
of Transportation (DOT), of which five were completely new routes. By
the end of 1995, 16 US and Canadian carriers were offering service within
the Open Skies Agreement. By August 1997, a total of 79 new routes
were established between Canada and the US (20 of these routes were
subsequently discontinued). Canadian carriers established 29 of these new
routes, and American carriers established 50 (see Dubey and Gendron,
1999).
Canada and the US signed an Air Service Agreement (ASA) on 24
February 1995.8 This agreement allowed Canadian and US carriers to
operate air services between any point in either country using any aircraft
and at any service frequency, and effectively removed price controls on
transborder routes.9 However, there were phase-in provisions over a three-
year transition period that: (a) limited the presence of US-based carriers
on routes involving Montreal, Vancouver and Toronto (all existing route
rights at that time held by US carriers were not affected); (b) restricted
Canadian airlines from accessing slots in US airports; and (c) reduced the
ability of airlines to engage in code-sharing services across the Canada–US
border.
At the time that the ASA was signed, it was expected that the removal
of constraints on the ability of carriers to freely serve transborder routes
would result in greater competition on these routes. As of 2012, this was
not the case.
border markets that existed from 1995 through to 1997. The restrictions
placed on entry by US carriers during the phase-in period at Vancouver,
Toronto and Montreal airports allowed Canadian carriers to increase
their dominance on transborder routes. Using flights as a measure for
calculating the degree of concentration, Pustay (1999) showed that Air
Canada increased their dominance across all transborder routes during this
period.11 At Toronto, Air Canada’s primary hub, AC increased its market
share of flights to 55 per cent, while the next largest carrier (American
Airlines) had a mere 10 per cent of the market for transborder flights in or
out of Toronto. Pustay (1999) noted that after the phase-in period ended,
the US carriers were able to expand the number of flights; however, Air
Canada’s dominance was still significant. Montreal experienced a differ-
ent outcome, with Air Canada’s dominance declining as Delta, American,
Continental and Northwest all added flights into Montreal.
In 2005, Canada negotiated a revised Open Skies Agreement with the
US that was signed in 2007. This agreement liberalized so-called ‘fifth’
and ‘sixth’ freedoms.12 With these new provisions in the Open Skies
Agreement, a carrier can now use a single through flight number to
operate from a foreign country to its home country and then fly to another
foreign country. However, even this further liberalization has not resulted
in greater competition on most transborder routes, as only a small number
of carriers have actually initiated service that takes advantage of these
additional freedoms.
What is striking when examining the data for the top eight airports in
Canada is the relatively flat growth in transborder passenger traffic in
nearly all markets. From 1997 through to 2010, the annual growth rate has
ranged from a low of 0.2 per cent in Vancouver to 6.8 per cent in Halifax;
Toronto has grown by only 0.7 per cent annually. Such low growth rates
could reflect a number of factors, including a lack of aggressive pricing
and lack of competition. Despite liberalizing the transborder market in
1995, and further moves to an open skies policy signed in 2007, the growth
in this market has been relatively small. The reasons are high fares due to a
lack of competition and the market dominance of Air Canada and its Star
Alliance partners. Evidence of this appears in a study of demand leakage
in the Canada–US transborder market: Elwakil et al. (2013) find an esti-
mated leakage of 4.7 passengers in 2008 who flew from/to US domestic
airports rather than take a transborder flight. In addition, after controlling
for route-specific factors, they find that fares are approximately 28 per cent
higher in the transborder market. Elwakil et al. (2013) conclude that ‘The
lack of low-cost carrier competition likely contributes to the higher trans-
border fares as might the cozy relationship between US and Canadian
carriers resulting from alliance membership and antitrust immunity for
that may make entry impossible or expensive and that may invite state
aid, which would create market distortions. What we observe is coun-
tries moving to higher allowable foreign ownership limits when capital
is needed – such as Brazil and India, who moved to 49 per cent foreign
ownership limits.
In 2016, under pressure from a small group of aviation entrepreneurs,
Canada lobbied the government for a change in the foreign ownership
restrictions. After much consultation, the government moved to change
the restrictions from 25 per cent to an upper limit of 49 per cent foreign
ownership. The new rules ensure, however, that no single foreign investor
or group of foreign investors can have more than a 25 per cent ownership
stake.
What are the differences between a merger between carriers, a joint
venture within a strategic alliance, and allowing foreign capital to invest
in a national airline either up to some limit or with no restrictions? Would
the strategic direction, pricing, product offering, route choice and network
configuration differ under each of these three different scenarios? One
significant difference between mergers, joint ventures and allowing [more]
foreign capital is that the foreign capital can come from anywhere and
not be limited to the airline business, as mergers or joint ventures would
be. Also, and importantly, with mergers or joint ventures a competitor is
removed from the market, which – as economic theory and empirical evi-
dence show – generally reduces economic welfare.
Airlines have invested in other airlines for many reasons. Lufthansa’s
decision to acquire a 20 per cent shareholding in British Midland may well
have been motivated by a desire to gain more slots at Heathrow Airport,
and they may have been similarly motivated with their 18 per cent pur-
chase of JetBlue, which has a major hub at slot-constrained JFK in New
York. Developed countries with small populations, or countries that are
isolated, have tended to have a more liberalized approach to their airline
markets in order to expand the networks of their national airlines. It could
be one of the key reasons that Australia and New Zealand formed a single
aviation market to allow their citizens to invest in each other’s airlines. It
is also why Korea and the Netherlands have taken a liberal approach to
aviation access. Cross-border airlines are springing up more frequently;
in express delivery, UPS, FedEx and DHL, in Europe, AF/KLM, LH/SR/
OS/SN and BA/IB. Airline families have emerged: Grupo TACA, LAN
Airlines, Air Asia and Tiger as well as mergers such as TACA1Avianca
and LAN1TAM, which produced LATAM.
SUMMARY
daily service means that an airline can provide this level of service if it sees
a business case in doing so. Airlines would not be constrained to fewer
weekly frequencies and would not be constrained to have symmetry with
Canadian and foreign carriers’ frequencies. The US, for example, pursues
a ‘single’ open skies policy with every country, with the ‘open skies’ being
as liberal as possible: unrestricted first to sixth freedoms with an optional
cargo seventh freedom, market-based pricing freedoms, extensive doing-
business protections, and strong safety and security articles are contained
in the agreements.
A new Canadian air policy would have an air service policy that lets
carriers of the two negotiating countries decide when and where to enter
a market, and how often. The carriers choose what is in their commercial
interests, rather than having government decide, subject to competition
laws.
The dynamics of the market come from innovation in products and
processes, and these outcomes generally come from outside the existing
market structure. How much competition is too much or not enough?
Most important markets in Canada are dominated by a small number of
firms, such as banking, telecommunications, large accounting firms and,
of course, air transport. As John Kay (2010) pointed out, ‘in all these
industries established businesses will tell you they welcome competition,
but they also explain enough is enough, that users already have a wide
array of products at competitive prices’. The firms argue that public
policy should be directed at investment and innovation and that it is global
competitiveness, not domestic competition, that matters. Unfortunately,
government seems to be all too accepting of these industry arguments. The
benefits of competition are not about today’s customers, but about future
customers. However, competition and competitiveness are different things:
tomorrow’s customers will have innovative products produced and deliv-
ered through innovative processes; without competition, the LCC business
model would not have evolved, and WestJet or Porter would never have
come into existence.
Fundamental to aviation policy and deregulation is to decide how
Canada defines the public interest for air policy, especially international
air policy. What weight is, or should be, placed on the interests of con-
sumer and aviation-dependent economic sectors vs. air carrier interest?
How explicit does this have to be? Do carrier interests include the interests
of the alliances that a Canadian carrier joins, or are they confined to the
Canadian carrier’s own flights only? Are we protecting Canadian carrier
capacity or its profits, including alliance-derived profits? Should Canadian
air policy protect one foreign carrier aligned with a Canadian carrier from
another foreign carrier?
NOTES
1. We use the term ‘cost-oriented carrier’ since WestJet is no longer a true low-cost carrier
(LCC); it has deviated significantly from its original LCC model based on Southwest
Airlines of the US.
2. The Canada–US market is referenced differently from other international aviation
markets given the proximity to the market over the entire US–Canada border and the
fact the US population is 316 million, nearly ten times that of Canada. The econo-
mies are also tightly linked as a result of various trade policies, particularly the North
American Free Trade Agreement between Canada, Mexico and the US.
3. In 2015 total passenger traffic in Canada was 133 354 859, with the top four airports
(Toronto, Vancouver, Montreal and Calgary) handling 70 per cent of traffic, and the top
eight airports handling 84 per cent of total traffic. Toronto Airport handled 31 per cent
of total traffic while Vancouver, the next closest, handled half that amount of passenger
traffic.
4. For rail, this only applies if the railway crosses provincial boundaries. For marine it
includes all waterways, even those wholly within a province.
5. Early in its life, Air Canada received some cash subsidies from the federal govern-
ment. These took two forms: the government paid for the airline’s deficits in its first
few years of operation, and Air Canada was given contracts to carry the mail at high
rates. The mail contracts represented 94 per cent of the airline’s operating revenues
in 1938, but this share fell to 37 per cent in 1943 and 1944, and to only 15 per cent in
1951.
6. The discussion in this section is based on the official Blue Sky policy document; see
http://www.tc.gc.ca/pol/en/ace/consultations/blueSky.pdf.
7. Metal neutral joint venture is a form of partnership between airlines. It is structured
so that partners in the venture are indifferent as to which one operates the ‘metal’ (air-
craft) when they jointly market services. Metal neutrality can be achieved through cost-,
revenue- and/or comprehensive benefit-sharing arrangements.
8. The United States considered the 1995 agreement with Canada to be an ‘open trans-
border agreement’ which, however, did not meet all the requirements of an ‘Open Skies
agreement’ as defined by the DOT.
9. Standing Committee on Transport ‘Air Liberalization and the Canadian Airports
System: Interim Report’, May 2005, available at: http://www.parl.gc.ca/content/hoc/
Committee/381/TRAN/Reports/RP1858323/TRAN_Rpt04/TRAN_Rpt04-e.pdf.
10. WestJet serves five US cities with its own aircraft (Chicago O’Hare (ORD), Los Angeles
(LAX), Las Vegas, New York (LGA) and Hawaii). It provides service to approximately
30 destinations in the US via its codeshare partners.
11. Measuring flights can underestimate the degree of concentration and hence poten-
tial market power. For example, Alaska Airlines increased flights into Vancouver
using turboprop aircraft, which seat 50–70 people, while other carriers may be
using 120–140 seat jet aircraft. Seats would be a better measure of input metric of
concentration.
12. The fifth freedom right is the right for the carrier to enplane traffic at one foreign point
and deplane it in another foreign point as part of continuous operation, also serving the
carrier’s homeland. For example, a Canadian carrier could fly from Vancouver to any
city in California, drop off and pick up passengers, and continue the trip to a point in
Mexico.
BIBLIOGRAPHY
Gillen, D., R. Harris and T. Oum (2002), ‘Measuring the Economic Effects of
Bilateral Liberalization on Air Transport’, Transportation Research, 28E(3–4),
May.
Gillen, D., T. Oum and M. Tretheway (1985), Canadian Airline Deregulation
and Privatization: Assessing Effects and Prospects, Vancouver, Centre for
Transportation Studies, University of British Columbia.
Gillen, D., T. Oum and M. Tretheway (1986), Airline Cost and Performance:
Implications for Public and Industry Policy, Vancouver, Centre for Transportation
Studies, University of British Columbia.
Gillen, D., T. Oum and M. Tretheway (1986), ‘Entry-Barriers and Anti-Competitive
Behaviour in a Deregulated Canadian Airline Market’, Transportation Research
Forum Proceedings, pp. 483–93.
Gillen, D., T. Oum and M. Tretheway (1986), Identifying and Measuring the Impact
of Government Ownership and Regulation on Airline Performance, Economic
Council of Canada, 1986.
Gillen, D., T. Oum and M. Tretheway (1988), ‘Duopoly in Canada’s Airline
Industry: Consequences and Policy Issues’, Canadian Public Policy, 14(1),
March, 15–31.
Gillen, D., T. Oum and M. Tretheway (1989), ‘Privatization of Air Canada; Why
it is Necessary in a Deregulated Environment’, Canadian Public Policy, XV(3),
September, 285–99.
Gillen, D., T. Oum and M. Tretheway (1990), ‘The Cost Structure of the Canadian
Airline Industry’, Journal of Transport Economics and Policy, 24(1), January,
9–34.
Gonenc, R. and G. Nicoletti (2001), ‘Regulation, Market Structure and Performance
in Air Passenger Transportation’, OECD Economic Studies, No. 32, 183.
Howe, C.D. (1938), Canadian Parliamentary debate, 21 March, available at http://
www.lipad.ca/full/1938/03/21/17/.
Iacobucci, E., M. Trebilcock and R. Winter (2007), ‘The Political Economy of
Deregulation in Canada’, in Martin Levin (ed.), Creating Competitive Markets:
The Politics of Regulatory Reform, Brookings Institute.
Kay, J. (2010), ‘Radical Innovation Rarely Comes from Within’, Financial Times,
24 November.
Levine, M. (1987), ‘Airline Competition in Deregulated Markets: Theory, Firm
Strategy and Public Policy’, Yale Journal on Regulation, 4(6), 393–466.
Mak, D. (2005), ‘Transborder Air Services since 1994’, in Canadian Aviation
Intelligence Report, InterVISTAS, August, available at http://www.intervistas.
com/downloads/CAIR/issues/2005/08_aug2005.pdf.
Mentzer, M.S. (2000), ‘The Impact of Discount Airlines on Domestic Fares in
Canada’, Transportation Journal, 39(4), Summer.
Morrison, S.A. (1993), ‘Deregulation and Competition in the Canadian Airline
Industry’. Paper prepared for the Royal Commission on National Passenger
Transportation.
Morrison, S.A. and C. Winston (2000), ‘The Remaining Role for Government
Policy in the Deregulated Airline Industry’, in S. Pelzman and C. Winston (eds),
Deregulation of Network Industries: What’s Next?, Washington: AEI-Brookings
Joint Center for Regulatory Studies.
Oum, T., A. Zhang and Y. Zhang (1993), ‘Inter-firm Rivalry and Firm Specific Price
Elasticities in Deregulated Airline Markets’, Journal of Transport Economics &
Policy, 27(2), 171–92.
INTRODUCTION
Australia has long been a reluctant liberalizer in air transport. This has
been true for both domestic and international markets. While being cau-
tious with air transport liberalization, this has not been the case with other
industries, such as energy, telecommunications and the finance sector.
Australia has been deregulating at about the same time as other industrial
countries other than the US, and it has privatized major government firms
earlier than most.
With domestic markets, Australia was a keen observer of the perfor-
mance of US deregulation. In the late 1980s it decided to deregulate, and
deregulated in 1990. While the market is now totally deregulated, this has
come about by accident rather than design. Although entrants were free to
enter from 1990, there was a significant barrier: the incumbent airlines had
control of the terminals and gates, and entrants were not successful. The
collapse of one of the incumbent airlines – Ansett – in 2001 and the pri-
vatization of the airports over 1997 until 2002 opened up terminal access,
and entrants were able to prosper.
In the international market Australia had a traditionally regulated
industry. However, by the late 1970s, there was strong pressure from the
newly established Asian airlines to enter the market; with significantly
lower costs, they were able to offer much lower fares to Asia and Europe
than the incumbent airlines such as Qantas. The government fought hard
to keep the sixth freedom airlines out, but eventually, responding to public
pressure for lower airfares, it relented and allowed some measure of liber-
alization. Since then it has gradually liberalized, though it still prefers to
impose (fairly lax) capacity controls. A measure of liberalization is that it
has been prepared to accept a low market share for the Australian airlines.
The Australian experience with the competition from Asian and now Gulf
airlines has been similar to that now faced by European and US carriers
facing competition from Gulf airlines.
This chapter begins with a brief background. It examines, first, domestic
51
BACKGROUND
exceptions being Sydney, Melbourne, Brisbane and Perth, which are now
subject to light-handed regulation. Some airports are listed on the stock
exchange (Sydney, for example) and others are owned by investors such as
Australian and overseas pension funds. Smaller airports are mostly owned
by local authorities.
As with other countries, international and domestic aviation are regu-
lated under different arrangements. International air transport is handled
by bilateral agreements, with the exception of routes to New Zealand,
which are handled under the Single Aviation Market. Domestic trunk ser-
vices are governed under the federal government, and regional services by
the states (some of which still impose regulation).
For most of the period between the end of WWII and 1990, Australia
imposed very tight regulation of domestic trunk airlines. The policy was
termed the Two Airline Policy. This policy was implemented rigorously
from about 1960 until the early 1980s, when it was loosened slightly
(Brogden, 1968; Forsyth, 1991). After a review in 1986 (May, 1986), in
1987 the government announced that it would deregulate the airlines in
1990, which it did. There was also tight regulation of regional services,
though the regulatory arrangements differed from state to state.
The underlying theme of the Two Airline Policy was a concern that com-
petition between a government-owned airline, Trans Australian Airlines
(TAA), and a private airline, Ansett, was inherently unstable – the govern-
ment airline would use subsidies to increase its market share, and drive
the private airline out of business. In the earlier, though not the later years
there were some grounds for this fear (Brogden, 1968). The only way of
eliminating this was to make the two airlines as similar as possible (and
thus preserve ‘competition’).
The core of the policy was rigorous control of capacity: the two airlines
were allowed to have exactly the same capacity in terms of seats. This was
achieved by the two airlines having exactly the same aircraft and additions
to capacity in the form of new aircraft arriving in Australia on the same
day. Capacity was carefully monitored, and additions to capacity were
allowed when target load factors were achieved. The outcome was that
actual load factors were constantly at 65 per cent, unless recessions meant
that actual load factors fell below this temporarily. Airfares were regulated
by the government, though in the 1980s an independent tribunal was set
up to regulate them (May, 1986). There was little diversity in airfares:
there were first and economy fares, plus some discounts for students and
pensioners. Fares were set on a distance basis (which meant that long-haul
flights cross-subsidized short-haul flights) (Forsyth, 1991).
The nature of the regulation in Australia was rather different from that
imposed in the US before deregulation. There was no capacity control in
the US, but there was control over airfares, along with control over routes
(see Bailey et al., 1985). In Australia there was no control over routes, and
the airlines could operate the network they chose. As it turns out, they
chose to operate very similar networks. One feature of the Australian
system was that the airlines tended to operate ‘parallel schedules’ (Gannon,
1979; Hocking, 1972), which meant that there were two flights per day
from Sydney to a destination, one from each airline, and they tended to
depart within five minutes of one another (economists noted the relevance
of the Hotelling model of spatial competition).
The Two Airline Policy was implemented for many years without much
discussion. However, in the late 1970s it started being discussed more
critically, reflecting the worldwide re-evaluation of regulation, which had
begun in the US (Forsyth, 1991). Australia was slightly behind the US,
but by the late 1970s banking, telecommunications and transport regula-
tion were all being subjected to critical scrutiny. It was not surprising that
the Two Airline Policy became a topical issue. There were a small number
of economic evaluations of it; some politicians advocated its removal
(notably those from Western Australia, the most remote state and the one
most affected by high airfares), and airlines other than the two chosen
trunk carriers began seeking access to the market.
The complaints about the policy were several. There were not many
complaints about the airlines’ networks or about the quality of services,
which by international standards were high. However, there were consider-
able complaints about airfares, concerning both their level and structure
(Holcroft, 1981). Australian airfares were significantly above those of the
trunk airlines in the US. Long-distance flights, such as Sydney–Perth, were
particularly high relative to those in the US, or even between Australia
and New Zealand. There was little diversity in airfares, although the air-
lines did make some discount fares available from the late 1970s, notably
standby fares. Fares were regulated, and profits were not excessive, though
they were very stable. This suggested that the productive efficiency of
airlines was low compared to that of the US carriers. A number of pro-
ductivity studies confirmed that this was so (Forsyth and Hocking, 1980;
Mackay, 1979). While most states continued to regulate regional airlines,
South Australia deregulated them; the results from this experiment were
encouraging (Starkie and Starrs, 1984).
While the Two Airline Policy became a policy issue, the two airlines were
very comfortable with it. The airlines and the government recognized that
there was pressure for change, but they did not wish to have major change,
let alone full deregulation. The government instituted a number of inquir-
ies (Holcroft, 1981; May, 1986), and there were several modifications to
the policy, both at the regulatory level (a new body to regulate airfares,
the Independent Air Fares Commission; see May, 1986) and at the airline
level (such as selected discount fares). By the middle of the 1980s the US
had deregulated its domestic airlines, and the results were regarded as
good, though there were some problem areas (for example, the failure of
some carriers). The US experience was very relevant for Australia, since it
became clear that there were alternatives to the Two Airline Policy.
Changes within the private carrier Ansett were significant. For many
years, ownership and control of Ansett was dominated by its founder, Sir
Reginald Ansett (who set up the airline in the 1930s as a way around crip-
pling road freight regulations). The airline was tightly held and difficult to
take over. However, by the 1980s a combination of the transport company
TNT and News Limited (Ansett owned some TV stations) managed to
take it over. The new owners were much less keen on regulation, and saw
opportunities in a deregulated environment (Forsyth, 1991). The upshot
was that the government decided to open up the domestic market quite
extensively in 1987, and implemented its new policy in 1990.
two LCCs: Jetstar, which is owned by Qantas, and Tiger Australia, which
is owned by Virgin Australia. The first LCC entry was that of Compass,
which commenced shortly after the market was opened. This airline faced
a number of problems; some of these were managerial failures (such as the
choice of very large, difficult to fill, aircraft) and other problems included
difficulties in obtaining gates (see below) (Nyathi et al., 1993). It only
lasted a year. It was resurrected, but Compass Mark II was short-lived.
After this, it appeared that Qantas and Ansett would continue to dominate
the market, and there were no further entries until 2000.
There were two entries in 2000: Impulse and Virgin Blue (for discussion
of Australian LCCs, see Forsyth, 2003; Gross and Luck, 2013; Budd and
Ison, 2014). The latter was founded by the Virgin Group, which also has
airline investments such as Virgin Atlantic. Impulse was soon taken over
by Qantas, but Virgin grew rapidly. Qantas established an LCC subsidiary,
Jetstar, which operates both domestic and international services, and has
been one of the most successful parts of the Qantas group. It was very
well placed when Ansett collapsed in September 2001. It has continued
to grow, to the extent that it has about one third of the domestic market.
Another airline, Tiger Australia, entered the market; recently, this airline
has been taken over by (the renamed) Virgin Australia, which has also
transformed itself to become an FSC (while there have been no entries by
FSCs, it is interesting that Virgin has seen a market for an FSC in addition
to Qantas). Thus, at the moment, Australia has two FSCs, each having
an LCC subsidiary. Since 2000 there have been some smaller operators,
such as Alliance Airlines, which have concentrated on (mining) charter
operations.
There were several problems facing the earlier entrants. In the early
days post-deregulation, access to funding was not much of a problem.
However, after the failure of Compass Mark II, potential new entrants
had great difficulty in persuading capital markets; it was argued that the
domestic market had reached its equilibrium of two airlines. However, the
two entrants in 2000 did have some advantages. Impulse was based in an
established regional carrier, and Virgin Blue was a venture of the interna-
tional Virgin group.
However, the really difficult problem which was faced by the early
entrants was access to terminals and gates (Nyathi et al., 1993). When the
government deregulated the airlines it protected the incumbent airlines by
signing long-term leases of their terminals with them. This had the (inten-
tional) effect of making it difficult for new airlines to operate. For most
of the major destinations, such as Sydney and Melbourne, there were two
domestic terminals, leased by the two airlines, Australian (formerly Trans
Australia Airlines) and Ansett, and there was also a common-user inter-
There are, however, two types of regulation which Australia still tends
to enforce: one of these concerns gateways, and the other is about capac-
ity. If a country seeks to establish air services with Australia, the default
position is that it may do so with no capacity controls except for flights to
Sydney, Melbourne, Brisbane and Perth. This is to encourage flights to
smaller centres, such as Adelaide, Cairns, the Gold Coast and Canberra.
If the country seeks permission for its airlines to access the four major
centres, Australia’s normal position would be that it would set capacity
controls (the ultimate outcome will depend on the bargaining power of the
country). Capacity would be set such that it is sufficient to handle expected
demand. As demand grows, additional capacity can be negotiated. There
are few flights which go to cities other than the big four.
On the face of it, this seems to be a fairly liberal arrangement. However,
there are some problems. For a start, the system relies on the government
regulators being able to forecast demand ahead; sometimes demand gets
ahead of capacity (for example, from time to time in the rapidly growing
China market). Secondly, and less obviously, this system can still result in
a degree of restriction which pushes fares above what they would be in a
competitive market – it is one thing to adjust capacity such that it is suf-
ficient to meet demand, but at what price? A good question would be ‘why
is it necessary to regulate capacity at all?’ Other airline markets in Australia
and elsewhere do not need this regulation and yet they perform well.
However, this regulation does have the effect of affording the Australian
carriers some degree of protection.
Given that there is limited capacity for the Australian airlines which
would like to service a specific route, an allocation problem arises. For
example, both Qantas and Virgin may wish to serve a route (either with
their own equipment or on a code-share basis), which could be potentially
profitable. To solve this problem, the government has set up an arm’s-length
body, the International Air Services Commission (IASC) (see Findlay and
Round, 1997), to allocate routes. Since this may pose some competition
policy issues, this body works closely with the main competition regulator,
the ACCC (Australian Competition and Consumer Commission).
The result of this incomplete deregulation is that while some routes
are very competitive, others are not, and airfares are moderately high.
The Europe route is very competitive; there is a wide variety in quality of
service and fares, and overall fares are very competitive. On some of the
less dense routes, such as those from Africa and South America, fares are
somewhat higher. Airfares depend on the tightness of regulation and on
competitive conditions (Forsyth and Seetaram, 2014). The US route is
not greatly regulated, though fares have been high sometimes (see below).
Currently they are low, reflecting the fact that there are several airlines
competing. However, ten years ago, with Qantas, United and Air New
Zealand accounting for most of the traffic, they were moderately high.
Like most airlines, the Australian international airlines have used alli-
ances to further their objectives. Qantas has been in alliances since the
1930s, when it had an alliance with Imperial Airways (forerunner of British
Airways) to operate flying boat services to the UK. It is a member of the
oneworld alliance, and as is common, it has alliances with individual air-
lines. For many years it had a joint venture (involving pricing and sharing
of revenues) with British Airways (this joint venture was approved by the
UK and Australian competition authorities). However, in 2013 it termi-
nated this arrangement in favour of an alliance with Emirates (which had a
very convenient hub for Australia–Europe traffic) (O’Sullivan, 2015); this
too was approved by the competition authorities. However, this was not
the case when it sought to develop an alliance with Air New Zealand, as it
was not able to obtain approval from competition authorities, which feared
lack of competition in the Trans Tasman market (O’Sullivan, 2015).
Virgin Australia has a small international operation and relies very much
on its alliances with larger carriers; it has an extensive Australian network,
but only operates a small number of flights to a range of international des-
tinations. As a result it is able to offer a wide variety of destinations. Thus,
it offers flights to Singapore with one of its owners, Singapore Airlines,
and New Zealand, with one of its former owners, Air New Zealand.
However, potentially its most important alliance is with Etihad, which is
creating an alliance of its own.
Given that issues and responses depend very much on routes and times,
a good way of analysing the issues is by brief case studies. Here five are
identified:
Until the 1970s, sixth freedom was not an issue for Australia or, for that
matter, most other countries. Qantas had flights to Europe (especially the
UK), and a range of European airlines had flights to Australia. Asian
airlines were in their infancy – they had few flights and were being given
substantial technical and financial help by the developed country carriers.
They were not in the position of competing on sixth-freedom markets.
This changed quickly over the 1970s. The airlines of several Asian
countries started services to Europe and to Australia. Initially they were
not seen as being very competitive, although this changed over time. The
capacity which they were able to devote to Australian routes was limited
by regulation. Some did offer flights to Europe, selling the flight with two
separate tickets; for example, one might have one ticket from Sydney to
Singapore, and another from Singapore to London. In addition, for a time,
ship/jet charters became popular: one would take a ship from Australia to
Singapore, and there board a flight to London. This was worthwhile, given
the high (regulated) fares of the Australia–Europe flights. The British LCC
Laker Airways repeatedly sought to enter the Australia–UK market, prom-
ising significantly lower fares, but it was never permitted to enter. Both the
government and the flag carrier were very concerned about ‘malpractices’
such as fare discounting, and the carrier was losing market share. The issue
of airfares became a live political issue.
The government set up an inquiry, the ICAP (International Civil
Aviation Policy) (Findlay, 1985; ICAP Report, 1978) committee, which
reported in 1978, and following this, the government changed its policy in
1979. The policy was one of more tight regulation. Some new fares were
introduced, such as advance booking fares, though stopovers were not
permitted on these. The intention was to concentrate traffic on hubs such
as Sydney to try to increase load factors. Capacity controls were tightened
with the intention that the Asian carriers would only have enough capacity
to serve their own market; that is, there was a conscious attempt to freeze
out sixth freedom traffic.
The Asian countries, led by Singapore, were very much opposed to this
policy, since they lost airline revenue and tourism revenue from stopovers.
The ASEAN (Association of South East Asian) countries were able to
negotiate as a bloc, and they succeeded in gaining some modifications
to the policy (Findlay, 1985). For a time, perhaps two or three years,
Qantas did well out of the policy. However, it was not really possible to
limit capacity of the Asian airlines to simply that warranted by third and
fourth freedom traffic. Essentially, the policy broke down, and discount-
ing became common again by the mid-1980s. The Australian experience is
relevant to the current debate in the US, and to a lesser extent in Europe,
about the impact of the Gulf carriers on US and European airlines. There
have been several suggestions that the Gulf carriers should be limited to
carrying only third and fourth freedom traffic, and that they should not
be allowed enough capacity to carry sixth freedom traffic (Partnership
for Fair & Open Skies, 2015). If the Australian experience is any guide, it
would be very difficult to police this (as it was not feasible even in a world
where regulation was the norm).
Initially, it was the Asian carriers that developed sixth freedom traffic
from Australia. More recently, it has been the Gulf carriers which have
become the pacemakers, and Emirates now is one of the two main carri-
ers on the Australia–Europe route (the other being Singapore Airlines).
Australia still controls capacity, but it has chosen to grant adequate
capacity for them to expand rapidly, even though these carriers are almost
entirely carrying sixth freedom traffic. Qantas was particularly concerned
about the growth of these carriers on Australian routes. However, the situ-
ation has changed, with the two Australian carriers forming close alliances
with Gulf carriers: Qantas with Emirates, and Virgin with Etihad.
Of the Asian routes, historically the most important has been the Japan
route, though its importance has been falling recently with a decline in
Japanese tourism to Australia. This route has also been the most tightly
regulated one. Airfares were very high from the time that Japanese tourism
started growing in the 1970s. There has been some liberalization, but it was
not until around ten years ago that the fares on the route fell to something
like those on other routes. The Japanese government preferred tight regula-
tion on this and all other international routes (Yamauchi, 1997). There was
some sixth freedom traffic, and for a time, Northwest Airlines from the US
operated on a fifth freedom basis, though without success.
A significant feature of this route is that the tourism industry first
became a major stakeholder in the international aviation market. The
tourism aspects of the Europe route were not discussed much: most of the
discussion was in terms of the benefits to the home consumer versus the
airline profit effects (though the ICAP report estimated both home and
foreign consumer surpluses in its evaluations). The Japan route was pri-
marily a tourism route, and the tourism industry developed a voice. Since
then, governments have been seeking a balance between airline industry
interests, consumer interests and now tourism industry interests in most
air transport discussions.
The route is an interesting one, since the two end-countries were acting
against their own best interests. By far most of the traffic was Japanese,
and the airline capacity was shared equally by airlines of the two countries.
This meant that the high fares were paid by Japanese nationals, and the
airline revenues were shared between Japanese and Australian airlines.
Since for much of the time the Japanese airlines had very high costs, the
profits from the route accruing to them were not large. This meant that
Japan was paying a very high price to protect its airlines.
By contrast, Australia kept pressing the Japanese to liberalize. This
the two countries for their airlines and liberalized and harmonized owner-
ship arrangements. In addition, the two countries introduced an open skies
agreement which commenced in 2002. This made it possible, for example,
for an Australian airline to fly to New Zealand and then pick up passengers
and freight to the US.
The SAM has resulted in the effective integration of the two aviation
markets. So far, Air New Zealand has not entered the Australian domes-
tic market, though it has considered doing so. On the other hand, the
Australian airlines have been active in the New Zealand domestic market:
both Qantas and Virgin Australia operate services through subsidiaries.
Qantas flights from Australia to New Zealand are operated by a New
Zealand subsidiary, using both New Zealand-registered aircraft and New
Zealand crews (costs in New Zealand are lower than in Australia). In the
past, Air New Zealand held a major shareholding in Ansett, but the Ansett
collapse has made it wary of Australia. For a period it held a share of
Virgin Australia, though it disposed of this in 2016. Qantas held a share-
holding in Air New Zealand for some time: it was planning an alliance and
joint venture (one which would cross alliance boundaries, since Qantas
is in oneworld and Air New Zealand is in Star Alliance). However, the
joint venture was rejected by competition authorities, and Qantas sold its
holding soon after.
Adaptation to Asia
Australia is close to Asia, and the aviation links to Asian countries are
important. One critical problem is that Australia’s costs are much higher
than the costs of most Asian countries, excluding Singapore and Japan
(Oum et al., 2000). This has made it difficult for Australia’s airlines to
compete. As a result, the Australian airlines, especially Qantas, have been
at the forefront of internationalizing their labour forces. Thus, Qantas
makes considerable use of foreign crews on international flights; it also
relies heavily on maintenance done offshore. It has established subsidiar-
ies, such as Jetstar Asia, based in Singapore so as to access Asian markets.
Most of these ventures have been successful. In spite of this, the Australian
airlines have not been able to gain a large share of the vast and rapidly
growing China market.
Australia lies in a region in which its airlines are not particularly cost-
competitive, at least in some key routes, such as those to Europe and Asia
(Oum et al., 2000). Most airlines in the Asian region (with the exception of
Singapore and Japan) and the Gulf countries face lower labour costs, and
if they are productive, can achieve higher cost competitiveness than the
Australian airlines. The airlines are cost competitive on North American
routes, but even with the New Zealand route, there has been pressure over
the years – since the 1970s – from consumer interests, and more recently
the tourism industry, for lower fares. There has been a succession of cases
where foreign airlines have been offering lower fares if they are permitted
to enter the routes. Governments have been prepared to allow these airlines
access, though rarely on an open skies basis. This has had implications for
the Australian airlines, and their market share has gradually fallen.
CONCLUSIONS
REFERENCES
Bailey, B., D. Graham and D. Kaplan (1985), Deregulating the Airlines, Cambridge,
MA: MIT Press.
Benns, M. (2009), The Men Who Killed Qantas, North Sydney: Heinemann
Australia.
Brogden, S. (1968), Australia’s Two Airline Policy, Carlton: Melbourne University
Press.
Budd, L. and S. Ison (eds) (2014), Low Cost Carriers: Emergence, Expansion and
Evolution, Farnham: Ashgate.
Bureau of Transport and Communications Economics (1991), Costs and Benefits
of a Single Australasian Aviation Market, Canberra: Australian Government
Publishing Service.
Department of Transport and Communications (Australia) (1988), Negotiating
International Aviation Rights, Canberra: Department of Transport and
Communications.
Findlay, C. (1985), The Flying Kangaroo: an Endangered Species? An Economic
Perspective of Australian International Civil Aviation Policy, Sydney: Allen and
Unwin.
Findlay, C. and D. Round (1997), ‘Multiple designation policy in Australia’, in
C. Findlay, Chia Lin Sien and Karmjit Singh (eds), Asia Pacific Air Transport
Challenges and Policy Reforms, Singapore: Institute for Southeast Asian Studies,
pp. 84–105.
Forsyth, P. (1991), ‘The regulation and deregulation of Australia’s domestic airline
industry’, in K. Button (ed.), Airline Deregulation: International Experiences
London, London: David Fulton Publishers, pp. 48–84.
Forsyth, P. (2001), ‘Total factor productivity in Australian domestic aviation’,
Transport Policy, 8(3), 201–207.
Forsyth, P. (2002), ‘Privatisation and regulation of Australian and New Zealand
airports’, Journal of Air Transport Management, 8, 19–28.
Forsyth, P. (2003), ‘Low cost carriers in Australia: experiences and impacts’,
Journal of Air Transport Management, 9, 277–84.
Forsyth, P. and R. Hocking (1980), ‘Economic efficiency and the regulation of
air transport’, Monograph M62, Committee for the Economic Development of
Australia.
Forsyth, P. and N. Seetaram (2014), ‘The future of Australian international avia-
tion: liberalisation, competition and the Dutch disease’, in D. Duval (ed.), Air
Transport in the Asia Pacific, Farnham: Ashgate, pp. 143–56.
Gannon, C. (1979), ‘Parallel scheduling’, Appendix 12 Department of Transport,
Domestic Air Transport Policy Review, Appendices, Canberra: Australian
Government Policy Review.
Gregan, T. and M. Johnson (1999), Impacts of Competition Enhancing Air Services
Agreements: A Network Modelling Approach, Productivity Commission, Staff
Research Paper, Canberra: AusInfo.
Gross, S. and M. Luck (eds) (2013), The Low Cost Carrier Worldwide, Farnham:
Ashgate.
Hocking, R. (1972), ‘Some economic aspects of Australia’s Two Airline Policy’,
Monograph M35, Committee for the Economic Development of Australia.
Holcroft, W. (Chairman) (1981), Domestic Air Fares: Report of the Independent
Public Inquiry, Canberra: Australian Government Publishing Service.
ICAP (1978), ‘Review of Australia’s International Civil Aviation Policy (ICAP):
Report of Review Committee’, Canberra: Australian Government Publishing
Service.
INTRODUCTION
The argument for liberalization in 1985 across Europe was that the closed
market access system of non-competing national airlines would produce
high fares, low productivity and a smaller overall market output and that
new market entry was required. The Irish case study vindicates the advo-
cates of liberalization. Ryanair is the largest new entrant to the liberalized
European market. The 2015 passenger numbers for Ryanair (106 million)
exceed those of traditional national airlines such as the International
Consolidated Airlines Group (IAG), based on British Airways and Iberia
69
London and New York and in aircraft leasing at Guinness Peat Aviation,
and the business model was based on that of Southwest Airlines in the
US; fares were low, leading to near immediate economies of density with
a rapid increase in traffic volume. Competitively, this also gave an advan-
tage on Ireland–UK routes because Ireland was in recession at the time of
the airline’s launch. Ryan recognized that another cartel airline was not
needed and that Ryanair had to differentiate itself from low-productivity
and high-cost legacy operations. Ryanair’s productivity could not have
been attained by adopting the traditional European national airline model.
Thus, Ryanair’s chief executive, Michael O’Leary, differentiated the airline
both from legacy airlines and from other start-ups.
Having achieved unprecedented productivity from its own staff com-
pared to incumbent airlines, Ryanair then targeted its suppliers. Discounts
were sought and achieved from aircraft manufacturer Boeing and from all
of the airports that Ryanair served. Airport competition was promoted to
create new routes and exert leverage on high-cost airports. For example,
the use of air bridges was declined and the design of terminals changed
to the low-cost model. Travel agents were removed from the selling of
Ryanair products and were replaced by online sales. Seats per aircraft were
maximized and load factors rose steadily to over 90 per cent. Free inflight
service was replaced by sales. Passengers readily took to the new product
because of the low fares on offer.
Ryanair achieved a 25-minute turnaround time at airports on the basis
that more round trips per day per plane would improve the economics
of the airline. Ryanair also located its fleet at 84 bases throughout the
network, an average of four per base. This contrasts with the old national
airline practice of locating aircraft at major hubs. All Ryanair planes
return to their home base each evening, in contrast to the expense incurred
by the legacy airlines of overnighting crews in hotels. Living away from
congested hubs reduced living costs for Ryanair crews and enhanced their
lifestyles. The traditional location of national airline fleets at busy hubs,
where they jointly operated the major air routes, left many second-city
markets open to new entrants such as Ryanair.
Irish airline financial instability affected Ryanair in its initial years, the
smaller airlines in Table 5.1 from time to time and Aer Lingus at intervals.
Weldon (2002) noted 28 years of loss-making between the establishment of
Aer Lingus in 1936/1937 and 2001, when it lost €140 million. Losses were
€317 million between 2006 and 2009 and €95.8 million in 2014. Ensuring
financial and employment stability at Aer Lingus was the motivation for
the government of Ireland to sell the airline to IAG in August 2015 and
delist it from the Irish Stock Exchange on 17 September. The alternative
bidder for Aer Lingus was Ryanair, which already held a 29.8 per cent
stake, but this option was opposed by the European Union, the UK com-
petition authorities and the Government of Ireland, which held a 25 per
cent stake.
Ryanair owes its existence to liberalization and has thrived in that
context. Its passenger numbers for 2017 to 2018 are predicted to be 125
million, and 135 million in 2019. Ryanair has annual profits of €250
million, a net margin of 15 per cent and a market capitalization of €15.5
billion. It employs 11 500 people and its dominant role in the Irish airlines
sector has secured overall profitability and employment growth in the
sector more than stability in its finances and numbers employed. The sug-
gested excessive instability does not seem to be present.
The second negative assessment cited by Goetz and Vowels is reduced
quality of service. The European airline that most prioritizes service
quality is CityJet, which offers leather seats, meals and the use of London
City airport with its proximity to the financial services district of London.
In 2016, CityJet has approximately the same passenger numbers as Aer
Lingus did before liberalization. Aer Lingus and British Midland dropped
their business class to compete with Ryanair. When food and bever-
age service on Aer Lingus was unbundled from high price tickets, it was
only on the Dublin–London and Dublin–Brussels routes that passengers
showed a willingness to pay for these services at the point of use. Many of
the services that were unbundled by price competition following liberaliza-
tion were inherited from the era of non-price competition by airlines. The
trade-off between price and non-price competition may change over time.
For example, in the European competition between Ryanair and easyJet,
Ryanair has sought to match its competitor by service improvements such
as flying to some hub airports, reserved seating and priority boarding,
while maintaining its claim to have the lowest prices. In 2017, Ryanair
secured a small presence at Frankfurt due to concerns over the dominance
of Lufthansa, but hubs such as Heathrow and Charles de Gaulle retain
grandfather rights in slot allocation.
It has also been argued that low-cost airlines will ultimately offer fewer
flights at higher fares to smaller places. In Ireland, this feared negative
result of liberalization has not occurred: fares have fallen and frequencies
increased on all routes. Both Cork and Shannon, the pre-liberalization sec-
ondary airports, gained from increased frequency and reduced fares. Two
additional regional airports, Knock and Kerry, shared this experience. The
Ryanair network is hardly small, serving 1800 routes to 200 airports.4 Put
simply, there were many airports in Europe that the legacy airlines chose
not to serve because they based their operations overwhelmingly at hubs.
Low-cost carriers have filled this void.
The new entrant airlines were restricted by the slot system and the grand-
father rights of legacy airlines at hub airports. Therefore, the new entrants
sought out secondary airports. In 2010, Ryanair served 153 airports in
Europe. Only two of these, Madrid and London Gatwick, were in the top 10
airports (by passenger numbers), while 129 (84 per cent) of the Ryanair air-
ports were outside the top 100. Thus, liberalization in Europe brought sig-
nificant gains to secondary airport locations. The higher population density
in Europe compared to the US meant that secondary airport services had
more potential passenger numbers and the new routes proved viable. The
results of liberalization at Europe’s smaller airports were the opposite of
those feared in the US. The success of low-cost airlines at secondary air-
ports in Europe has prompted some hub airports to seek service from these
airlines at airports such as Brussels, Amsterdam and Copenhagen because
growth at such airports is reduced by reliance on legacy airlines alone.
Goetz and Vowles (2009) viewed industry bankruptcy as an undesirable
result of liberalization in the US. In stressing the role of freedom of entry
in the theory of contestable markets, William Baumol (1982) stated that
‘perhaps a bit newer is the emphasis on freedom of exit which is as crucial
a requirement of contestability as is freedom of entry. Thus, we must
reject as perverse the propensity of regulators to resist the closing down of
unprofitable lines of activity’.
Bankruptcy removes the control of assets from a management that has
failed the market test. Many long-established airlines, both in Europe and
the US, have either left the market or have been absorbed by other airlines,
as was the case with Aer Lingus. Having grown up in a tradition of pro-
tectionism, they were unable to survive in a liberalized market. Mergers
and takeovers were the alternative to bankruptcy for many airlines. The
industry is now more concentrated and this should be considered by com-
petition authorities dealing with further takeovers, both in the US and in
Europe. This aspect of liberalization is considered below.
Goetz and Vowles (2009, p. 259) also refer to ‘the renewed debate about
the efficiency of deregulation and liberalization policies, particularly at a
time when the global financial crisis has cast a harsh spotlight on the (un)
desirability of these polices’. The Irish example does not show any conta-
gion linking the severe financial crisis and successful aviation liberaliza-
tion. Regarding airline liberalization in Ireland, the only regulatory change
was the removal of the restrictions on new entrants. All other standards as
applied to Aer Lingus were applied to the new entrants. Contagion may
well exist but as a contextual matter that did not apply to Ireland.
The Goetz and Vowles view of airline deregulation in the US is signifi-
cantly more pessimistic than the Irish experience since 1986. Kahn’s (1988,
p. 321) assessment of US deregulation is closer to the results in Ireland,
namely,
The last ten years have fully vindicated our expectations that deregulation
would bring lower fares, a structure of fares on average in closer conformity
with the structure of costs, an increased range of price-quality options and great
improvements in efficiency – made possible by the abandonment of regulatory
restrictions and compelled by the greatly increased intensity of competition – all
this along with a 35 per cent or so decline in accident rates.
The Irish Aviation Authority (IAA) has responsibility for safety, jointly
with the United Kingdom and Canada, of a large section of the North
Atlantic. The IAA is also responsible for the safety regulation of some 500
civil commercial airliners in the Irish fleet, of which 350 belong to Ryanair
and 50 to Aer Lingus. IAA also regulates safety at each destination flown
by Irish airlines. It is itself regulated by the International Civil Aviation
Organization (ICAO).
According to the chief executive of the IAA, ‘ICAO ranks Ireland
consistently among the world’s top four regulators ahead of the US which
came sixth in 2015. The FAA categorizes Ireland as a category 1 state.
They take our certificates unreservedly’ (Irish Times, 8 July 2016). The
ICAO 2015 Universal Safety Oversight Audit ranked Ireland second in
Europe for effective implementation of almost 11 000 ICAO standards and
recommendations. Ireland’s compliance rate was 94 per cent, compared to
93.63 for the UK, 91.35 per cent for USA, 89.8 per cent for Germany, 94.4
for France and 95.3 for Canada. Ryanair has never had a passenger fatal-
ity, while the last Aer Lingus passenger fatality was in 1968.
The IAA website states that the IAA ‘regulates legacy and low-cost
airlines to the same safety and operating standards which conform to all
European Union and European Aviation Safety Agency regulations and
requirements’. Some early opponents of liberalization in Ireland claimed
that new-entrant airlines would use inferior aircraft and would be poorly
have aircraft and crew wait there while passengers disembarked to engage
in duty-free shopping for an hour. On eastbound journeys, the stopover
required the planes to land at Shannon in the early morning, thus disrupt-
ing the sleep patterns of those on the overnight flight before flying on to
Dublin. Given the absence of direct flights to Dublin, many passengers
chose voluntarily to stop over at London instead of Shannon. In 2006,
when Shannon’s North Atlantic passengers peaked at 781 000, they were
37 per cent of the Dublin and Shannon combined market on the routes. By
2012 the Shannon share had declined to 15 per cent.
In 2015 Dublin was the fifth-largest airport in Europe in terms of
traffic to North America. The requirement to serve both Dublin and
Shannon had improved the competitiveness of Dublin routes serving
North America, winning back traffic that had been diverted over London.
Dublin has also engaged in competition with London by becoming a hub
that feeds its North Atlantic services from its large UK and European
mainland network of routes. In 2016, one million passengers from Aer
Lingus’s European routes transferred to services to North America from
Dublin. US customs and border protection services conduct customs and
immigration pre-clearance procedures at Dublin and Shannon so that pas-
sengers are treated as domestic passengers on arrival at their US destina-
tion airports.
While the Shannon stopover was a burden imposed by Ireland on
itself, and could have been removed at any time, its demise in 2008 as
part of Open Skies was a further benefit of liberalization. The end of the
Shannon stopover brought a Blue Skies air agreement between Ireland
and Canada, permitting services between the countries and allowing the
countries to use each other as platforms to serve other destinations. In the
summer of 2016 there were 20 daily flights from Dublin to the US and
four daily to Canada (to Toronto, Montreal, Vancouver and St John’s).
Added to this, a fifth freedom service was opened in 2015 by Ethiopian
Airlines, serving Dublin in both directions on its Addis Ababa–Los
Angeles route.
Norwegian Air’s attempt to enter the Cork–Boston route in May 2016
was delayed by opposition in the US to the granting of a Foreign Air
Carrier Permit. The service is now scheduled to begin in 2017. Norwegian
Air is registered in Ireland and sought to serve the US under the Open
Skies agreement between the EU and USA. After lengthy objections from
airlines in the USA, and aviation trade unions in the US and those rep-
resenting pilots at Aer Lingus, Norwegian was permitted on 3 December
2016 to serve the Ireland–US market. It has 55 aircraft on the Irish register
and plans to feed its Cork–Boston service from Barcelona. Further plans
in Ireland include serving Shannon and New York.
those at Cork and Shannon were 37 and 43 per cent lower, respectively,
with Cork declining by 1.2m passengers and Shannon by 1.3m passengers.
The period from 2008 to 2015 also saw a decline at the new airports, other
than Knock, which entered the market after liberalization. Galway, which
had 266 000 passengers in 2008, closed in 2011. Sligo had 42 000 passengers
in 2008 but also closed in 2011. Waterford was the base for Ryanair’s first
route, to London Gatwick in 1985, but when it chose to concentrate solely
on the Boeing 737 in 1991, Ryanair moved to other airports. Waterford
had 144 000 passengers in 2008 but lost passenger service in 2016 when
Belgian airline VLM left the Waterford–London route. The VLM chief
executive stated that ‘Waterford is close enough to Dublin and Cork, so the
catchment area will always be sort of limited’.6
Turning to matters of ownership, Shannon’s traffic fell steeply after
2008, falling from 3 million to a trough of 1.3 million in 2012 but rising to
1.7 million in 2015. It has had independent governance since 2013, with the
aspiration that this will enable it to compete with Dublin more effectively
than if remaining part of the Dublin Airport Authority.7 Independence for
Cork Airport will be re-examined in 2019. The Airport Authority narrowly
declined the option of independence in 2008 because of fears about the
airport’s debt. Its decline has been continuous since 2008, from 3.3 to 2.1
million passengers, but turned around in 2016.
Cork and Shannon are legacy state airports that predate airline liberaliza-
tion in 1986. In that era, all state airports were run by Aer Rianta. Airlines
did not compete on price and passed on higher airport charges in the form
of higher fares. In that era, there were no new entrant airlines choosing
between competing airports and there was no independent regulation of
airport charges. There were likely inefficiencies in such a market structure,
with dangers to the efficient allocation of both capital and labour resulting
in ‘Averch–Johnson effects’ (basically gold-plating) and low labour produc-
tivity.8 Capital projects were prestige buildings that embodied the region or
neo-Keynesian development measures, rather than commercial investments
where a wrong decision could harm the competitiveness of an airport.
Shannon invested £40 million in a terminal in 2000 when it had 2.4 million
passengers, a number that declined to 1.7 million in 2015. Cork invested
€120 million in a new terminal in 2006 and the Airport Authority took
responsibility for the debt in 2008, thus reducing the competitiveness of the
airport; however, as we have seen, passenger numbers declined until 2016.
Criticisms by the Commission for Aviation Regulation of capital spend-
ing at Irish airports, as proposed in 2001, included:
The regulator disallowed €50 million of the €106 million investment pro-
posed at Shannon and €79 million of the €136 million proposed at Cork.
Labour productivity at Cork and Shannon was examined by the
Infrastructure Management Group (2001). While Cork was 15 per cent
above the average of its UK peers, Shannon was only 37 per cent of that
average. Part of the justification for the Shannon stopover discussed
above was the creation of employment in the region. However, creating
employment by over-manning an airport undermines the competitive-
ness of a region. Cork’s entry to the North Atlantic market in 2017 with
Norwegian Air is a new market for the airport. Shannon will also be served
by Norwegian with the advantage of US immigration and pre-clearance.9
The distribution of Irish airport passengers towards Dublin was also
influenced by the development of motorways and the reduction of subsi-
dies to regional air services. As we have seen, the motorway network, with
connections from Dublin and Dublin Airport to Cork, Galway, Limerick
and Waterford, had an impact on air travel. Internal air routes to Dublin
peaked at 885 000 passengers in 2007 but declined to 61 000 in 2012 after
the completion of the surface network. The air services affected com-
prised a mix of commercial routes, such as Cork and Shannon to Dublin,
and subsidized services to smaller centres. Cork to Dublin declined from
half a million passengers in 2007 to under 5000 in 2012 and by 2015 had
no service. Similarly, Shannon to Dublin declined from 100 000 in 2006
to 1393 in 2012 and had no service in 2015. The subsidized routes from
Dublin to Galway, Sligo and Knock were withdrawn from the public
service obligation subsidy scheme in 2011, leaving the services to Kerry
and Donegal as the only remaining parts of the domestic route network.
These routes had 84 000 passengers in 2015, which accounted for only
0.3 per cent of passengers at Dublin. The PSO subsidies to Donegal and
Kerry are scheduled to end in 2018. Unless actions by airlines and airports
improve the economics of these routes, there will be no remaining internal
routes in Ireland.
Regional airports in Ireland gained from the initial stages of liberaliza-
tion when turboprop aircraft could undercut the high fares that the incum-
bent airlines were charging on services to the UK and mainland Europe,
and Aer Lingus responded by operating turboprop feeder services to its
hub at Dublin. The lower fares charged by Ryanair and a restructured Aer
Lingus subsequently undermined the economics of both direct turboprop
services to Irish regional airports and the Aer Lingus feeder services to
Button (2012) stated that ‘the time has come to begin experimenting with
more infrastructure reforms in the US’ and called for ‘more market-based
approaches to the supply and use of air transportation infrastructure’. At
airports, ‘fees do not reflect the prevailing congestion levels at airports and
the congestion caused by different types of aircraft at different times of
the day.’ Poole and Butler (1988) proposed that ‘airports should be free to
increase their capacity and to charge congestion pricing’. They also pro-
posed the privatization of air traffic control and the use of new technology
to increase capacity. The success of airline liberalization in generating extra
revenues for airports and air traffic control would fund infrastructure, in
contrast to the congestion that is currently being experienced.
The allocation of slots at hub airports based on airline seniority is an
obvious barrier to new entrant airlines. While, in economic terms, airlines
are customers of airports, airlines have acquired property rights to slots
over time and a market in slots has developed. An example at Heathrow
is the sale of a slot in February 2015 by SAS to Turkish Airlines for $22
million. The allocation of slots at hub airports by airlines rather than by
airport management is a barrier to new entrant airlines. The allocation of
slots at hub airports by grandfather rights rather than by price or admin-
istrative intervention has remained the practice. Competition authorities
have not intervened and airport management have abdicated the role. In
the takeover of Aer Lingus by IAG in 2015, no provision was made for slot
reassignment on the Heathrow routes from Dublin (1.7 million passengers
per year), Cork (400 000 passengers) and Shannon (200 000 passengers).
IAG formerly faced competition from Aer Lingus on the three routes and
from British Midland on the Dublin route but, since the takeovers, has had
a monopoly on Ireland–Heathrow routes.
Therefore, secondary airports in Europe were crucial to successful
airline liberalization. Europe’s large number of empty airports provided
the vital market access needed by new entrant airlines that were typically
excluded from hub airports. The new entrant airlines and the secondary
airport managements took the entrepreneurial risk in opening new air
routes from virtually unused airports. Passengers responded initially to
avail themselves of low airfares but they also grew to like low-cost airports
with less crowding, less complexity, less walking, cheaper car parking, and
nearness to local airports.
Charleroi airport, 29 miles from Brussels, is an example of the success
of a secondary airport in a liberalized market. Charleroi lost its service
to London by Sabena during the 1970s because of poor results. Ryanair
commenced a service to Dublin in 1997 and an early objector was the main
Brussels Zaventem airport. Charleroi increased its passenger numbers
from less than 20 000 in 1996 to 5.5 million in 2015. The European
CONCLUSIONS
Given the tradition among Europe’s legacy airlines of not competing with
one another, low-cost airlines and new entrants are key to retaining the
gains from liberalization. New entrant airlines remain largely excluded
from major hub airports. The success of Ryanair in finding substitutes for
hub airports has occurred despite the opposition of the European Union,
notably at Charleroi.
NOTES
1. In Ireland, the capture of the Department of Transport by Aer Lingus was so successful
that the Department was described as the ‘downtown office of Aer Lingus’.
2. This followed the pattern described by Tullock (1967) and Krueger (1974), who warned
of the costs to society from successful rent-seeking and lobbying as a substitute for
competitive enterprise.
3. Before liberalization, airlines overbooked flights in anticipation of ‘no-shows’ and
denied boarding to those checking in late. The low-cost model imposes the cost of no-
shows on the passenger who fails to check in.
4. More broadly, Fewings (1998) looked at 431 airports in 13 European countries; exclud-
ing the special case of Iceland, with a ratio of 49 airports per million people, the average
is 1.9 airports per million people. David Starkie (2008) listed 56 airports on the UK
mainland serving 58 million people.
5. For an assessment of the network effects of this deregulation, see Button (2009).
6. Irish Independent, 19 June 2016.
7. Cahill et al. (2017) provided more details on the relative efficiency of Ireland’s airports.
8. See Averch and Johnson (1962).
9. The devaluation of sterling following the Brexit referendum in mid-2016 has increased
the cost of visiting Ireland for UK residents. Ryanair has decided to reduce its capacity
at Shannon in 2017 by 100 000 seats.
10. The airport authority stated that ‘airline operators have a powerful incentive to oppose
large scale expansions of airport facilities, given that such expansion will frequently
allow the entry of further competitors’ and ‘so-called “low-cost” operators generally
attach little or no value to the provision of facilities at airports other than essential facili-
ties such as checking in, baggage handling and passenger embarkation’ (Barrett, 2009).
REFERENCES
INTRODUCTION
Civil aviation in India has come a long way since it took wings when the
Indian State Air Services, in collaboration with Imperial Airways, opened
the first domestic air route between Karachi and Delhi in December of
1912 (Association of Private Airport Operators, 2013). Currently, India
is the ninth largest and the fastest-growing aviation market in the world
and has the capacity to cater to more than 220 million passengers and 4.6
million tonnes of air cargo per annum. India is expected to become the
third-largest aviation market in the world, with 450 million passengers by
2020. Most of this growth has come about during the last 25 years, post-
liberalization of the Indian economy in the 1990s.
At the time of India’s independence in 1947, nine air transport compa-
nies carried air cargo and passengers within and beyond the boundaries of
the country, with Air India – owned by the Tata Group – being the largest
of these carriers. In 1948, the Government of India acquired 49 per cent
of Air India and permitted it to operate international services under the
name Air India International (ATIC, 1950). Soon after, with the enactment
of the Air Corporation Act 1953, the Government nationalized the airline
industry and the assets of nine existing companies were assigned to the two
new corporations: Air India International and Indian Airlines. These cor-
porations served on international and domestic routes, respectively, under
the control of the Director General of Civil Aviation India (DGCA), the
Indian governmental regulatory body.
However, after a period of national carrier monopoly, the economic
reforms in the 1990s saw a re-privatization of the airline sector with the
entry of prominent airlines such as Jet Airways and Air Sahara following a
relaxation of passenger, cargo and foreign direct investment (FDI) norms.
Significant growth in air traffic in India finally arrived with the advent of
low-cost carriers in the early to mid-2000s, as these airlines brought down
fares and expanded the airline network. These airlines have also given rise
to price wars, bringing profits under immense pressure. While private air-
92
lines were allowed to operate, there were several restrictions with respect to
route decisions, pricing, funding and MRO facilities. In the mid-2000s, the
airport infrastructure was given a boost through partnerships with private
players and FDI. The Greenfield Airport Policy is expected to boost
further the airport infrastructure in areas that currently suffer from poor
connectivity. In addition, the Airports Authority of India has endeavoured
to upgrade air navigation systems in line with global standards to ensure
smooth operations over the airspace.
This chapter examines the major policy changes over the last 30 years
and their impact on airlines, airports and passengers. It also characterizes
the market structure post-deregulation and the fate of various players in
the course of these changes. Regulatory changes in airline operations,
airport and air navigation services are discussed in separate sections, fol-
lowed by a discussion of the environmental impact of the aviation sector.
Finally, the section on future outlook describes the most recent policy
changes that aim to improve regional connectivity and the ease of doing
business in the airline industry.
operate in India since 1953 – took to the skies with a fleet of three Boeing
737–200s. Soon after its maiden flight from Bangalore to Mumbai, the
airline was asked by the Civil Aviation Minister to bring in more aircraft to
counter the impact of a disabling Indian Airlines pilot strike. This marked
a beginning of the re-privatization of the civil aviation industry in India.
Another significant move in the same direction was that Air Taxis were
permitted to obtain up to 40 per cent of equity from foreign sources, with
100 per cent equity participation for NRIs and Overseas Corporate Bodies
(Hooper, 1997). With this option and 20 per cent funding from each of
Kuwait Airways and Gulf Air, Jet Airways leased a fleet of four Boeing
737–300 aircraft and started operations in May 1993. The airline would
later go on to become a fully-fledged private airline, operating 300 daily
flights to 68 destinations worldwide. Damania Airways, ModiLuft and
Air Sahara were other prominent operators who took off in 1993 under
the Air Taxi system. As can be seen from Figure 6.1, the years following
the open skies policy marked a steady increase in the number of domestic
passengers handled by private carries.
The ‘open skies policy for air cargo’, adopted in 1990 for a period of
three years, was extended in 1992 on a permanent basis (Aeronautical
Information Circulars, 1992). This allowed both Indian and foreign carri-
ers to operate cargo services (scheduled and non-scheduled) to and from
60
50 56.92
40 48.93 49.08 49.14
41.35
30 36.1
20
20.92 Indian Airlines
10 Private Airlines
4.11
0
1992 1993 1994 1995 1996 1997 1998 1999
Source: Created using data from the website of the Directorate General of Civil Aviation.
The Indian economy witnessed a GDP growth rate of 6 per cent in the year
2000, which further propelled the demand for travel for business as well as
leisure. An important development in this context is the emergence of the
Indian middle class – a section of working professionals and businesspeo-
ple who fall between the upper and working classes. Sensing an opportu-
nity, Air Deccan entered the domestic market in 2003 in an attempt to take
air travel to the masses in India. This was based on the ‘low-cost’ model
of cutting out frills from airline operations, popularized by Southwest
Airlines in the US and Ryanair in Europe. The airline excluded services
such as free meals, limited the staff, offered point-to-point flights and sold
tickets only through the internet at throwaway fares. Competition increased
with the entry of low-cost carriers such as SpiceJet (2005), GoAir (2005)
and IndiGo (2006), with greater focus on cost and efficiency. Full-Service
Airlines (FSAs) like Kingfisher Airlines and Paramount started operations
in 2005. The advent of these private carriers brought down ticket fares
and helped not only to increase the frequency of travel, but also to convert
train and bus travellers to airline passengers (ASCI, 2009). The combined
fleet of scheduled carriers grew from 162 to 402 from 2004 to 2015, and
Others, 1.2
Spicejet, 15.2 Air India, 17.2
Go Air, 9.3
Jet Airways,
15.4
Indigo, 37.2
Source: Created using data from the website of the Directorate General of Civil Aviation.
tered outside India were exempted from levy of all taxes and duties on
fuel and lubricants (Aeronautical Information Circulars, 2003), domestic
fares were no longer subject to the Inland Air Travel Tax and international
passengers no longer had to pay a Foreign Travel Tax (The Hindu, 2004).
Along with an easing of international tourist charter norms, these poli-
cies gave much-needed impetus to a growing aviation sector. In 2007, the
government announced the regional airlines policy, which allowed airlines
to operate between airports of designated regions. This policy was aimed
at increasing connectivity to Tier II and III cities – these airlines had lower
fleet and paid-up requirement (Bangalore Aviation, 2008). Such regionally
focused airlines were allowed to operate between one metro and all other
non-metro airports within that region.
The growth of the airline industry in the last decade has brought far-flung
destinations closer and generated great value for consumers by provid-
ing speed, reliability and reach. Train journeys spanning two days were
brought down to a few hours at affordable costs, creating high-speed access
to business and holiday destinations. According to Oxford Economics
(2011), in 2009 the monetary value of consumer benefit to Indian citizens
was estimated at $17.2 billion from air travel and $0.64 billion from air
cargo. This figure could be much higher when the catalytic impact on
tourism, employment and manufacturing is taken into account. A by-
product of the increased consumer travel was the rise of e-commerce in air
ticket purchase, allowing customers to choose from all available ticketing
options and select the most convenient flight at the lowest cost – often
competitive relative to air-conditioned railway fares.
Passengers benefited from improved frequency of service between major
Indian cities and connectivity to economically important destinations
around the world. In 2010, Indian carriers operated more than 59 flights
per day between Delhi and Bombay, a major trunk route for business trav-
ellers, and eight flights per day between Delhi and Dubai, an important
Middle East destination. IndiGo airlines, a low-cost carrier that com-
menced operations in 2006, grew its network to 40 destinations in India
and abroad, with 679 daily flights by 2015. Other low-cost airlines such
as SpiceJet and GoAir have established a number of direct links between
metros and hitherto unconnected cities. Cheaper and faster movement of
people and goods has prompted Indian companies to expand their cus-
tomer base, exploit economies of scale and eventually earn higher revenues
7
6
5
4
3
2
1
0
19 –96
19 –97
19 –98
19 –99
20 –00
20 –01
20 –02
20 –03
20 –04
20 –05
20 –06
20 –07
20 –08
20 –09
20 –10
20 –11
20 –12
20 –13
20 –14
5
–1
09
95
96
97
98
99
00
01
02
03
04
05
06
07
08
10
11
12
13
14
19
Source: Created using data from the website of the Directorate General of Civil Aviation.
to $200. Air Deccan was the first airline to introduce dynamic pricing
(Business Standard, 2015), changing prices based on demand and duration
from the departure date. As the ticket prices decreased due to increasing
competition from 2005, and expenses increased due to high fuel and main-
tenance costs, the average operating revenue earned per revenue passenger
kilometre (RPK) became less than the average operating expense incurred
per RPK at an aggregate level for all airlines combined, after the entry of
Low Cost Carriers (LCCs) from 2005 to 2014 (Figure 6.3). This suggests
that airlines spent more than they earned to fly each passenger over a dis-
tance of one kilometre. Intensive price competition, expensive taxation,
price-sensitive customers and the global recession in 2008 were some of the
factors that contributed to continued losses for most airline operators. As a
result, many carriers struggled to remain financially viable and were either
shut down or merged with other airlines.
By the year 2013, three airlines – Kingfisher, Air Deccan (Kingfisher
Red) and Paramount – had shut shop, and no major airline had entered
the market since the launch of IndiGo in 2006. The government gave the
sector a ‘shot in the arm’ by permitting foreign airlines to hold up to 49 per
cent equity in Indian companies operating scheduled and non-scheduled
air transport services (Aeronautical Information Circulars, 2013). The
policy also allowed Indian carriers to access pilot training, maintenance,
repair and overhaul facilities operated by foreign airlines. Abu Dhabi-
based Etihad Airways picked up a 24 per cent stake in the financially
strained Jet Airways, and paid $70 million to buy Jet’s airport slots at
London Heathrow Airport (The Hindu, 2013). Air Costa, a regional airline
that focused on connectivity between Tier II and Tier III cities, com-
menced operations in October 2013. The airline obtained a regional opera-
tor’s permit according to the policy announced in 2007. The Tata Group,
an industrial conglomerate, launched two start-up airlines, Tata AirAsia
and Tata SIA, and marked a return to the aviation business. AirAsia India
is a low-cost carrier focused on providing connectivity to Tier II and Tier
III cities, and Vistara (Tata–Singapore Airlines joint venture) is a premium
full-service carrier to cater to the demands of high-end business travellers.
The performance of these carriers in a competitive market against the
backdrop of changing regulation is yet to be seen. However, the success
of low-cost airline IndiGo, which managed to post consistent profits for a
continuous stretch of eight years, demonstrates the potential for exploit-
ing operational efficiencies despite the changing regulatory environment
(Saranga and Nagpal, 2016).
The rapid expansion of the air transport network in India since 1991 has
been heavily dependent on airport infrastructure. The capacity of airports
can be a bottleneck for steady growth of passenger traffic. Although the
construction of civil airports began, way back in 1924, with airports in
Calcutta and Allahabad, most airstrips in India were used for military
aviation. After Indian independence, many military airfields known as
‘Aerodromes’ came to be used for civil aviation under the control of the
Civil Aviation Department (Mehta, 2011). However, it was only in 1995,
with the formation of the Airports Authority of India (AAI) through an
Act of Parliament, that the management, upgrading and maintenance of
the civil aviation infrastructure in India were given a major thrust. This Act
also enabled the AAI to allow private players to operate a subsidiary busi-
ness for developing, financing and operating airports managed by the AAI
(Ministry of Civil Aviation, 2003a). Today, the AAI manages 125 airports,
provides ground handling and air traffic management services and ensures
safety of aircraft operations.
The Airport Infrastructure Policy (Ministry of Civil Aviation, 1997)
announced in November 1997 envisaged the development of international
and regional hubs with world-class facilities to provide capacity ahead
of the burgeoning volume of air travel. The policy also conceived of the
infusion of private capital for building new airports or for the moderniza-
tion of existing ones. This was in line with the global transition towards
2008). However, as per the new policy, the DGCA was given authority
to grant licences for the operation of Greenfield airports beyond 150 km
from an existing civil aviation airport, ‘no prior central government per-
mission is required’ (iGovernment, 2013), provided the airport is compliant
with the Aircraft Rules, 1937. Greenfield airports set up by the AAI were
to be structured and financed with private participation via open com-
petitive bidding. The policy outlined incentives such as concessional land,
airport connectivity, development rights and fiscal exemptions provided by
respective state governments. For Greenfield airport projects, 100 per cent
FDI was permitted under automatic approval. However, functions such
as traffic control, security, customs and immigration would be managed
by the central agencies. The government has since then given ‘in-principle’
approval to 14 Greenfield airports in the country under public–private
partnership, made available the necessary land and in most cases the forest
and environmental clearances required for construction. The Airport
Economic Regulatory Authority was established in May 2009 by an Act
of Parliament and was responsible for fixing and regulating tariffs for
aeronautical services and user fees levied by the airport operator. The
Authority also keeps a check on the quality standards of amenities pro-
vided by airport developers.
According to the Centre for Asia Pacific Aviation (2014), as of 2014 the
AAI earned $1.7 billion in revenue share from public–private partnership
(PPP) airports. The major PPP airports at Delhi, Hyderabad, Mumbai and
Bangalore accounted for 53 per cent of passenger traffic during the finan-
cial year 2013–14. As shown in Figure 6.4, the passenger traffic since 2004
has seen a steady growth, and even faster growth rates after 2013, with the
introduction of Greenfield airports. Moreover, this has led to a dramatic
improvement in the flyer experience through modern amenities, state-of-
the-art infrastructure and efficient handling of passenger and cargo. As
per a report by the Working Group on Civil Aviation Sector (National
Transport Development Policy Committee, 2012), in the financial year
2011, 75 per cent of domestic and 72 per cent of international cargo was
handled by the joint-venture international airports with fewer infrastruc-
ture woes. The share of private sector investment in airport infrastructure
increased to 69.85 per cent in the 11th five-year plan (2007 to 2012) from
43.36 per cent in the 10th plan. Revenues generated at airports in Delhi and
Mumbai increased dramatically after their privatization. In the financial
180 168.9
160
139.3
140 121.5 116.4 122.3
120 105.5
100 87.1 89.4
77.3
80 70.65
54.7
60 51.04
43 46.6 50.8
39.9 37.9 40.8
40 23.13 25.61 29.8 31.6 34.4
19.45 22.25
20
0 16.6
4
6
–0
–0
–0
–0
–0
–0
–1
–1
–1
–1
–1
–1
–1
03
04
05
06
07
08
09
10
11
12
13
14
15
20
20
20
20
20
20
20
20
20
20
20
20
20
International passenger traffic (million)
Domestic passenger traffic (million)
Source: Created using data from the website of the Directorate General of Civil Aviation.
year 2014–15, close to half of the $321.14 million bottom-line came from
these airports (Swarajya, 2016).
Greenfield airports constructed through the PPP model will prove
essential in building connectivity in underserved Tier II and Tier III
cities. Expert private players can help to build new capacity in cities where
current airports are over-utilized. The government has decided to increase
the number of airports to 500 in the next 20 years, with 367 of these being
Greenfield airports. As part of the latest National Civil Aviation policy, the
government has proposed the development of ‘no-frills’ airports at desti-
nations with low air traffic in the central and north-eastern parts of the
country. Airlines operating from these airports will be given concessions
on landing and parking charges. To boost traffic at these airports the ticket
prices have been capped at about $50 for a one-hour flight.
ENVIRONMENTAL IMPACT
FUTURE OUTLOOK
It is beyond doubt that the Indian airline industry has seen tremendous
growth in passenger traffic in the last 30 years. The first wave of liber-
alization in the 1990s brought back private players to the Indian skies,
and the twenty-first century saw the entry of low-cost carriers. Today,
passengers are beneficiaries of the price wars between various public and
private sector airlines and a wider airline network. This is largely due to
privatization and globalization of the airline industry. However, to unleash
the growth potential of the market, enhance the ease of doing profitable
business, build a global aviation hub and make air travel a reality for the
common man in India, positive regulatory measures will have to be revised
and made relevant to the current scenario.
The National Civil Aviation Policy, 2016 (PIB, 2016) outlines steps in
this regard. The policy scrapped the 5/20 rule, allowing start-up airlines
to fly to international destinations without requiring them to be operating
in the domestic circuit for a minimum of five years, provided 20 aircraft
or 20 per cent of capacity (whichever is higher) is utilized for domestic
operations. This allows new airlines to provide services on profitable inter-
national routes and gives Indian travellers more options to fly abroad.
The government will enter into reciprocal open sky agreements with coun-
tries located beyond 5000 km from Delhi and remove restrictions on the
number of seats, flights and airports.
Additionally, regional connectivity has been given a boost by capping
the charge for a one-hour flight at Rs. 2500 (approx. 50 USD) on regional
routes (North East, Jammu and Kashmir, etc.). Liberal measures include
reduced airport charges, service tax, value added tax and excise duty on
fuel used on these routes. The government will also provide financial assis-
tance to carriers flying on unprofitable regional routes.
While the Indian government has enabled private participation in
airport infrastructure, the growth in capacity seems to be lagging behind
the demand. Major airports at Mumbai, Delhi and Bangalore continue
to struggle with passenger traffic and over-utilized runways. The Policy
envisages development of additional airports through the PPP route
REFERENCES
Administrative Staff College of India (ASCI) (2009), Competition Issues in the Air
Transport Sector in India, Khairatabad.
Aeronautical Information Circulars (1992), AIC No. 18/1992, Director General of
Civil Aviation, Government of India, New Delhi.
Aeronautical Information Circulars (2003), AIC No. 03/2003, Director General
of Civil Aviation, Government of India, New Delhi.
Aeronautical Information Circulars (2013), AIC No. 12/2013, Director General of
Civil Aviation, Government of India, New Delhi.
Air Transport Action Group (2014), Aviation Benefits beyond Borders, Geneva.
Air Transport Inquiry Committee (ATIC) (1950), Report of the Air Transport
Inquiry Committee, University of California.
Ministry of Civil Aviation (MoCA) (2003a), The Airports Authority of India Act,
1994, New Delhi.
Ministry of Civil Aviation (MoCA) (2003b), Report of the Committee on a Road
Map for the Civil Aviation Sector, New Delhi.
Ministry of Tourism, Government of India (2014), India Tourism Statistics, New
Delhi.
Nathan Economic Consulting India (2012), Research Study of the Civil Aviation
Sector in India, Chennai.
National Transport Development Policy Committee (2012), Report of Working
Group on Civil Aviation Sector, New Delhi: Ministry of Civil Aviation.
National Transport Development Policy Committee (2013), Civil Aviation, New
Delhi.
Oxford Economics (2011), Economic Benefits from Air Transport in India, Oxford.
Parliament of India (1994), The Air Corporations (Transfer of Undertakings and
Repeal) Bill, 1994, New Delhi.
Press Information Bureau (PIB) (2016), National Civil Aviation Policy, 2016:
Salient Features, 15 June.
Saranga, H. and R. Nagpal (2016), Drivers of operational efficiency and their
impact on market performance in the Indian airline industry, Journal of Air
Transport Management, 53, 165–76.
Swarajya (2016), FDI alone won’t help – major restructuring needed for AAI take-
off, 22 June.
The Hindu (2004), Air traffic boom expected from Kochi, 11 January.
The Hindu (2006), Flying high, 5 November.
The Hindu (2013), A year of surprises for aviation industry, 29 December.
The Hindu (2015), Satellite-based navigation system launched, 14 July.
Varkkey, B. and G. Raghuram (2001), Public private partnership in airport develop-
ment: governance and risk management implications from Cochin International
Airport Ltd, India Infrastructure Report 2002: Towards Better Governance for
Commercialization, Oxford: Oxford University Press.
INTRODUCTION
Since 2007 China has been the second largest aviation market in the world
(after the United States) in terms of volume of passengers and air cargo
moved in its domestic market. In 2015 the whole industry handled 435.6
million passengers and 6.3 million tons of air cargo, an increase of 11.1
per cent and 5.2 per cent from the previous year. China’s airline market is
a growing market underpinned by a huge population and rapid economic
growth. The International Air Transport Association (IATA) forecasts
that China will overtake the US as the largest air passenger market by
around 2030 as measured by traffic to, from and within a country.
India’s civil aviation industry is also on a high-growth trajectory.
Passenger traffic in 2015 increased from 67.4 million in 2014 to 81.1
million the next year, at a rate of 20.3 per cent. It is expected that India
will become the third largest aviation market by 2020. Both India and
China started the deregulation process in the aviation sector in the 1980s.
However, 30 years on, the outcomes of the two aviation markets are stun-
ningly different. In particular, in China, most of the aviation reforms have
put the interests of the state-owned airlines first, which results in the four
state-owned airline groups dominating the domestic market with an aggre-
gate market share of 90 per cent in 2015. In contrast, in India the private
airlines commanded a market share of more than 80 per cent in recent
years, with low-cost carriers (LCCs) commanding a market share of about
60 per cent.
This chapter attempts to outline the differences between the two markets
and evaluate airline efficiencies in the two countries, which will help build
the momentum for further reforms in the aviation sector.
112
LITERATURE REVIEW
China and India are two of the fastest-growing economies in the world,
operating under very different political systems. India is the world’s largest
democracy, while China fails to develop Western-style democratic institu-
tions and practices, but has been pursuing the establishment of a market
economy. Comparative studies between China and India are abundant in
academic research and even more popular in public media, given that they
are the two most populated countries in the world (Liu and Jayakar, 2012).
There has been much research examining the air transport develop-
ment in the two countries, including the topics of aviation reforms and
their impacts. In the case of China, a pioneer work by Zhang (1998) first
described China’s airline deregulation process and documented its impact
on traffic growth, market structure, and airline operation and competi-
tion. This study was extended by Zhang and Chen (2003), who argued
that any reforms targeting the following four constraints would acceler-
ate the air traffic growth in China: (1) the protection of non-competitive
state-owned airlines; (2) low-income and restrictive policy on its citizens
travelling abroad; (3) underdeveloped infrastructure including airport
capacity constraints and air traffic control inefficiencies; and (4) the
lack of human capital and management skills. Zhang and Round (2008)
provided a detailed analysis on the driving forces that led to the airline
consolidations in China’s aviation markets in 2002, and argued that the
aviation market concentration is a natural response to the uncertainties
and changes that accompanied airline deregulation in China. Wang et al.
(2016) also agreed that the aviation market has become more concentrated
as a result of airline deregulation. Airline concentration and market power
issues in China’s airline market were further studied by Zhang and Round
(2009, 2011) and Zhang (2012). The general conclusion is that airline con-
solidations in the early 2000s did not grant China’s three airline groups
(China Eastern, China Southern and Air China) any significant market
power in both the short and longer term, largely owing to the implementa-
tion of other forms of deregulation in the last two decades – including the
relaxation of entry to and exit from markets, inviting private capital into
the aviation sector and the emergence of LCCs. However, after examining
the merger case between China Eastern and Shanghai Airlines in 2009,
which resulted in a rise in departure-day prices by 22 per cent, Zhang
(2015) warned that China’s anti-trust authorities should remain vigilant in
handling airline mergers when numerous parallel routes are involved. The
market power issue was also noted in Zhang et al. (2014), whose Lerner
indices for the period 2010–11 confirmed the existence of a certain degree
of market power in China’s airline markets.
Before 1980, China was politically and economically isolated from the
outside world. The airline industry was a semi-military organization under
the dual leadership of the Air Force and the State Council, engaging in
1990s and early 2000s (Zhang and Round, 2008). This has led to the
airline consolidations that occurred in October 2002. China’s nine state-
owned airlines merged into three airline groups: the Air China Group,
the China Eastern Group and the China Southern Group. These mergers
faced no anti-trust challenge at the time due to the absence of any effective
anti-trust laws. Under a constant expansion policy, Hainan Airlines had
become the fourth largest group in China in the early 2000s.
In the years following the 2002 consolidations, competition in the
markets associated with Beijing, Guangzhou and Shanghai remained
strong; these markets are of critical importance to Chinese airlines due
to their large market potential (Fu et al., 2015b; Zhang et al., 2014). The
routes in and out of these three cities are a significant source of revenue
for almost all domestic airlines. Airlines never stopped increasing frequen-
cies and launching new routes to these cities. As a result, price wars broke
out regularly, which was one of the reasons for the merger between China
Eastern and Shanghai Airlines in 2009: both airlines were in deep financial
distress prior to the merger (Zhang, 2015).
Entry to and exit from a route for existing airlines has been easier after
2006. Apart from the routes associated with the eight busiest airports,
prior approval for entry and exit is no longer required: a simple notifica-
tion by an airline to the regional civil aviation bureau prior to an action
is sufficient (Zhang and Round, 2008). Further deregulation of the route
control occurred in 2010, with prior approval only being needed when
airlines operate a flight in and out of Beijing, Shanghai and Guangzhou.
Although airfares had been largely deregulated since the late 1990s, to
prevent ‘irrational’ pricing behaviour the government promulgated ‘The
Scheme of Domestic Airfare Reform’ in 2004. This reform scheme set a
benchmark price (0.75 yuan per kilometre)2 and imposed a price ceiling
and a price floor for domestic airfares. However, from the day that this
scheme took effect, the price floor limit was largely ignored and competi-
tion in prices remained strong as the state-owned airlines had gained more
autonomy and power in doing business. There were no effective measures
to enforce the pricing rules. The government realized this and publicly
noted in 2006 that the price control would be abandoned in the next few
years. The 2004 scheme and all restrictions on pricing were formally abol-
ished in 2013.
China’s private airlines emerged in 2005, with more than ten being
licensed in that year following the release of the ‘Regulation on Domestic
Investment on Civil Aviation’ in 2004, which allowed private sector
participation in the civil aviation industry – including setting up new
airlines. Shortly prior to this, the limit on foreign ownership participa-
tion in Chinese airlines was raised from 35 per cent to 49 per cent. As a
Tata, renamed ‘Air India’ in 1946, was the most successful private airline
in India before 1953. Since the independence of India, the government
has maintained a heavy presence in the aviation activities, with many
cumbersome rules and regulations imposed on almost every sector of
the industry. In 1947 the Indian Government acquired 49 per cent of Air
India’s shares. In 1953, under the Air Corporation Act (1953), the whole
industry was nationalized. At the same time, the government established
Indian Airlines by merging 11 domestic carriers on the grounds that there
were too many airlines in the domestic market. After that and until 1994
the operation of scheduled air transport services was largely dominated by
the two government-owned corporations. In the intervening years between
1953 and 1994, Indian Airlines mainly served the domestic routes, while
Air India primarily served the international market. The government con-
trolled airfares and regulated the carriers’ entry to and exit from the routes.
India initiated a move towards liberalization in the 1980s. The Minister
of Tourism and Civil Aviation announced that private airlines were
allowed to operate charter and non-scheduled services under the Air Taxi
scheme in 1986. This move came at a time when the poor performances
of Air India and Indian Airlines were blamed for stagnant international
tourism (Hooper, 1998). However, this initial liberalization was inade-
quate, as many conditions were imposed on the private airlines: they could
not publish time schedules or issue tickets to passengers, and they could
only use 15- to 50-seat aircraft. The government then declared in 1989
that the air taxi operators would eventually be given the right to provide
scheduled services. East West Airlines, established in 1992, was the first
private airline to become a substantial competitor against Indian Airlines.
Together with Jet Airways, Damania Airways and ModiLuft, this airline
added capacity and ‘stole’ the national airline’s market share (Hooper,
1998).
Deregulation in air transport was part of a broader economic liberali-
zation agenda in India in the early 1990s that opened many industries to
private domestic and foreign investors. The Air Corporations Act 1953
was repealed in 1994 and replaced with the Air Corporations Act of the
same year. The 1994 reform corporatized Air India and Indian Airlines,
and allowed the new private companies to provide services as fully-fledged
airlines (Findlay and Goldstein, 2004; Hooper, 1998). Six air taxi opera-
tors were given permits to operate scheduled services in October 1994.
However, the government’s policy at the time did not intend to remove
restrictions on private investment in airlines. For example, Tata’s proposal
for launching a domestic carrier was rejected in 1996–98. In 1997, Kuwait
Airways and Gulf Air had to sell out their shares in Jet Airways to a non-
resident Indian as required by the domestic service ownership rules.
In 2003, the Civil Aviation Ministry set up a high-level committee
headed by the former Cabinet Secretary Shri Naresh Chandra to prepare
a roadmap for the civil aviation sector; this roadmap was to provide the
basis of a new National Civil Aviation Policy. This milestone report,
released by the Naresh Chandra Committee, recommended that foreign
entities be allowed to pick up to a 49 per cent equity stake in the domestic
and international scheduled air transport services with the approval of the
Foreign Investment Promotion Board (FIPB). This recommendation was
immediately adopted by the government in 2004, with the limit on foreign
equity in airline companies being raised from 40 per cent to 49 per cent.
However, equity from foreign airlines was not allowed – either directly or
indirectly – in domestic transport services.
In 2004 the Indian government introduced the 5/20 rule, which stated
that new airlines (mainly private airlines) are permitted to fly international
routes only if the airline has five years of domestic flying experience and at
least 20 aircraft in its fleet. Despite the restrictions of serving international
markets, the private carriers have flourished in the domestic markets. In
1992 Indian Airlines commanded a market share (domestic) of 95 per cent
(Jain, 2006); only one year later, its market share dropped to 76 per cent. In
2003 there were only three private carriers in India: Jet Airways, Air Sahara
and Air Deccan, but the private airlines had already carried more than
half of the passenger traffic in the domestic market. The private airlines
have not only taken market share, but their presence is also associated with
lower prices. They have led the significant change in competition in India’s
domestic market. In 2015, major private carriers present in India’s domes-
tic market included Jet Airways, SpiceJet, IndiGo, Vistara, AirAsia India
and Go Air. By the end of 2015 their market share had increased to 84 per
cent, according to the Directorate General of Civil Aviation.3
Low-cost airlines have also made a significant contribution to lower-
ing airfares. Established in 2003, Air Deccan was India’s first low-cost
airline based at Bangalore, famous for its software exports. O’Connell and
Williams (2006) noted that Air Deccan’s strategy for low fares provided
many low-income Indians with their first opportunity to fly. Deccan also
established an effective distribution system by taking advantage of India’s
software technology, thereby substantially lowering its distribution cost
(O’Connell and Williams, 2006). The airlines’ rapid growth sparked the
entry of many other LCCs in India between 2005 and 2007, including
SpiceJet, IndiGo, GoAir and JetLite. However, the quick expansion was
not accompanied by healthy financial positions. Deccan was acquired
by Kingfisher Airlines in 2008; Kingfisher failed in 2012 due to financial
An Overview
Over the past decade the growth of revenue passenger kilometres (RPK)
in the airline industry has remained at an annual rate of about 5 per cent
among the advanced economies such as the US, Australia, Canada and the
European Union. On the other hand, the emerging and developing econo-
mies (EDEs) such as Russia, Brazil, China and India have enjoyed a higher
RPK growth rate of about 10 per cent. Overall, these eight economies
accounted for 71.6 per cent of the world’s RPK in 2014, and of this, the
US, Australia, Canada and the EU accounted for 54.5 per cent, whereas
25%
20%
15%
10%
5%
0%
–5%
–10%
–15%
–20%
1990 1995 2000 2005 2010 2015
US AU CA E6 RU BR CN IN
Notes:
1. E6 5 the UK, France, Germany, Netherlands, Italy and Spain; E6 account for 55 per cent
of the EU’s RPK.
2. Because of the disintegration of the USSR, USSR’s RPK data is used for Russia before
1992.
Source: ICAO.
Russia, Brazil, China and India accounted for only 17.1 per cent. China’s
RPK share was far greater than that of India.
Figure 7.1 depicts the ten-year geometric average growth rates of RPK
for the eight economies. The aviation industries of the EDEs have experi-
enced much faster growth than that of the advanced countries in recent
years, and growth has been particularly impressive in China and India. As
shown in Figure 7.1, the growth of China’s RPK has been maintaining a
rate of over 10 per cent. India has had a similarly fast growth pattern since
2002.
Tables 7.3 and 7.4 provide data on air passenger traffic carried by the
major players in the two countries’ domestic aviation markets from 2001.
It can be seen that LCCs, India’s IndiGo and China’s Spring, enjoyed the
fastest growth from 2007, although SpiceJet and GoAir also grew fast.
However, it should be noted that even the largest Indian carrier, IndiGo, is
smaller than China’s fourth largest airline group, Hainan Airlines, in terms
of the volume of passengers carried. In fact, compared with their Chinese
counterparts, most Indian carriers operate at a much smaller scale. China’s
Air India Alliance Air Jet Airways Air Sahara/JetLite SpiceJet IndiGo GoAir Kingfisher
M4371-FINGER_9781786431851_t.indd 126
2000–01 3 288 000 1 670 000 5 925 000 688 000
2001–02 3 126 000 1 490 000 5 820 000 613 000
2002–03 3 394 000 1 420 000 6 402 000 1 302 000
2003–04 3 772 000 1 550 000 6 898 000 1 924 000
2004–05 4 391 000 1 450 000 9 015 000 2 606 000
2005–06 4 362 000 1 010 000 9 104 000 2 767 000 1 108 000 166 700
2006–07 4 342 000 820 000 10 739 373 3 180 694 2 598 479 912 962 1 263 048 3 260 714
2007–08 3 839 000 540 000 11 411 494 3 296 680 4 094 632 3 890 112 1 806 121 5 785 627
2008–09 2 834 000 320 000 11 030 474 3 369 012 4 099 458 4 917 360 1 128 475 8 651 773
126
2009–10 3 424 000 455 000 12 184 743 3 596 664 5 692 049 6 554 908 2 438 167 11 063 400
2010–11 3 843 000 503 000 14 643 462 4 328 524 7 384 010 9 466 897 3 393 984 12 011 807
2011–12 13 614 000 467 000 17 282 287 4 791 324 9 572 842 13 094 766 3 785 881 10 649 941
2012–13 14 183 200 393 000 16 818 737 3 867 556 11 681 066 16 887 969 4 386 914 1 211 925
2013–14 15 405 800 363 000 17 175 949 3 509 646 13 537 000 19 568 603 5 252 416
Annual growth 6.4% since −11.0% 6.9% 1.4% 26.6% 54.9% 22.6%
since 2007 2011–12
Notes:
1. For the 2011–12 period, the figures are from the merged data of Air India and Indian Airlines.
2. Air Sahara was taken over by Jet Airlines in 2007 and renamed Jetlite, which was rebranded as JetKonnect in 2012.
3. Kingfisher ceased operation in 2013 due to financial difficulties.
14/11/2017 11:14
Table 7.4 Number of passengers carried by Chinese airlines
M4371-FINGER_9781786431851_t.indd 127
2001 9 288 941 10 371 399 19 120 855 3 728 786 1 589 814 1 708 794 1 527 746
2002 10 586 674 10 385 521 14 074 768 3 706 595 2 560 566 2 037 485 4 737 062 2 172 512
2003 18 053 749 10 534 242 12 850 280 3 217 348 3 502 843 2 221 933 4 906 639 2 278 144
2004 24 499 955 15 273 214 17 675 604 4 357 095 4 818 808 3 655 026 6 232 405 3 009 114
2005 27 694 749 2 672 112 31 510 998 5 503 823 5 733 814 4 790 071 6 921 895 3 800 996 180 533
2006 31 504 292 29 707 707 34 641 367 7 298 060 7 178 267 5 881 210 7 784 997 5 040 415 1 137 062
2007 34 841 224 33 233 100 41 562 421 8 877 748 9 518 511 6 755 946 9 249 413 5 359 869 2 353 059
127
2008 34 256 741 31 686 627 43 967 758 8 976 447 11 951 918 6 701 470 9 618 860 5 420 085 2 943 775
2009 39 830 010 39 961 339 49 416 431 11 907 599 15 120 153 9 164 113 11 125 700 6 621 557 4 312 889
2010 46 245 017 43 063 011 56 636 882 11 508 391 16 487 503 10 667 547 13 561 359 8 071 913 5 859 697
2011 48 672 216 45 704 878 58 504 693 13 780 307 18 280 805 12 948 760 15 316 509 9 887 030 7 150 814
2012 49 334 466 48 068 308 61 809 891 15 238 122 19 805 010 14 445 233 16 843 870 11 246 342 9 110 993
2013 52 019 775 52 073 143 64 510 555 18 515 494 21 356 761 16 716 342 18 572 251 12 855 107 10 550 788
2014 54 637 234 54 275 647 70 611 294 20 937 194 23 214 385 19 191 395 20 380 159 14 125 476 11 446 984
Annual 6.6% 7.3% 7.9% 13.0% 13.6% 16.1% 11.9% 14.8% 25.4%
growth
since 2007
14/11/2017 11:14
128 Air transport liberalization
three major state airlines – Air China, China Eastern and China Southern
– recorded a steady growth rate in the last 15 years, while Air India, owned
by the Indian Government, was stagnant for many years and has only
managed to achieve a stable growth in recent years.
The two outputs used are passenger traffic and cargo, measured by
RPK and revenue ton-kilometres. The inputs included in this research
are the number of employees and the number of aircraft, reflecting the
fact that the airline industry is both labour-intensive and capital-intensive.
The choice was also determined by the availability of data. The input and
output data for Indian airlines were drawn from air transport statistics
provided by the Directorate General of the Ministry of Civil Aviation of
the Government of India, while the data for Chinese carriers come from
the yearbooks of Statistical Data on Civil Aviation of China.
Much of the work measuring airline efficiency uses the input-oriented
model, given that airlines have a greater influence on the inputs than on
the outputs, as macroeconomic conditions tend to have a strong impact on
consumer demand for air services (Merkert and Hensher, 2011). This seems
appropriate because a carrier’s management has control over how inputs
are used to generate outputs in most instances. The rule of thumb when
establishing a sample size in DEA studies is that it be at least equal to the
product of input and output or three times the sum of inputs and outputs
(Cooper et al., 2001). The two inputs and two outputs used here satisfy the
minimum requirement, as there are 14 airlines in the sample.
The technical efficiency scores as a percentage for 2007–14 under
various assumptions are seen in Tables 7.5 and 7.6. The main findings are:
M4371-FINGER_9781786431851_t.indd 130
2007 2009 2011 2013 2014
OTE PTE OTE PTE OTE PTE OTE PTE OTE PTE
(CRS) (VRS) (CRS) (VRS) (CRS) (VRS) (CRS) (VRS) (CRS) (VRS)
Air India 100 100 100 100 100 100 100 100 100 100
Jet Airways 83.26 83.3 90.98 100 71.3 100 100 100 93.49 100
SpiceJet 99.27 100 91.76 92.75 100 100 74 78.07 80.85 82.67
IndiGo 56.86 73.92 100 100 100 100 100 100 100 100
130
GoAir 84.95 100 42.01 64.45 75.26 95.61 95.78 100 91.87 100
Air China 69.25 100 64.37 100 65.74 100 79.48 100 80.38 100
China Eastern 49.91 69.71 47.95 73.9 51.88 78.24 55.89 82.51 55.26 70.19
China Southern 48.06 87.3 46.01 72.63 51.76 100 65.82 100 68.17 100
Hainan Airlines 76.82 93.32 71.76 94.26 65.01 76.88 87.94 100 100 100
Sichuan Airlines 100 100 65.2 100 98.68 100 90.64 100 90.14 100
Spring Airlines 100 100 100 100 100 100 100 100 100 100
Juneyao Airlines 63.74 88.84 67.35 75.73 75.06 81.87 79.79 84.12 84.93 90.93
Okay Airways 54.32 81.88 44.38 68.47 69.56 88.34 51.87 79.62 50.71 85.39
China Express 17.81 100 16.66 100 23.95 100 19.04 100 30.75 100
14/11/2017 11:14
Table 7.6 Scale efficiency and technical efficiency under the NIRS assumption
M4371-FINGER_9781786431851_t.indd 131
2007 2009 2011 2013 2014
CE TE CE TE CE TE CE TE CE TE
(NIRS) (NIRS) (NIRS) (NIRS) (NIRS)
Air India 100 100 100 100 100 100 100 100 100 100
Jet Airways 99.95 83.26 90.98 100 71.3 100 100 100 93.49 100
SpiceJet 99.27 99.27 98.93 91.76 100 100 94.8 74 97.8 80.85
IndiGo 76.92 56.86 100 100 100 100 100 100 100 100
131
GoAir 84.95 84.95 65.19 42.01 78.72 75.26 95.78 95.78 91.87 91.87
Air China 69.25 100 64.37 100 65.74 100 79.48 100 80.38 100
China Eastern 71.59 69.71 64.88 73.9 66.31 78.24 67.73 82.51 78.73 70.19
China Southern 55.05 87.3 63.35 72.63 51.76 100 65.82 100 68.17 100
Hainan Airlines 82.32 93.32 76.13 94.26 84.57 76.88 87.94 100 100 100
Sichuan Airlines 100 100 65.2 100 98.68 100 90.64 100 90.14 100
Spring Airlines 100 100 100 100 100 100 100 100 100 100
Juneyao Airlines 71.74 63.74 88.94 67.35 91.67 75.06 90.78 62.56 93.41 84.93
Okay Airways 66.34 54.32 64.82 44.38 78.74 69.56 76.8 60.95 59.39 50.71
China Express 17.81 17.81 16.66 16.66 23.95 23.95 24.31 24.31 30.75 30.75
14/11/2017 11:14
132 Air transport liberalization
●● IndiGo, the largest Indian carrier, and Jet Airways have been operat-
ing at a very efficient level in recent years in using their resources. In
general, these private airlines outperformed their Chinese counter-
parts such as Okay and Juneyao in terms of OTE and PTE.
●● In 2014 Air China, China Southern, Sichuan Airlines and Jet
Airways exhibited diminishing returns to scale, implying that they
would be more technically efficient if they could shrink their size.
Apart from Air India and Spring Airlines, the other carriers were at
the point where increasing returns to scale applied in 2014.
The analysis has shown that private airlines are dominant players in the
Indian airline market, and that they were generally more efficient than
their Chinese counterparts. State-owned Air India, despite being more
efficient than Chinese state-owned carriers, constantly lost market shares
to private airlines. This is not surprising, as India opened the aviation
sector to private investors in the mid-1990s, whereas China only did this
halfway through 2005. In India, airfares have been largely determined by
market forces since the 1990s. In contrast, China’s first private airline did
not appear until 2005, and the government had attempted to influence
the price-setting in various ways until 2013, when the 2004 Scheme of
Domestic Airfare Reform was formally repealed. Although price discounts
were allowed as early as 1997, the CAAC constantly sought to re-regulate
airfares to avoid price wars and industry losses by putting a limit on the
maximum discounts. The Chinese government also introduced a ‘revenue
pooling’ programme and guided the airline consolidations to restrict com-
petition from the late 1990s to the early 2000s (Zhang and Round, 2008).
Additionally, before China introduced its anti-trust laws in 2008, price-
fixing was tacitly allowed and sometimes encouraged by the CAAC.
Zhang and Round (2011) noted that airfare collusion in China was not a
secret and had been widely reported by newspapers. Major Chinese airline
groups such as Air China, China Southern and China Eastern held talks
from time to time to prevent airfares sliding down to train-fare levels.
Zhang and Round (2011) reported that since the deregulation of airline
prices in 1997, Chinese passengers have constantly witnessed overnight
across-the-board increases in airfares. Although most of the collusive
agreements were short-lived, some markets had consistently higher prices
for a relatively long time, especially on some business routes associated
with Beijing, Guangzhou and Shanghai, which are vital to the profits of
the state-owned airlines. This was still the case even after the introduction
the guideline still explicitly considers the interests of the new carriers,
which is lacking in China’s system of slot allocation.
Private airlines in India were also subject to some constraints. For
example, the five-year operation requirement prior to being eligible for
operating international services protected the interests of the experienced
airlines such as Air India and Jet Airlines, but is unfair for the newly
established low-cost carriers which could not launch international services
until five years later. This policy closed the door for new airlines to grow
through operating international flights, even though they may have met
all the security and safety standards. There have been constant calls for an
equal treatment for private and state-owned airlines (e.g. Nathan, 2012),
while such voice is much weaker in China.
India’s state-owned carriers appeared to have served more social goals
in addition to commercial performance. For example, the government
required that all scheduled carriers flew at least 10 per cent of their capac-
ity on trunk routes in the north-eastern sector, Jammu and Kashmir,
Andaman and Nicobar Islands and Lakshadweep (Category II routes).
However, the state-owned national airline was forced to put 17 per cent
of its flights on these routes, causing huge losses every year. On Category
III routes (non-trunk and non-Category-II routes, mainly associated
with small cities), the airlines were mandated to deploy 50 per cent of
the number of aircraft deployed on Category I routes. Again the national
airline was forced to undertake a higher percentage of its flights in these
markets than the private airlines. It is inefficient for the public carrier to
operate with such constraints (Jain, 2006). This might be one of the factors
that led to Indian Airlines constantly losing its domestic market share to
private competitors.
CONCLUSION
China and India have attracted the attention of the world in the last two
decades as powerhouses of economic growth. There has been a plethora
of comparative studies on various aspects of the economies between China
and India; however, research into the development of air transport poli-
cies and market outcomes for the two countries is lacking. This study has
reviewed the evolution of the air transport industry in the two countries,
and then assessed and compared the efficiency performances of the state-
owned and privately owned airlines in the two markets. It has found that
Indian carriers tend to be more efficient than their Chinese counterparts.
Air India and Spring Airlines are the leaders in terms of pursuing techni-
cal efficiency, and the dominant status of the private airlines in the Indian
market does not come by chance. The analysis of this research shows
that China’s aviation policy has long been hostile to the private carriers
and overly protective of the state-owned airlines. Various obstacles have
constrained the expansion and growth of the private airlines. In contract,
Indian private airlines, especially the LCCs, operate in a relatively more
liberal environment. With the private carriers dominating the market, they
could exercise a stronger influence on government policy, which would in
turn promote and accelerate their growth.
NOTES
1. This intuition was formalized in a recent theoretical paper by Jiang and Zhang (2016).
2. One US$ is about 6.6 yuan (Chinese currency, also known as RMB) as of August 2016.
3. See http://dgca.nic.in/reports/Traffic-ind.htm.
REFERENCES
Jain, R.K. and R. Natarajan (2015), A DEA Study of Airlines in India, Asia Pacific
Management Review, 20, 285–92.
Jain, S. (2006), Competition Issues in Transportation Sectors, in P.S. Mehta
(ed.), Towards a Functional Competition Policy for India, Haipur: Academic
Foundation in Association with Consumer Unity and Trust Society (CUTS
International).
Jiang, C. and A. Zhang (2016), Airline Network Choice and Market Coverage
Under High-speed Rail Competition, Transportation Research, Part A: Policy
and Practice, 92, 248–60.
Le, T.T. (1997), Reforming China’s Airline Industry: From State-owned Monopoly
to Market Dynamism, Transportation Journal, 36, 45−62.
Li, M. (2014), Spring Airlines Approved for IPO After Ten Years of Strict Cost
Control. Jiemian, available at http://www.jiemian.com/article/202427.html.
Liu, C. and K. Jayakar (2012), The Evolution of Telecommunications Policy-
making: Comparative Analysis of China and India, Telecommunications Policy,
36, 13–28.
Mazumdar, A. (2008), Deregulation of the Airline Industry in India: An Analysis
of the Government’s Policy, Rationale and Strategy, PhD thesis, Department of
Political Science, Miami University.
Merkert, R. and D.A. Hensher (2011), The Impact of Strategic Management and
Fleet Planning on Airline Efficiency – A Random Effects Tobit Model Based on
DEA Efficiency Scores, Transportation Research Part A: Policy and Practice, 45,
686–95.
Nathan (2012), Research Study of the Civil Aviation Sector in India, Nathan
Economic Consulting India, submitted to the Ministry of the Corporate Affairs,
Government of India.
O’Connell, J.F. and G. Williams (2006), Transformation of India’s Domestic
Airlines: A Case Study of Indian Airlines, Jet Airways, Air Sahara and Air
Deccan, Journal of Air Transport Management, 12, 358–74.
Ray, S.C. (2004), Data Envelopment Analysis: Theory and Techniques for Economics
and Operations Research, Cambridge: Cambridge University Press.
Saranga, H. and R. Nagpal (2016), Drivers of Operational Efficiency and its
Impact on Market Performance in the Indian Airline Industry, Journal of Air
Transport Management, 53, 165–76.
Saraswati, S.K. (2001), Operating Environment for a Civil Aviation Industry in
India, Journal of Air Transport Management, 7, 127–35.
Scheraga, C.A. (2004), Operational Efficiency Versus Financial Mobility in the
Global Airline Industry: A Data Envelopment and Tobit Analysis, Transportation
Research Part A, 38, 383–404.
Taneja, N. (2004), Simpli-flying, Optimizing the Airline Business Model, Aldershot:
Ashgate Publishing.
Wan, Y., H-K. Ha, Y. Yoshida and A. Zhang (2016), Airlines’ Reaction to High-
speed Rail Entries: Empirical Study of the Northeast Asian Market Working
paper, Hong Kong Polytechnic University.
Wang, J., D. Bonilla and D. Banister (2016), Air Deregulation in China and its
Impact on Airline Competition 1994–2012, Journal of Transport Geography, 50,
12–23.
Wang, K., Q. Gong, X. Fu and X. Fan (2014), Frequency and Aircraft Size
Dynamics in a Concentrated Growth Market: The Case of the Chinese Domestic
Market, Journal of Air Transport Management, 36, 50–58.
Wang, W. (2015), Big Carriers Dominate First Airport Slot Auction, China
Daily USA, 31 December, available at http://usa.chinadaily.com.cn/epaper/201
5–12/31/content_22885816.htm.
Yuen, A. and A. Zhang (2009), Effects of Competition and Policy Changes
on Chinese Airport Productivity: An Empirical Investigation, Journal of Air
Transport Management, 15, 166–74.
Zhang, A. (1998), Industrial Reform and Air Transport Development in China,
Journal of Air Transport Management, 4, 155−64.
Zhang, A. and H. Chen (2003), Evolution of China’s Air Transport Development
and Policy towards International Liberalization, Transportation Journal, 42,
31−49.
Zhang, A., S. Hanaoka, H. Inamura and T. Ishikura (2008), Low Cost Carriers in
Asia: Deregulation, Regional Liberalization and Secondary Airports, Research
in Transportation Economics, 24, 36–50.
Zhang, Q., H. Yang, Q. Wang and A. Zhang (2014), Market Power and its
Determinants in the Chinese Airline Industry, Transportation Research Part A:
Policy and Practice, 64, 1−13.
Zhang, Y. (2012), A Retrospective Analysis of the Effects of Airline, Mergers on
China Eastern’s Lifeline Routes: A Difference-in-differences Approach, Journal
of the Asia Pacific Economy, 17, 464−75.
Zhang, Y. (2015), Merger Between Airlines in Financial Distress: Does the Merger
Save Them?, Competition and Regulation in Network Industries, 16, 66−81.
Zhang, Y. and C. Findlay (2010), Air Transport Liberalization, in E. Ghani (ed.),
The Service Revolution in South Asia, New York: Oxford University Press.
Zhang, Y. and Z. Lu (2013), Low Cost Carriers in China and its Contribution to
Passenger Traffic Flow, Journal of China Tourism Research, 9, 207−17.
Zhang, Y. and D.K. Round (2008), China’s Airline Deregulation Since 1997 and
the Driving Forces Behind the 2002 Airline Consolidations, Journal of Air
Transport Management, 14, 130−42.
Zhang, Y. and D.K. Round (2009), The Effects of China’s Airline Mergers on
Prices, Journal of Air Transport Management, 15, 315−23.
Zhang, Y. and D.K. Round (2011), Price Wars and Price Collusion in China’s
Airline Markets, International Journal of Industrial Organization, 29, 361–72.
Zhang, Y. and A. Zhang (2016), Determinants of Air Passenger Flows in China
and Gravity Model: Deregulation, LCCs, and High-speed Rail, Journal of
Transport Economics and Policy, 50, 287–303.
INTRODUCTION
The creation of the single airline market in the European Economic Area
(EEA) is a major achievement of the European Union. The establish-
ment of this market has created incentives for new airlines to appear and
grow, made air travel affordable to millions of people, and has (generally)
contributed to deeper integration of the European countries. This brief
chapter will review the main changes that have occurred in European com-
mercial passenger aviation since the 1990s (with a more detailed focus on
the last decade), compare outcomes of the European liberalization to what
happened in the United States, and outline the threats and opportuni-
ties for the future of European aviation, especially in light of the United
Kingdom’s forthcoming departure from the EU.
Overall, the main developments in the European passenger airline
market since deregulation have been the following. First, we have witnessed
an explosive growth of a new breed of airline – the so-called low-cost
carrier or LCC. While industry observers and experts would not, gener-
ally speaking, be able to point to the threshold in, say, cost per available
seat mile (CASM), below which an airline becomes an LCC, the business
models of these airlines do share several common traits. Interestingly, the
LCCs that emerged in Europe post-liberalization are different in a number
of ways from the airlines that are considered to be their counterparts in the
United States. The key hereto unexplored question related to these airlines
is whether the LCCs pose a genuine threat to the so-called legacy carri-
ers, or whether they simply expand the market to the segments the legacy
airlines did not cover. The second important development is consolidation
of the European legacy carriers. I believe that we are on the path towards
a set-up that includes three large EU network carriers. Interestingly, this
setting mirrors what is happening in the United States, transatlantic and
mainland China markets. Last but not least, while the market within the
EEA is truly liberalized and open; the same cannot be said about most of
138
Eastern European countries. Those airlines did not have significant (if
any) experience of competing in a liberalized market, and they suddenly
ended up playing against both the larger network airlines that have been
working in a deregulated environment, sometimes for over a decade, as
well as against the new breed of point-to-point low-cost carriers that have
only known a competitive environment and have never been under any
regulatory protection. While not all of the Eastern European flag carri-
ers have fared well in this new environment, it would be incorrect to call
European market liberalization a failure for the region. I will now examine
the outcome of the EU market deregulation process in more detail.
LIBERALIZATION OUTCOMES
costs, and also started following suit by reducing in-flight service quality
on short-haul intra-European flights (service adjustments on long-haul
flights have been less notable: if anything, premium class services on long-
haul flights have actually improved over the last two decades).
Market consolidation seems to be a natural outcome of airline market
liberalization. The United States has seen two merger waves since deregula-
tion (the first occurred in the 1980s and the second in the new millennium).
The Chinese domestic market has also seen some consolidation recently.
Mergers on the European market have generally been less conspicuous, as
they rarely result in disappearance of the airline brands. It would be more
correct to say that mergers between legacy carriers from different coun-
tries have not yet resulted in the airline brands disappearing. Air France
and KLM are now parts of one company; British Airways, Iberia and
Aer Lingus form the entity called the International Airline Group (IAG);
and Lufthansa owns Austrian Airlines, Swiss International Airlines and
Brussels Airlines. Consolidation events involving the new low-cost carri-
ers, however, have always resulted in one of the merger partners’ brands
disappearing. This somewhat unusual arrangement is partly driven by the
regulatory structure of most of the EU-outside-Europe markets. While
the European Union is striving to conclude multilateral air services agree-
ments with the outside countries, little progress has been made on this
front. With the exception of the United States and several North African
countries, air travel between the EU countries and outside states is still
driven by bilateral treaties, typically containing nationality clauses. This
necessitates retaining the airlines’ brands after the mergers, which is obvi-
ously not an issue for the United States airline industry.
Both markets are currently experiencing the emergence of a more or
less similar structure. Following the last US merger wave and consolida-
tion events in Europe, both sides of the Atlantic are left with three large
network carriers and a number of so-called low-cost airlines, which gener-
ally focus on point-to-point rather than connecting traffic. Interestingly,
there is a similar structure on the domestic market in Mainland China. Of
course, bankruptcies are as much a feature of liberalized markets as the
entry of new players and mergers between the incumbent firms. Notable
bankruptcies in the European airline industry have included the disappear-
ance of Swissair and Sabena (the former Belgian flag carrier) in 2001, as
well as Malev Hungarian Airlines and Spanair, both within one week in
February of 2012. If we trace the bankruptcy events on both sides of the
Atlantic since 2000, we will notice that bankruptcies are more common on
the US market. In fact, most of the large US network carriers have gone
bankrupt at least once since 2000 (with Southwest Airlines being a notable
exception). At the same time, larger airline brands on the US market
For instance, it was American Airlines, not Southwest or JetBlue, that first
introduced fees for a passenger’s first checked bag. In fact, Southwest cur-
rently has the most generous checked baggage policy on the US market:
the airline does not impose charges on the first two checked bags. Also, all
US low-cost carriers not only offer point-to-point, but also connecting ser-
vices, and even the self-proclaimed ultra-low-cost Spirit Airlines operates
a frequent flier programme (something that none of the top-five European
LCCs offer).
We will now examine the dynamics of the European airline industry over
recent years. Passenger volume in the European Union has recovered from
the slump caused by the 2008 financial crisis. Overall, about 918 million
passengers travelled on commercial passenger flights within the 28 EU
member states in 2015, according to Eurostat. This represents a 15 per cent
increase compared to 2008.
Traffic growth (both over the last dozen years and since the latest finan-
cial crisis) has been uneven across the European countries, as Figure 8.1
clearly demonstrates. The cumulative growth reported here is for passenger
traffic within the EU/EEA.
Looking at the time since the 2004 EU enlargement, four Eastern
European countries (Latvia, Poland, Lithuania and Romania) emerge
as clear leaders in air passenger traffic growth, with Iceland and Estonia
being the two other countries that have been able to more than double
their passenger numbers between 2004 and 2015. However, it would not
140
120
100
80
60
40
20
0
Iceland
Lithuania
Romania
Luxembourg
Poland
Malta
Croatia
Portugal
Belgium
Latvia
Norway
Switzerland
Netherlands
Greece
Sweden
Denmark
Hungary
Italy
Estonia
Bulgaria
Finland
Germany
France
Austria
United Kingdom
Spain
Cyprus
Ireland
Czech Republic
–20
Slovenia
Slovakia
–40
250
200
150
Millions
100
50
0
2002 2004 2006 2008 2010 2012 2014 2016
35
30
25
20
Millions
15
10
0
2002 2004 2006 2008 2010 2012 2014 2016
4
Millions
0
2002 2004 2006 2008 2010 2012 2014 2016
Estonia Latvia Lithuania
shows that in the three Baltic States (Latvia, Estonia and Lithuania),
the divergence across the countries has been even more dramatic. The
countries entered the EU in 2004 with nearly equal annual passenger
volumes. Eleven years later, Latvia’s traffic volume nearly quintupled and
Lithuania’s quadrupled, while Estonia’s ‘merely’ doubled.
What are the key players on the European airline market, and has any-
thing changed on this scene over the last decade? The short answer is that
while not much has changed, there have been some quite tectonic shifts.
Indeed, the top five European airlines in 2016 are the same as in 2006. At
140
120
100
80
60
40
20
0
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Note: Passenger volume is in millions per annum and only includes intra-European traffic.
the same time, only one carrier in the top five preserved the same rank in
2016 as it had in 2006. Air France-KLM dropped from the top spot in 2006
to fourth in 2016, while Ryanair rose from fourth to first. Lufthansa and
the IAG (with British Airways a key airline here) swapped second and third
positions, while easyJet remained Europe’s 5th largest carrier. Figure 8.5
illustrates this situation and also demonstrates the extent to which Europe’s
largest carriers have grown over the last decade. In 2006, none of the top
airlines carried over 80 million passengers. In 2016, however, all but one
exceeded 90 million.
While Ryanair and easyJet – Europe’s leading low-cost airlines – firmly
occupy spaces in the top five, with the former having become Europe’s
largest carrier in 2016, the big three network carriers have not lost their
place on the market, managing healthy growth rates. Among the next five
largest airlines (Figure 8.6), we see two legacy network carriers (SAS and
Alitalia) and three airlines that are normally associated with the low-cost
business model (Norwegian, Air Berlin and Wizz Air). At the same time,
the business models of Norwegian and Air Berlin tend to be somewhat
different from that of Wizz Air, as the former operates more of a hybrid
network (offering limited connecting services while still mostly relying
on point-to-point traffic). Note also that the airlines in this group carry
about one-third of the passenger traffic as compared to the top five carri-
ers. Overall, the picture is clear: the development of the European airline
market post-liberalization has seen explosive growth of LCCs; however,
it appears that the legacy carriers have been able to find ways to compete
35
30
25
20
15
10
0
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Note: Passenger volume is in millions per annum and only includes intra-European traffic.
with them. We will take a closer look at the LCCs’ business model in the
next section.
increase the quality of its product. Would easyJet be able to run a proper
network at a lower cost than Lufthansa? It probably would. However, the
difference in conventional cost measures between the two carriers would
definitely become less dramatic.
LOOKING AHEAD
At first, my intention for this part of this chapter was to present my
vision of the European airline industry 10–15 years from now. However,
the events of the second half of 2016 brought about some more urgent
concerns. Specifically, on 23 June 2016 voters in the United Kingdom sup-
ported the idea of their country leaving the European Union. While the
referendum was an advisory one, the UK Government looks determined
not only to act upon the recommendation, but also to take the country out
of the EU Single Market. As of the time of this writing, it appears that the
United Kingdom will leave the EU by the spring of 2019. This has created
significant challenges for the European market and beyond.
Essentially, if the UK decides not to leave the Single Market, nothing
will change for UK–Europe (and the transatlantic) traffic. Consequences
of the United Kingdom’s departure from not only the European Union
but also from Europe’s common aviation market will, of course, depend
on whether and what kind of agreement the UK will be able to reach
with the EU. In the worst-case scenario, we can return to the network of
bilateral air service agreements between the UK and the corresponding
European countries. These agreements will likely come with nationality
clauses, creating difficulties for the point-to-point LCCs and their custom-
ers. As I noted above, former flag carriers have not been able to establish
an appreciable presence outside of their home countries. Thus most, if not
all, of the flights by those carriers are operated from their home countries,
which means that satisfying the nationality clauses of any future bilateral
agreements should not be a problem for them. The situation will be even
easier if the UK and EU manage to agree an open-skies deal similar to the
one the EU currently has with some countries in North Africa (such as
Morocco).
However, the situation with the point-to-point LCCs can get rather
messy. For instance, Ryanair is an Irish-registered airline. So, without
an open-skies deal with the EU, the airline might be restricted to flying
between the UK and the Republic of Ireland. Ryanair has already
announced that it has no intention of registering a UK subsidiary, so
even if a UK–EU open skies deal happens, Ryanair will most likely dis-
continue serving the UK domestic market (unless, of course, the UK
general problem with EU bureaucracy. Last but not least, for a liberalized
market to thrive, everyone must play by the same rules. Therefore, the
European Union must continue working to ensure a level playing field for
all competitors. This includes absence of subsidies and adherence to fair
competition practices by all market players.
REFERENCES
Albalate, D., G. Bel and X. Fageda (2015), Competition and cooperation between
high-speed rail and air transportation services in Europe, Journal of Transport
Geography, 42, 166–74.
Bachis, E. and C. Piga (2011), Low-cost airlines and online price dispersion,
International Journal of Industrial Organization, 29, 655–67.
Behrens, C. and E. Pels (2012), Intermodal competition in the London–Paris pas-
senger market: high-speed rail and air transport, Journal of Urban Economics,
71, 278–88.
Bel, G. and X. Fageda (2010), Privatization, regulation and airport pricing: an
empirical analysis for Europe, Journal of Regulatory Economics, 37, 142–61.
Bergantino, A.S. and C. Capozza (2014), Airline pricing behaviour under limited
intermodal competition, Economic Inquiry, 53, 700–713.
Bilotkach, V., J.A. Clougherty, J. Mueller and A. Zhang (2012), Regulation, privati-
zation, and airport charges: panel data evidence from European airports, Journal
of Regulatory Economics, 42, 73–94.
Bilotkach, V., A. Gaggero and C. Piga (2015a), Airline pricing under differ-
ent market conditions: evidence from European Low-Cost Carriers, Tourism
Management, 47, 152–63.
Bilotkach, V., S. Gitto, R. Jovanovic, J. Mueller and E. Pels (2015b), Cost efficiency
benchmarking of air navigation service providers, Transportation Research Part
A, 77, 50–60.
Bilotkach, V., J. Mueller and A. Nemeth (2014), Estimating the consumer welfare
effects of de-hubbing: the case of Malev Hungarian Airlines bankruptcy,
Transportation Research Part E, 66, 51–65.
Burghouwt, G. and J. de Wit (2005), Temporal configurations of European airline
networks, Journal of Air Transport Management, 11, 185–98.
Button, K. and R. Neiva (2014), Economic efficiency of European air traffic
control systems, Journal of Transport Economics and Policy, 48, 65–80.
Calzada, J. and X. Fageda (2012), Discounts and public service obligations in
the airline market: lessons from Spain, Review of Industrial Organization, 40,
291–312.
Dobruzskes, F. (2006), An analysis of European low-cost airlines and their net-
works, Journal of Transport Geography, 14, 249–64.
Dobruzskes, F. (2011), High-speed rail and air transport competition in Western
Europe: a supply-oriented perspective, Transport Policy, 18, 870–79.
Dobruzskes, F. (2013), The geography of European low-cost airline networks: a
contemporary analysis, Journal of Transport Geography, 28, 75–88.
Fageda, X. (2006), Measuring conduct and cost parameters in the Spanish airline
market, Review of Industrial Organization, 28, 379–99.
INTRODUCTION
Demand for air transport in the Latin America and Caribbean (LAC)
region has increased significantly since the 1990s, and even more so in the
twenty-first century. The increase has had its basis in the internationaliza-
tion of production, economic growth and, possibly, efforts to deregulate
air transport markets throughout the region. This chapter is concerned
primarily with an assessment of evidence in existing academic studies and
industry reports on the effects of deregulation of air transport markets in
the region. To narrow the geographic scope, only developments in sover-
eign countries are highlighted. This list is included in Table 9.1.
With the exception of Chile, LAC countries have found air transport
liberalization challenging, even after having been long exposed to the US’s
trade policies in the region. Indeed, the LAC’s most important trading
156
partner, the US, has pushed air markets’ liberalization in the form of Open
Skies Agreements (OSA) since the early 1990s. To illustrate the slow pace
of liberalization in the region, only 12 OSAs were signed between the US
and the 33 LAC sovereign countries in the Americas by September 2009;
by 2016 this had only increased to 18 OSAs.
OSAs in the region and other efforts to liberalize the provision of air trans-
port services have helped remove economic regulations, but the details of this
are poorly documented. Using a case study approach, this chapter consoli-
dates the information available to expose the implications of liberalization of
the LACs’ transport markets. Descriptive statistics are used to assess the state
of these markets in the region. The focus is mostly on market and economic
regulations. Less emphasis is given to the liberalization of technical and
safety oversight because these functions have largely remained in the hands
of the governments. Mexico, Chile and Colombia – partly because of the
scale of their markets – are used as the case studies. Additionally, in selecting
these case studies, consideration was given to the traffic composition split
between domestic and international passengers, as was the fact that Chile
was a pioneer in liberalizing its air transport sector whereas Mexico and
Colombia have had a lukewarm relationship with liberalization.
M4371-FINGER_9781786431851_t.indd 158
Mexico 15 364 400 14 968 800 21 857 658 45 560 063 6.6
Colombia 5 737 300 7 863 300 9 984 424 30 742 928 14.5
Chile 825 200 3 197 200 5 939 020 15 006 762 57.3
Argentina 4 713 200 6 641 700 6 938 436 14 245 183 6.7
Peru 1 563 900 2 508 300 4 332 223 13 907 948 26.3
Panama 390 300 661 200 1 795 958 12 018 103 99.3
Venezuela 4 966 500 4 445 000 5 043 028 6 456 854 1.0
Ecuador 664 200 1 670 900 2 011 004 5 762 485 25.6
Trinidad and Tobago 1 299 500 1 727 000 1 055 106 2 617 843 3.4
158
El Salvador 395 000 1 698 000 2 540 564 2 597 649 18.6
Bolivia 1 343 300 1 223 700 1 892 343 2 578 960 3.1
Costa Rica 310 000 870 400 953 217 1 617 076 14.1
Cuba 894 400 823 700 812 781 1 294 458 1.5
Antigua and Barbuda 500 000 1 050 000 778 260 1 039 810 3.6
Belize 935 604 110.0
Bahamas, The 860 500 936 900 1 020 000 587 517 −1.1
Paraguay 212 000 105 200 445 854 452 004 3.8
Suriname 108 000 161 800 315 410 259 682 4.7
Honduras 451 000 251 150 −1.5
Guatemala 108 400 300 000 93 129 −0.5
Total 54 110 000 71 049 200 105 377 019 260 064 567
14/11/2017 11:14
Latin America and the Caribbean 159
was to link its sparsely inhabited areas with its major economic centres (US
Department of Commerce, 2006).
As in other regions of the world, the data presented in Table 9.2 exposes
the key role that air transport plays in facilitating population movements
and the integrating of people in spatially dynamic regional and global
labour markets. As pointed out by Button and Vega (2008), air transport
has not only made long-distance initial permanent migration less costly,
but has also contributed to limit social costs by allowing regular visits to
friends and family.
The markets for air transport services have developed rapidly since
the mid-1990s, to the point where the provision of air services is a major
enabler of global supply chains. Connectivity of employees operating
from different hubs around the world and the integration of the electron-
ics, high-tech goods, pharmaceuticals, fashion products, perishables and
exotics (P and E) are quite unimaginable without access to air transport.
Vega (2008 and 2010) investigated in detail trade flows and air transport
costs of high-tech goods and P and E and provided definitions and char-
acteristics of their commerce. The list of exotics is extensive and includes
the legal trade of items such as tropical fruits and vegetables, ornamentals,
fresh herbs and spices, wild mushrooms, fish and seafood.
In LAC countries, because of economies of scope, airfreight plays a role
in allowing trade among countries with different capacities for production
but with similar tastes. For instance, air transport has allowed countries
such as Colombia and Ecuador, Chile and Peru to engage in the trade
of exotics and other perishable products in which these countries have a
comparative advantage. Similarly, in MERCOSUR countries, Martínez-
Zarzoso and Nowak-Lehmann (2004) have found that air transport offsets
the effect of geographic distance having only a low or statistically non-
significant effect on these countries’ exports of fish, meat, vegetables and
fruits to the European Union (EU). Evidently, this is partly also the result
of more diverse populations, mainly in the US and the EU, demanding and
willing to pay for more expensive exotic groceries such as starfruit, jicama,
pitahaya, tropical flowers and so on.
Large islands and landlocked countries have benefited from air trans-
port, as they have seen their production feasibility frontiers expand with
access both to remote suppliers, domestic and international, and to cus-
tomers. Table 9.3 shows the magnitude of the flows of goods that are
moved by air from the LAC region to the EU.
As expected, Brazil and Mexico, the LAC region’s largest economies,
register the highest freight exchanges with the EU. Argentina, the region’s
third largest economy appears in fifth place. Colombia and Ecuador, due
to their exports of flowers and other P and E, appear in third and fourth
Table 9.3 LAC–EU freight and mail, loaded and unloaded, metric tons
Source: Eurostat.
place, respectively. From the data, it also seems that the annual growth rate
in freight levels was linked to the region’s economic expansion with the
commodity boom after the year 2005, followed by collapsing levels related
to the financial crisis of 2008.
High freight costs are often the result of imbalanced freight flows, since
a relevant factor determining airfreight rates is the peak load factor. While
passenger business is generally bidirectional, cargo is not. This implies
that when transporting goods from point A to point B, the freight rate
charged must also cover the return trip from B to A. When the demand for
transport services is unidirectional, freight rates are simply higher on one
direction, as the shipper pays for forgone capacity on either the inbound
or outbound flight. When the trade imbalance is high (import/export
ratio higher than one), transport costs for exports tend to be higher. LAC
countries have struggled to unmask the factors behind the region’s notori-
ously high airfreight rates. Vega (2010), for example, found that there is
not a clear relationship between distance and the cost of airfreight ser-
vices. Vega (2010) also found that liberalization of air transport markets,
through OSAs, reduced airfreight charges for countries exporting P and E.
Most interestingly, the costs of air shipping varied widely across countries,
regardless of proximity to the US market. Previously, Vega (2008) had
found that neighbours Ecuador and Colombia had very different freight
rates for their exports of fresh flowers: transport costs tended to be 10 to
20 per cent higher for Ecuador compared to Colombia.
In the LAC region, for economic growth reasons, reducing air transport
costs should have been a policy priority many years ago. Unfortunately, it
has not and this has had a negative effect on the movement of P and E,
which make up a high proportion of air cargo exports from South America
to the US (347 000 tons in 2006) and to the EU (100 000 tons in 2006)
(Vega, 2008).
Table 9.4 Top 15 airports in the LAC region, all flights and international
flights only (2015)
or about 0.6 per cent of total infrastructure investment (Kohli and Basil,
2011).
Airports are largely run by the central government in the LAC region.
They are at times in poor condition and provide inadequate commercial
air transport services. Many have runways that are short or not properly
asphalted, lack adequate passenger and cargo handling facilities, and
provide only limited control and landing aids. In other cases, airports
are used to meet wider government objectives such as a source of foreign
revenue or even to fund general expenditures.1 The case of Chile, where
airport charges were used for general expenditures, is one case cited in
the academic literature. Chile increased passengers’ boarding charges for
international flights to finance campaigns for promoting the country as a
tourist destination and to support a ‘Fight Against Poverty’ fund (Gómez-
Lobo and González, 2008).
To lessen the problems, a common practice in running airports in LAC
countries in recent years has been to hire a concessionaire to manage
them, particularly large ones. In general, concessions have contributed to
improvements in specific aspects of operating an airport, often only the
financial aspects – as a third party is arguably more agile at collecting fees
than governments. In Argentina, for instance, the concessionary system
proved defective, with a single bidder that offered an unrealistic contract
that needed to be restructured afterwards (Lipovich, 2008). Ultimately,
the government of Argentina was forced to commit additional resources
to subsidize the management of smaller rural airports. Another case that
deserves attention is that of the Mexican airports, where only those air-
ports with high passenger and cargo volumes and the best growth expecta-
tions were selected for privatization.2
Regarding productivity, Perelman and Serebrisky (2010) found that in the
LAC region the largest airports increased their productivity between 1995
and 2007, especially those airports handling between 7.5 and 10 million pas-
sengers per year, which posted an annual growth of 5.4 per cent. The authors
also reported that public airports have fared better, on average, than private
ones, experiencing an increase in productivity of 2.9 per cent compared to
0.7 per cent of private airports. However, when the authors looked at the
period between 2003 and 2007, both public and private airports experienced
positive productivity increases. On a more recent variation of the study of
airports’ productivity, Perelman and Serebrisky (2012) found that private
airports performed better than public airports in the period between 2000
and 2007, experiencing an increase in productivity of 2.8 per cent per year.
On the other hand, during the same period publicly-owned airports experi-
enced a decrease in productivity of 0.9 per cent. In contrast to their earlier
research results, the authors also found that it was medium-sized airports –
those handling between 5 and 8 million passengers per year – that experi-
enced the fastest growth in total factor productivity: 1.5 per cent on average.
Another aspect of the concession process over which no definite
outcome can be expected is that of an independent vs. non-independent
regulatory authority. Serebrisky et al. (2011) have reported that evi-
dence on the functioning for independent vs. non-independent airport
regulators did not provide definite conclusions, with each model having
its advantages and disadvantages. Independent authorities scored better
in the transparency of the rule-making and the establishment of a quality
bureaucracy. When it came to autonomy, some government departments
performed better than independent regulatory agencies.
A final issue that has been a major challenge in the history of conces-
sions is renegotiation. Sirtaine et al. (2005) have stated that ‘the short
interval between the granting of a concession and its renegotiations,
about two years, and the outcome of the renegotiation process, makes
the resulting regime a hybrid of price caps (high-powered contracts) and
rate of return (lower-powered contracts)’. More recently, Moore et al.
(2014) have investigated the contracting conditions in the LAC region,
presenting evidence of how private concessionaires ultimately maximize
their profits by increasing their leverage in the regulatory process of the
concession.
Full liberalization has been a utopia in the negotiations for air access
between LAC countries, and especially between LAC countries and their
northern neighbour, the US. Even though bilateral ASAs must be anchored
in reciprocal rights, unfortunately in the North–South trade, reciprocity is
not perceived as a negotiating point, since one party may be required to
make more concessions than the other simply because air transport indus-
try sizes vary across countries. Policy-makers in LAC countries have often
referred to this as an asymmetry of negotiation condition, an asymmetry
in which they base arguments against liberalization with politically charged
statements such as ‘air carriers in developing countries cannot compete
with those of developed ones’.
Within the LAC region, the CAN adopted a decision to integrate its
air transport sector back in 1991. Members of the Caribbean Community
(CARICOM) signed a multilateral ASA (MASA) in 1996, followed by
additional negotiations. Members of the MERCOSUR signed the Sub-
regional ASA (Fortaleza Agreement) in 1996. In all cases, however, only
limited gains have been achieved, mainly due to the relatively small size
of the intraregional aviation market. Often, new routes made possible by
these agreements do not have the potential to generate economically viable
traffic to guarantee economically sound operations.3 Overall, efforts to
liberalize the airspace of member countries of the CAN, and to a lesser
degree MERCOSUR countries, have gone unnoticed or have been halted
since individual country members are allowed to request exemptions or
can apply their own rules in accordance with other policy priorities, such
as the negotiation of a bilateral trade agreement with a third country.
At the bilateral level between LAC countries and the US, the most liberal
approach has been the negotiation of Open Skies. A major development
of this negotiation is the optional provision for the so-called 7th Freedom
of the Skies or all-cargo rights.4 Overall, OSAs are a welcome development
for users of air services in a region that depends on air transport for its
exports of P and E. Vega (2010), for example, found that developing coun-
Notes:
* Optional 7th Freedom All-Cargo Rights provides authority for an airline of one country
to operate all-cargo flights between the other country and a third country, via flights that
are not linked to its homeland.
** Open Skies Agreements currently enforced that include 7th Freedom all-cargo rights.
tries that have negotiated an OSA with the US benefit from an increase
in their volume of exports of P and E of 85.7 per cent, while the value of
their exports increases by 46.9 per cent. In the LAC region, 18 out of 33
countries have signed bilateral Open Skies agreements with the US as of 14
November 2016. Of those 18 countries, 11 have an Open Skies agreement
fully in force that also includes 7th Freedom all-cargo rights, as shown in
Table 9.6. The OSA between the US and Brazil has yet to go into effect.
In contrast to some advances in liberalizing air markets in the Americas,
progress in liberalizing intercontinental markets between the Americas
and other continents is basically nil. Perhaps the European Commission
has made marginal progress in negotiating horizontal aviation agreements
with LAC countries.5 Progress in negotiating comprehensive aviation
a greements, on the other hand, has been met with multiple obstacles; no
such agreement currently exists between any LAC country and the EU.6
As an alternative, the EU has promoted horizontal aviation agreements
in the region, but they do not replace the existing bilateral agreements in
place between EU member states and a third country. Chile was the first
country to successfully negotiate a horizontal agreement with the EU,
in September 2005; however, negotiations of a comprehensive aviation
agreement, that were expected to start right afterwards, have not reached
any major milestone. Uruguay was the second country in the region to
sign a horizontal aviation agreement with the EU, in November 2006. In
February 2007, Paraguay became the third LAC country to sign a horizon-
tal aviation agreement with the EU. Peru and the EU started negotiations
in September 2009. More recently, there was much expectation about the
negotiations of the first comprehensive ASA agreement between the EU
and a South American country, Brazil, in March 2011. This agreement
would have allowed all ‘EU airlines to operate direct flights to any des-
tination in Brazil from any point in the EU (and vice versa for Brazilian
carriers) – without restrictions on routes, prices or the number of weekly
flights’ (European Union, 2017). According to press reports, the agreement
is currently being renegotiated.
Notes:
* Countries not serving the US at the time of the assessment.
As of 8 March 2013, countries are removed from the list after four years if they do not
provide air transport service to the US, have no code-share arrangements with US air
carriers, and have no significant interaction with the FAA.
Mexico has not been active at negotiating OSAs. For example, although
the traffic between Mexico and the US is the most important traffic for
Mexico, Mexico has used a restrictive policy towards opening its air market
to competition from US carriers. In 2015, both countries concluded nego-
tiations of a lightweight OSA that entered force in August 2016. The main
development is the removal of the limitation of two carriers per route, with
the aim of increasing competition. The previous bilateral ASA between
the countries dated from 1960: it took these neighbours roughly 55 years
to reach some level of liberalization in their joint air transport market.
Although some features of OSAs are part of the deal, overall the agree-
ment is not an OSA.
The agreement will allow scheduled airline flights between the two coun-
tries to travel from any point in one country, via an intermediate country,
to any point in the other country and beyond. However, cabotage will
still not be allowed. There will be no restrictions to schedule cargo flights
for any number of airlines in any city pairs. Airlines from both countries
From 1920 to 1988, the Mexican federal government imposed strict price
controls and entry barriers to Mexico’s airlines market. The consequence
of that was the persistence of legal government-owned monopolies for
decades. Valdés and Ramírez (2011) have illustrated how the liberalization
process that started in 1988 – aimed at encouraging regional airlines to
serve as feeders of passenger traffic to national route carriers – never took
off.
As of today, Mexico’s two main airlines, Aeroméxico and Mexicana,
have been privatized. The story of privatization goes back to 1988, when
the Government of Mexico publicly announced that Aeroméxico was
bankrupt. The company Aerovías de México subsequently bought it
out. The same year, Mexicana entered into a public–private model of
ownership in which the Mexican government owned about 40 per cent
of the capital. By 1989, the Mexican government had authorized private
capital investments in Aeroméxico and Mexicana de Aviación. In 1993,
Aeroméxico acquired 55 per cent of the shares of Mexicana. According
to Vargas-Hernandez and Noruzi (2010), at the time the merged company
represented two-thirds of the Mexican air travel market. In 1995, the
Mexican government released its stake in Mexicana. It has been argued
that the privatization process of the two Mexican legacy carriers was
simply a change in the ownership of a monopoly, an action which provided
only limited opportunities to introduce competition and enhance the com-
petitiveness of the broader Mexican economy. The introduction of LCCs
took place in 1991, partly triggering the financial debacle of Mexicana and
Aeroméxico which ended up in these two airlines being statized once again
in 1994–95 (Valdés, 2013).
In July 1991, new air carriers were authorized to serve national routes,
and control over tariffs was eliminated in routes with more than one com-
petitor. The 1995 Civil Aviation Law formalized the liberalization process.
Valdés and Ramírez (2011) have assessed the impact of deregulation on
the performance of the Mexican airline industry using long sets of data
available from 1980 to 2007. Specifically, the authors sought to isolate the
effect of different variables from the effect of deregulation on the number
of passengers and airfares using a series of econometric models. They
found that the effect of deregulation was only marginal and smaller than
statistically more significant variables such as increases in income per
capita. Ros (2011) has also suggested that annual growth in domestic pas-
senger traffic between 1989 and 2008, at a rate of 5.4 per cent, coincided
with a period when the Mexican economy was growing at approximately
3 per cent per year. Regarding the effect of deregulation on the price of
airfares, Ros (2011) did not find evidence of increases or decreases attrib-
utable to deregulation between 1996 and 2004. While there were periods
when competition seemed to lower the cost of flying, during other times
airfares remained high despite additional players entering the market; for
instance, LCCs. Regarding the effect on airfares, existing research has
found that between 1996 and 2004 there were periods with greater contest-
ability in airfares, but in other periods the opposite occurred. Thus, the
effects of deregulation on the Mexican airline industry have been mixed at
best. There is no evidence that the growth in the number of passengers can
be directly attributed to deregulation only. There is also no evidence of a
decrease in the cost of airfares.
Today, the Mexican government continues to set limits when it comes
to approving new routes or establishing limits on foreign ownership of
airlines. Nevertheless, there have been welcome market liberalization dis-
ruptions. For example, Aeroméxico is the last legacy airline in the Mexican
aviation market, as low-cost carriers have come to dominate the market.
Overall, it took Mexico’s institutions nearly 30 years to unenthusiastically
approach liberalization strategies.
Chile’s liberalization of the air transport sector started in 1979 with the
enactment of the Commercial Aviation Law, DL No. 2564. Chile has been
known as a pioneer in the deregulation of the aviation sector using an
approach known as ‘Open Skies with reciprocity’. There might be several
reasons for Chile being a pioneer, including its geography and political
history, but assessing those would be outside the scope of this chapter.
Since the focus here is regulation, it is instead more relevant to mention
the four major principles that guided the development of the Aviation Law
and which remain relevant today. These principles can be summarized as
follows: (1) expand OSAs with as many partners seeking free entry to the
market as possible; (2) grant freedom to set fares for air carriers; (3) mini-
mize intervention of the authority; and (4) liberalize ownership require-
ments and control over the air carriers.
In the last ten years, Chile has pursued an ‘Open Skies with reciprocity’
policy that has meant that Chilean authorities have granted foreign car-
riers the same freedoms of the air that foreign countries have granted to
Chilean carriers. Since 2011, with a new policy further liberalizing the
sector, cabotage rights were granted unilaterally to foreign airlines wishing
to operate domestic flight segments. According to Chile’s Civil Aviation
Board the objective of this latest wave of liberalization is to create more
competitive conditions through cabotage, facilitating the access of foreign
operators to the internal transport of passengers and cargo – that is, ‘that
all the companies of the world can enter the domestic Chilean market
without restrictions’. As expected, foreign airlines must still meet all the
established technical and insurance requirements as domestic carriers.
Villena et al. (2008) attempted to quantify the effects of market liberali-
zation through OSAs (Table 9.9). There seems to be evidence of a positive
impact on the flow of passengers and a tendency to market concentration
after a treaty is signed. Connectivity seems not to be affected. Regarding
concentration in the industry, there is a tendency for the Herfindahl–
Hirschman Index to increase following the finalization of an OSA.
Available academic and industry studies support the claim that Chilean
authorities have sought to give air carriers – both domestic and international
– freedom to set service frequencies, fares and capacity without prior
approval of the authority, except for technical aspects and insurance provi-
sions (Alvarado and Marcos, 2012). There are no legal restrictions with
respect to establishing an airline, its ownership or its management.
M4371-FINGER_9781786431851_t.indd 178
$million
Argentina 1990 Steady increase in Increase from 0.14 34 per cent additional 300.9 4 pairs in 1991 to
passengers 1990–99; in 1990 to 0.57 in passengers in 1991– 14 in 1995, 10
greater rate since 2003 2001 2006; 221 371 more in 2004–06
annual passengers
Brazil 1989, Large increase since 0.3 in 1989, 0.45 in 6 per cent extra 65.6 3 pairs in 1994, 8
1993, 2002 1992, 0.35 in 1995, passengers in 1998– in 2000, 3 in 2003,
1996 0.2 in 2006 2006, 21 038 more 7 in 2006
178
annual passengers
Spain 1991, Large increase since 0.3 in 1990, 0.5 in 54 per cent additional 110.6 1–3 pairs, no
1997, 1997 1995, 0.5 in 2004, passengers in 1997– apparent influence
2005 0.35 in 2006 2006, 79 662 more of the OSA
annual passengers
US 1999 Large decrease following 0.3 in 1999, 0.35 in Not estimated Not 5 pairs in 1992, 9
2001, recovery since 2006 estimated in 1996
2004
Peru 1997 Constant growth until 0.27 in 1997, 0.7 in 6 per cent additional 12.3 1 pair in 1995, 6 in
2000, recovery starting 2000, 0.5 in 2005, passengers in 1998– 1997, 1 after 1999
in 2004 0.7 in 2006 2006, 12 904 more when cabotage is
annual passengers enforced
14/11/2017 11:14
Latin America and the Caribbean 179
that is, the system is financed through the collection of fees paid by air-
lines for use of the services provided by the DGAC, by passengers using
the network of air terminals and by aeronautical and non-aeronautical
concessions. The Aviation Law requires that the income obtained by the
collection of these rates be reinvested in the system itself (Agostini, 2008).
Through time, the Chilean concession model has been adjusted using a
case-by-case strategy.
CONCLUSIONS
NOTES
1. Oum et al. (2006) provided evidence that airports with government majority ownership
and those with multi-levels of government involvement are significantly less efficient than
private airports.
2. In 1998 the Mexican government held 85 per cent of the shares, with a private cor-
poration holding the remaining 15 per cent. As of February 2006, the 85 per cent of
government shares had been auctioned in the domestic and international stock markets,
completing the privatization process (Rico Galeana, 2008). Overall, Mexican airports
have experienced steady growth of traffic. Smaller airports continue to be subsidized by
central and local governments.
3. A major accomplishment in Central America, although not an aviation market liberaliza-
tion, has been the structuring of the Central American Corporation for Air Navigation
Services, a regional consortium that provides air traffic services for its members.
4. This provision allows a domestic carrier to operate all-cargo services between the foreign
country and a third country, via flights that are not linked to its homeland.
5. Horizontal agreements are agreements negotiated by the European Commission on
behalf of EU member states, in order to bring all existing bilateral ASAs between EU
member states and a given third country in line with EU law, removing nationality restric-
tions of existing bilateral ASAs and therefore allowing European airlines to operate
flights between any EU member state and a third country.
6. According to the EU’s website, the so-called comprehensive aviation agreements that
the European Commission negotiates on behalf of the EU and its member states
are not limited to ‘Open Skies’ provisions only, but instead ‘the EU model also seeks
the establishment of a process of liberalization of ownership of airlines and a process
of regulatory convergence in matters of safety and security, competition, environment,
passengers protection, labor, etc. – which could not be obtained at national levels’.
7. The main distinction between failures in safety and security is the issue of intentional-
ity. While safety failures have unintentional causes, security failures are intentional in
nature.
REFERENCES
Agostini, C.A. (2008), ‘The industrial organization of air transport in Chile (La
organización industrial del transporte aéreo en Chile)’, Revista de Análisis
Económico, 23, 35–84.
ALTA (2016), The Latin American and Caribbean Capacity Analysis. Available at
https://www.alta.aero/la/upload/report/_9945.pdf.
Alvarado, A. and G. Marcos (2012), Evaluación del Impacto Económico de un
Acuerdo de Cielos Abiertos Chile-UE, Informe Final, Junta de Aeronautica Civil,
Santiago de Chile.
Button, K. and H. Vega (2008), ‘The effects of air transportation on the movement
of labor’, GeoJournal, 71, 67–81.
capital structure: theory and evidence from Latin American transport conces-
sions’, Journal of Regulatory Economics, 45, 209–32.
Oum, T.H., N. Adler and C. Yu (2006), ‘Privatization, corporatization, owner-
ship forms and their effects on the performance of the world’s major airports’,
Journal of Air Transport Management, 12, 109–21.
Paredes, R.D. and J.M. Sánchez (2004), ‘Government concession contracts in
Chile: the role of competition in the bidding process’, Economic Development
and Cultural Change, 53, 215–34.
Perelman, S. and T. Serebrisky (2010), Measuring the Technical Efficiency of Airports
in Latin America, Policy Research Working Paper No. 5339, Washington, DC:
World Bank.
Perelman, S. and T. Serebrisky (2012), ‘Measuring the technical efficiency of air-
ports in Latin America’, Utilities Policy, 22, 1–7.
Poole Jr, R.W. (2016), Annual Privatization Report 2016: Air Transportation,
Washington, DC, Reason Foundation.
Rico Galeana, O.A. (2008), ‘The privatisation of Mexican airports’, Journal of Air
Transport Management, 14, 320–23.
Ros, A.J. (2011), ‘The determinants of pricing in the Mexican domestic airline
sector: the impact of competition and airport congestion’, Review of Industrial
Organization, 38, 43–60.
Serebrisky, T., S.L. Azumendi and L.A. Andrés (2011), ‘Institutional design and
governance of airport regulators: the case of Latin America’, Journal of Air
Transport Management, 17, 207–10.
Sirtaine, S., M.E. Pinglo, J.L. Guasch and V. Foster (2005), How Profitable are
Infrastructure Concessions in Latin America? Empirical Evidence and Regulatory
Implications, Washington, DC: World Bank.
US Department of Commerce (2006), Brazil, Washington, DC: Western
Hemisphere Diversification and Defense Market Guide. Accessed on 22 July
2006 at http://www.bis.doc.gov/DefenseIndustrialBasePrograms/OSIES/Export
MarketGuides.
Valdés, V. (2012), Recomendaciones de Política Aeronáutica para México,
USAID.
Valdés, V. (2013), ‘Fallas del mercado y del gobierno en el sector aeronáutico
Mexicano’, Revista de Economía Institucional, 15, 253–83.
Valdés, V. and J.C. Ramírez (2011), ‘Una evaluación sobre la desregulación
del mercado de aerolíneas en México’, Economía Mexicana. Nueva Época, 20,
5–35.
Vargas-Hernandez, G.J. and M.R. Noruzi (2010), ‘From entrepreneurial state to
state of entrepreneurs: ownership implications of the transformation in Mexican
governance since 1982’, Journal of Public Administration and Policy Research,
2(1), 7–19.
Vega, H. (2008), ‘Air cargo, trade and transportation costs of perishables
and exotics from South America’, Journal of Air Transport Management, 14,
324–8.
Vega, H.L. (2010), Developing Countries and their Airborne Export Flows of
Perishable and High-tech Goods, Fairfax: George Mason University.
Villena, M.J., R. Harrison and M.G. Villena (2008), ‘Impacto Económico de la
Política de Acuerdos de Cielos Abiertos en Chile’, Revista de Análisis Económico,
23, 107–49.
Wanke, P. and C.P. Barros (2016), ‘Efficiency in Latin American airlines: a
INTRODUCTION
The bodies that are generally viewed as the most authoritative sources
of aviation forecasts – Boeing, Airbus and the International Civil
Aviation Organization – consider Africa as having the potential for a
rapid increase in its air transport sector.1 This sector is characterized by
generally favourable conditions for a significant expansion of its aviation
industry, partly due to the improved political and economic climates in
many African countries, particularly those in the sub-Saharan region.
The situation in North Africa is also somewhat more stable than it was a
few years ago. Adjusted estimates by the International Monetary Fund
of the national incomes of countries like Nigeria suggest faster economic
growth than previously assumed, while peaceful, democratic transitions
of power in countries like Gambia and Ghana indicate greater political
stability.
Africa is also a continent with large areas of open space and river net-
works that, unlike those of Europe, are often not conducive to trade and
are difficult terrain for constructing surface transportation. It also has a
number of large urban areas that are remote from each other and from
non-African cities, making movement between them difficult. Given these
geographical and demographical characteristics, and the significant growth
of many other continental economies seeking the resources of Africa, air
transport appears to be the ideal catalyst for the economic development of
the continent.
Partly because of its strategic importance in the colonial and immediate
post-colonial periods, and partly because the necessary private institutional
structure to finance and operate an air network did not exist, African air
transportation has traditionally been heavily regulated. However, fol-
lowing the experiences of aviation policies elsewhere, particularly in the
US and Europe, coupled with the manifest failure of many state-owned
airlines in Africa, this attitude changed and market liberalization became
seen as a way forward. This led to the beginning of Africa’s air transport
185
Our analysis of the African airline industry is mainly based on the Official
Airline Guide (OAG) database, from which are collected scheduled operat-
ing direct flights covering 1997 to 2013.4 Button et al. (2015b) point out
200
150
100
50
0
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Figure 10.1 Number of airlines operating to, from and within Africa
(1997–2013)
1997 2013
Routes Monopoly 76% 68%
Duopoly 13% 16%
More than 2 airlines 11% 16%
Seats Monopoly 37% 28%
Duopoly 17% 18%
More than 2 airlines 48% 54%
2035
Africa–Southeast Asia 2015
2008
Africa–North America
Africa–Middle East
Africa–Europe
Africa–Africa
There are several reasons for such cost inefficiencies, mainly related to
disadvantages in fuel purchasing, mismanagement, fleet ageing, low load
factors, and expensive ticket distribution and complementary services.
Fuel transportation (across countries that are landlocked and/or have poor
infrastructure) and fuel purchasing (due to insufficient scale to hold bar-
gaining power with suppliers), along with the fact that fuel hedging is still
not a very common practice, put many African airlines at a disadvantage.
A second category of costs is the heritage of a long-standing policy of
government control of airlines that have traditionally been operated more
for political objectives rather than for economic efficiency. In particular,
many countries have consistently subsidized the operations of their inef-
ficient flag carriers, often creating costs associated with mismanagement.
Also, the ageing of fleets implies not only higher costs of maintenance but
also increased fuel consumption, poor reliability and increased downtime.
The inability to exploit the full economies of optimal hardware use is com-
pounded by poor scheduling on the part of the carriers and night flying
restrictions by the authorities which leads to one of the lowest aircraft
utilization rates in the world.
Also, African nations have the lowest average load factors in the world
(Figure 10.3). This not only reflects high fares but also lack of reliability
in services and the late emergence of a middle-income class, the group that
tends to make the most use of aviation.
Furthermore, the low penetration of both the internet and credit cards
in Africa makes ticket distribution more complicated and expensive.9
Moreover, because they were state-owned monopolies, airports and air
14%
12%
% of Total Seats
10%
8%
6%
4%
2%
0%
01
02
03
04
05
06
07
08
09
10
11
12
13
)
ay
20
20
20
20
20
20
20
20
20
20
20
20
20
–M
an
(J
14
20
Source: CAPA (2014).
navigation service providers used to charge higher fares than their counter-
parts in the rest of the world.
These problems faced by African airlines have significantly hampered
the development of low-cost carriers (LCCs) in Africa. Unlike most
regions in the world, where LCCs have boomed followed the liberaliza-
tion of the aviation industry, only slightly more than 10 per cent of avail-
able seats in the intra-African market were offered by low-cost carriers
(Figure 10.4), compared to one-third of the global average of 26 per cent.
These numbers are even more notable when compared to those of other
developing regions such as South Asia and South-East Asia, which has
36 per cent and 53 per cent respectively of LCC market shares (Boeing
Commercial Aeroplanes, 2014). LCC services were provided mainly in
large domestic markets, such as South Africa and Kenya.
Furthermore, despite the improvements discussed in the previous
section, it seems that Africa’s airline network is not yet sufficiently devel-
oped. Ssamula (2008) suggested that Africa’s unique situation means it
must move toward a hub-and-spoke network structure based on four
main hubs, one for each African region: Northern, Western, Eastern and
Southern Africa. Such a structure should help airlines minimize their oper-
ating and network costs. However, many airports are still competing for
this position, especially in the Northern and Eastern parts of Africa, and
this struggle may slow down the development of the African air transport
network.
A further persistent problem with the African airlines has been their
safety record; in 2005, close to 25 per cent of the world’s aircraft crashes
occurred in Africa, while African flights only make up 5 per cent of the
global traffic. However, there has been some improvement in recent years,
partly due to the pressure of the International Civil Aviation Organization
but also because of a general improvement in aviation safety. Despite the
improvements achieved in recent years, poor maintenance and a lack of
monitoring and control led to an increasing trend of blacklisting airlines
(Ssamula, 2008). As a result, several airlines have been prevented from
flying to the European Union and the United States, with a detrimental
effect on the growth of the African air transport sector. Moreover, the
infrastructures of most African airports are not good enough to meet
the new technological standards and the security demands imposed by the
industry.
INFRASTRUCTURE DEVELOPMENTS
Each country controls its own air space and, as such, is responsible for its
air navigation system (which includes air traffic control). However, there
has been some coordination in their operation, with the 19 members of the
Agency for Aerial Navigation Safety in Africa and Madagascar manag-
ing 620 000 square miles of airspace covering Antananarivo, Brazzaville,
Dakar Oceanic and Terrestrial, Niamey and N’Djamena from its control
centres located at the international airports in each region. There have
also been changes in the ways countries provide their own systems. Most
remain state-owned, but some are now institutionally separate from gov-
ernment; for example, South Africa’s Air Traffic and Navigation Services
Ltd is a not-for-profit joint-stock corporation, although a ministerial com-
mittee regulates its rates.
The quality of the air traffic control systems varies widely between and
even within countries. In general, North Africa, Morocco and Egypt have
reasonable infrastructures; west sub-Saharan Africa, Senegal, Guinea,
Ivory Coast and Ghana are adequately served, as are Kenya, Rwanda
and Tanzania on the east coast, and South Africa in the south. Indeed,
South Africa’s air traffic control is ranked 15th in the world by the World
Economic Forum in terms of its air transport infrastructure, while Tunisia,
Mauritius, Ethiopia, Morocco, Seychelles and Namibia are all in the top
60.10 The African interior, on the other hand, especially the vast expanses
of the Central African Republic and the Republic of the Congo, is largely
devoid of radar and good radio communications. In such areas, inter-
aircraft communication is the main way of maintaining separation.
One of the major problems in operating all infrastructures in most of
Africa, including air traffic control, is labour supply. The implications of a
shortage of suitable equipment in many parts of Africa are worsened by a
shortage of skilled labour, and this is a problem that extends to other ele-
ments of the aviation supply chain. For example, if African aviation does
grow as predicted by Boeing and others, there will be a need for over 20 000
additional pilots over the next two decades. Problems of poor pay hinder
retention and the ‘poaching’ of qualified personnel by non-African airlines
adds to the issue of limited training capacity.
Most airports in Africa are relatively small in international terms.
No African airport even approaches the top 50 airports in the world in
terms of passenger numbers. In 2014, the continent’s three largest air-
ports in terms of passengers handled were, OR Tambo International
Airport Johannesburg (19 164 000 passengers), Cairo International Airport
(14 684 892), and Cape Town International Airport (8 636 294). In terms
of airport quality, there are also vast differences between countries and
between facilities within countries, with many airports being outdated and
poorly equipped, particularly in sub-Saharan Africa. Some of these facili-
ties have long since become incapable of efficiently handling the thousands
of passengers flowing through them. Added to this, there is considerable
variation in efficiency, even among the older airports. In many cases they
are also poorly used, and in some cases there is demonstrable underutiliza-
tion (Stephens and Ukpere, 2011).
Recently, there have been significant efforts to foster public−private
partnerships (PPPs) of various kinds, and joint ventures to both finance
investments in airports and to enhance their subsequent operational effi-
ciency. Despite the challenges of developing PPP contracts, there have been
some successes that, in conjunction with more conventional government-
based initiatives, have led to increases in actual and projected airport
expenditures. International agencies have also become more engaged in
airport development, both in terms of developing new facilities and signifi-
cantly upgrading the existing capacity. Overall, the Center for Asia Pacific
Aviation (CAPA) has found that at least US$33.8 billion has been invested,
or is earmarked to be invested, in construction and associated projects at
existing airports in 77 projects in more than nine African countries, with
an average price tag per project of almost $440 million. Examples include
Angola (30 projects costing $2.16 billion), Nigeria ($870 million on multi-
ple passenger and cargo terminal projects), Senegal (a new airport at Dakar
costing $700 million), Rwanda (a new airport at Kigali, $650 million), and
Kenya (new terminals and a second runway at Nairobi, $1 billion).
China has also become a major player in developing airports in Africa,
and although the exact scale of the commitment is unclear, the Chinese
Ministry of Commerce in 2014 suggested the net annual flow of Chinese
direct investment into Africa increased eight-fold between 2005 and 2014,
to $3.2 billion, with a significant part going to infrastructure of one kind or
CONCLUSION
The African air transport industry has experienced a process of partial lib-
eralization in recent years. The principles that inspired the Yamoussoukro
Decision have been implemented heterogeneously across the continent.
As a result, despite the growth in traffic flows – registered on both intra-
African and intercontinental connections – and the increases in both the
number of airlines and the level of competition, air transportation in
Africa is still suffering from some long-standing weaknesses, including a
lack of interconnectivity, an unbalanced traffic distribution, higher (com-
pared to the rest of the world) fares and airline costs, and sparse demand.
In sum, the market for air services in Africa is, overall, clearly healthier
than in the past but, with a few notable exceptions, still lacks the capacity
and services levels found in most other parts of the world.
NOTES
1. The forecasts are somewhat optimistic. Boeing Commercial Aeroplanes (2016), for
example, predicts that intra-Africa revenue passenger miles flown will grow by an
annual average of 6.7 per cent between 2015 and 2034, and those between Africa and the
Middle East and Asia by 7.3 per cent and 7.1 per cent, respectively. Similarly, air cargo
is projected to grow by 6.6 per cent annually compared to a global growth rate of 5 per
cent. The UN’s International Civil Aviation Organization (2014) indicates steady growth
with passenger traffic to, from and within Africa, increasing at an annual average rate of
5.3 per cent until 2030, while cargo traffic is projected to grow at 5.6 per cent. Africa–
Middle East traffic is forecast to expand at 9.9 per cent, the most rapid growth in the
world. Forecasts produced by Airbus offer a very similar picture.
2. The pre-war civil aviation market was extremely limited. For example, the total number
of passengers flown by British Imperial Airways throughout the 1930s was only about
50 000, and Africa involved only part of this network.
3. Many of them failed in the following years, mainly due to low demand and
mismanagement.
4. Where not otherwise specified, tables and figures are the result of authors’ elaboration
of OAG data.
5. The main problems at that time were about low security standards and unstable political
situations in many countries, which had led to the bankruptcy of many African airlines
in recent years, including relevant national carriers such as Air Nigeria and Air Gabon.
6. Ryanair and EasyJet, namely the two most important European LCCS, provide direct
flights toward Morocco and Egypt.
7. British Airways also operated indirectly in the domestic South African air transport
market through its Comair and kulula.com brands.
8. O’Connell (2006) points out that almost 11 per cent of the Emirates traffic coming from
Europe and stopping in Dubai has a final destination located in Africa.
9. Commission payable to the travel agencies is typically about 7 per cent of the ticket price
(Heinz and O’Connell, 2013).
10. Button and McDougall (2006) also offered a detailed comparison of 11 major air traffic
control systems, including South Africa’s.
REFERENCES
Njoya, E.T. (2016), ‘Africa’s single aviation market: the progress so far’, Journal of
Transport Geography, 50, 4–11.
O’Connell, J.F. (2006), ‘The changing dynamics of the Arab Gulf based airlines
and an investigation into the strategies that are making Emirates into a global
player’, World Review of Intermodal Transportation Research, 1, 94–114.
O’Connell, J.F. (2011), ‘The rise of the Arabian Gulf carriers: an insight into the
business model of Emirates Airline’, Journal of Air Transport Management, 17,
339–46.
O’Connell, J.F. and D. Warnock-Smith (2012), ‘Liberalization and strategic change
in air transport: an examination of current and future variations in tourist
traffic to and from Egypt resulting from policy changes at EgyptAir’, Tourism
Economics, 18, 845–70.
Schlumberger, C.E. (2010), Open Skies for Africa: Implementing the Yamoussoukro
Decision, Washington, DC: World Bank.
Ssamula, B. (2008), Strategies to Design a Cost-effective Hub Network for Sparse
Air Travel Demand in Africa, PhD. dissertation, University of Johannesburg.
Stephens, M.S. and W.J. Ukpere (2011), ‘Airport capacity utilization in Nigeria: a
performance and efficiency analysis’, African Journal of Business Management,
5, 11104–15.
Topical issues
205
studies found that airline safety records had continued to improve during
the deregulated era (Oster and Zorn, 1987). However, safety records for
new entrant airlines and for regional/commuter carriers were sometimes
significantly worse than for established carriers (Savage, 1999). Even with
this segment’s poorer safety performance, Savage (1999) noted that overall
intercity travel safety improved as passengers shifted modes from auto
travel to airlines.
Liberalization in Canada and Australia differed from the United States
because of the prior dominance of two airlines (one state-owned) in each
country. As such, privatization went hand-in-hand with liberalization and
proceeded on much more of an evolutionary track. Studies found no sig-
nificant effect on airline safety following liberalization in these countries,
perhaps in part because changes in industry structure and the regulatory
environment were quite different from those in the United States (Forsyth,
1991; Oster et al., 1992).
Liberalization has also had different characteristics across countries and
regions. Some countries adopted evolutionary approaches to deregulation,
as in the three liberalization packages in Western Europe between 1987 and
1993. Others adopted a more complete change all at once, such as occurred
in India and South Africa. In some settings, larger changes in the macro
environment encompassed the aviation sector, as occurred in Russia, the
countries of the former Soviet Union, China, India and Brazil.
Many of the concerns that airline deregulation might adversely impact
safety stem from four factors: industry growth, new entrants, financial
stress, and changes in aviation markets (Moses and Savage, 1989; Oster
et al., 1992). First, industry growth following liberalization has been a
common experience worldwide. Pricing deregulation and new entrants
(especially low-cost airlines) have resulted in lower fares and spurred
demand, with passenger volumes sometimes growing quite dramatically.
This growth creates challenges for safety. Rapid growth and the emergence
of new mainline and regional carriers have also dramatically increased the
workload on aviation regulatory authorities, who have not always been
able to respond quickly in terms of budgets and other resources (Button,
2004; Finger and Piers, 2005; General Accounting Office, 1996; Yadav and
Nikraz, 2014). Airport and aviation infrastructure may face capacity con-
straints that may induce congestion and operational challenges, and affect
workloads of air traffic control organizations.
Second, new entrant carriers may pose challenges to aviation safety, as
they have often operated in new markets, with older equipment and less
experienced personnel. New entrants frequently begin service with older,
leased aircraft, which can present maintenance challenges (Kanafani and
Keeler, 1990). Industry growth also increases demand for skilled labor,
especially pilots and mechanics, which can create a situation where the
least experienced personnel are operating in more challenging environ-
ments. This growth has consequences for flight crew and maintenance
training, which also has the potential to affect safety.
Third, lower fares and legacy costs often produce financial stress in the
older airlines as well. There is always a concern that a more competitive,
market-driven environment will create incentives to cut costs and take risks
that could result in decreases in safety (Dionne et al., 1997; Golbe, 1986;
Madsen, 2013; Rose, 1990).
Fourth, the consequences of liberalization have varied as well. While
lower fares and rapid passenger growth have generally resulted, other
industry structure effects have been more varied. For example, in some
countries liberalization has resulted in a boom of new entrant carriers,
some of which have succeeded, others of which have not. As a result of
increased competition, liberalization has also brought both financial and
operating pressures to legacy, state-owned carriers. More recently, some
regions have seen an increase in industry consolidation. The merger of two
carriers involves combining different maintenance, inspection, training
and safety programs. Even if the original programs were well designed and
well executed, the consolidation to a single program has the potential to
impact safety negatively.
However, liberalization may also have improved safety by changing
the nature of aviation markets. In general, scheduled services have grown
faster than, and partially substituted for, non-scheduled charter operations.
Scheduled services have long had a better safety record. Industry growth
has spurred the development and adoption of a new generation of smaller
jets, replacing older and less safe piston and turboprop aircraft (Barnett,
1990). There also is the potential for learning: countries that liberalized in
more recent periods had the advantage of the lessons from earlier deregu-
lation experiences. In addition, aviation has been able to learn from safety
analysis and management in other industries that have similar risk char-
acteristics in terms of key personnel, technology and systems (Biermann,
2016; Drax et al., 2014; Stolzer and Goglia, 2016).
century. These accident rates have continued to decline in the past three
decades of liberalization. But in terms of safety, the airline industry is far
from homogeneous and far from static. Both accident rates and underlying
causes have changed over time, and have varied across regions of the world
and across different types of carriers.
Assessing safety performance pre- and post-liberalization is difficult for
a number of reasons. First, determining the pre- and post- periods is chal-
lenging. Many countries experienced liberalization as a series of stages that
moved in an evolutionary way from a heavily regulated to a largely deregu-
lated industry. The best example of this is Western Europe, in which three
liberalization ‘packages’ occurred from 1987–93, with phase-in periods
to each. In addition, liberalization can take place in both domestic and
international settings, which may have different effects on safety perfor-
mance. Second, liberalization has had different characteristics in different
settings. Third, data limitations have become more prevalent as liberaliza-
tion has spread to emerging markets and regions. Many safety records are
incomplete. Operational data is often inconsistently reported, making it
more challenging to construct and interpret accident rates. Moreover, lack
of data consistency and availability makes it very hard to analyze specific
industry segments, let alone individual airlines.
In this section, to classify causal factors and the type of service, we
analyze the comprehensive set of 771 worldwide commercial accidents in
which passengers were killed between 1990 and 2015. Several data issues
have shaped our approach in this chapter. First, our focus is on commercial
accidents in which passengers were killed, and the fatality rates for those
accidents. Data reporting requirements and effectiveness vary across coun-
tries, regions and types of service, and non-fatal accidents are often unre-
ported or lack key data about what occurred. Passengers are also killed in
private or general aviation aircraft, but here again, reporting of those acci-
dents is inconsistent and operations data are often unavailable, particularly
in developing countries. In contrast, fatal commercial accidents generally
involve more extensive reporting and investigation. Second, accidents are
classified in terms of carrier and home country, rather than location. (For
most cases, these are within the same country or region.) Third, we define
pre- and post-liberalization periods in broader, multi-year terms. Fourth,
where data is missing or inconsistent, we discuss what effect this might
have on our analysis and conclusions, as well as the implications for future
research.
Our primary source for fatal accident data is the World Air Accident
Summary (WAAS) (Ascend Flight Global, 2016). Researched and pub-
lished on behalf of the UK Civil Aviation Authority, WAAS includes
detailed descriptions for 8000 accidents involving jet and turboprop and
250
Index 1990–1994 = 100
200
Passengers Index
150
Fatalities Index
Fatality Rates Index
100
50
0
1990
1995
2000
2005
2010
2015
fatalities per 1 million passengers carried fell over 80 percent from 1.01 in
the 1990–94 period to 0.18 in the 2011–15 period. At the global level, it
appears that the airline safety has continued to improve during the exten-
sion of liberalization across the world since 1990.
For our safety index, we use passenger fatalities per million passengers
carried. It would also be possible to use fatal accidents per million aircraft
departures; patterns seen using that measure would be similar to Figure
11.1. It is important to realize, however, that in spite of the impression
one might get from looking at media coverage of aircraft accidents, many
passengers survive crashes in which one or more passengers is killed.
Historically, in about 51 percent of fatal crashes, all passengers on board
are killed. In 49 percent of fatal crashes, there are survivors. Overall, about
68 percent of passengers on board in fatal crashes are killed, and about 32
percent survive.
Table 11.1 shows the distribution of commercial accidents with passenger
fatalities by type of service. During the entire 1990–2015 period, with more
than 48 billion passengers carried, there were 771 fatal accidents involving
commercial passenger airlines, involving 21 727 fatalities. The global fatal
accident rate for the entire period was 0.45 passenger fatalities per million
passengers. As can be seen in the table, scheduled service dominates pas-
senger service worldwide, accounting for 96 percent of passengers carried,
but accounting for only 80 percent of passenger fatalities. Non-scheduled
service, especially non-scheduled domestic service, accounts for a dispro-
Table 11.2 shows the distribution of fatal accidents by region. There are
marked variations. The fatal accident rate for sub-Saharan Africa
is by far the highest, at 7.17 fatalities per million passengers. Other
regions with rates much higher than the worldwide average include the coun-
tries from the former Soviet Union, Central America and the Caribbean,
South Asia, and Southeast Asia. These regions all have fatal accident rates of
more than about 1 per million, well over twice the worldwide average. There
are also individual countries which have statistically significantly higher rates,
including Russia (1.77) and Indonesia (1.60). It should be noted, though,
that even these higher rates represent improvements in aviation safety in these
countries and regions relative to their performance in the 1970–89 period
(Oster et al., 1992).
Why did these airplanes crash? The approach for assigning causes to
accidents which is taken in this chapter and in the authors’ prior work, is
to select as the cause the factor that initiated the sequence of events that
culminated in the accident (Oster et al., 1992; see also Ancel et al., 2015).
The assumption behind this approach is that, in the absence of the factor
These two primary causes are relatively unchanged from the 1970–89
period. Terrorism and criminal activity were a more common cause in the
1990–2015 period than earlier.
Also of concern is the large share of accidents in non-scheduled service
where the cause was ambiguous. These were cases where there simply was
not enough information about the crash to determine what happened.
Many non-scheduled accidents, particularly in developing countries, simply
do not receive the investigative attention that the scheduled service crashes
do. Also, many of the aircraft used in non-scheduled service are smaller,
turboprop or piston driven aircraft that are not equipped with flight data
recorders or cockpit voice recorders. In domestic scheduled service, for
example, jet aircraft account for 37 percent of accidents and piston powered
aircraft account for 7 percent of accidents. In contrast, for domestic non-
scheduled service, jet aircraft account for only 12 percent of accidents while
piston powered aircraft account for 47 percent. Many non-scheduled air-
craft also crash in remote locations where it can be days (or longer) before
safety officials can examine the site of the crash. This represents an impor-
tant challenge in how to devote resources to improve safety in the future.
Note: *Because of data limitations the rates for China and India were calculated for four
years before and after rather than five years.
In one case, Indonesia, it improved only slightly more than the worldwide
rate. In only one case, Turkey, did the fatality rate rise. In Turkey, however,
the five-year period prior to liberalization had one accident in which only
one passenger was killed. In the five-year period following liberalization,
Turkey also had only one accident, this time in which 50 passengers were
killed. While Turkey’s fatal accident rate per million passengers increased
following liberalization, the number of fatal accidents was unchanged from
the prior period.
Excluding Turkey, all of the countries experiencing liberalization had a
lower passenger fatality rate in the period following liberalization than in
the similar period prior to liberalization. Also, again with the exception of
Turkey, all of these countries saw their passenger fatality rate drop by more
than the worldwide passenger fatality rate over the same periods.
Recall from Table 11.4 that many of the liberalizations happened before
1990 (when our data begins) or so soon after 1990 that it wasn’t possible to
construct multi-year before and after comparisons. Figure 11.2 compares
what has happened to passenger fatality rates for four of those liberaliza-
tions, Canada, Western Europe, the United States and Australia, during
the 1990 to 2015 period. In all four cases, fatality rates were lower than the
worldwide rate at the beginning of the period and continued to decline,
remaining below the worldwide rate.
1.0
Fatalities per Million Passengers
0.8
Worldwide
Canada
0.6
Western Europe
United States
0.4
Australia
0.2
0
1990
1995
2000
2005
2010
2015
Figure 11.2 Fatality rates for Canada, Western Europe, the United States
and Australia
CONCLUDING OBSERVATIONS
Starting with airline deregulation in the United States in the late 1970s and
continuing through the recent liberalizations, there have been continuing
concerns that liberalizations might cause safety to degrade through some
combination of the pressur es on personnel and infrastructure of rapid
industry growth and the emergence of new entrant airlines. But these liber-
alizations have occurred against a backdrop of steadily improving aviation
safety because of lessons learned from investigating accidents, improved
equipment, and improved approaches for training pilots, mechanics and
others involved in aviation (Cui and Li, 2015; Evans, 2016). Also, each
liberalization brought forth lessons for future reforms, so that countries
could shape and be better prepared as their own version of deregulation
unfolded.
In examining the safety records of countries and regions both before
and after liberalization, our analysis finds that safety has generally
improved following liberalization and often at rates that were greater than
the worldwide improvement in safety rates. These results suggest that lib-
eralization has not prevented or even hampered improvements in aviation
safety. Indeed, to the extent that liberalization has sped up the substitution
of scheduled service for the generally less safe unscheduled service, liberali-
zation may even have helped improve aviation safety.
While overall aviation safety has continued its improvement through
the age of liberalization, it remains true that variations in safety perfor-
mance persist across regions, carriers and segments of the industry. Rapid
growth has been especially challenging in developing countries and in set-
tings where new entrant carriers are operating older piston and turboprop
aircraft. Looking ahead, a major challenge to improving safety is that a
significant proportion of airline accidents now involve situations where
information about what happened is limited or non-existent, especially in
non-scheduled services in developing aviation markets.
REFERENCES
Ancel, E., A.T. Shih, S.M. Jones, M.S. Reveley, J.T. Luxhoj and J.K. Evans (2015),
‘Predictive safety analytics: inferring aviation accident shaping factors and cau-
sation’, Journal of Risk Research, 18(April), 428–51.
Ascend Flight Global (2016), ‘World air accident summary’, database available at
http://www.ascendworldwide.com/what-we-do/ascend-data/accident-and-loss-
data/.
Ballard, S. (2014), ‘The U.S. commercial air tour industry: a review of aviation
safety concerns’, Aviation, Space, and Environmental Medicine, 85(2), 160–66.
Barnett, A. (1990), ‘Air safety: end of the golden age’, Chance: New Directions for
Statistics and Computing, 3, 8–12.
Bier, V., J. Joosten, D. Glyer, J. Tracey and M. Welsh (2003), Effects of Deregulation
on Safety: Implications Drawn from the Aviation, Rail, and United Kingdom
Nuclear Power Industries, Norwell, MA: Kluwer Academic.
Biermann, T. (2016), Safety Management in Aviation – and Beyond, Wildau: Wildau
Verlag.
Button, K. (2004), ‘Conforming with ICAO Safety Oversight Standards’, Journal
of Air Transport Management, 10(4), 249–55.
Cui, Q. and Y. Li (2015), ‘The change trend and influencing factors of civil avia-
tion safety efficiency: the case of Chinese airline companies’, Safety Science, 75
(June), 56–63.
Dionne, G., R. Gagné, F. Gagnon and C. Vanasse (1997), ‘Debt, moral hazard, and
airline safety’, Journal of Econometrics, 79, 379–402.
Drax, C., A. Wittmer and R. Muller (2014), Aviation Risk and Safety Management:
Methods and Applications in Aviation Organizations, Cham: Springer.
Evans, J.K. (2016), Differences in Characteristics of Aviation Accidents During
1993–2012 Based on Flight Purpose, Hampton, VA: NASA Center for Aerospace
Information.
Finger, M. and M. Piers (2005), ‘Air transport regulation under transformation: the
case of Switzerland’, Journal of Air Transportation, 10(2), 20–32.
Forsyth, P. (1991), ‘The regulation and deregulation of Australia’s domestic airline
industry’, in K. Button (ed.), Airline Deregulation: International Experiences,
New York: New York University Press, pp. 48–84.
General Accounting Office (1996), Airline Safety: New Airlines Illustrate Long-
Standing Problems in FAA’s Inspection Program, Report GAO/RCED-9 7–2,
Washington, DC: US Government Printing Office.
Golbe, D.L. (1986), ‘Safety and profits in the airline industry’, Journal of Industrial
Economics, 34(3), 305–18.
International Civil Aviation Organization (2016), ICAOData1 database, available
at https://www4.icao.int/NewDataPlus/.
Kanafani, A. and T. Keeler (1990), ‘Air deregulation and safety: some econometric
evidence from time series’, Logistics and Transportation Review, 26(3), 203–209.
Madsen, P.M. (2013), ‘Perils and profits: a reexamination of the link between
profitability and safety in U.S. aviation’, Journal of Management, 39(3), 763–91.
Moses, L. and I. Savage (1989), Transportation Safety in an Age of Deregulation,
New York: Oxford University Press.
Moses, L.N. and I. Savage (1990), ‘Aviation deregulation and safety: theory and
evidence’, Journal of Transport Economics and Policy, 24(2), 171–88.
Oster, C.V. and C.K. Zorn (1987), ‘Deregulation’s impact on airline safety’, Journal
of the Transportation Research Forum, XXVIII(1), 3–12.
Oster, C., J. Strong and C.K. Zorn (1992), Why Airplanes Crash, New York: Oxford
University Press.
Rose, N. (1990), ‘Profitability and product quality: economic determinants of
airline safety performance’, Journal of Political Economy, 98(5), 944–64.
Savage, I. (1999), ‘The economics of commercial transportation safety’, in J.
Gomez-Ibanez et al. (eds), Essays in Transportation Economics and Policy: A
Handbook in Honor of John R. Meyer, Washington, DC: Brookings Institution
Press.
Stolzer, A.J. and J.J. Goglia (2016), Safety Management Systems in Aviation,
London: Routledge.
Yadav, D.K. and H. Nikraz (2014), ‘Implications of evolving civil aviation safety
regulations on the safety outcomes of air transport industry and airports’,
Aviation, 18(April), 94–103.
INTRODUCTION
220
Detailed reliable air traffic data covering all of the world’s commercial air
transport activity was sourced from the Official Airline Guide. The OAG
Max Historical Plus databases give ex post daily schedules of all of the
commercial air transport services offered for sale. Using these databases,
annual traffic capacity data series were compiled and matched with popu-
lation data and basic geographic information on cities and countries. The
analysis presented in the chapter draws from these data series and presents
a supply-side view of air transport activity, since it is capacity rather than
actual passenger volumes that are analysed. The analysis is restricted to jet
services (including regional jets).
Passenger air transport activity, as measured by non-stop departure
movements, has grown at an annual average rate of 3.2 per cent through-
out the 1996 to 2015 period, and a 3.4 per cent average annual growth in
available seats, despite two major global recessions in 2001 to 2002 and
2008 to 2010 being observed. The deployment of significant fleets of small
regional jets during the 2000s, particularly in the North American market,
reduced the average aircraft size during this decade. In the more recent
period, average aircraft size has increased and the trend is particularly
influenced by the strong growth in the Asian market since 2008, where
typically larger aircraft are used.
To compare trends in the major global markets, a regional nomenclature
is utilized: the regional classification is IATA-based and does not align
with economic trading blocs. For example, Europe (EU) is defined as all
areas west of the Ural Mountains, so that Russia is split between Europe
and Asia (AS); North America consists of Canada and the USA but not
Mexico, which is included in Latin America (LA). The major regions are
further subdivided into smaller country groupings (for example, there are
four Latin American and four Asian sub-regions), but these will not be
presented in this chapter.
700
600
NUMBER OF AIRPORTS
500
400
300
200
100
0
1996 1998 2000 2002 2004 2006 2008 2010 2012 2015
AF AS EU LA ME NA SW
Considering traffic shares for the major continental regions over the period
from 1996 to 2015, the North American (NA) region share fell from 46 per
cent of global departure movements in 1996 to 29 per cent in 2015. The
Asian region share rose from 15 per cent in 1996 to 29 per cent in 2015,
while Europe slightly increased its share (22 per cent in 1996 to 24 per cent
in 2015), reflecting slightly higher growth than the average annual rates
cited earlier. The Middle East (ME) region increased its share from 2 per
cent in 1996 to 2.9 per cent in 2015, while Latin America (LA) and the
Southwest (SW) have maintained their shares.
Figure 12.1 shows the number of airports receiving passenger jet ser-
vices in each region over the analysis period. There are dramatic increases
in the number of Asian airports (from 447 to 655) and in the number
of European airports (from 419 to 553). For North America, there was
a significant drop in the number of airports receiving jet air transport
services after deregulation in the US in 1978, and this was sustained until
after 2010; since then there has been a significant increase in the number
of airports receiving scheduled jet services, from 361 in 2010 to 444 in
2015. Smaller increases are observed in the numbers of airports in Latin
America, Africa (AF), the Middle East and Southwest regions.
To examine the distribution of traffic across these continental systems
of airports over the analysis period, two approaches were used. The first
0.88
0.86
GINI INDEX SCORE [0,1]
0.84
0.82
0.8
0.78
0.76
0.74
0.72
1996 1998 2000 2002 2004 2006 2008 2010 2012 2015
AF AS EU LA ME NA SW
Figure 12.2 Gini Index scores for air traffic movements for each major
region, 1996–2015
approach uses the Gini Index, which gives a measure of the extent of con-
centration in a traffic distribution and summarizes the deviations from an
equal share traffic distribution. A score of 0 indicates an equal share across
all airport communities; a score close to 1 indicates a highly-concentrated
distribution focused on a relatively small number of communities. Using
traffic movement and seating capacity shares at each airport across the
system of airports in each year and for each region, the Gini Index was
computed and adjusted to take account of changes in the number of air-
ports receiving services (for a discussion on these adjustment factors see
Reynolds-Feighan, 2007). Figure 12.2 shows the Gini Index scores for the
regional movements traffic distributions over the analysis period.
The Gini scores range between 0.74 and 0.87, which reflects a high degree
of concentration in the traffic flows in all regions. The Gini index scores for
seating capacity are consistently about 2 per cent higher but show a similar
trend to the departure movement traffic trends shown in Figure 12.2,
reflecting the fact that larger aircraft are deployed at the largest airports
and therefore account for higher shares of seating capacity compared
with movements. Traffic concentration declined in the European, North
American and Asian regional markets, while concentration increased in
the Middle East, African, Southwest, and Latin American markets.
226
2 4 16 5 2 6 33 3 10 18 6 5 27 1 70
3 1 2 3 1 3 1 5 1 11
4 1 1 2 1 3
5 1 1 1 1
6 1 1
Total 238 443 380 318 75 324 94 1872 261 643 514 346 106 404 104 2378
14/11/2017 11:14
Small community impacts of liberalization 227
been an increasing average aircraft size at all classes of hubs and non-hubs
in all regions. Because of congestion at several of the large hubs, some
have reduced the numbers of short-haul or smaller aircraft services and
focused on the development of longer-haul traffic and facilitating larger
equipment. The medium and small hubs play an important role in provid-
Population Characteristics
The air traffic communities were linked to urban population data gathered
from the UN Population Division reports, the World Bank and from the
Tableau databanks.5 For smaller communities, internet searches were con-
ducted to find up-to-date urban population information.6 The average pop-
ulation of each category of hub is presented by region in Table 12.5 for 2015.
The large hubs generally serve large cities with populations exceeding 1.5
million. For the Asian and Latin American markets, the average large hub
city size is significantly greater than in other regions. In Asia, the medium
hubs are also very large urban centres, with an average population of 2.5
million. The relatively low propensity to travel by air in Asian markets is
associated with lower incomes, low rates of private consumption and a small
middle-class population. The leading industry forecasts produced by Airbus
and Boeing point to changes in these factors as the key drivers behind their
forecast annual growth rates of between 4.5 per cent and 4.8 per cent until
2035 (Airbus Industries, 2016; Boeing, 2016), with strong growth in private
consumption in Asia, particularly in the Chinese domestic market expected
over the forecast period. For Africa, it is anticipated that there will be 22
cities with populations of at least 4 million by 2025 (Airbus Industries, 2016).
While intraregional traffic has increased significantly in the last five years,
the propensity to travel by air is well below rates in other global regions. For
the North American market, the smaller average size of the hub populations
is an indicator of the high propensity to travel and higher incomes in this
region. The small hubs and non-hub communities have substantial popula-
tions, particularly in Asia, Latin America, the Middle East and Africa. By
contrast, the North American and Southwest non-hubs have populations of
under 100 000, and less than 30 000 in the case of the Southwest.
1996
Region City Pair No. City Pair Distance Percentage of Regional Number of Carriers
Airports (km) Traffic
M4371-FINGER_9781786431851_t.indd 229
AF Johannesburg Cape Town 2 1271 6.3 14
Durban Johannesburg 2 501 4.9 12
Benghazi Tripoli 2 669 1.6 1
AS Kaohsiung Taipei 3 301 2.7 8
Busan Seoul 2 336 1.2 4
Seoul Jeju 2 451 1.1 2
EU Dublin London 6 464 1.1 9
229
Rome Milan 5 498 1.0 13
Madrid Barcelona 2 483 1.0 11
LA Rio de Janeiro Sao Paulo 4 366 3.4 27
Guadalajara Mexico City 2 459 1.8 5
Mexico City Monterrey 2 713 1.3 6
ME Riyadh Jeddah 2 853 3.3 3
Dharan Riyadh 2 373 3.2 4
Bahrain Doha 2 145 1.8 10
NA New York Chicago 6 1166 0.8 17
Kahului Honolulu 2 163 0.7 6
Houston Dallas/Fort Worth 4 368 0.7 7
SW Sydney Melbourne 2 703 9.4 21
Sydney Brisbane 2 740 6.5 10
Wellington Auckland 2 480 4.4 2
14/11/2017 11:14
Table 12.4 (continued)
2015
Region City Pair No. City Pair Distance Percentage of Regional Number of Carriers
M4371-FINGER_9781786431851_t.indd 230
Airports (km) Traffic
AF Johannesburg Cape Town 3 1271 5.1 12
Durban Johannesburg 3 501 3.0 9
Cairo Jeddah 2 1217 3.0 9
AS Jeju Seoul 3 450 0.9 8
Sapporo Tokyo 4 801 0.6 9
Tokyo Fukuoka 4 909 0.6 8
EU Istanbul Izmir 3 336 0.5 9
230
Dublin London 7 463 0.5 9
London Amsterdam 7 341 0.5 11
LA Sao Paulo Rio de Janeiro 5 366 3.7 19
Brasilia Sao Paulo 4 841 1.5 10
Belo Horizonte Sao Paulo 5 505 1.4 7
ME Riyadh Jeddah 2 850 2.4 6
Dubai Doha 3 370 2.3 5
Kuwait Dubai 3 853 1.6 7
NA New York Chicago 12 1175 0.7 16
Boston New York 11 302 0.5 10
San Francisco Los Angeles 2 544 0.5 5
SW Sydney Melbourne 4 712 9.4 7
Brisbane Sydney 4 750 5.5 4
14/11/2017 11:14
M4371-FINGER_9781786431851_t.indd 231
Table 12.5 Average population of air transport communities for major regions and hub types in 2015
Region Large Medium Small Non-hub ‘A’ Non-hub ‘B’ Non-hub ‘C’ Average for
Region
AF 2 576 220 1 011 142 372 920 209 307 70 433 587 116
AS 7 749 732 2 681 584 85 734 439 460 314 056 124 408 962 677
231
EU 2 358 831 813 727 382 665 219 958 136 902 105 486 389 538
LA 3 921 544 849 151 427 915 183 557 194 455 64 580 598 274
ME 1 904 106 828 383 325 665 161 303 234 413 173 228 632 931
NA 1 333 028 461 868 158 670 68 781 48 144 90 624 199 773
SW 1 150 177 75 945 24 220 29 864 23 335 198 467
Average per hub type 3 104 298 1 131 526 456 197 267 258 187 128 92 019 566 948
14/11/2017 11:14
232 Air transport liberalization
Over the period 1996 to 2015, the North American, Latin American and
Southwest regions saw very little change in the number of carriers, includ-
ing those externally registered, serving the markets. There was an increase
in North America during the 2000s, but numbers have fallen again since
2008. The European market has seen the most dramatic change over the
period, with a rapid and substantial increase in the number of carriers
between 1996 and 2004/05. Since then, numbers have fallen equally dramat-
ically due to consolidation, code-sharing arrangements through alliance
partners and withdrawal by some international carriers. The Asian, Middle
East and African markets have seen increases in the number of carriers
since the mid-2000s, though numbers stabilized in the most recent period.
The numbers of city pairs departing from each region are shown in
Figure 12.3. Europe has by far the most extensive network of city-pair
routes and this has continued to increase over time. Many of the routes
have a low frequency of service, but facilitate very high levels of connectiv-
ity and accessibility compared to other regions. It can be noted that the
Asian route network is expanding significantly in the most recent period.
The North American market has a relatively small number of city pairs
that typically have high frequency of service. The route networks in the
African and Southwest markets are relatively small by contrast.
In terms of the average number of carriers operating on city-pair routes
for the different hub classes by region in 2010 and 2015, the large hubs in
18 000 50
Thousands
45
Number of city pairs per region
16 000
Total global city pair markets
14 000 40
35
12 000
30
10 000
25
8000
20
6000
15
4000 10
2000 5
0 0
1996 2000 2006 2010 2015
SW ME AF LA
NA AS EU Worldwide
Figure 12.3 City pairs served by region, and world city pair routes,
1996–2015
Region Year
1996 2000 2006 2010 2015
AF 71% 67% 63% 63% 67%
AS 58% 59% 55% 54% 56%
EU 61% 60% 62% 63% 61%
LA 58% 54% 59% 57% 55%
ME 64% 66% 64% 62% 62%
NA 62% 65% 66% 60% 60%
SW 55% 59% 63% 60% 62%
Worldwide 61% 61% 62% 60% 59%
all regions typically have two carriers operating on routes departing from
these communities. Asian markets have more carriers per route compared
with other regions, while the North American, Latin American, Middle
East and African markets have between 1.8 and 1.9 carriers per route from
their large hubs. As the community size gets smaller, the number of carriers
per route also declines, with most non-hub routes having just one carrier.
This trend is observed in all regions. The average number of carriers per
route has increased marginally over the 2009–15 period for medium and
small hubs and for non-hub types A and B communities.
The extent of competition on city-pair routes is further explored in
Table 12.6, where the percentage of all routes with just a single carrier pro-
viding service is presented, while Table 12.7 shows the percentage of depar-
ture movements on single carrier and two-carrier routes for 1996, 2000,
2006, 2010 and 2015. While most routes in every region are single-carrier
routes, these routes account for just below 20 per cent of all traffic world-
wide in 2015. The African market has the largest share of single-carrier
routes at 67 per cent in 2015, and this accounted for 28 per cent of move-
ments. The number of single-carrier routes has declined over time, as the
network of city pairs has expanded in every region. New routes tend to be
single-carrier routes initially, but as traffic expands, the number of carriers
increases. Air transport markets are generally becoming more competitive,
particularly those routes operating from large hubs.
The simple average stage length for city-pair routes from each type of
hub and non-hub was computed for each period, and gives an indication of
the range of services operated from each hub type and is not weighted by
the share of movements. Ignoring the small number of type-‘C’ non-hubs
in each region, as the hub size increases, the average route stage length
increases.7 The Southwest region has the highest average stage length, as
many of the long-haul services operated from the large hubs in Australia
and New Zealand connect to Europe, North America and Asia. By con-
trast, in other regions there are high numbers of short- and medium-haul
routes operating from the large hubs. The small hub and non-hub commu-
nities in most regions have shorter average stage lengths.
The growth of traffic movements was examined across hub classes and
average changes are presented for the 2009 to 2015 period in Table 12.8.
The number of communities losing air service and the number of new
Table 12.8 Average percentage change in departure movements from 2009
to 2015, and number of non-hub communities losing service
and receiving new service between 2009 and 2015
communities receiving service are recorded in the table. The large hubs
experienced strong growth in all regions over the period, and averaged
24.5 per cent increase in movements, with the North American market
experiencing the lowest growth. The Asian, Latin American and Middle
East regions had the highest growth. Growth was even stronger in these
regions at the medium and small hubs. The non-hubs in all regions had a
more varied experience. For European, Latin American and Asian non-
hubs type A, very strong traffic growth was recorded and many new com-
munities began receiving air services over the period (402 in total). Over
the period 111 non-hub communities lost jet air services, and these were
distributed across all regions.
markets where such services may help to drive the development of new
tourism products in the Least Developed Countries (LDCs) (ICAO, 2005).
The criteria set out in the framework are based on analysis of social air
service provision in developed-country markets. In their study, ICAO and
WTO set out a template for the design of social air service schemes under
the following criteria:
United States
Canada
Canada does not operate an essential air service programme, but funds 13
remote airports through the National Airports Policy. Since the mid-1990s,
the federal government has moved to transfer ownership and operation
of regional and local airports to locally based authorities who take on the
responsibility for funding their maintenance (see Metrass-Mendes et al.,
2011 for a more detailed description of the process). At the same time,
the nationally funded Airports Capital Assistance Program funds projects
aimed at protecting the airport assets at about 200 regional and local
airports. A consultant report in 2015 to the Canadian Assembly recom-
mended against introducing an essential air service process (and thereby
subsidizing air carrier operations) in Northern Canada (RP Erickson &
Associates, 2015).
Australia
The Remote Air Service Subsidy Scheme (RASS) was introduced in 1983,
though the Australian government had been subsidizing air transport ser-
vices to remote regions since 1957. It is part of the federal government’s
European Union
fully commercial operations; and (3) start-up aid to airlines to launch new
air routes with the aim of increasing the connectivity of a region. These
guidelines regularize and update the 2005 guidelines introduced to harmo-
nize the public financing of airports and new route development funding.
Member states can also devise their own initiatives, as long as they
comply with EU regulations. For example, the UK government intro-
duced its Regional Air Connectivity Fund12 in November 2014 with a
view to expand UK regional airport routes within the UK and Europe,
while avoiding linkages to the large London airports because of capacity
constraints. These new routes are expected to be commercially viable after
three years, and 15 routes were initially selected.
India
Under the 1992 Route Dispersal Guidelines (RDG), the Indian Ministry
of Civil Aviation set out requirements for carriers to distribute their capac-
ity across three different categories of routes and thus cross-subsidize
air services to small and more remote communities. The 2016 National
Civil Aviation Policy rationalized the RDG policy and categorization of
routes, and will take effect from 2017. The route categories at present are:
(1) 12 high-density (Category I) routes (heavily trafficked routes between
the major urban centres of Mumbai, Kolkata, Hyderabad, Bangalore,
Trivandrum and Chennai); (2) routes to more remote parts of the country
in the Northeast, Jammu and Kashmir and island territories (Category
II routes); (3) routes within Northeast India and Jammu and Kashmir
(Category IIA); and (4) all other routes (Category III). The guidelines
require that 10 per cent of a carrier’s Category I capacity be deployed
on Category II routes, 1 per cent on Category IIA and 35 per cent on
Category III routes.13 This policy has forced carriers to connect small and
remote communities within their regions and to the main urban centres.
last ten years. In 2012 the Brazilian government set out plans to build or
adapt 70 airports for commercial use, as part of a strategic infrastructure
programme (Logistics Investment Program PIL). The plan envisaged an
investment of US$2 billion to develop a network of airports that would
serve remote regions (with the support of government funding) as well
as larger cities and tourist destinations (where private investors and con-
cessionaires would be sought). The plan aims to have 96 per cent of the
Brazilian population living within 100 km of an airport. In 2015, the
plan was renewed and extended to fund the development of regional hubs
through the concession of four state airports. An additional six state air-
ports were expected to be concessioned to the private sector in 2016 with
either 20- or 30-year terms (USTDA, 2015). Couto et al. (2015) described
the national network structure characteristics and identified regional
sub-networks with relatively low traffic volumes and connectivity. In
January 2015 the government enacted legislation to set up the Program
of Development of Regional Aviation (Programa de Desenvolvimento
da Aviação Regional – PDAR). This act relates to airports with less than
600 000 annual passenger throughput (embarking and disembarking),
which are designated as regional airports for the purposes of the pro-
gramme, with the exception of airports located in the Amazon, where the
passenger threshold is 800 000. There were 689 local and regional airports
in Brazil in 2016. The aims of the PDAR are to:
I – increase access of the population to the air transport system, with priority
to those living in less developed regions of the country, considering both the
increase in the number of municipalities and routes served by scheduled air
transport;
II – integrate isolated communities to the national civil aviation network to
facilitate the mobility of its citizens; and
III – facilitate access to areas with tourist potential, subject to the provisions of
section I. (Câmara Dos Deputados, 2015)
CONCLUSIONS
Air transport activity has grown steadily in all global regions over the last
two decades, and is expected to continue this growth trend as Asian, Latin
American and African regions develop economically. The urbanization of
the world’s population is leading to the spatial concentration of people
and of economies in relatively small spaces, and these act as anchor points
for the global air transport system. Changes in domestic and international
regulations governing air transport activities have led to a more market-
oriented industry and approach to providing air services.
The number of air transport communities has expanded substantially,
with the airports at the world’s largest population centres handling the
majority of air transport activity. Medium and small communities are
enjoying increasing levels of service with jet aircraft that enable them to
connect to national and international centres. The size of communities
with regular jet air services is still substantial in most regions, so that
many communities with populations under 100 000 do not have services.
For these communities and for much smaller and more remote locations,
social air service policies have been devised to enable access to national
air transport systems, and it can be expected that these kinds of policies
will be deployed in developing countries over the next couple of decades.
Based on experiences to date, these programmes can provide cost-effective
means of enabling accessibility for small communities, many of which may
become commercial over time. The approach to social air service provi-
sion makes explicit the need for cross-subsidization and for government
intervention to identify and provide transportation services in certain
circumstances.
NOTES
1% or More Large
At least 0.25%, but less than 1% Medium
At least 0.05%, but less than 0.25% Small
More than 10 000, but less than 0.05% Non-Hub Primary
At least 2500 and no more than 10 000 Non-Hub Non-Primary
Primary Airports are Commercial Service Airports that have more than 10 000 passen-
ger boardings each year. Hub categories for Primary Airports are defined as a percent-
age of passenger boardings within the United States in the most current calendar year
ending before the start of the current fiscal year.
REFERENCES
Although oligopoly is often associated with the airline industry, the indus-
try is characterized by many firms, different business models and a strong
international dimension. For this reason, any analysis must answer the
question of whether the industry has strong or weak oligopolistic charac-
teristics, and if these characteristics differ between domestic and interna-
tional markets and influenced by strategic groups.
Since deregulation of the US airlines in 1978 there have been several
waves of mergers in the US airline industry, usually followed by concerns
about intensified oligopolization in the market. Underpinning these con-
cerns is the common definition of an oligopoly. Practitioners have sug-
gested a rule of thumb that says if four or fewer companies have 50 per
cent or more of a market then there is effectively an oligopoly. Using this
benchmark for the US domestic market, the top four airlines had 60.2
per cent passenger market share in 2015 (Southwest Airlines 20.4 per
cent, Delta Air Lines 16.5 per cent, American Airlines 13.4 per cent and
United Airlines 9.9 per cent), telling us that indeed the industry has the
characteristics of an oligopoly. However, in the US international market
the story is altogether different, with the top four airlines having only 39.9
per cent market share (United Airlines 12.7 per cent, American Airlines
12.4 per cent, Delta Air Lines 11.3 per cent, and British Airways 3.5 per
cent).
Nonetheless, simple rules of thumb and even more sophisticated con-
centration measures can be misleading as to the true state of competition
intensity in an industry. Different strategic groups in an industry may
behave differently within the group and across the groups. In this context
it is important to point out that the leader in the US domestic market
is a low-cost airline that has grown slowly but surely since deregulation
(Gudmundsson, 1998), while the other three are legacy carriers that have
maintained their position mostly through mergers and acquisitions.
244
STRATEGIC GROUPS
INTERDEPENDENCE
legacy airlines follow the golden rule, avoiding price changes in one route in
order to avoid competitors’ retaliation in other contested routes. However,
this relationship does not necessarily apply across strategic groups.
Southwest Airlines has a large proportion of monopoly routes but
refrains from charging monopoly prices for two reasons: first, the need to
create new demand for its services; second, to compete with other trans-
port modes such as cars and buses, rather than with other airlines. For
this reason, a price change by a dominant airline in a nearby hub airport
is unlikely to affect Southwest’s pricing on a nearby point-to-point route.
In support of this hypothesis, Brueckner et al. (2013) showed that legacy
airline competition had weak effects on average fares, while low-cost airline
competition, no matter how it occurs, had major impacts on legacy car-
riers’ average fares, implying different behaviour of strategic groups in
oligopolistic markets.
New entry and exit to the industry has fluctuated considerably and has
been subject to various internal and external conditions. In an attempt
to explain some of the influencing factors, Gudmundsson and Van
Kranenburg (2002) studied entry rates into the US airline industry. They
found that political pro-entry interventions and announcements of inter-
ventions were significantly related to entry rates. Other important positive
factors were reduction in acquisition and leasing costs of aircraft, and the
rate of GDP growth.
To give an idea as to the entry and exit dynamics in the industry, 176
operating certificates for jet services were issued in the US from 1979 to
1999; of these, 123 (70 per cent) started operations and 37 (21 per cent)
airlines were still operating in 1999 (Gudmundsson and Van Kranenburg,
2002). Similarly, ATA (2008) examining exit (Chapter 11 and 7 bankrupt-
cies) of air carriers from 1979 until 2008, reported that 185 carriers had
disappeared.2
Thus, we can talk of two sources of concentration, namely mergers and
acquisitions and exit. However, if entry of new airlines to the industry is
taking place, it acts to some extent as a countervailing force against con-
centration. At the same time, industry concentration and new entry have
occurred in waves. In other words, natural market equilibrium appears to
be non-existent (Button, 1996). Thus, the common assumption that only
a few airlines will survive in the market is not an inevitable conclusion of
market deregulation unless the focus is on the legacy carriers, but exclud-
ing new entrants (Swan, 2005).
PRICING IMPLICATIONS
Airfares in the US have fallen in constant dollars since the start of deregu-
lation. However, annual change in airfares has fluctuated considerably over
the same period (see Figure 13.1). The average fare decreases over longer
periods only give us a bird’s-eye view, as concentration in the industry may
penalize certain market segments and benefit others; it may also change
pricing behaviour of the oligopolist. Finally, temporal fare effects may take
place with short periods of increases following waves of concentration
without changing the overall downward trend in average fares, assuming
there is no long-run stability following periods of concentration.
In other words, the fluctuating concentration index in the industry causes
temporary benefits at best when the index is high; these diminish gradu-
ally as efficiency-driven mergers and acquisitions open space for emerging
airlines to gain a foothold in new markets while unprofitable routes are
shed by the merged airlines trying to extract merger benefits. As previously
mentioned, this relationship should hold for merger-driven concentration,
M4371-FINGER_9781786431851_t.indd 251
Delta Air Lines / Pan Am Airlines (Shuttle and Atlantic Routes) 1991 L/L
Air Wisconsin / Aspen Airways 1991 R/R
United Airlines / Pan Am: Latin and Caribbean Routes 1992 L/L
Southwest Airlines / Morris Airlines 1993 LC/LC
AirTran Airways / Valujet 1997 LC/LC
American Airlines / Reno Air 1999 L/LC
Delta Air Lines / Atlantic Southeast Airlines 1999 L/R
Delta Air Lines / Comair 1999 L/R
251
American Airlines / TWA 2001 L/L
US Airways / America West 2005 L/LC
Republic Airways / Shuttle America 2005 L/R
SkyWest / Atlantic Southeast 2005 R/R
Pinnacle Airlines / Colgan Air 2007 R/R
Delta Air Lines / Northwest 2009 L/L
United Airlines / Continental 2010 L/L
Southwest Airlines / AirTran 2010 LC/LC
US Airways / American 2013 L/L
14/11/2017 11:14
252 Air transport liberalization
decrease dramatically both at the low end and the high end of the fare
dispersion (Tan, 2016). Thus, there is evidence that the proportional
pricing impact of smaller players in the industry – most of which are
low-cost or low-fare airlines – is proportionally more important than
their market share indicates, due to the strong reaction of legacy carriers,
but not vice versa.
Thus, concentration among low-cost airlines versus legacy airlines
may have markedly different impact on airfares. Dobson and Piga (2013)
argued that the EasyJet–Go Fly and Ryanair–Buzz acquisitions had
impacted pricing on the affected routes. In fact, they showed that early-
booking fares decreased, although late-booking fares increased. In other
words, the acquisitions had a net benefit for price-conscious consumers
who booked early. Conversely, research focusing on mergers among legacy
airlines in the US showed that mergers, on average, caused short-run fare
increases of 2.3 to 5.9 per cent (Jain, 2015). Although few studies have been
performed on the difference in price effects between legacy carrier mergers
and low-cost airline mergers, it nevertheless raises the question of whether
the pricing effects are different depending on strategic groups.
Figure 13.1 shows the general trend in airfares in the US domestic
market from 1995 to 2016. The figure clearly shows that, on average, fares
increased substantially during the financial crisis period in 2009, but then
began to decline again (3.8 per cent in 2015). Although a global measure
like this does not unravel all the nuances of individual route impacts, it is
nevertheless an indicator of the general trend in the industry.
Jain (2015) showed that although fares increased following mergers, such
effects are just a temporary halt in the long-run downward trend. Thus,
fluctuation in industry concentration does not have a negative long-run
impact on average price trends. On the contrary, the industry shows sub-
stantial decrease in long-run average airfares despite periods of concen-
tration. This being said, the price impact of concentration on individual
routes may differ based on the close-substitutes hypothesis (Lijesen, 2004).
In this regard, close substitutes constitute two routes that exhibit high
cross-elasticity of demand.
There is another unique airfare related development in air transport
markets, namely the increased separation of the ticket price from growing
ancillary revenues. For example, in 1990 fares composed 88.5 per cent of
the airlines’ total revenue from passenger transport, whereas in 2015 this
component represented only 74.7 per cent and 73.8 per cent in 2016 (BTS,
2016). For this reason, the fare component is likely to continue to decrease
in the long run as airlines develop more sophisticated sources for ancillary
revenues.
500 8.0
450 6.0
400 4.0
350 2.0
300 0.0
250 –2.0
200 –4.0
150 –6.0
100 –8.0
50 –10.0
0 –12.0
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Average Fare ($) Annual Change (%) Linear (Average Fare ($))
Note: 2016 constant $ using Bureau of Labor Statistics Consumer Price Index.
MEASURING CONCENTRATION
of these measures, the following section will cover the most important
methods and their limitations.
The most notable industry concentration measures are the Lorenz–Gini
index, the concentration ratio (C4) (see Finkelstein and Friedberg, 1967)
and the Herfindahl–Hirschman Index (HHI) (Hirschman, 1969). The
latter has generally replaced the former two, although the concentration
ratio is considered useful to measure concentration, and the Lorenz–Gini
index useful to measure industry dispersion. The Lorenz–Gini approach
works well when applied to a large number of observations, but is less
practical for industry concentration as it only measures disparity in market
shares and fails to take into account the size of a firm’s market share
(Roberts, 2014). However, the industry concentration ratio (CR) provides a
more straightforward measure of concentration, making up for this short-
coming. The CR is expressed as follows:
CR m 5 a si
m
i51
where si is the market share and m defines the mth firm in the industry.
Using the CR, cut-off values were established as between 0 and 50 per
HH805toa100
cent market share for a low-concentration N industry, 50 to 80 per cent for
medium concentration, and from s2i 3per100cent for high concentra-
i51
tion. Since the CR has been criticized for revealing little about the inequal-
ity among the size distribution of firms, researchers sought other measures
a si a(1969),
to aid analysis for anti-trust court cases.
m
n
One solution was suggested CR mby5Hirschman f (i) jt,k
who further devel-
i51
j i51
oped the work of Herfindahl, SL t,k 5 1
culminating 2 in the HHI. The HHI has
nt,k
become the standard in applied concentration analysis, and is expressed
as follows:
i51
where si is the market share of firm i in the market, and N is the number
0B 0B 0B
M4371-FINGER_9781786431851_t.indd 255 , 0, , 0,and s
,0 14/11/2017 11:14
256 Air transport liberalization
Overall
500 US Domestic and International HHI 8.0
450 6.0
Using the DoJ and FTC definition of industry concentration, the US
400 4.0
airline industry is competitive both in the domestic and the international
markets.
350 Figure 13.2 shows that the HHI index (HHI-DOM) for domestic 2.0
air transport
300
in the US ranges from 1026 in 1990 to 1061 in 2015. However,
0.0
both the beginning (1990) and the end of the period (2015) have high HHI
250
compared to 2007 when the HHI went down to 673. Using the HHI–2.0 cut-off
values
200we can see that the domestic US industry has shown moderate –4.0 con-
centration
150
in two periods: until 1992, and following the Financial –6.0 Crisis in
2008. For the rest of the period since deregulation, the domestic industry
has 100
shown an unconcentrated index. –8.0
The50 level of concentration is even lower when examining the –10.0HHI
index for international air transport (HHI-INT) from and to the US,
0 –12.0
ranging from 414 in 1990 to 518 in 2015, with the highest index values
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
in 2012, at 564. In other words, the industry – according to the HHI
index forAverage
the international
Fare ($) marketChange
Annual – is (%)
far fromLinear
being concentrated.
(Average Fare ($))
However, the international trend line is rising whilst the trend line for
domestic concentration is falling, regardless of a substantial rise in the
HHI-DOM from 2009 until now. In other words, although the fluc-
tuations in the HHI-DOM are greater, the HHI-INT is more stable but
gradually rising.
1400
1200
1000
800
600
400
200
0
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
US Domestic HHI
Table 13.2 and Figure 13.3 show clearly the remarkable rise of the low-cost
airlines in the US domestic industry (HHI-LC10 and HHI-LC4), effec-
tively challenging the legacies (HHI-L10 and HHI-L4). However, in the
latter part of the period the legacies (HHI-L10) have somewhat increased
the concentration through mergers and acquisitions, although far from the
1990 level (HHI-L10 in 1990 5 973; HHI-L10 in 2010 5 433; HHI-L10
in 2015 5 571). What is clear is that the concentration in the early 1990s
is different from today, due to a stronger position of low-cost carriers and
Southwest Airlines in particular (HHI-LC10 in 1990 5 42; HHI-LC10 in
2015 5 466). What is more, based on the past trend Southwest is likely to
increase its market share due to its sustainable competitive advantage over
the legacy carriers, while the latter are likely to maintain their position
through mergers and acquisitions.
US International HHI
Table 13.3 and Figure 13.4 show the HHI values of US carriers and inter-
national carriers in the top four (HHI-4, only one international carrier is
present in this group), top 10 (HHI-10) and top 20 (HHI-20) carriers in
the international market. It is clear that US carriers have increased their
share in this market (HHI-US10 in 1990 5 299; HHI-US10 in 2015 5 451),
M4371-FINGER_9781786431851_t.indd 259
50.00% 600
40.00%
30.00% 400
20.00% 200
10.00%
0.00% 0
1990 2000 2010 2015 1990 2000 2010 2015
(a) Top 4 Top 10 (b) HHI-4 HHI-10
90.00% 1200
259
80.00%
1000
70.00%
60.00% 800
50.00%
600
40.00%
30.00% 400
20.00%
200
10.00%
0.00% 0
1990 2000 2010 2015 1990 2000 2010 2015
(c) Legacy-4 LC-4 Legacy-10 LC-10 (d) HHI-L4 HHI-C4 HHI-L10 HHI-LC10
14/11/2017 11:14
260 Air transport liberalization
CONCLUSION
In this chapter, concentration was measured by separating domestic and
international markets, revealing different portrayals of the level of con-
centration. In the calculations, HHI market share for passenger traffic was
used (instead of the usual available seat-miles or total revenues) as a more
viable method to account for different business models of low-cost and
legacy carriers.
M4371-FINGER_9781786431851_t.indd 261
20.0%
100
10.0%
0.0% 0
1990 2000 2010 2015 1990 2000 2010 2015
60.0% 500
50.0%
261
400
40.0%
300
30.0%
200
20.0%
10.0% 100
0.0% 0
1990 2000 2010 2015 1990 2000 2010 2015
Top 10 US Top 10 International HHI-US10 HHI-INT10
(c) Top 20 US Top 20 International (d) HHI-US20 HHI-INT20
14/11/2017 11:14
262 Air transport liberalization
NOTES
1. American Airlines’ Crandall says, ‘Airlines make up world’s most inefficient oligopoly.’
Marketing News, San Diego, 23 April 1976, p. 9.
2. At the dawn of EU liberalization (1993 to 1999) there were 131 start-up airlines in the
15 EU countries; of these, 74 had disappeared by 1999 (77 per cent failure rate). Of
124 incumbent airlines operating at the beginning of 1993, 55 had disappeared by 1999
(44 per cent failure rate) (Lyons, 2000). When market shares are examined between the
incumbent carriers and low-cost carriers, we see that in Europe, between 2003 and 2006,
low-cost carriers’ market share increased by 40 per cent to 20 per cent overall share, but
the incumbent airlines lost 10 per cent, bringing it down to 55 per cent overall market
share. Another parallel development, stimulated by the advent of low-cost airlines, was
the reduction in premium intra-EU passenger traffic that declined by 33 per cent between
2001 and 2005 (AEA, 2007).
REFERENCES
Air Transport Association of America (ATA) (2008), ‘Airline bankruptcy and
service sessations’. Available at: www.airlines.org/economics/specialtopics.
Alderighi, M., A.A. Gaggero and C.A. Piga (2015), ‘The effect of code-share
agreements on the temporal profile of airline fares’, Transportation Research
Part A: Policy and Practice, 79, 42–54.
Association of European Airlines (AEA) (2007), ‘Operating Economy of AEA
Airlines’, December.
Bachman, J. (2016), ‘Where are all the startup airlines to save you money?’,
Bloomberg Technology News, 20 April.
Bailey, E.E. and J.C. Panzar (1981), ‘The contestability of airline markets during
the transition to deregulation’, Law and Contemporary Problems, 44(1),
125–45.
Bain, J.S. (1954), ‘Economies of scale, concentration, and the condition of entry in
twenty manufacturing industries’, The American Economic Review, 44(1), 15–39.
Borenstein, S. (1992), ‘The evolution of U.S. airline competition’, Journal of
Economic Perspectives, 6(2), 45–73.
Brueckner, J.K. (2001), ‘The economics of international codesharing: an analysis of
airline alliances’, International Journal of Industrial Organization, 19, 1475–98.
Brueckner, J.K., D. Lee and E.S. Singer (2013), ‘Airline competition and domestic
US airfares: A comprehensive reappraisal’, Economics of Transportation, 2(1),
1–17.
Bureau of Transportation Statistics (BTS) (2016), Airfares P-1.2.
Button, K. (1996), ‘Liberalising European aviation: is there an empty core
problem’, Journal of Transport Economics and Policy, 30(3), 275–91.
Caves, D.W. (1962), Air Transport and its Regulators, Cambridge, MA: Harvard
University Press.
Caves, D.W., L.R. Christensen and M.W. Tretheway (1984), ‘Economies of density
versus economies of scale: why trunk and local service airline costs differ’, The
Rand Journal of Economics, 15, 471–89.
Caves, R.E. and M.E. Porter (1977), ‘From entry barriers to mobility barriers: con-
jectural decisions and contrived deterrence to new competition’, The Quarterly
Journal of Economics, 91(2), 241–61.
Dobson, P.W. and C.A. Piga (2013), ‘The impact of mergers on fares structure:
evidence from European low-cost airlines’, Economic Inquiry, 51(2), 1196–217.
Edwards, C.D. (1955), ‘Conglomerate bigness as a source of power’, in the NBER
conference report, Business Concentration and Price Policy, Princeton, NJ:
Princeton University Press.
Evans, W.N. and I.N. Kessides (1994), ‘Living by the “golden rule”: multimarket
contact in the US airline industry’, The Quarterly Journal of Economics, 109(2),
341–66.
Swan, W.M. (2005), ‘Consolidation in the airline industry’, working paper, accessed
10 September 2016 at www.sauder.ubc.ca/. . ./consolidation%20paper.ashx.
Tan, K.M. (2016), ‘Incumbent response to entry by low-cost carriers in the US
airline industry’, Southern Economic Journal, 82(3), 874–92.
Tan, K.M. and A. Samuel (2016), ‘The effect of de-hubbing on airfares’, Journal of
Air Transport Management, 50, 45–52.
US Department of Justice and the Federal Trade Commission (DoJ and FTC)
(2010), Horizontal Merger Guidelines, 19 August.
INTRODUCTION
266
number of about ten airports were heavily capacity constrained, that is,
during most daytime hours, in 2008 (the year before the global economic
downturn) there was a high degree of average capacity utilization among
the largest 100 airports, which accounted for more than 50 per cent of
flights globally. About 15 per cent of all flights were operated under
capacity-constrained conditions (Gelhausen et al., 2013). Key measures
to mitigate capacity constraints at airports include increasing capacity by
adding new runways, reorganization of traffic either to off-peak times or
less-congested airports and using aircraft with higher seat capacity. The
most straightforward option is investing in additional runways to increase
airport capacity. However, in particular at large airports this requires
lengthy planning procedures because of the involvement of the public,
who are typically opposed to airport capacity enlargements, especially
in highly developed countries. These planning procedures can take up to
between 10 and 20 years at large hub airports, and in some cases adding
new runways to an airport is an insurmountable task. Examples include
Frankfurt and Munich airports and London Heathrow. Long-term capac-
ity bottlenecks are therefore to be expected, especially at some large hub
airports (Gelhausen, 2013).
The other three options rely more on airline behaviour to mitigate the
effects of airport capacity shortages. However, reorganizing traffic to off-
peak times might conflict with the airlines’ wave system of transfer con-
nections and at some airports, like London Heathrow, there is a shortage
of available slots during most daytime hours. Moving flights with a high
share of transfer passengers to less congested airports is unfavourable
from the viewpoint of economics of density, while this might be an option
in the case of flights dominated by OD passenger volumes. Employing
larger aircraft is a trend that has been observed for a while at both large
and small airports. However, employing larger aircraft is not only caused
by airport congestion, but also – and mainly – because average flight
length has increased over time. Here, technical and economic reasons play
an important role. Yet this trend threatens airports of low-demand routes
(Berster et al., 2015).
The compound annual growth rate of global flights over the period was
1.70 per cent, which means that flights grew on average by 1.70 per cent
a year. However, if we split the period into two sub-periods (2000–07 and
2007–14) we detect a rapid decline of the compound annual growth rate:
while the compound annual growth rate is 2.44 per cent from 2000 to 2007,
it decreases to only 0.97 per cent from 2007 to 2014. Looking at the cor-
responding annual growth rate of global gross domestic product, we find
a close relationship: compound annual growth rate of global GDP being
2.79 per cent from 2000 to 2014, rises to 3.43 per cent from 2000 to 2007
and then decreases to 2.15 per cent from 2007 to 2014 (World Bank, 2016).
Turning to the traffic distribution between world regions in terms of
take-offs, North America was still the largest market in 2014, but its global
share fell throughout from 48 per cent to 31 per cent. This was mainly
due to a stagnating US market. On the other hand, the share of the Asian
market more than doubled: in 2000, 11 per cent of global take-offs were
handled at Asian airports, but this share increased to 18 per cent in 2007
and 25 per cent in 2014. The third-largest market in 2014 was Europe,
and its share fluctuated between values of 23 per cent (2000) and 25 per
cent (2007). The remaining four regions had a fairly small share of global
flights; however, the airports in Africa and in particular the Middle East
showed a continued positive traffic growth.
Considering the compound annual growth rate for take-offs by world
region between 2000 and 2014, as well as the sub-periods 2000–07 and
2007–14, with the exception of the rather small market of South America,
we find a stagnating or a decreasing growth rate across all world regions.
While growth rates remain positive in four out of seven regions from
2007 to 2014, we find a negative trend in Europe and North America.
Correspondingly, there is weak economic growth in North America and
the European Union: while GDP grew by 2.45 per cent and 2.29 per cent a
year on average during the period from 2000 to 2007, these rates fell to 1.01
per cent and 0.12 per cent a year for 2007 to 2014.
In summary, air traffic growth was strong in most world regions from
the year 2000 until 2007; however, growth rates have declined considerably
since then. This appears in many cases to be due to weak economic growth
since the global financial crisis in 2007, as GDP is typically a main driver
of air traffic development (IATA, 2008). In particular, well-developed
markets such as North America and Europe suffer from restricted growth,
while emerging markets such as Asia and the Middle East are still going
strong, which again coincides with economic development. However, in
the case of the US market a number of airline mergers due to economic
problems also play an important role in explaining the traffic development
there.
Figure 14.1 shows the Lorenz curve of global take-offs for 2000, 2007 and
2014. Airports are ranked in descending order relating to their number of
flights in the appropriate year and their shares are cumulated so that the
last value ends up at 100 per cent. The curve is based on more than 3000
airports for 2000, 2007 and 2014 that have at least one scheduled flight
a year. Concentration of traffic at very large airports, and especially at
hubs, is large: in 2000 there were 3076 airports and the value of the Gini
coefficient was 0.807, in 2007 with 3258 airports it was 0.815, and in 2014
with 3349 airports it was 0.812. Thus, traffic concentration increased only
marginally on a global scale. The largest four airports in 2014 were in the
US (Chicago O’Hare, Atlanta Hartsfield-Jackson, Dallas/Fort Worth and
Los Angeles), followed by Beijing Capital airport. The largest European
airport is London Heathrow, in ninth place. The world’s 30 largest air-
ports handled more than 21 per cent of the global take-offs, and 50 per
cent of the flights globally were handled by the 120 largest airports in
2014.
If we limit the sample to airports that have more than 1000 take-offs
100
90
80
Share of global take-offs (%)
70
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100
Share of airports (%)
0.86
0.84
0.82
y = 4E-05x + 0.7673
R2 = 0.0841
0.80
0.78
0.76
0.74
0.72
0 100 200 300 400 500 600 700 800 900 1000
Pacific region, although traffic has decreased over the past 14 years. We
now briefly discuss the Lorenz curve for each world region.
The Lorenz curve for North America comprised 880 airports for 2014,
which is the highest number of airports and, with a Gini-coefficient of
0.847, it exhibits the greatest degree of concentration of any world region.
However, concentration increased little between 2000 and 2014, although
take-offs decreased by 1.30 per cent a year. South America, which had 454
airports in 2014 and is one of the smaller world regions, has the lowest
degree of concentration of any of the seven world regions, with a Gini-
coefficient of 0.734. Concentration increased between 2000 and 2014, and
especially before 2007. Furthermore, take-offs increased by 1.75 per cent a
year between 2000 and 2014, and even by 3.31 per cent a year from 2007 to
2014. However, concentration increased less between 2007 and 2014. The
622 European airports in 2014 was the third-largest number in the world
and, with an associated Gini-coefficient of 0.780, was in the middle of the
global range. This would seem to reflect the high degree of airline allianc-
ing in the region and large LCC market. The airline alliances prefer hub
airports for their operations but LCCs tend to choose regional airports,
although they are increasingly offering flights from larger or some hub
airports. As a result, concentration remained roughly constant between
2000 and 2014. There has been a significant rise in take-offs, however, of
3.58 per cent a year between 2000 and 2007, although this stagnated over
the next seven years.
There were 684 airports in Asia in 2014, making this the second-largest
of the global regions. The degree of concentration is akin to Europe’s with
a Gini-coefficient of 0.789, with concentration increasing between 2000
and 2007 as traffic grew annually by 10.31 per cent, although the concen-
tration declined in subsequent years, with growth in take-offs falling to
6.13 per cent a year.
The Lorenz curve for the Middle East, which with just 104 airports is by
far the smallest, reflects a mid-range concentration with a Gini-coefficient
of 0.805. Concentration increased between 2000 and 2007, which was
accompanied by significant annual traffic growth of 6.37 per cent, and
even greater concentration increased thereafter, with traffic growth of 6.63
per cent a year.
The Southwest Pacific had 298 airports for 2014, the second-smallest in
the world. However, associated with a fall of 0.18 per cent a year in take-
offs between 2000 and 2014 concentration increased consistently, and for
2014 the region exhibited the second-largest Gini-coefficient (0.824).
Africa in 2014 comprised 307 airports and is similar in size to the
Southwest Pacific region. Its take-offs increased by 4.11 per cent a
year over the period from 2000, but the concentration remained almost
The process of airline liberalization has fostered two very different airline
business models: on the one hand, there are the full-service network carri-
ers (FSNCs) with a complex hub system, so that airlines enjoy economics
of density, making thin routes viable, especially in long-haul travel. On
the other hand, there are the LCCs with a less complex business model,
mainly serving point-to-point routes on denser routes. At the beginning
of liberalization, LCCs mainly used secondary and regional airports,
but in recent years they have also been serving larger airports as well
as hub airports that are closer to agglomerations. This development is
supported by the tendency of FSNCs to establish their own, largely inde-
pendent LCC subsidiary, such as Germanwings/Eurowings in the case of
Lufthansa, Vueling (International Airlines Group, IAG) and Transavia
(Air France/KLM). Membership in an airline alliance is a big competitive
advantage, giving airlines access to larger markets. Nevertheless, there
are still large carriers, such as Emirates, that do not belong to an airline
alliance.
In terms of these business models, while the shares of global take-offs
of FSNCs not belonging to any airline alliance decreased rapidly between
2000 and 2014, airlines of the three strategic alliances and LCCs grew
considerably. The share of regional and charter airlines decreased slightly
during this period.
The positioning and success of the carrier types vary across various
30
2000
2007
Share of take-offs at European airports (%)
25 2014
20
15
10
0
N t
ce
rld
er
t
na
FS den
gh
LC
ea
rt
an
C
wo
io
ei
ha
yt
n
lli
Fr
eg
ne
pe
Sk
C
A
R
O
de
ar
In
St
60
2000
2007
Share of take-offs at Asian airports (%)
50 2014
40
30
20
10
0
ld
N t
ce
er
t
na
FS den
gh
LC
r
ea
rt
an
C
wo
io
ei
ha
yt
n
lli
Fr
eg
ne
pe
Sk
C
A
R
O
de
ar
In
St
40
Share of take-offs at North American airports (%)
2000
35 2007
2014
30
25
20
15
10
0
t
ld
ht
na
FS den
rte
nc
LC
or
ea
ig
C
io
ha
lia
e
w
N
n
yt
eg
Fr
ne
pe
Al
C
Sk
R
O
de
ar
St
In
Rank Airport name IATA Code Region Take-offs CAGR CUI (2014)
M4371-FINGER_9781786431851_t.indd 278
(2014) (2000–2014)
1 Chicago O’Hare International ORD North America 446 213 –0.23% 0.66
2 Atlanta Hartsfield-Jackson International ATL North America 427 603 –0.23% 0.64
3 Dallas/Fort Worth International DFW North America 333 328 –1.47% 0.71
4 Los Angeles International LAX North America 299 946 –1.73% 0.69
5 Beijing Capital PEK Asia 292 313 9.22% 0.75
6 Denver International DEN North America 276 065 0.84% 0.60
7 Charlotte Douglas International CLT North America 257 002 1.93% 0.74
278
8 Houston George Bush Intercontinental IAH North America 243 278 0.48% 0.65
9 London Heathrow LHR Europe 236 295 0.13% 0.84
10 Frankfurt International FRA Europe 230 031 0.60% 0.70
11 Paris Charles de Gaulle CDG Europe 219 016 –0.70% 0.69
12 Amsterdam A MS Europe 211 795 0.62% 0.62
13 Tokyo Haneda HND Asia 211 649 3.87% 0.68
14 Istanbul Ataturk 1ST Europe 209 395 8.71% 0.75
15 New York J F Kennedy International JFK North America 208 964 0.82% 0.68
16 Shanghai Pudong PVG Asia 205 541 16.98% 0.40
17 San Francisco International SFO North America 203 656 0.03% 0.67
18 Guangzhou CAN Asia 203 159 9.51% 0.75
19 Toronto Lester B Pearson International YYZ North America 201 002 0.24% 0.62
20 Jakarta Soekarno-Hatta CGK Asia 200 371 10.45% 0.71
14/11/2017 11:14
M4371-FINGER_9781786431851_t.indd 279
21 Philadelphia International PHL North America 197 791 –0.52% 0.70
22 Detroit Wayne County DTW North America 191 413 –1.59% 0.60
23 Newark Liberty International EWR North America 190 801 –0.88% 0.70
24 Phoenix Sky Harbor International PH× North America 189 229 –1.64% 0.67
25 Minneapolis International MSP North America 188 906 –1.46% 0.57
26 Hong Kong International HKG Asia 188 539 5.69% 0.69
27 New York La Guardia LGA North America 187 098 –0.45% 0.76
28 Mexico City International ME× South America 185 964 2.22% 0.68
279
29 Munich International MUC Europe 177 107 1.66% 0.60
30 Kuala Lumpur International KUL Asia 173 575 4 8.83% 0.67
North America 4 042 295 0.665
South America 185 964 0.680
Europe 1 283 639 0.704
Asia 1 475 147 0.668
Top 30 airports 6 987 045 0.674
14/11/2017 11:14
280 Air transport liberalization
Reichmuth et al., 2011). Therefore, the CUI is the ratio of the mean hour
volume divided by the 5 per cent peak hour volume. Airports with CUI
values in the range of more than 0.65 to 0.70 can be considered as con-
gested already, while values of more than 0.80 indicate very high levels of
congestion. However, everything else being about equal, Asian airports
typically tend to have slightly higher CUI values compared with other
regions because traffic tends to be more evenly distributed.
Table 14.2 shows the world’s 30 largest airports with their number of
take-offs and CUI value for 2014, as well as the compound annual growth
rate between 2000 to 2014. The top 30 are dominated by North American
airports; however, while many North American and European airports
basically stagnated between 2000 and 2014, Asian airports grew strongly
during the same period. The top 30 airports have an average CUI value
(weighted by take-offs for 2014) of 0.67, along with the North American
and Asian airports (CUI values were rounded by two decimal places). The
European airports have on average a CUI value of 0.70, and the average
CUI value of South America only includes Mexico City International
airport. However, on the airport level, peripheral conditions play a major
role: for instance, London Heathrow and New York LaGuardia have CUI
values of 0.84 and 0.76. These are large airports close to large agglomera-
tions with typically high noise and pollution emissions. Frankfurt airport
opened the fourth runway in late 2011 and traffic has slightly decreased
since then, so that the capacity situation has eased. In 2012, Frankfurt
airport had a CUI value of 0.77.
Overall, the relatively weak traffic growth of recent years in Europe
during the Great Recession, but in particular in North America, has helped
to prevent an even tighter capacity situation at many large airports. On
the other hand, emerging regions such as the Middle East and Asia have
overall traffic levels that are still lower than North America and Europe
and airport expansion plans are easier to implement in case of capacity
bottlenecks (e.g. Beijing and Dubai airports). However, the capacity situ-
ation at some airports may worsen again if traffic growth rates increase in
the long term. Then again, hubbing helps to utilize overall airport capacity
more efficiently.
CONCLUSIONS
especially since 2007. This ties in with weak economic development glob-
ally, but especially in highly developed regions such as North America and
Europe. As a result, traffic growth has also differed markedly between
regions: while highly developed world regions grow only slowly or are
even decreasing, emerging regions such as Asia and the Middle East are
growing strongly. However, these are regions whose economies have also
developed quite dynamically in recent years. Nevertheless, the largest
market remained North America in 2014, but Asia replaced Europe as the
second-largest market in terms of number of take-offs in 2014.
Traffic concentration at airports remains at a very high level globally:
the value of the Gini-coefficient for 2014 is 0.812 and it only increased
marginally from 2000. There are some differences in traffic concentration
between different world regions; however, these differences are relatively
small: Gini-values ranged from 0.734 for South America to 0.847 for North
America. While concentration in highly developed regions remained
roughly the same between 2000 and 2014, concentration tended to increase
for emerging regions such as the Middle East and Asia. Furthermore,
airline structure tends to influence traffic concentration: the higher the
share of airline alliance traffic, the higher the traffic concentration of a
region. On the other hand, LCC and regional airlines tend to dampen con-
centration, since they are more focused on point-to-point traffic. During
the period from 2000 to 2014, airline alliances and LCCs achieved massive
gains in traffic share, while independent FSNCs, regional and charter
carriers either stagnated or decreased their traffic shares. Therefore, liber-
alization has fostered two very different airline business models: FSNCs
with their hub-and-spoke system on the one hand, and LCCs with mainly
point-to-point traffic on the other hand. Airline alliances accounted for
more than half of the global traffic in 2014, while LCCs had a share of
about 20 per cent of global take-offs. Furthermore, LCCs increasingly
try to arrange some degree of hubbing (‘self-help hubbing’), and they are
serving a greater number of larger and hub airports. This trend is fos-
tered by FSNCs establishing their own largely independent LCC subsidi-
ary, such as Germanwings/Eurowings (Lufthansa), Vueling (International
Airlines Group) and Transavia (Air France/KLM).
Hub airports are prone to congestion and capacity constraints, but this
is not a particular hub airport problem, as large airports with attractive
catchment areas close to dense agglomerations are generally more affected
than smaller airports that are remotely located. In these cases, airport
expansion plans are difficult to implement and take a long time, since
the surrounding population is typically opposed to such plans because of
further noise and pollution emissions, especially in economically highly
developed countries with a high level of participation. Furthermore, it
is unlikely that a shift away from hubbing will resolve airport capacity
problems in the long term, since a pure point-to-point network needs more
flights to connect a given network.
REFERENCES
Berster, P., M.C. Gelhausen and D. Wilken (2012), Demand and supply develop-
ment patterns of Low Cost Carriers in Africa, America, Europe, Australia and
Asia, paper presented at the Air Transport Research Society World Conference
2012, Tainan.
Berster, P., M.C. Gelhausen and D. Wilken (2015), Is increasing aircraft size
common practice at congested airports?, Journal of Air Transport Management,
46, 40–48.
Burghouwt, G. and J. de Wit (2005), Temporal configurations of European airline
networks, Journal of Air Transport Management, 11, 185–98.
Button, K. (2002), Debunking some common myths about airport hubs, Journal of
Air Transport Management, 8, 177–88.
Button, K., K. Haynes and R. Stough (1998), Flying into the Future: Air Transport
Policy in the European Union, Cheltenham, UK and Northampton, MA, USA:
Edward Elgar Publishing.
Daft, J. and S. Albers (2012), A profitability analysis of low-cost long-haul flight
operations, Journal of Air Transport Management, 19, 49–54.
Dobruszkes, F. (2006), An analysis of European low-cost airlines and their net-
works, Journal of Transport Geography, 14, 249–64.
Ehmer, H., P. Berster, J. Basedow and C. Jung (2000), Liberalisierung im Luftverkehr
Deutschlands – Analyse und Wettbewerbspolitische Empfehlungen, Cologne:
German Aerospace Center.
Gelhausen, M.C. (2013), A model of airport capacity expansion delays, paper pre-
sented at the Air Transport Research Society World Conference 2013, Bergamo.
Gelhausen, M.C., P. Berster and D. Wilken (2013), Do airport capacity constraints
have a serious impact on the future development of air traffic?, Journal of Air
Transport Management, 28, 3–13.
Goetz, A.R. and T.M. Vowles (2009), The good, the bad, and the ugly: 30 years of
US airline regulation, Journal of Transport Geography, 17, 251–63.
Grimme, W. (2011), The evolution of the low cost carrier business model –
connections, hubbing and interlining, presentation at the 5th Annual Airneth
Conference, The Hague.
Hooper, P. (2005), The environment for Southeast Asia’s new and evolving airlines,
Journal of Air Transport Management, 11, 335–47.
IATA (2008), IATA Economics Briefing No 9: Air Travel Demand, Montreal: IATA.
Independent (2016), Frankfurt airport opens up to no-frills flights as Ryanair
moves in, 2 November, available at http://www.independent.co.uk/travel/news-
and-advice/cheap-flights-frankfurt-ryanair-budget-airlines-germany-holidays-
spain-brexit-a7393231.html.
Malighetti, P., S. Paleari and R. Redondi (2008), Connectivity of the European
airport network: ‘Self-help hubbing’ and business implications, Journal of Air
Transport Management, 14, 53–65.
Moreira, M.E., J.F. O’Connell and G. Williams (2011), The viability of long-haul,
low cost business models, Journal of Air Transport Studies, 2, 69–91.
Morrell, P. (2008), Can long-haul low-cost airlines be successful?, Research in
Transport Economics, 24, 61–7.
O’Connell, J.F. and G. Williams (2005), Passengers’ perceptions of low cost airlines
and full service carriers: a case study involving Ryanair, Aer Lingus, Air Asia and
Malaysia Airlines, Journal of Air Transport Management, 11, 259–72.
Official Airline Guide (OAG) (2015), Market Analysis, Dunstable: Reed Travel
Group.
Pels, E. (2008), Airline network competition: full-service airlines, low-cost airlines
and long-haul markets, Research in Transportation Economics, 24, 68–74.
Redondi, R., P. Malighetti and S. Paleari (2011), Hub competition and travel times
in the world-wide airport network, Journal of Transport Geography, 19, 1260–71.
Redondi, R., P. Malighetti and S. Paleari (2012), De-hubbing of airports and their
recovery patterns, Journal of Air Transport Management, 18, 1–4.
Reichmuth, J., P. Berster and M.C. Gelhausen (2011), Airport capacity constraints:
future avenues for growth of global traffic, CEAS Aeronautical Journal, 2, 21–34.
Reynolds-Feighan, A. (2001), Traffic distribution in low-cost and full-service
carrier networks, Journal of Air Transport Management, 7, 265–75.
Sabre Airport Data Intelligence (ADI) (2015), Data based on Market Information
Intelligence Tapes (MIDT), Southlake: Sabre Airline Solutions.
Wensveen, J.G. and R. Leick (2009), The long-haul low-cost carrier: a unique busi-
ness model, Journal of Air Transport Management, 15, 127–33.
Wilken, D., P. Berster and M.C. Gelhausen (2011), New empirical evidence on
airport capacity utilisation: relationships between hourly and annual traffic
volumes, Research in Transport Business & Management (Airport Management),
1, 118–27.
Wilken, D., P. Berster and M.C. Gelhausen (2016), Analysis of demand structures
on intercontinental routes to and from Europe with a view to identifying poten-
tial for new low-cost services, Journal of Air Transport Management, 56, 79–90.
Williams, G. (2001), Will Europe’s charter carriers be replaced by ‘no-frills’ sched-
uled airlines?, Journal of Air Transport Management, 7, 277–86.
The World Bank (2016), World Development Indicators, Washington, DC: The
World Bank.
Regulation has gone astray. . . Either because they have become captives of
regulated industries or captains of outmoded administrative agencies, regula-
tors all too often encourage or approve unreasonably high prices, inadequate
service, and anticompetitive behavior. The cost of this regulation is always
passed on to the consumer. And that cost is astronomical.
Senator Edward Kennedy, opening remarks to the Subcommittee on
Administrative Practice and Procedure, 6 February 1975.
INTRODUCTION
284
tries that comprise the air transportation sector – airlines, airport and air
navigation systems – are no exception (Button, 2005b).1 The various forms
of instability in air transportation tend to ripple through time and across
markets in an interactive way that affects its component industries. In the
context of this chapter, these ripples are made more complicated because
of the nature of market liberalization that has taken place. Unlike many
other sectors characterized by several component industries, airlines, air-
ports and air navigation systems have historically largely been supplied in
individual markets and regulated separately. Furthermore, the liberaliza-
tions of air transportation markets that have occurred since the 1970s have
come at different times and in different ways. Given the network nature of
air transportation, the diversity of timings and natures of these liberaliza-
tions have, as we see below, added to the inherent instability of the sector.
This chapter provides some guidance as to the implications that liber-
alization of air transportation markets has had for both the suppliers and
their customers, particularly focusing on the challenges associated with
varied types of instability that have emerged. In doing this it draws both
upon empirical evidence and theoretical arguments to highlight the nature
of the problems, their scale and their underlying causes. It also looks at the
variety of ways the suppliers of air transportation have acted to mitigate
adverse effects of market instabilities, and the responses of governments
that often see instability of any kind as an adverse element of their desired
social welfare function.
The writing examining the liberalization of the world’s various air trans-
portation markets is gargantuan. Some of the more important work can
be found in Levine (1987), Kahn (1988), Bailey (2010) and Morrison
and Winston (1995). Here we limit the description to a few paragraphs
as a brief aide-memoire, or as a simple introduction as the case may be,
for readers. It is far from comprehensive, but further analysis is readily
available.
Since its inception, the airline industry has been subject to economic
regulation, the first being the patent right afforded to the Wright broth-
ers for their method of steering a heavier-than-air flying machine.
Following that, regulations grew at the national and, later, international
levels. Governments sought ways to make use of civilian air transporta-
tion for mail services initially, and subsidies became common, and then
to provide the infrastructure for aviation with direct involvement in
airport and air traffic control supply for local development and safety
reasons. The 1944 Chicago Convention drew up the framework for
agreements on international air transportation supply by creating a basis
of bilateral international negotiations and establishing an institutional
structure to oversee it (the International Civil Aviation Organization).
This, while it was not the intent, led to a mass of restrictive bilateral
national agreements that limited capacity on each route, the fares
charged, the carriers that could operate and often the ways that revenues
were split. The institutional regime had the potential for creating wider,
multilateral structures, but national self-interest and an inclination for
protecting national carriers prevented a more global perspective being
taken by individual states.
National domestic regimes varied in the regulatory structures they
imposed, with most giving monopolies to generally subsidized state-owned
carriers whilst at the same time controlling fares and routes. The US was a
notable exception to this in terms of ownership. Fares were inevitably regu-
lated. There was some flexibility for charter carriers offering unscheduled
services, but these had to conform to stringent criteria, for example, to be a
component of a package vacation, with the travel company de facto hiring
the entire aircraft.
While the rationale for regulations varied between countries and over
time, there were some common features. One was quasi-military in that a
country wished to have adequate air-lift at times of national emergency;
another was that countries often saw their international carriers, even if
not fully efficient, as a source of foreign exchange earnings (Button, 2010).
In larger countries, such as the US and Canada, air transportation was
viewed as an instrument of national political integration and, as such,
serving a key public interest.
Liberalization was initially introduced in 1977 for US inter-state freight
aviation and in the following year for passenger airlines as a way of reduc-
ing costs in the sector and inducing greater innovation in technology and
management. It was gradually taken up at the international level, nudged
along by the US’s Open Skies policy and the Three Packages of reforms
in the European Union in the 1990s that were part of a much wider Single
European Market initiative, as well as domestically by individual states.
There were also changes in the ownership of airlines in many countries,
as privatization of former state-owned carriers occurred; with this, there
has been a shift away from state subsidization, except for strategic reasons
(as with the US Civil Reserve Air Fleet) or in the case of social policy, and
even in the latter case competitive tendering has become the norm to make
more efficient use of public funds.
The liberalization of aviation infrastructure has been much slower, which
to an airline’s, where labour and fuel are the main cost considerations. This
inevitably means that it takes longer to adjust to a new institutional struc-
ture when a large part of a business’s costs are fixed. This is one reason
why airline markets were liberalized before airports and air navigation
service providers. But there is another element to this: some costs are sunk
and have to be left ‘stranded’ if an undertaking changes its activities; for
instance, an airport runway has few alternative uses if airlines take their
operations elsewhere (Baumol and Willig, 1981).
The challenge with airline liberalization is that there was little prior
experience of how a market-oriented air transportation system would
operate, or how long it would take for adjustments to take place. Indeed,
the end of the Civil Aeronautics Board in 1985 after carrying through the
reforms of the 1978 Airlines Deregulation Act was the first major regula-
tory body of any form to be disestablished in the US. While there were
some general ideas of what would occur, there were also surprises (Kahn,
1988). For example, the greater use of hub-and-spoke networks was antici-
pated once airlines were free to choose the routes and airports they would
serve, but the extent and the rapidity of their adoption was not known,
leaving companies like Braniff International Airways in the US with an
inappropriate fleet and network, resulting in bankruptcy. Other companies
experimented with route, fleet and service combinations as they sought to
fill the information gap and this was costly both to them and to their pas-
sengers as the market adjusted.
The approaches to handling short-term problems of adjustment insta-
bility stemming from institutional changes have differed across markets.
The US approach with its liberalization of air cargo transportation in 1977
and passengers in 1978 was to initiate the change rapidly and with very
little phasing in, expecting large short-term disruption followed by a fairly-
rapid move to a significantly higher equilibrium level of welfare: the ‘big-
bang approach’ (Button and Johnson, 1998). In contrast, although it was
probably more out of political expediency than economic efficiency, the
European Union adopted a phased approach to liberalization, bringing
in three packages of reforms in 1987, 1990 and 1992, with some phasing
in of each. The idea of this is to give the European carriers time to adjust
their operations and thus to minimize their stranded costs. However, the
challenge of this gradual approach is the possibility of the carriers captur-
ing the system and stymying the effectiveness of the change by slowing
or modifying it at each stage to their advantage, the extended use of
European restructuring aid being an example of this.
M4371-FINGER_9781786431851_t.indd 290
4
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
–2
290
–4
–6
–8
–10
Notes:
1. A lack of a bar indicates a missing observation and not a zero operating margin.
2. Memberships of the various reporting bodies vary over time and thus the reported margins reflect the associated carriers at the time of
reporting.
14/11/2017 11:14
Market instability 291
130
120
First year of data = 100
110
100
90
80
70
60
50
1997 1998 1999 2000 2001 2002 2003 2004
Note: All data indexed to 100 based on 1997 except for NATS UK 20015100.
Figure 15.2 En route unit rates levied by air service navigation providers
for en route and terminal services in constant 2004 prices
air navigation service providers for the period around 2001. The pro-
viders vary in their corporate structure: some (such as NATS UK, the
British supplier) were purely commercial undertakings, while others
(such as Airways New Zealand and Airservices Australia) were govern-
ment corporations or government agencies (such as LVNL, the Dutch
provider), but they all exhibit an upturn in the rates they levy for a period
after 2001.
The reasons for the differences are in part due to the institutional nature
of the undertaking involved, the reaction of governments (a matter we
return to later) and the degree to which their local markets were affected by
the events in New York and Arlington. In particular, many of the privat-
ized or corporatized systems had the remit of cost recovery often over a
period as short as one year, basically implying a zero rate-of-return regula-
tion. In consequence, they did not have sufficient reserves to cover their
extensive fixed costs as airline traffic fell after 2001 and they had to raise
their fees. This is a situation running somewhat counter to normal market
logic, whereby one would have lowered fees to attract more traffic after the
September 11th attacks.
We could further extend this descriptive analysis by looking at cycles
in airline bankruptcies and mergers, the geographical pattern and size of
airport hubs, air service withdrawals, market entry, and so on that accom-
pany economic recessions or sudden, unsystematic shocks to the aviation
market on either the demand or the supply side. But these sorts of events
are features that have counterparts across many sectors, and with some
minor peculiar characteristics fall within the generic umbrella of industrial
economics rather than being air-transportation-specific.
There are also – and of more interest – market structures that can be inher-
ently unstable simply because of their intrinsic characteristics, a point
made over 125 years ago but often overlooked by more recent economists
(Edgeworth, 1881).2 These markets are essentially competitive in nature,
but many of the assumptions of Marshall’s perfect competition do not
apply. There may be an ‘empty core’ with no equilibrium when there are
relatively large fixed costs, avoidable (set-up) costs, indivisibility, network
effects or severe fluctuations in demand. An unsustainable market may
also exist when some suppliers enjoy a degree of institutional or financial
protection and when there are significant variations in the costs of sup-
pliers. In practice, many public utilities, transportation industries and
some manufacturing industries meet these criteria. A general survey of
the theory is found in Telser (1978), with Button (2003; 2005b) linking its
relevance specifically to airline markets.
While developed in the context of a time when manufacturing concerns
dominated thinking about market structures, the idea of empty cores
is germane to the scheduled airline industry, not because there are high
physical fixed costs – although there may be – but because of the nature of
the service provided. An airline has to establish its schedule many months
in advance and, in doing so, commits to costs of having a plane, fuel, crew,
take-off slots, airport gates, check-in and baggage handling capacity, at
the very least, available at the time of take-off. These are the de facto fixed
costs of airline services.3 This is true even if the carrier is a virtual entity
that leases planes and crew; it has to have contracts in place. Because all
the competitors have similar fixed costs, there is a tendency to compete for
business; in effect, they compete to fill their seats by discriminate pricing
down to a marginal cost that is very low once a commitment to a schedule
has been made. Any revenues above this marginal cost can contribute to
the fixed costs of the service. But with competition, the tendency is for
these extra marginal cost revenues to be small.
This can be illustrated – albeit with some rather strong assumptions
for pedagogic reasons – with a simple example based on Telser (1987).
Suppose that an economy has three individuals (B1, B2 and B3) who wish
to fly between two points. B1 is prepared to pay a fare of $100, B2 of $80
and B3 of $75 for the trip. There is one aircraft owned by A1 that can
carry up to two passengers. The plane can make the trip at a cost of $140
whether it carries zero, one or two passengers. All of the sustainable out-
comes have the property that B1 and B2 travel on the plane, whilst B3 fails
to make a contract. The outcomes can differ in the fare that B1 and B2 pay
for the journey. For example, the allocation in which B1 and B2 both have
a flight and pay a fare of $75 is in the ‘core’. In all the outcomes in the
core, airline A1 makes positive profits of at least $10. If the carrier lowers
its fare further, below $75, the third individual (B3) would also wish to fly,
potentially disturbing the coalition of A1, B1 and B2. There would be an
excess demand at this rate and hence the core would be empty. If another
identical aircraft, owned by A2, also with a maximum capacity of two
and the same costs as A1 enters the market, there would also be no core.
Because the two aircraft are identical and because equilibrium requires
their owners to receive the same profit, there is not enough demand to
sustain both planes. They would compete prices down until one is driven
from the market. But the remaining carrier can again make supernormal
profits attracting a new entrant, and so on.
One may suppose that businesspeople are not entirely myopic, and
potential entrants during the times of high profits would recognize the risk
they take in entering the market, but as Knight (1921) pointed out nearly a
century ago – and which has certainly not been disproved by the specula-
tive behaviour that continues to abound – where there are conditions of
risk there are winners and losers and there is always a tendency for some
optimists to be around.
The fact that this economic conundrum applies to the scheduled airline
industry has been appreciated for some time, but it has been largely ignored
in policy formulation. The complexity of the underlying economic model
has hindered the communication of the issue to decision-makers. The situ-
ation also runs counter to an often ingrained element of traditional, often
ideological views of competition policy that hold that there can ‘never be
too much competition’. The idea that there can be ‘too much competition’
is something of an anathema in such circumstances.
Air carriers are not the only ones that have trouble recovering costs due
to what is often called excessive competition, nor are they the only case
where this leads to market instability. Ronald Coase (1946) argued that the
conditions for an empty core are the norm rather than the exception, and
that multipart pricing in various forms is the most practical way of cost
recovery, one that is widely adopted in industry. He goes on to say that
there is no neat solution to the problem, but what happens is that manage-
ment becomes inventive in finding ways to give their company an edge.
Many airline managers are certainly not unimaginative within this context;
indeed, in many ways they have been very creative in devising ways to try
to recover their full costs. However, the fact is that in gaining a temporary
edge they only add to the instability in the long term as others copy the
first mover.
Generally, industries suffering from an empty core problem often employ
a wide variety of techniques to recover their full costs, with the adoption
of a particular practice or measure depending upon a combination of the
technical characteristics of the industry and the institutional environment
in which it operates (Button, 2005a). These actions largely revolve around
efforts at creating a degree of monopoly power to allow full cost recovery.
Many airlines function for a considerable time without recovering their
full costs of capital. In some cases, this may be possible because capital
debts are written down through institutional means. The US Chapter 11
bankruptcy laws, for example, effectively allow an airline to restructure and
write down its capital without being broken up. Virtually all US carriers
have used this mechanism (some, such as US Airways, more than once).
From a wider perspective, competing firms often view this as unfair com-
petition, because the costs of their rival have been reduced through legal
means rather than through strict efficiency improvements. But Chapter 11
can be expensive not only in terms of legal fees, but also because while in
bankruptcy – and often for a period after – the costs of purchasing inputs
(fuel, aircraft leases, etc.) tend to be higher and can require immediate
payment.
Stability can be increased by de facto coordination down the air trans-
portation supply chain. As we have seen, airlines are at the end of a value
chain and, as such, other industries such as airports, air navigation service
providers, airframe and aero-engine manufacturers and global distribu-
tion systems rely on them to generate the revenues on which they in turn
rely. Because some of these industries are more successful – often because
they have monopoly power – in recovering their costs of capital, they may
‘invest’ further down the value chain in airlines to ensure their solvency; it
is in their own interest to do so.
Airlines have also attempted to reduce competitive pressures by encour-
aging customer loyalty though frequent-flyer programmes that reward
regular passengers with free flights and other bonuses such as upgrades to
higher classes of service and access to airport lounges. The ‘miles’ earned
on carriers within alliances are normally interchangeable, albeit not per-
fectly, providing passengers with an extensive range of services for redemp-
tion. It is now also possible to collect miles through non-airline purchases
such as credit card use, car rentals and dining. The airlines effectively sell
their miles to other industries that then give them as rewards to their own
customers: in 2010 Delta Air Lines earned $1.6 billion from such sales.
Since liberalization, the greater flexibility afforded to airlines has led
to new carriers entering the market that provide lower fares and more
basic services. This has been the strategy of airlines like Southwest in the
US and Ryanair and easyJet in Europe, and many more. In some ways,
this business model represents the ‘blue ocean’ approach to market pen-
etration by seeking out new categories of customers as well as picking
up business from the edges of existing markets (Kim and Mauborgne,
2005). It generally entails considerable standardization in its operations
(e.g. the use of a common family of aircraft and a homogeneous, usually
short-haul, network of services), maximizing the use of its labour, serving
less-congested, smaller airports, providing a ‘no-frills’ service on the plane
and at the airport, limiting methods of booking, charging for non-core
services such as refreshments and checked bags and offering only one class
of service. Such measures can reduce costs by 30 per cent or so compared
to those of traditional airlines.
Liberalization has also allowed airlines flexibility in their price setting,
and the advent of computer booking systems has allowed carriers to fine-
tune this with prices being set according to customers’ willingness to pay.
This price discrimination allows an airline to offer seats at various prices,
and to continue to vary these offers as seats are purchased. The conditions
pertaining to a ‘seat’ can also differ; for example, the ticket may be refund-
able, it may be upgradeable to a higher level of service if combined with
frequent-flyer miles or it may be at a location on a plane (for instance, a
seat in an emergency exit row), and prices are adjusted according to these
quality factors. Yield management, being related to the willingness of cus-
tomers to pay, results in passengers who are less sensitive to price paying
more and contributing to the capital cost of the service, while those who
are less willing are charged lower fares that at least cover their marginal
costs. While it can be used to generate large profits – and this has been
done in many industries – its main purpose in air transport is to generate
sufficient revenue to earn an acceptable return after all costs, including the
cost of capital, have been covered.
Instead of focusing on reducing costs, some airlines seek greater stabil-
ity by making revenue flows more certain. Charter airlines effectively do
this. They sell capacity well in advance to tour operators and therefore
know months ahead how much capacity needs to be delivered to custom-
ers, thus removing some risk from their operations. Scheduled airlines
cannot follow this pattern because they guarantee a service ahead of time,
and then effectively become common carriers of the traffic willing to pay
for flights. However, in some US cities, groups of businesspeople have
tried to ensure regular air services with guarantees of adequate patronage
for an initial period: ‘travel banks’. In Wichita, Kansas, for example, 400
businesses raised $7.2 million for this purpose, and similarly in Pensacola,
Florida, 319 businesses raised $2.1 million, while companies and indi-
viduals in Stockton, California bought $800 000 of prepaid tickets from
American West (Nolan et al., 2005).
The problem with all these measures is that while they can afford first-
mover advantages to those who initiate them and thus reduce the competi-
tive pressures confronting these carriers, the ability to rapidly emulate in
a liberalized market means that the short-term stability enjoyed by first
movers is soon evaporated. Just consider discriminatory pricing. This
afforded early adopters a degree of monopoly power because these carriers
sold many of their tickets through their own retail outlets, and then their
own computer reservation systems control the flow of fare information
to potential passengers. It was time-consuming for potential customers
to search for the cheapest ticket. The emergence of global distribution
systems has largely removed the asymmetric information advantage that
the airlines enjoyed: customers can easily get details of fares from a range
of airlines and the associated services and restrictions that go with them.
This may involve looking at a few pages of information, but sites such as
Orbitz and Kayak keep this to a minimum. It has become much harder for
any airline to differentiate between customers and to extract the highest
possible fares from them.
When it comes to low-cost carriers, while the underlying model clearly
was advantageous and remains so for some carriers today, the most suc-
cessful carriers adopting this approach have tended to be the first in the
market, allowing them to enjoy a ‘first-mover advantage’. The list of
subsequent failed low-cost airlines is long. One problem is that as they
have expanded, they have gradually moved into increasingly thin and less
suitable markets for their style of operation, and diminishing returns have
set in. Additionally, as more carriers have emerged, competition between
low-cost airlines has grown, hitting all their bottom lines (Button et al.,
2007). The traditional airlines have also become leaner and more skilled
at resisting the challenges associated with low-cost carriers trying to enter
their routes (Morrison, 2001; Pitfield, 2008). Equally, loyalty programmes
have been reduced in their potency as any currency does when its supply
expands excessively, with ever-increasing numbers of miles being required
to buy flights and the number of flights for sale shrinking.
Public policy has always played a key role in the air transportations sector,
and still does, even though the intensity of market intervention has signifi-
cantly declined in most countries since the 1970s and the instruments used
are somewhat less blunt than in the past. The concerns of policy-makers
has also changed somewhat, with less focus on anti-trust issues as their
understanding of the nature of competition in air transportation evolved.
Light-handed economic regulation, where regulation continues, is more
the norm. Similarly, state ownership has declined in its traditional forms,
although public–private partnerships have grown, often in the form of con-
cessionary structures. All these trends have removed some of the former
cushions against instability, but have resulted in greater efficiency in sup-
plying aviation services and in matching those services with the demands
of customers.
Subsidies have long been used to recover capital costs. The argument
often used is that once an investment has been made, it becomes economi-
cally efficient to maximize its use, subject to the willingness of users to pay
the incremental costs of their decisions. The original argument couched
150 years or so ago related to a bridge: once built, if the demand had been
overestimated it would be wasteful of the asset not to let people use it,
provided they paid the maintenance and repair costs associated with their
actions. The current trend to unbundle attributes of a service – such as
charging for food and second checked bags by some airlines – attempts to
isolate the activities in which the fixed costs are concentrated and to charge
explicitly for the incremental costs. The fixed costs in this sense can then be
isolated, and the other attributes – the food and bag service – sold in the
market at competitive prices. This concept of unbundling is quite common
in the service sector and is being increasingly used by airlines, where a basic
service can be augmented with add-ons.
Direct subsidies in this context are designed to cover the fixed costs
that cannot be recovered from customers. In the case of airlines, however,
where the fixed cost is that of a commitment to a schedule, it is difficult to
isolate the fixed cost in the traditional sense. Further, there is the generic
CONCLUSIONS
NOTES
1. We have adopted the standard definition of economic stability here as meaning a system
that is stable when it dissipates market imbalances that arise endogenously, as a result of
significant adverse and unforeseeable events, or as an endogenous consequence of the
intrinsic nature of the market. A stable system absorbs shocks largely via self-corrective
mechanisms, preventing adverse events from disrupting its internal economic efficiency,
or spreading over to other economic systems.
2. There are other arguments put forward as to why scheduled airlines cannot cover
their full economic costs. For example, Borenstein (2011) puts it down to demand factors
and the dichotomous market, with high-cost carriers competing with low-cost airlines.
3. One could argue that an airline could change the gauge of equipment used for a flight
as the levels of booking become apparent, but many low-cost carriers only use a single
broad type of plane (for example, Southwest and Ryanair both only use Boeing 737 air-
craft to reap economies of fleet).
REFERENCES
Antoniou, A. (1998), ‘The status of the core in the airline industry: the case of the
European market’, Managerial and Decision Economics, 19, 43–54.
Bailey, E.E. (2010), ‘Air-transportation deregulation’, in J.J. Siegfried (ed.), Better
Living Through Economics, Cambridge, MA: Harvard University Press.
Baumol, W.J. and R.D. Willig (1981), ‘Fixed costs, sunk costs, entry barriers, and
sustainability of monopoly’, Quarterly Journal of Economics, 96, 405–31.
Borenstein, S. (2011), ‘Why can’t US airlines make money?’, American Economic
Review, Papers and Proceedings, 101, 233–7.
Button, K.J. (1996), ‘Liberalising European aviation: is there an empty core
problem?’, Journal of Transport Economics and Policy, 30, 275–92.
Button, K.J. (2003), ‘Does the theory of the “core” explain why airlines fail to
cover their long-run costs of capital?’, Journal of Air Transport Management, 9,
5–14.
Button, K.J. (2005a), ‘The economics of cost recovery in transport’, Journal of
Transport Economics and Policy, 39, 241–57.
Button, K.J. (2005b), ‘How stable are scheduled air transport markets?’, Research
in Transportation Economics, 13, 27–49.
Button, K.J. (2007), ‘The implications of the commercialization of air transport
infrastructure’, in D. Lee (ed.), The Economics of Airline Institutions, Operations
and Marketing 2, Oxford: Elsevier.
Button, K.J. (2010), ‘Air transportation services: both a traded commodity and
a transactions cost in international trade’, Journal of International Commerce,
Economics and Policy, 1, 105–20.
Button, K.J. and K. Johnson (1998), ‘Incremental versus trend-break change in
airline regulation’, Transportation Journal, 37, 25–34.
Button, K.J. and G. McDougall (2006), ‘Institutional and structural changes
in air navigation service-providing organizations’, Journal of Air Transport
Management, 12, 236–52.
Button, K.J. and H. Vega (2006), ‘Airlines competing with themselves: a note on
the temporal pattern of fare setting prior to departure’, International Journal of
Transport Economics, 31, 341–50.
Button, K.J., A. Costa and C. Cruz (2007), ‘Ability to recover full costs through
price discrimination in deregulated scheduled air transport markets’, Transport
Reviews, 27, 213–30.
Coase, R.H. (1946), ‘The marginal cost controversy’, Economica, 13, 169–82.
Edgeworth, F.Y. (1881), Mathematical Physics, London: Kegan Paul.
Kahn, A.E. (1988), ‘Surprises of airline deregulation’, American Economic Review,
Papers and Proceedings, 78, 316–22.
Kim, C.W. and R. Mauborgne (2005), Blue Ocean Strategy: How to Create
Uncontested Market Space and Make the Competition Irrelevant, Cambridge,
MA: Harvard Business Review.
Knight, F.H. (1921), Risk, Uncertainty and Profit, Boston, MA: Hart, Schaffner
and Marx.
Levine, M.E. (1987), ‘Airline competition in deregulated markets: theory, firm
strategy, and public policy’, Yale Journal on Regulation, 4, 393–494.
Meyer, J.R. and W.B. Tye (1981), ‘The regulatory transition’, American Economic
Review, Papers and Proceedings, 71, 268–72.
Morrison, S.A. (2001), ‘Actual, adjacent and potential competition: estimating the
full effect of Southwest Airlines’, Journal of Transport Economics and Policy, 35,
239–56.
Morrison, S.A. and C. Winston (1995), The Evolution of the Airline Industry,
Washington, DC: Brookings Institution.
Nolan, J., P. Ritchie and J. Rowcroft (2005), ‘Small market air service and regional
policy’, Journal of Transport Economics and Policy, 39, 363–78.
Pitfield, D.E. (2005), ‘A time series analysis of the pricing behaviour of directly
competitive “low-cost” airlines’, International Journal of Transport Economics,
32, 15–38.
Pitfield, D.E. (2008), ‘The Southwest effect: a time-series analysis on passengers
carried by selected routes and a market share comparison’, Journal of Air
Transport Management, 14, 113–22.
Future challenges
INTRODUCTION
BACKGROUND
In the last 45 years trade, technology and economic growth have created
an age of globalization in which the welfare of people, firms and nations
has become ever more interconnected. During this period, civil avia-
tion has evolved from a heavily regulated system of national airlines and
government-operated airports to a much larger and more competitive
305
100
1970s
90 1980s
73%
80 1990s
Total number of attacks
70 2000s
67%
60
50
40 84%
30
19%
20 17% 14%
43%
10 5% 11% 24% 24% 3%
2% 5% 0% 1% 2% 1% 0% 5%
0
Hijacks Bomb Cargo Bomb Hijack-Crash Other
global industry in which private airlines and airports compete along with
publicly-owned counterparts and hybrid organizations under diverse regu-
latory regimes. There has been remarkable growth in the number of global
air passengers over the last 55 years, a long-term trend which has been
largely impervious to the negative shocks of macroeconomic recessions,
health crisis, military conflict and acts of terror. In 2016, IATA World
Air Transport Statistics show there were approximately 3.7 billion air
passengers.
However, civil aviation has been a visible target for acts of violence
and terrorism throughout this period. Figure 16.1 shows the number
of attacks inside planes worldwide by attack type over the four decades
between 1970 and 2009, and indicates the percentage of attacks in each
decade accounted for by each attack type. In the 1970s, attacks on aircraft
were heavily skewed towards hijackings. However, this mode of attack has
declined both in number and in relative importance over time. The figure
also shows a sharp rise in the number of bomb attacks that occurred in the
1980s, falling again in the subsequent decades. Thus, the data captures the
evolutionary nature of aviation security. As authorities implement security
measures to nullify a given mode of attack, terrorists adapt their strategies
and the preferred mode of attack evolves.
Overall, the number of attacks has declined significantly over time, with
111 attacks in the 1970s but just 21 attacks between 2000 and 2009. In
terms of fatalities, there were 557 deaths because of attacks inside planes
in the 1970s. This number rose to 1115 in the 1980s, mainly as a result
of a small number of attacks that inflicted a large number of casualties,
including Air India flight 182 in 1985 (329 fatalities) and Pan Am flight
103 in 1988 (270 fatalities). In the 1980s the objectives for attacking air-
craft had evolved from attention-seeking through prolonged live media
coverage of a hijacking to the shock and terror generated by the sudden
and unexpected mass killing of innocent civilians. By the 1990s, aviation
security had responded and was evolving into a complex (and expensive)
system combining intelligence agencies, security personnel at airports and
investments in scanning equipment to detect bombs, weapons and other
prohibited items. In the 1990s the number of fatalities from all terror
attacks inside planes declined to 160, but the following decade will forever
be defined by the 2938 deaths resulting from the attacks in New York and
Washington on September 11th, 2001. Excluding the 9/11 attacks, globally
there were just 94 fatalities as a result of terrorist attacks inside aircraft
from 2000 to 2009.
The events of September 11th, 2001 represent by far the biggest and
most shocking realization to date of the ever-evolving threat of terror
attacks against aviation. The attacks demonstrated how civilian aircraft
could be used as weapons to kill large numbers of civilians and destroy
assets on the ground. The attack created mass panic over the vulnerabili-
ties of the civil aviation system and led to sweeping and significant changes
in the design, provision and financing of aviation security throughout the
world. Since 2001, governments have created new organizations to imple-
ment airport security systems and there have been massive investments
in both technology and the hiring and training of security personnel.
Through these changes and increases in security costs, airports and airlines
have faced new challenges in managing passenger throughput, minimizing
delays and negative passenger experiences resulting from elevated levels
of security effort. The general public and the travelling public have borne
both the direct and indirect economic costs of these investments.
Despite the elevated efforts in the first decade of the 21st century, avia-
tion has remained a target for terror attacks. Most recently, in 2016, attacks
in Brussels and Istanbul involving suicide bombers and shootings resulted
in 77 fatalities. In Brussels suicide bombers detonated explosives in a
departure hall, while in Istanbul attackers opened fire at the international
terminal of Atatürk airport, detonating suicide bombs in the car park and
in the departure and arrivals halls. These attacks demonstrated that a rela-
tive focus in preventing potential attackers or weapons from entering the
airside of an airport left vulnerabilities on the landside (prior to screening
checkpoints) both inside and outside airport terminal buildings.
Following Gillen and Morrison (2005) one can view a civil aviation
network as a ‘public franchise’. Similar to private franchise organizations,
the overall value created by a public franchise is the result of the com-
bined efforts of individual organizations in the franchise. The glue that
binds public franchises together is the vertical rules and regulations that
define the relationship between each organization and the government. In
this regard, the government (as franchisor) attempts to create incentive-
compatible rules that guide each member organization to voluntarily
provide effort levels (in whatever dimensions are applicable) that will maxi-
mize the value of the public franchise. However, each organization in the
network also has a horizontal relationship with other member organiza-
tions and mal-incentives can exist because each organization draws value
from the overall success and value created by the public franchise, which
is due to the combined efforts of all the members. The benefit of a public
franchise network is that if implemented correctly, the overall objectives
of the franchise can be met while taking account of differences in the
circumstances of individual members. In other words, given heterogeneity
in member characteristics and circumstances, a public franchise approach
potentially avoids the problems of one-size-fits-all policy rules. However,
asymmetric information, measurement problems and monitoring costs
make the design of incentive-compatible franchise rules difficult and
potentially complex. The analytics of the franchising problem in a public
network setting (such as aviation) focuses our attention on the extent to
which the government as ‘franchisor’ can design a governance structure
that mitigates the horizontal externality problems and maximizes the
public welfare. Thus, a network of independently operated airports within
a nation or jurisdiction can be characterized as a public franchise network
in which the government – as franchisor – sets rules, regulations and a gov-
ernance structure to maximize the public value of the airport system. More
specifically, aviation security can be viewed in the same way, whereby the
government seeks to create an institutional structure that recognizes differ-
ences in the circumstances of individual airports in the system, while at the
same time ensuring that each airport contributes a level of effort in the pro-
vision of aviation security that achieves national (and international) objec-
tives. Canada’s National Civil Aviation Security Program seems to embody
this approach by stating a preference for ‘outcome-based’ approaches to
security and less reliance on prescriptive regulatory rules: ‘Prescriptive
regulations and “one-size fits all” solutions may not always account for the
wide variability in industry operations and skills, and can place unneces-
sary burdens on stakeholders resulting in ineffective security outcomes.
Therefore, the most effective way to address security gaps is often through
outcome-based rules.’ (Government of Canada, 2013).
Viewed in this way, global aviation security becomes a network of nation-
ally controlled public franchise networks. To visualize this, consider an
international environment with just two countries (A and B). International
laws and agreements create a common international standard for the
provision of minimum security levels at all airports. Each nation is then
responsible for the implementation of these standards. If it is assumed that
each country has two airports: (a1, a2) and (b1, b2), that there are domestic
flights between a1 and a2 and between b1 and b2, and international flights
a1
V H
A a2
V
International
Standards H H
b1
V H
B
V b2
between a1 and b1 and between a2 and b2. Even with this simple model,
there is a complex three-tier set of vertical and horizontal relationships, as
illustrated in Figure 16.2. At the top tier there is a horizontal relationship
between countries A and B. Formally, the governments of these countries
have agreed to the same international standards; however, in their vertical
relationships with their own airports, A and B can vary the level of enforce-
ment which does not guarantee that international standards are met, since
lower enforcement levels are less costly. These choices form a strategic
interaction between A and B that can be summarized as the first stage of
a two-stage game, as illustrated in Figures 16.3a and 16.3b. In Stage 1 of
the game governments decide on their level of enforcement, then in Stage
2 each airport makes decisions concerning its level of security effort. The
strategic interaction between airports occurs domestically between a1 and
a2 and between b1 and b2, but there is also strategic interaction internation-
ally between a1 and b1 and between a2 and b2.
In these strategic encounters, it is assumed each airport knows the level
of enforcement implemented by its domestic government. Figure 16.3b
shows the strategic interaction between international airports ai and bi,
and here there is a potential for a prisoner’s dilemma pay-off structure
(Coughlin et al., 2002). If it is assumed that citizens’ willingness to fly is a
function of the average perceived security at airports, then if one airport
Strong B Weak
enforcement enforcement
Strong Weak
Strong enforcement enforcement
1 , 1 2, 2
enforcement
A Strong
1 , 1 2, 2
enforcement
Weak
3, 3 4, 4
A enforcement
a Weak
3, 3 4, 4
enforcement
Figurea16.3a Stage 1 bi-matrix game between jurisdictions
bi
implements a higher level of security and bears the full costs of that action,
other airports will benefit from an increase in the average level of security
in the system.
As the number of airports in the system gets large, the average increase
in overall security resulting from the actions of one airport becomes small
relative to the cost it bears to implement higher security standards. In this
way, the implementation of higher security is discouraged. Theoretically,
this incentive problem can be avoided if there is strong enforcement of
international standards in the vertical relationship between a nation’s
government and its airports. Thus, a nation that strongly enforces security
standards changes the pay-offs in the Stage 2 games such that choosing
high security becomes a dominant strategy for its airports. However, the
prisoner’s dilemma can still prevail if the strategic interaction between
where pt,−s represents the probability of attacks in year t without the imple-
mentation of the security measure; pt,s represents the (reduced) probability
of attacks in year t after the implementation of the security measure and
k represents the value of lives and assets expected to be destroyed in suc-
cessful attacks annually. These probabilities must be conjectured from
historical evidence and intelligence. Nevertheless, despite these difficulties,
it is possible for benefit–cost analysis to provide insights in support of
improved resource allocation decisions, providing that informed assump-
tions can be made about the risk reductions associated with any given
security measure. Stewart and Mueller (2008) used such an approach to
assess specific aviation security measures in the US in the aftermath of the
September 11th attacks. A key element is the use of sensitivity analysis to
provide the ranges of probabilities required for a given security measure
to generate net benefits. This step is essential as a robustness check on
the reliability of the estimates. In evaluating the PreCheck (trusted trav-
eller) programme in the US, Stewart and Mueller calculate the benefits
of the current aviation security system with and without the PreCheck
programme. Under conservative assumptions concerning the ability of a
trusted traveller system to detect terrorists and the overall probability that
any citizen is a terrorist, they showed that the net effect of the PreCheck
What exactly has been produced from all the efforts to provide avia-
tion security? One way of answering is to ask ‘what is the probability
of someone being killed as a result of an aviation security incident?’
Measuring this requires a numerator, which is the fraction of people killed
in incidents, and the denominator, which measures the ‘opportunity’ for
such incidents – the number of flights in a given period in a given juris-
diction. In the case of aviation safety, a reasonable measure would be the
proportion of people who do not survive a flight in comparison to the
number of flights.1 The ‘reasonableness’ of this measure can be improved
by considering differences in the severity of an incident: a large proportion
of people dying from an incident is different from when a small proportion
of people die. In addition, we wish to account for the majority of aviation
incidents that take place during landing or take-off.
One metric of aviation security associated with passenger security ser-
vices at an airport could be ‘1 minus the ratio of the proportion of fatali-
ties (or injuries) over the number of domestic plus international scheduled
commercial flights in the jurisdiction being considered’. This metric could
be calculated for a region or country.2 The metric could include a broader
number to measure the impact in the numerator; for example the number
of people killed or the number of people killed or injured, or the number
m 5 acalculations
of non-airline passengers killed mor the number of air and non-air pas-
CRThese
sengers killed (or injured). si could take account of the
different potential for a threat byi51considering, for example, where flights
originate. Are they from high-risk jurisdictions? Would there be greater
risk with a flight of a large aircraft than a small aircraft? For example, are
than5ona
terrorists more likely to exert greater
N effort to inflict damage on a long-
haul, wide-body aircraft HH s2i 3 100
a shorter-haul, narrow-body aircraft?
To formalize an output measure i51
of aviation security, consider the
following:
SL j
t,k 512 (16.2)
nt,k
T^
Matthias Finger and Kenneth Button - 9781786431851
Downloaded from Elgar Online at 01/13/2018 08:55:07AM
via University of Durham
0C/0T^
M4371-FINGER_9781786431851_t.indd 314 14/11/2017 11:14
Economic perspectives on aviation security 315
where S(a) is the likelihood that the security incident will be at an airport
involving passenger screening, a(M) is the likelihood the incident will be of
type M and f(i(M)) is the consequence an incident of type M has on the
likelihood a plane will crash. For example, Jackson et al. (2012) reported
the security incident locations and the type of attack on the airplane. As
Figure 16.1 illustrated, in absolute terms hijackings have dominated air-
craft terrorism attacks, but the pattern of attacks has changed significantly
over time. From 1970 through to 2008, hijackings went from over 75 per
$12.00
Canada
U.S.
$10.00
$8.00
$6.00
$4.00
$2.00
$0.00
2003 2004 2005 2006 2007 2008 2009 2010 2011
provides a comparison of cost per passenger where costs are operating and
capital costs. From this data, costs in the US exceed those in Canada in
every year, and the spread is widening. These cost differences can reflect
several factors, such as the larger US market (number of airports) and
differences in governance and public finance between the two countries.4
Figure 16.5 displays the cost per passenger for checking baggage and
personal screening. Again, data limitations only allow for a comparison of
three countries: the US, Canada and New Zealand. Canada has the lowest
costs, which are approaching those of the US in 2010. New Zealand’s costs
are considerably higher than the other two countries. All three countries’
costs level off in 2010 and have tended to have the same spread.
To investigate further, two types of costs are examined: operating plus
capital, and passenger screening and bag check costs (expressed in real
terms). In the former, data from Canada, US and Australia is employed
and in the latter data from Canada, US and New Zealand expressed in real
Canadian dollars. In terms of total cost, no significant difference in costs
emerge between Canada, US and Australia that would be explained by
some other variables or that are inherent to those respective countries, and
the incremental cost of providing security services to a passenger is found
to be $11.79 CAD.5
The role of scale in the production of security could be an important
factor in determining the costs of providing aviation security. There
are significant variations in the size of airports both within and across
national borders, and it is possible that larger airports (with higher annual
Europe has a different model to fund and provide security wherein the
national government sets security standards and each airport in a country
(or member state) provides the security services (either through producing
them themselves or contracting out) levying a charge on airlines and/or
passengers.6 Utilizing data from 2011 we were able to construct measures
of security costs for a sample of airports, as shown in Figure 16.6. Average
security cost per passenger varies from $6.28 at Salzburg Airport, Austria,
to $0.73 at Sabiha Gokcen Airport in Istanbul, with an average over all
airports of $2.88 per passenger. These numbers could be somewhat dis-
torted if transfer passengers are not screened when they connect and to
the extent that an airport handles a significant amount of cargo; this will
also upward bias the averages. To correct for these potential biases, a cost
model is estimated that includes variables for cargo volumes, the percent-
age of international passengers, whether the airport was a large hub and/
or gateway and dummy variables to examine differences across countries.
The results indicate that marginal cost per passenger is $2.59, which is
substantially less than our estimates using data for the US, Canada and
Australia. However, the elasticity of cost with respect to passengers is
0.94, which is like the US/Canada/Australia estimate. Security costs are
higher for large hubs, while cargo adds $36.33 per tonne to security costs
SZG
CGN
FRA
BUD
MUC
VIE
ADP
ORY
CDG
TLV
DAA
DUB
STR
HAJ
LHR
ZRH
LUX
AMS
DUS
GVA
Finavia
HEL
LIU
OSL
Avinor
HAM
TAV
MAG
WAW
MXP
SEA Milano
LIN
BTS
NCE
MAN
ATH
Berlin
TXL
GLA
ADR
CIA
FCO
BGY
MAD
BHX
MAP
AGP
EDI
BCN
VCE
ANA
LIS
ALC
IST
LPA
BRS
PMI
SAW
Average
320 Air transport liberalization
ing this number to X − t?’ Several forces will influence this value. First, the
(average) flow rate and the (average) flow time will influence the inventory.
The flow rate is the average number of passengers screened per unit time,
and this is governed by the amount of resources used to produce screening
services or the capacity of the system. An additional factor is the arrival
rate of the passengers: sometimes it may be high and at other times it may
be low. The flow time will be determined by the level of screening, and
how long it takes to complete a screening process. Rigorous screening will
increase flow time, while less rigorous screening reduces it. This time will
include waiting time in the queue. Thus, flow time is measured from the
time the passenger joins the queue to the time the passenger completes the
screening process and exits to the terminal. However, the flow time will also
depend on the ‘nature’ of the passenger: people who travel less often, the
elderly, foreign individuals, families and people with carry-on luggage all
take more time for a given level of screening.
Another factor is the size of the list of prohibited items. As the list grows
longer there is more inspection time as well as increased demands on pas-
sengers (shoes off, belts off, computers out of cases, etc.). The inventory
will also be affected by the processing capacity of the system. Combining
machines and humans produces security screening: the machines can
operate at a given processing speed, but the human agents create variability
in the time for processing. Different agents may have different experiences,
differing levels of risk tolerance or different training. The combination of
agents may affect processing speed.7
Developing an efficient screening service time is a matter of identifying
cost and benefit components and linking changes in those components to
changes in service times. The measure of costs would be the resource cost
in total costs to the airport, or to the security agency in charge. The same
is true on the benefit side; it is the full economic benefit to all parties that
needs to be measured. The benefit measure raises a question: who are the
beneficiaries of security screening? Some jurisdictions take the position
that it is air passengers, while for others it is air passengers and the general
public.8 This could raise a problem of how to measure benefits to society
from changes to security screening times, although this question is more to
do with screening levels than with screening times.
The economic cost of providing security screening can be represented
as:
CR m 5 a si
i51
m
CR m 5 a si
i51
m
HH 5 a s2i 3 100
N
i51
322 Air transport liberalization
of a
i51
N
tables in the screening area; and T is amount HH 5 times 3
2
itaken
100for the screen-
a i51 f (rigour
where tS is the amount of time required toSL n complete
t,k 5 1 2
j
the screening and tS 5
i) jt,k (shoes nt,k
f(S′) where S′ is the defined level of screening on or off, belt
t,k 5 1
SLetc.).
on or off, laptops in or out, j s is the variability of time and
ts 2 T^ is a func-
nt,k
tion of N/K, which is the ratio of passengers (N) to machines (K) (a volume
capacity ratio). T^
Equation 16.4 can be rewritten to represent private costs for a defined T
which is the service time T^ : T^
which states that marginal benefits are decreasing with increases in service
time, queuing time and variance in queuing time.
To put the problem in context, consider a passenger taking a flight
from point A to point B who must complete several transactions: getting
from home/work to the airport, checking in and obtaining a board-
ing pass and checking any luggage, passing through security screening,
queuing at the boarding gate and finally boarding the aircraft. There
can be delays in any of these processes. It is assumed all other processes
do not change if the screening service time changes, meaning if security
invests resources to reduce waiting times in screening queues, airlines will
not reduce check-in resources to capture some of the added time now
available. It is also assumed that passengers do not change their behav-
iour, except that any time savings can be used to produce income for busi-
ness passengers or leisure. Further it is assumed that the screening service
level does not change; that is, the level of security does not change. This
allows investigation of how service times, and hence costs, can change as
the screening process changes, such as excluding more items and requir-
ing passengers to remove shoes, belts, laptops from cases and liquids
from carry-on bags. Costs and benefits are assumed to accrue to air pas-
sengers and not to the general public. Security screening for baggage or
freight is not considered.
Delays and variations in delays in completing the security screening
process will impose costs on two economic agents: airlines and passengers.
The added cost of delay is the opportunity cost of a passenger’s time plus
any disutility associated with waiting in line longer than is needed to com-
plete the screening process. For example, if the agreed service standard is
‘screening should take no more than five minutes’, the flow time should
be no more than five minutes; it is the time in excess of five minutes that
is the basis of cost measurement. In addition, if there is variability in the
flow time the system is not considered robust by passengers; therefore,
they will be forced to self-insure and arrive at the airport sooner than they
would have otherwise, using up valuable time. Thus, the time measure is the
expected flow time composed of the processing time including the average
time in the queue plus the insurance time, which is the variance in process-
ing time. These time quantities can be monetized by using value of time
and value of reliability measures.
The data requirements for this type of study would be:
connections as well. If the airline was not able to fill the seat, one might
argue there is an opportunity cost to the airline of an unused seat.
For airports, there are costs of space used in processing passengers and
the incremental costs of space if additional screening machines and per-
sonnel are added. One must take account of any costs to an airport related
to contracting security services and how these costs vary by number of
screeners and machines, plus any other direct and indirect costs associated
with security screening.
Most countries around the world recover at least a portion of the costs of
aviation security through charges levied on airlines, which are passed on to
passengers through ticket prices. We uncover, in a comparable way, avia-
tion security charges for a sample of 60 airports (Table 16.1). To do this, a
proprietary data set on airport charges, compiled by RDC Aviation Data
Insight is accessed.9 This data set is principally used by airlines for budget-
ing purposes, but it allows calculation of stated aviation security charges
while controlling for equipment used, load factor, dwell time, date (charges
were all calculated for the same date in May 2014) and exchange rates.
The airports in our sample can be compared by country, annual passenger
volume and governance structure. The data in Table 16.1 shows the high
variation in published aviation security charges between countries, with
Canada having by far the highest charges, at over $25 per international
passenger.
China, Spain and the US have roughly uniform charges, while variation
in charges between airports is evident in Canada and Germany. Figure 16.7
shows the relationship between stated airport security charges and annual
passenger volumes. The figure indicates no discernible pattern linking
the airport size and aviation security charges. For example, San Diego
and Atlanta levy the same security charge but differ widely in passenger
volumes. Frankfurt and Vienna also have comparable security charges but
also differ significantly in passenger volumes.
An alternative explanatory variable for security charges is governance
structure, that is, airports with the same ownership structure may set
similar security charges. Figure 16.8 compares security charges by govern-
ance structure and it is evident that there is still variation across different
ownership classes. Here two types of variation emerge: within-country
(Canada’s airport authorities) and between-country variations (US public
Airport Country Annual Pax Governance Stated Pax Stated Govt Total Stated Unspecified
Vol 2013 Security Security Security Pax Charges
Charge Charge Charges
SYDNEY, AU (SYD) Australia 38 254 039 Private $ – $ 52.66
M4371-FINGER_9781786431851_t.indd 325
BRISBANE (BNE) Australia 21 791 997 Private $ –
VIENNA, AT (VIE) Austria 21 999 820 Priv/Pub $ 10.76 $ 10.76 $ 72.50
BRUSSELS (BRU) Belgium 19 133 222 Priv/Pub $ 8.99 $ 8.99 $ 28.15
SAO PAULO (GRU) Brazil 36 460 923 Priv/Pub $ – $ 12.22
TORONTO ON, CA Canada 36 037 962 Airport Authority $ 25.91 $ 25.91
(YYZ)
VANCOUVER BC, Canada 17 971 883 Airport Authority $ 1.48 $ 25.91 $ 27.39
CA (YVR)
325
CALGARY (YYC) Canada 14 316 074 Airport Authority $ 25.91 $ 25.91
MONTREAL (YUL) Canada 14 095 272 Airport Authority $ 25.91 $ 25.91
EDMONTON (YEG) Canada 6 983 229 Airport Authority $ 2.80 $ 25.91 $ 28.71
OTTAWA (YOW) Canada 4 578 591 Airport Authority $ 2.63 $ 25.91 $ 28.54
HALIFAX (YHZ) Canada 3 585 864 Airport Authority $ 3.18 $ 25.91 $ 29.09
WINNIPEG (YWG) Canada 3 483 898 Airport Authority $ 25.91 $ 25.91 $ 12.96
SANTIAGO (SCQ) Chile 2 100 000 Public $ 5.06 $ 0.78 $ 5.84 $ 11.24
BEIJING (PEK) China 83 712 355 Public $ 2.03 $ 2.03 $ 11.86
HONG KONG, China 59 609 414 Public $ 4.44 $ 4.44 $ 16.14
HK (HKG)
SHANGHAI (PVG) China 47 189 849 Public $ 2.03 $ 2.03 $ 11.86
XIAMEN, CN (XMN) China 19 753 016 Public $ 2.03 $ 2.03 $ 11.86
COPENHAGEN (CPH) Denmark 24 067 030 Priv/Public $ 8.08 $ 1.12 $ 9.20 $ 8.21
CAIRO (CAI) Egypt 15 741 000 Public $ – $ 28.18
14/11/2017 11:14
Table 16.1 (continued)
Airport Country Annual Pax Governance Stated Pax Stated Govt Total Stated Unspecified
Vol 2013 Security Security Security Pax Charges
M4371-FINGER_9781786431851_t.indd 326
Charge Charge Charges
HAMBURG (HAM) Germany 13 484 628 Crown Corp $ 1.48 $ 4.99 $ 6.47 $ 13.80
FRANKFURT, DE Germany 58 036 948 Crown Corp $ 1.69 $ 9.18 $ 10.87 $ 34.16
(FRA)
BERLIN, DE (TXL) Germany 19 591 849 Public $ 0.92 $ 7.47 $ 8.39
DUSSELDORF (NRN) Germany 21 228 226 Priv/Pub $ 5.08 $ 5.08 $ 8.04
STUTTGART (STR) Germany 9 577 551 Priv/Pub $ 1.93 $ 6.58 $ 8.51 $ 15.86
MUNICH (MUC) Germany 38 672 644 Public $ 1.09 $ 7.44 $ 8.53 $ 25.19
ATHENS (ATH) Greece 12 944 041 Crown Corp $ 6.70 $ 6.70 $ 16.29
326
JAKARTA (CGK) Indonesia 59 701 543 Public $ – $ 13.97
TEL AVIV (TLV) Israel 14 227 612 Public $ – $ 32.11
ROME (FCO) Italy 36 165 762 Private $ 3.56 $ 3.56 $ 43.23
MILAN (LIN) Italy 9 229 890 Private $ 3.52 $ 3.52 $ 26.39
TOKYO (NRT) Japan 35 341 341 Crown Corp $ 5.56 $ 5.56 $ 22.36
KUALA LUMPUR (KUL) Malaysia 37 704 510 Crown Corp $ 1.82 $ 1.82 $ 17.94
SCHIPOL (AMS) Netherlands 52 569 250 Crown Corp $ 17.91 $ 17.91 $ 21.07
AUCKLAND, NZ (AKL) New Zealand 14 862 485 Crown Corp $ 9.89 $ 9.39 $ 25.42
OSLO (OSL) Norway 22 956 540 Crown Corp $ 5.17 $ 5.17
MOSCOW (DME) Russia 30 760 000 Private $ 6.30 $ 6.30
RIYADH (RUH) Saudi Arabia 18 300 000 Public $ 1.10 $ 1.10 $ 13.34
SINGAPORE (SIN) Singapore 35 840 000 Public $ 6.26 $ 6.26 $ 15.58
CAPETOWN (CPT) South Africa 8 434 799 Crown Corp $ 1.57 $ 1.57 $ 48.37
14/11/2017 11:14
MADRID (MAD) Spain 39 710 903 Public $ 0.81 $ 5.28 $ 6.09 $ 31.27
PALMA MALLORCA Spain 22 768 082 Public $ 5.28 $ 0.81 $ 6.09 $ 13.82
M4371-FINGER_9781786431851_t.indd 327
(PMI)
BARCELONA (BCN) Spain 35 210 735 Public $ 5.06 $ 0.78 $ 5.84 $ 19.71
BANGKOK (BKK) Thailand 51 363 451 Crown Corp $ – $ 22.06
ISTANBUL (IST) Turkey 37 406 025 Private $ – $ 15.00
DUBAI (DXB) UAE 50 977 960 Public $ 1.42 $ 1.42 $ 21.30
LONDON, GB (LHR) UK 72 368 030 Private $ – $ 62.30
MANCHESTER (MAN) UK 20 751 581 Public $ 7.27 $ 7.27
LONDON, GB (LGW) UK 34 200 000 Private $ – $ 19.26
GLASGOW (GLA) UK 7 363 764 Private $ 0.03 $ 0.03 $ 18.42
327
BIRMINGHAM (BHX) UK 9 120 201 Private $ 0.55 $ 0.55 $ 17.87
ATLANTA (ATL) USA 94 431 224 Public $ 2.61 $ 2.61 $ 35.88
LOS ANGELES (LAX) USA 66 667 619 Public $ 2.61 $ 2.61 $ 35.88
DALLAS (DFW) USA 60 436 739 Public $ – $ 35.88
NEW YORK (JFK) USA 50 423 765 Public $ 2.61 $ 2.61 $ 35.88
SAN FRANCISCO (SFO) USA 45 011 764 Public $ 2.61 $ 2.61 $ 35.88
MIAMI FL, US (MIA) USA 40 563 071 Public $ 0.64 $ 2.61 $ 3.25 $ 35.88
SAN DIEGO CA, US USA 17 710 241 Public $ 2.61 $ 2.61 $ 35.88
(SAN)
14/11/2017 11:14
328 Air transport liberalization
$35.00
Canadian airports
$30.00
Total delineated charges (2014 $US)
$25.00
$20.00
Schiphol
$15.00
Vienna
San Diego Frankfurt
$10.00
$5.00 Atlanta
$–
0 20 000 000 40 000 000 60 000 000 80 000 000 100 000 000
Annual passenger volume
$35.00
security charge (2014 $US)
Total delineated airport
$30.00
$25.00
$20.00
$15.00
$10.00
$5.00
$0
Airport Authority
Airport Authority
Airport Authority
Airport Authority
Crown Corp
Crown Corp
Crown Corp
Crown Corp
Crown Corp
Priv/Pub
Priv/Pub
Private
Private
Private
Public
Public
Public
Public
Public
Public
Public
Public
Public
Public
Governance Structure
airports all levy essentially the same charge, but this differs from the
charges levied by public airports in other countries).
The figures provide some indication of the variance in stated passen-
ger charges for airport security across airports, and they appear to show
Canada’s airports as charging significantly more than airports elsewhere.
However, many airports do not explicitly state their security charges,
and even those that do have explicit security charges also have ‘passen-
ger service fees’ that could include additional security-related fees. For
example, London Heathrow Airport (like almost all UK airports) does not
specify a passenger security charge; however, the airport does levy a pas-
senger service charge of $62.30. We have been unable to determine which,
if any, security-related charges are included in this fee. Similarly, Brussels
International Airport delineates a security charge of $8.99, but also has
unspecified passenger service charges of $28.15.
Canada’s airports are the only ones that do not have any unspecified
passenger service charges. Therefore, it seems likely that at least part of the
seemingly high charges for airport security in Canada is overstated relative
to charges in other countries. The lack of transparency in data on airport
charges for security worldwide is a real problem that prevents any proper
analysis and comparison.
Sp
Ssub
P1
Subsidy
P2
Ds
Dp
Qp Qs Q
CONCLUSIONS
ACKNOWLEDGEMENTS
initiated through the efforts of victim families of the Air India tragedy of
1985, in which 329 innocent people lost their lives.
NOTES
1. Using something like ASMs (available seat miles) as a measure of the opportunity for
incidents would bias (upward) the measure of safety.
2. It is unlikely that the metric could be calculated for an airport, since data is too thin.
3. A flight would not be counted that may be damaged or blown up due to an air-to-
ground rocket attack.
4. In the US, the Transportation Security Administration is part of the Federal Department
of Homeland Security, whereas the Canadian Air Transport Security Agency is an inde-
pendent agency established by, but at arm’s length from, the federal government. In the
US most security screening agents at airports are federal government employees, with
some airports using contractors under the Screening Partnership Program, whereas
in Canada they are employees of private security companies contracted by individual
airports.
5. Details of the econometric results can be found in Gillen and Morrison (2015).
6. Aviation Security in the European Union is set out in EC Regulation No. 300/2008,
which came into full effect on 29 April 2010.
7. A good example of differences in variability in screening processes is screening people
versus screening baggage; baggage moves at a constant and higher speed.
8. It is often pointed out that many more people died on the ground than on aircraft as a
result of the 9/11 attacks.
9. Source: http://www.rdcaviation.com/.
10. Airports Council International Europe report that publicly owned airports handled 52
per cent of Europe’s passenger traffic. This in part indicates how private investment has
been drawn to larger airports in Europe.
11. In addition to the indirect costs listed in Table 16.2 we can add negative economy-wide
effects such as stock market reaction, insurance rates and potential suppression of
foreign direct investment through changes to perceived risk.
12. The diagram does not proscribe governance issues, that is, there is no argument pro-
vided here to support government provision (separate from government financing) of
aviation security.
13. IATA (October 2013). Available at: https://www.iata.org/publications/ceo-brief/oct-
2013/Pages/aci-iata-mou.aspx.
REFERENCES
Air navigation services provision and its main components, notably air
traffic management (ATM), is one of the three elements of civil (and mili-
tary) aviation, along with air transport and airports. Together, these three
elements form the complex and dynamic socio-technical aviation system.
In this chapter, we will focus on the core function of air navigation ser-
vices provision, namely ATM. ATM is typically performed exclusively at a
national level by a so-called air navigation services provider (ANSP) that,
with very few exceptions, is a national entity with a level of autonomy that
varies from country to country. However, our focus is not on the different
ANSPs and their reforms and transformations. Instead, we look at the core
function of air navigation services provision, namely ATM, and examine
how ATM evolves, could evolve, or would need to evolve, both in parallel
to and in support of rapidly growing air transport, transforming airport
operations, and the changing entire socio-technical aviation system.
There is a need for ATM to evolve. Solely from an economic point of
view, it can easily be seen that almost all elements of the aviation value
chain have been optimized, ranging from aircraft producers (airplanes are
now standardized products) to highly efficient air transport operations,
to optimized aircraft maintenance, to highly successful yield management
techniques, to increasingly efficient airport operations, and so on. ATM is
probably one of the last, if not the last element of the aviation value chain
waiting to be optimized, that is, being looked at from an economic perspec-
tive rather than just from safety and national sovereignty points of view.
This argument has been made by many authors who have subsequently
tended to advocate for various forms of liberalization of ATM and air
navigation services provision, yet generally end up in simplistic and, in our
view, ideological solutions when calling for the privatization of ANSPs or
some other free market constructs (see ATM Policy Institute, 2016).
339
Here we do not take this point of view and instead take a (socio-
technical) systems approach. Such an approach not only helps us see that
air transport (civil and military), airports, and air navigation services form
a single system, but, more importantly, recognizes the increasingly inte-
grated global or at least regional (for example, North America, Europe,
South East Asia) nature of such a system, to the point that current,
national air navigation services provision is not only anachronistic, but
also inefficient and sub-optimal. Accordingly, we focus on Europe as an
integrated aviation region and on the efforts of the European Commission
(EC) since 2000 to create a so-called ‘Single European Sky’ (SES), mainly
by way of rules and regulations, as the EC typically does. Such efforts have
been a direct response to the recognition of Europe’s overly fragmented
airspace in light of increasingly integrated (European) air transport and
subsequent traffic growth. In this sense, Europe – particularly its SES
initiative – clearly constitutes the only laboratory of evolving ATM.
However, as we will also see, the EC’s SES initiative has ended up in grid-
lock (Baumgartner and Finger, 2014), leaving the problem unsolved, but
also, because of the pressing need to evolve ATM, opening up new and
more technologically oriented avenues for adapting ATM to the needs of
the European and global aviation systems.
Therefore, this chapter aims to offer a comprehensive and systemic
understanding of the challenges facing ATM today. We look at the case
of Europe, where the differing interests of the main actors, namely the EC
and some member states, have led to a trial-and-error approach, making
Europe an interesting laboratory of evolving ATM thanks to an elaborate
system of European-wide rules and regulations. The remainder of the
chapter is structured as follows. In the first section, we define the issue and
explain the main functions of ATM. In a second section, we will briefly
outline the conceptual framework underlying our approach to evolving
ATM. In the third section, we describe how ATM evolved up to the year
2000, which marked the beginning of the Single European Sky project. The
fourth section presents the EC’s SES initiative, which we call the ‘labora-
tory of ATM’, along with its gridlock today. In the fifth section we identify
the emerging, more technology-focused attempts as a way out of this grid-
lock. The final section concludes.
WHAT IS ATM?
From the moment the doors of a commercial flight close at the departure
gate until they open again at the arrival gate, the crew is controlled by
an air traffic controller who guides the aircraft through the airspace. A
times a minute. The core function here, as performed by the air traffic
controller, is to separate air traffic, that is, to separate the different
airplanes from one another within a given control sector and, by
doing so, to guide them safely from one control sector to the next
one. Over deserted regions and oceans, radar equipment is not avail-
able and other more procedural means of separation will be applied.
●● Air traffic flow management (ATFM) has been put into place
because each control sector or ATC unit – and each air traffic
controller for that matter – can only handle a limited number of
airplanes (typically 10–30) simultaneously. This leads to the defini-
tion of so-called ‘air traffic control slots’, of which there is a limited
number within each control sector. Furthermore, each airport also
has (physical) limits in terms of the number of airplanes that can
either take off or land, thus defining what is called ‘airport slots’.
This is in addition to the fact that airplanes can only remain in the
air for a limited amount of time for safety reasons and should fly
as directly as possible from A to B for both economic and environ-
mental reasons. Consequently, the capacity of both airspace (air
traffic control slots) and airports (airport slots) must be balanced
against the demands of aircraft wishing to fly. Such balancing is pre-
cisely the function of ATFM, which is performed simultaneously at
various levels. At the highest level, this is currently the responsibility
in Europe of the so-called ‘network manager’, that is, it is handled
operationally by an organization called Eurocontrol. However, bal-
ancing between capacity and demand also takes place at the FIR
level by the ANSP and at the control center level by the ATC unit.
While air traffic has increased exponentially, these three functions – defin-
ing the airways, separating aircraft, and balancing capacity and demand
– have remained constant. So too has the conceptual approach to, the prac-
tice of, and the technology for ATM. What has changed is the amount of
air traffic, which will continue to increase with the forecasted rapid growth
of unmanned air vehicles (UAVs). As ATC units (can) handle more or
less the same number of aircraft given the current (radar) technology, the
size of the control sectors has decreased – causing the number of control
sectors to increase steadily – to be able to handle this growing traffic. This
has led to exponentially growing interfaces between ATC units, diminish-
ing the overall performance of ATM per flight. Therefore, we will inevita-
bly reach a point where a qualitative change in the way ATM works will
become necessary. The EC had understood this during the late 1990s, but
has mainly looked at changing the rules governing ATM thanks to its so-
called Single European Sky (SES) project (see EC, 1996 and EC, 1999),
CONCEPTUAL FRAMEWORK
This section will briefly outline our conceptual framework. This framework
is based on the idea of a co-evolution between institutions (rules) and tech-
nology (Crettenand and Finger, 2013; Finger et al., 2005), as well as on the
assumption that there are certain critical technical functions that must be
fulfilled for this socio-technical system to work (Finger and Künneke, 2007).
In any socio-technical system, rules evolve – in principle, at least – in par-
allel to technology so as to make the system perform. Performance, in turn,
depends upon a series of critical technical functions that must be fulfilled.
For each of these critical technical functions there is a need for a certain
degree of alignment or coherence between the technology used and the rules
applied. The degree of coherence or alignment among these critical technical
functions will determine the overall performance of the system. This system
is also dynamic and its dynamism results from the (strategic) behavior of
the various actors involved, whether they be market actors, institutional
actors (such as regulators or policy-makers) or technological actors (such as
technology developers). Overall, institutions evolve from government (top-
down rules) to governance (negotiated rules among the relevant involved
actors). Similarly, and in parallel, technology evolves from more centralized
technologies to ever more distributed technologies, notably also thanks to
the pervasive nature of the information and communication technologies.
Figure 17.1 summarizes this broad conceptual framework.
We apply this conceptual framework to ATM. As we have seen in the
first section, there are three critical technical functions that must be ful-
filled for ATM to perform: defining the airways; separating aircraft; and
balancing capacity and demand. Each function is fulfilled via a combi-
nation of rules and technology. Both the rules and the technologies that
have been applied so far will be discussed in the next section. Here, we will
simply summarize the main drivers of the dynamics of this ATM socio-
technical system (see Figure 17.2):
Institutions
(rules)
Institutional
Actors
Market
Coherence
Actors
Performance
Technology
Actors
Technology
Push: ATM
Market Market &
- Air traffic growth performance
Actors demand pull
- Efficiency concerns
Technological changes:
Technology SESAR
Actors
navigation, standards are needed for: (7) dedicated technology; and (8)
performance-based navigation. It is the presence or absence of such stand-
ards that determines the technological evolution in these eight areas, as
well as, more generally, the performance of ATM resulting from interoper-
ability and seamlessness at the global level (Sudarshan, 2003).
To achieve global interoperability and a globally seamless sky, it is very
important that all the signatory states to the ICAO follow its Standards
and Recommended Practices as closely as possible. With the advent
of the Single European Sky and its regulations and standards, it has
become crucial that ‘regional’ (European) standards match the global ones.
Therefore, the European Commission coordinates between ICAO, EASA,
the SESAR JU and Eurocontrol.4 If, under ICAO, a state has an opt-out
possibility, this is no longer possible under the SES. Therefore, it is essen-
tial that SES standards meet at least the ICAO standards and that, when
the SES standards go beyond the ICAO standards, they do not jeopardize
the global harmonization efforts of ICAO.
In this section we will present the SES process, that is, the various steps
and measures by which the EC is trying to create a Single European Sky.
This is done in three steps: SES I, SES II, and the Performance Regime. We
will also discuss the most recent developments, specifically the European
Commission’s ambition to present a new so-called SES II1 package.5
specific areas: (1) performance; (2) technology; (3) safety; and (4) capacity.
For technology a SESAR (Single European Sky ATM Research) master-
plan was drawn up, while EASA (the European Aviation Safety Agency) was
declared responsible for safety.6 As it does in all other sectors, the European
Commission periodically reviews the application of its actions (EC, 2007).
Consequently, it requested the newly created PRC (Performance Review
Commission) to assist in this evaluation. The PRC delivered its first report
in December 2006, recommending, among other things, the acceleration of
the FABs, corresponding technology development (SESAR), as well as the
empowering of Eurocontrol, especially in its pan-European functions and
ATM network design. As a result, in 2009 the Commission proposed a SES
II package. This package identified fragmentation as the major bottleneck
in improving the performance of the European aviation system.
The SES II adopts a total system, gate-to-gate approach for air navigation
services via the performance scheme, a strengthening of FABs, as well as a
central network management function. It is based on closely interrelated
pillars that all converge towards the keystones of aviation performance and
sustainability. The safety pillar extends the competence of the EASA to
ATM and airports. The technological pillar, SESAR, aims to develop state-
of-the-art technology, supporting member states in their efforts to improve
performance, safety and sustainability. To that effect, a new organization
called SESAR-JU (joint undertaking) has been established. The airport
pillar, comprising the creation of the Airport Capacity Observatory, is
complemented by terminal ANS performance indicators as part of the
performance scheme. Finally, the human factor pillar is considered as
the ‘overarching SES pillar’, as it is acknowledged by all stakeholders that
the SES cannot be successful without the full buy-in and support of the staff
concerned.7
The performance targets set out under the so-called ‘SES performance
scheme’ should reflect and foster improvements in all five SES pillars, as
well as from individual FABs, member states, ANSPs, airports and air-
space users. There are three specific targets: (1) the environmental target,
which will ensure carbon-neutrality of aviation growth insofar as en-route
ANS is concerned; (2) the capacity target, which is designed to avoid major
disruptions and indirect costs for airspace users and their passengers from
ANS delays; and (3) the cost-efficiency target, which will secure progres-
sive improvement in unit rates and maintain route charges nearly constant
over the period. A second reference period started in January 2015 and will
last until the end of 2019.9 The first year of RP2 (Reference Period 2) has
passed and the corresponding monitoring report was published during the
first quarter of 2017.10
Virtual air traffic control (VATC) and cloud-based services (CBS) are still
treated as two different technological developments. This may be mainly
because they are being developed by different technology providers, each
pushing for their own technological ends. In reality, VATC is a tool, spe-
cifically a tool for standardization, that constitutes the basis for services
offered via the cloud. Sooner or later, both may well merge into something
called a cloud-based ATM infrastructure.
More precisely, VATC is a method of providing air traffic control
services through common and standardized interfaces from a location-
independent virtualized environment. VATC is composed of two main
technologies: the common controller cockpit and a protected ATM infor-
mation cloud.13 Both enable functional and geographical decoupling of
ATM data service providers (ADSPs) from air traffic service units (ATSUs)
by introducing open and modular service architectures. As conceptualized
in the SESAR Master Plan, these services include surveillance, flight
data, trajectory, correlation and coordination.14 Currently, a consortium
of seven ANSPs and four industry partners are working on the maturity
and demonstration phases.15 Skyguide, the Swiss ANSP, has progressed to
realize VATC, in direct response to overcoming the SES gridlock, and is
now implementing it in between its two ATC units at Geneva and Zurich,
thus standardizing the technologies and operational procedures of the two
ATCs. Consequently, ATC can be provided independently of location and
the airspace user has only a single interface. In short, VATC is expected to
provide common operational requirements, cross-border service provision
and, of course, better performance.
CBS are air navigation services that use the principle of shared alloca-
tion of computing resources, such as processing power, storage and ser-
vices. This technology, originating from the IT industry, enables the service
provision of dynamic computing resources regardless of location. Future
ATM systems will indeed rely on integration and automation, using the
ATM information cloud as the backbone. Currently, large ANSPs are
migrating from legacy Flight Data Processing (FDP) systems to cloud-
based FDP services. These services also make standardization and con-
solidation processes possible among ANSPs (Kampichler and Eier, 2012).
A contemporary example of this technology is co-flight cloud services
(CCS).16 Co-flight e-FDP is a flight data processing system that enables
remotely accessible flight data.
VATC and CBS are both expected to bring benefits that are aligned
with SESAR’s objectives by integrating FDP systems among ANSPs and
strengthening the common vision of aircraft trajectories. Reducing the
Remote Tower
ardize the technology, EASA and EUROCAE have been working on speci-
fications and regulation for RTWR.
Centralized Services
Because the SES project has not delivered on its original promises,
European ATM is now passing through a challenging process of new tech-
nological developments, driven by a variety of actors. Advancements in the
three ATM technologies are opening up a new frontier for European and
probably also global ATM. All three of the technologies presented above
have the potential to defragment European ATM, but they will also lead to
consolidation of ANSs and ANSPs because of economies of scale (cost-
effectiveness) and broader performance improvements. Their adoption,
the pace of their implementation, and ultimately their disruptive nature
will depend largely upon current and future regulations, and especially
upon how the different stakeholders will react and behave.
The main actors (stakeholders) in European ATM all face these three
and even other different technological options and, consequently, face dis-
ruption. However, their reactions will vary depending on how they antici-
pate winning or losing in the process of such technology adoption. For
example, Eurocontrol is already trying to centralize air navigation support
services, whereas ANSPs are trying to consolidate some of these services
among themselves. Regulations play a key role in determining the outcome
of the European ATM laboratory. With proper regulations, the European
technological ATM experiment could be successful without consuming
too much time and too many resources. In this regard, in-depth analysis of
the consequences of such potentially disruptive technological innovations
should be carried out before moving on with separate experimentations by
the different actors. Moving towards a defragmented and technologically
intelligent Pan-European ATM is a gradual process, where the alignment
of interests of the actors with these technologies is required (Cook et al.,
2015).
Still, there is general agreement that, to obtain performance improve-
ments in European ATM, future ATM services should be more standard-
ized, more integrated, and probably more centralized, and that technology
NOTES
1. The ‘Q’ code is a standard collection of three-letter codes that all start with the letter
‘Q’. It was initially developed for commercial radiotelegraph communication and later
adopted by other radio services, such as for maritime communications to promote radio
clarity, particularly for international speakers. The code set was adopted by a conven-
tion in London in 1912, in which codes in the range QAA–QNZ were reserved for
aeronautical use. Although their use is quite rare these days, a few terms persist, such as
‘QNH’ in aviation to promote unambiguous communications.
2. Not to be confused with Secondary Surveillance Radar (SSR), the first of which in
Europe was installed in France in 1962.
3. Automatic dependent surveillance–broadcast is a surveillance technology installed on
board equipped aircraft reporting in a ‘dependent’ (from the aircraft) way, different
parameters, like position and altitude. Compared to an independent ground surveil-
lance system, however, it is limited by its dependence on the aircraft’s technological
capabilities.
4. http://ec.europa.eu/transport/modes/air/single_european_sky/co-operation_icao_en.
5. This section is based on Baumgartner and Finger (2014).
6. EASA was created in 2002 and came into full functionality in 2008. The responsibili-
ties of EASA include to conduct, analyze, and research safety; provide authorization
for foreign operators; advise regarding the drafting of EU legislation; implement and
monitor safety rules (including inspections in the member states); provide certification
of aircraft and components; and provide approval for organizations involved in the
design, manufacture and maintenance of aeronautical products.
7. PRB: Proposed regulatory approach for a revision of the SES Performance Scheme
addressing RP2 and beyond, Version 1, 1.3.2012.
8. Performance Review Commission and Eurocontrol: Evaluation of the Impact of the
Single European Sky Initiative on ATM Performance, 21 December 2006.
9. EC 390/2013 article 8.
10. https://ec.europa.eu/transport/modes/air/single_european_sky/ses-performance_en.
11. http://europa.eu/rapid/press-release_SPEECH-1 2–711_en.htm.
12. http://fsr.eui.eu/event/8th-florence-air-forum/.
13. Skyguide. The Virtual Centre Model. Skyguide, Geneva, 2014.
14. http://ec.europa.eu/transport/sites/transport/files/modes/air/sesar/doc/eu-atm-master-
plan-2015.pdf.
15. http://www.eurocontrol.int/news/demonstration-world-first-virtual-centre.
16. DSNA. DSNA Services Coflight as a Service. DSNA, 5 March 2015. Accessed 20
January 2017 at http://dsnaservices.com/portfolio/coflight-as-a-service/.
17. Skyguide. Blueprint: The 2017 Skyguide Technology Outlook. Skyguide, Geneva, 2017.
18. Directive (EU) 2016/1148 of the European Parliament and of the Council of 6 July 2016
concerning measures for a high common level of security of network and information
systems across the Union.
19. http://www.frequentis.com/fileadmin/content/Brochures/ATM/2016/RVT_whitepaper.
pdf.
20. Eurocontrol. Centralised Services CONOPS. Eurocontrol, Brussels, 2014.
21. Eurocontrol. The challenges and premise of implementing centralised services. Skyway,
Eurocontrol, Brussels, 2015.
22. http://ec.europa.eu/research/participants/portal/desktop/en/opportunities/h2020/topics/
sesar.ir-vld.wave 1–1 8–2015.html.
23. M. Baumgartner and A. Smoker, ‘Will unmanned vehicles modernise Air Traffic
Management?’, in Regulating Drones – Creating European Regulation that is Smart
and Proportionate, European Transport Regulation Observer 3/2015.
24. The era of space-based surveillance is here! Air Traffic Management, 14 January 2017.
25. ICAO Separation And Safety Panel (SASP) 1, 16 November, Information Paper 04.
26. Tech giants explore connectivity potential of unmanned platforms, IHS Jane’s Airport
Review, February 2017.
REFERENCES
ATM Policy Institute (2016). The case for liberalising air traffic control. Available at
http://atm01.drack-design.ch/wp-content/uploads/2016/12/The_ATM_Policy_
Institute_Report.pdf.
Baumgartner, M. (2007). The organisation and operation of European airspace. In
A. Cook (ed.), European Air Traffic Management. Farnham: Ashgate, pp. 1–34.
Baumgartner, M. and M. Finger (2014). The Single European Sky gridlock: a dif-
ficult 10-year reform process. Utilities Policy, 31, 289–301.
Cook, A., H.A.P. Blomb, F. Lillo, R.N. Mantegna, S. Miccichè, D. Rivas,
R. Vázquez, et al. (2015). Applying complexity science to air traffic management.
Journal of Air Transport Management, 42, 149–58.
Crettenand, N. and M. Finger (2013). The coherence between institutions and
technologies in network industries. Competition and Regulation in Network
Industries, 14, 106–29.
INTRODUCTION
LIBERALIZATION INITIATIVES
Overview
359
New Zealand
New Zealand has become a well-studied case on the field of ANS liberali-
zation, as it was the first country to do so in 1987 with the creation of a
for-profit government corporation, Airways New Zealand. It was also the
first country to separate ANS operations from safety oversight, a principle
that both the ICAO and the EU now recommend. While a small system
in terms of traffic (around one million flights per year), it has a very large
continental and oceanic airspace of around 30 million km2.
The commercialization of ANSs in New Zealand came at a time when
the country was going through macroeconomic difficulties. The govern-
ment at the time decided to go through a number of reforms, which
included the creation of Airways New Zealand. The company is owned
by the Minister of Finance and the Minister of State-Owned Enterprises,
while the Minister of Transport performs safety oversight duties.
The company is a for-profit commercial enterprise, paying dividends
to the national government as a result. Its funding comes from user fees
charged to airspace users (including some charges for general aviation),
and it has access to private capital markets for financing needs. Economic
regulation is performed by the Commerce Commission of New Zealand.
In the almost three decades since it has been created, Airways New Zealand
has always had an operating profit (NZD$15.1 million in 2015, with rev-
enues of NZD$186.3 million), and has paid around NZD$40 million
(nominal values) in dividends to the New Zealand government since 1987,
including NZD$4 million in 2015.
Australia
While there are a few studies that deal with the issue of ANSP benchmark-
ing,4 the literature dealing specifically with the effects of ANSs’ liberaliza-
tion is somewhat scarcer. Additionally, most of the works in this field are
case study approaches, with little economic analysis. With those caveats
present, this section will provide an overview of the works in the area.
The case study approaches include, among others, Majumdar (1995),
Dempsey et al. (2005), United States Government Accountability Office
(2005), mbs Ottawa Inc. (2006), Oster and Strong (2007) and Neiva
(2015).5 The first of these works, Majumdar, was the most sceptical about
commercialization. The study noted that the pressure to reduce costs
would inevitably lead to safety issues, which was a common theme at the
time. As the later works reported, time would assuage this fear, and com-
mercialized ANSPs have followed the overall trend of the industry with
ANSs becoming safer regardless of governance status. Overall, these works
concluded that institutional changes were mostly positive, and at worst
neutral. Positive outcomes include improvements in customer service,
financial stability with great control of increasing costs and more capabil-
ity to finance and implement modernization projects. However, it has been
argued that some of these changes might not result directly from all aspects
of commercialization, but simply because of greater managerial and finan-
cial independence, which might happen without commercialization. Still,
taking political realities into consideration, commercialization appears to
be the better way to achieve such independence.
On the works that provide an econometric approach to the problem,
Button and Neiva (2014a, 2014b) and Neiva (2015) performed several ana
lyses of the European system using Data Envelopment Analysis (DEA) and
Stochastic Frontier Analysis (SFA). The results showed that, overall, more
governmental control was associated with better economic efficiency. The
authors offered three explanations for the results. One possibility is that
since those commercialized entities are still subjected to full cost recovery
every year, they are subjected to rate-of-return regulation6 (with a return of
zero in this case, as they have full cost recovery requirements). This form of
regulation has been associated with the Averch–Johnson effect (Averch and
Johnson, 1962), which leads to lower efficiency through suboptimal high
capital to labour ratios. Another explanation considered by the authors
was that the ANSPs that were first commercialized were the ones under-
performing, and were still ‘catching up’ to their peers in terms of efficiency.
Additionally, Neiva (2015) considered that perhaps some of the corpora-
tized ANSPs operated de facto like governmental agencies, clouding the
distinction between commercialized and non-commercialized entities.
Conclusion
Privatization
branch, and funded by taxation. Back then, many in the industry, includ-
ing airlines and labour unions, were displeased with the state of affairs; the
airlines felt that the system was not keeping pace with the growth in traffic
and labour unions were unhappy with salary freezes lasting for several
years.
After a consultation process that discussed several options (including
full privatization and contracting out), it was decided to create a non-profit
private entity to run the system. That company would be totally independ-
ent from the Canadian government, having full control of its manage-
ment and budget, including the ability to charge user fees and issue debt.
Governance would be provided by a number of aviation stakeholders,
including airlines, general aviation7, labour unions, the government and
independent members.
Following this consultation process NAV CANADA was created in
1996, with 6300 federal employees being transferred to the new company.
The government received a payment of CAD$1.5 billion in the privatiza-
tion process. User fees were established after 18 months, and since 1999
they have to provide full cost recovery to the company, a principle that still
exists today. To deal with downturns in traffic, NAV CANADA has a rate
stabilization fund that is replenished during periods of growth in traffic
and is used during downturns. This allows the ANSP to avoid a ‘death
spiral’ in which less traffic leads to higher fees, leading to less traffic. After
the September 11 attacks in the US in 2001, NAV CANADA used all of
the money available in the fund and user fees had to be raised six times over
a period of three years. In the aftermath of the 2008 financial crisis the
fund was used but never depleted, allowing user charges to be maintained
at the same levels. Since 2012 the fund has been maintained at 7.5 per cent
of revenues.
Given its board structure, it makes the company in essence a non-profit
user cooperative, a relatively common structure in North America for utili-
ties. The board structure, with several stakeholders in the governing board,
but none with a majority, also reduces the need for economic regulation.
With the mandate to fully recover costs every year, and with those stake-
holders governing, there is an incentive to keep costs and user fees in check.
Still, every change in user fees requires an extensive consultation process,
and there is an ex post right to appeal user fee changes to the government.
After its creation NAV CANADA had a few tumultuous years, with one-
sixth of the workforce being laid off, most of them being middle managers.
An extensive restructuring and consolidation plan was also put in place, as
well as a capital expenditure programme. Ultimately, that initial transition
period had the help of an expanding economy and aviation market, which
allowed NAV CANADA to fund these restructuring programmes and
please its labour force with more attractive compensation packages. When
the first real test to the system came in 2001 with the September 11 attacks
in the US, the company was much better prepared to withstand the shock.
Conclusion
In the US, ANSs are provided by a federal agency, the Federal Aviation
Administration (FAA). This agency controls the biggest airspace in the
world, with 10.4 million km2 of continental airspace and 65 million km2
of oceanic airspace. Besides being the biggest in area, US airspace is also
the biggest in terms of traffic, with 23.1 million flight hours controlled
in 2015. The agency employs around 46 000 people, from which less than
11 000 are fully certified air traffic controllers.12 In terms of governance
and funding, the system is a holdout among developed nations, as it is still
directly operated by the national government and is funded by taxation,
not user fees. This section will present an overview of the current system,
along with a discussion of the most important attempts that have been
made at reforming it.
In fiscal year 2015, these taxes collected around $14.3 billion, rep-
resenting 92 per cent of FAA’s budget for that year (Federal Aviation
$12.00
$10.00
$8.00
Billions
$6.00
$4.00
$2.00
$–
1990 1995 2000 2005 2010 2015
NextGen
2015 the FAA scaled back its intentions significantly, focusing the agency’s
resources on some key projects that had the most potential to deliver ben-
efits in the short term. Adding to those implementation issues, many air-
space users have been delaying equipping their airplanes with the necessary
technologies for the FAA to take advantage of the NextGen technologies
already in place. Many of the benefits touted by the FAA will only materi-
alize once a critical mass of airspace users is equipped with those technolo-
gies. But until that critical mass comes, it will be hard to prove to the users
that equipage is advantageous.19
Attempts at Reform
had failed because the FAA created rules for itself that mimicked the rest
of the government, ending up in a situation similar to the starting point
(United States Government Accountability Office, 2003).
In 2000 came the most relevant reform to be enacted: the internal
spin-off of the ANS functions within the FAA into a new department,
the Air Traffic Organization (ATO). This reform followed a 1997 report
from the so-called Mineta Commission that recommended the creation
of a performance-based organization within the FAA to provide ANSs
(National Civil Aviation Review Commission, 1997).23 However, the
organization envisioned by the Mineta Commission never came to be. The
report recommended the creation of an entity separated from the federal
budget process, with access to capital markets and financed by user fees.
None of this was enacted. Instead, the ATO was created as a department
of the FAA, but remained part of the federal budget process and funded
by taxes on the aviation system as well as general funds, and ultimately very
little changed in ANS provision.
After the creation of the ATO, the focus has been on the implementation
of NextGen. But that changed in 2013 when political rows over the federal
budget led to a ‘sequester’; that is, a mandatory budget cut throughout the
government, and a government shutdown, in which non-essential func-
tions of government are stopped. These events had significant impacts
for the FAA, as it was forced to stop training new air traffic controllers.
Air traffic controllers were furloughed, leading to traffic delays, and many
capital projects were put on hold. Because of this, since 2013 an interest for
institutional reform resurged in the US, with many in the industry – includ-
ing the majority of the airlines and the air traffic controllers union – openly
supporting reform. Those efforts culminated in early 2016 with a proposal
from the House of Representatives, the ‘Aviation, Innovation, Reform and
Reauthorization Act’ (AIRR), to spin-off ANSs into a non-profit entity.
AIRR would create a federally chartered non-profit corporation, the
ATC Corporation, modelled after NAV CANADA and to start opera-
tions on October 2019. It would be funded by user fees on commercial
airlines (general and business aviation would be completely excluded from
payment) and totally independent of the government. Its governance
would be assured by a board of directors with members from airlines, the
government, labour, general aviation, business aviation and manufactures.
In contrast to what happened in Canada, there would be no payment for
the transference of assets to the new corporation.24
Despite the strong support of the air traffic controllers union and the
majority of the airlines,25 the bill failed in Congress. (In 2017, a similar
bill was once again introduced in the House of Representatives.) As with
other previous efforts, the proposal failed to gain support from the general
aviation community. Given that the bill pre-empted their fear of having to
pay user fees by excluding them from that payment, their opposition was
based on concerns about access to airspace, and that a private provider
would eventually impose restrictions for general aviation operators to fly
as they pleased. Together with reluctance from some members of Congress
in giving up control of the system, the bill suffered the same fate of the
USATS proposal 22 years earlier.
Conclusion
Since the AATF was created in 1970, FAA governance and funding have
remained essentially the same. Despite a myriad of research produced in
the last four decades pointing to the problems with ANS provision in the
US, the system is still provided by a governmental agency funded by taxa-
tion on the aviation system and general funds, appropriated on a short-
term basis. The reforms that have been implemented (procurement reform
in 1996, ATO in 2000) ended up being mostly cosmetic, without significant
improvements.
Today, the focus of the agency is on the deployment of NextGen.
While there have been some successes in its implementation, namely since
the focus was shifted in 2015 to the projects with the most benefits in
the shorter term, serious challenges remain. Overall, the programme is
still underfunded. Even if everything went smoothly at the FAA’s end,
Congressional fights over budgets mean that every so often the agency is
faced with externally imposed delays on the implementation of NextGen.
Every time that happens, a few million dollars are added to the costs of
the ongoing projects, and future projects get delayed. While this status quo
remains, it is hard to envision a future where NextGen delivers at the level
that allows the system to accommodate the expected growth in traffic.
CONCLUSION
NOTES
1. In the case of Canada and the UK, commercialization also resulted in a payment to the
government when the respective commercialized ANSPs were created.
2. Space-based satellite navigation will be discussed in greater depth in the section con-
cerning Canada.
3. For a study on the separation of provision from safety oversight and its outcomes, see
Brown et al. (2014).
4. See, for example, the various works produced by the Eurocontrol Performance Review
Commission (available at http://www.eurocontrol.int/prb/publications) or the Civil Air
Navigation Services Organisation (available at https://www.canso.org/publications).
Additional works include NERA Economic Consulting (2005); Helios Economics and
Policy Services (2006); Button and Neiva (2013); Arnaldo et al. (2014) and Standfuß et
al. (2015).
5. For a more complete bibliography on the issue, see Neiva (2015).
6. For more on rate-of-return regulation, as well as other forms of regulation, see, for
example, Train (1991).
7. General aviation agreement with privatization came with a cost for the company:
general aviation pilots only have to pay between CAD$67.64 and CAD$225.84 annu-
ally for unlimited en-route and terminal access (some larger airports impose a $9.95
surcharge).
8. For an analysis of the relevance of the use of this metric, see Button and Drexler (2006).
9. However, NAV CANADA still reports low usage of this technology. Still, from fewer
than 50 000 messages exchanged in the first quarter of fiscal year 2013, more than
300 000 messages were exchanged in the fourth quarter of fiscal year 2015.
10. Inmarsat is another company intending to offer space-based satellite navigation
commercially.
11. For example, Aireon estimates that on the North Atlantic routes the use of space-based
satellite navigation will allow lateral separation to be cut by half to 15 nautical miles.
12. This is the lowest number in over three decades, and is creating serious concern among
aviation stakeholders, who fear that the agency will be unable to increase those numbers
(they have failed their hiring targets multiple times and around 25 per cent of controllers
are already eligible for retirement and have mandatory retirement by age 56), leading
to delays in the future because of a controller shortage. In many installations across
the US, controllers already have mandatory six-day workweeks because there are not
enough controllers. In some instances, as with New York Terminal Radar Approach
Control (TRACON), those mandatory six-day workweeks have been in place since the
late 1990s.
13. The first efforts of what are now known as ANSs were by a group of airlines that came
together to form a non-profit user cooperative, the Aeronautical Radio, Incorporated
(ARINC), a company that still exists today. It started by providing air/ground commu-
nications, and eventually evolved into providing terminal control and en-route separa-
tion before the federal government took over that role. For more on the history of ANSs
in the US, see Nolan (2010) and Neiva (2015).
14. Until the mid-1990s the agency was also responsible for the promotion of the airline
industry, but by seeing the conflict of interest of having an agency regulating the safety
of an industry and having to promote it, Congress would eventually remove that obliga-
tion from the FAA charter.
15. The FAA, like other agencies in the US federal government, does not have a capital
budget per se; it simply has an account where expenses related to maintaining and
buying equipment on a given year are accounted for.
16. Of costs totalling USD$12.9 billion, fees in the US at the same level as NAV CANADA’s
fees would only raise USD$8.8 billion.
17. While it might be argued that the overall complexity of the US system is higher, par-
ticularly in some high traffic areas like New York City or Chicago, the system is also
much bigger, with seven times the traffic, thus allowing for bigger economies of scale.
Additionally, the Civil Air Navigation Services Organisation, in their benchmarking
report for 2014, also reports a much higher cost per continental flight hour controlled at
the FAA compared to NAV CANADA: USD$465 for the former and USD$321 for the
latter (Civil Air Navigation Services Organisation, 2015).
18. For an extensive list of reports by government overseers on the implementation of
NextGen, see Neiva (2015).
19. For general aviation users, the FAA created a rebate programme where the agency pays
up to $500 of the price of equipping an aircraft with ADS-B Out capabilities.
20. For a more complete discussion of US attempts at governance reform, see Neiva (2015).
21. This initiative, the National Performance Review, was spearheaded by US Vice
President Al Gore. Its goal was to streamline governmental functions and it would make
recommendations in many areas, from public housing to defence and use of informa-
tion technologies throughout the government. For more information on the issue, see
National Partnership for Reinventing Government (2001).
22. Two good examples of this congressional meddling are Amtrak, the medium- and long-
distance rail operator, and the Postal Service. The former has been known to try to end
service to underperforming locations, and the latter has tried to end the delivery of
regular mail on Saturdays. Both initiatives have been stopped by Congress multiple times.
23. The Mineta Commission was very critical about FAA’s operation and congressional
oversight. Among other things, it noted that the federal budget rules were not appropri-
ate for the provision of ANSs and there were ‘too many cooks’ trying to influence the
FAA in the different layers of government (from Congress to the government overseers).
24. This was justified on the basis that those assets were mostly paid by the users of the
system via taxation, so if the ATC Corporation had to buy those assets from the govern-
ment, users would be paying twice.
25. The holdout was Delta Airlines, which opposed the intent to privatize the system. Delta
would eventually quit Airlines for America, the main trade association for the airline
industry, because of the association’s support of privatization.
26. Naturally, the number one focus on any ANSP, liberalized or not, is and should always
be safety.
27. This is despite FAA’s budget for maintenance and modernization being roughly 20 times
bigger than NAV CANADA’s capital budget, while the US system is only seven times as
large, in terms of traffic, as the Canadian one.
28. The FAA, and other ANSPs, would gain even more if they had adopted space-based
satellite navigation from the beginning for both their oceanic and continental airspaces,
instead of installing hundreds of ground stations to collect the ADS-B Out signals that
aircraft emit. But the FAA cannot be faulted for not using a solution that was not avail-
able when ADS-B ground stations started to be installed in 2007, and as such they are
now stuck in paying for a system potentially much more expensive to maintain.
REFERENCES
Air Traffic Management (2016) Nav Canada 2017 Fees to Save Airlines $100m,
accessed 1 August 2016 at http://www.airtrafficmanagement.net/2016/07/
nav-canada-2017-feestosave-airlines-100m/.
Aireon (2016) Aireon and ATNS Sign First Regional Commercialization
Agreement for African Air Traffic Surveillance Data, accessed 1 August 2016
at http://aireon.com/2016/05/03/aireon-and-atns-sign-first-regional-commercial
ization-agreement-for-african-air-traffic-surveillance-data/.
Arnaldo, R.M., V.F.G. Comendador, R. Barragan and L. Pérez (2014) European Air
Navigation Service Providers, Efficiency Evaluation Through Data Envelopment
Analysis (DEA), 29th Congress of the International Council of the Aeronautical
Sciences, St Petersburg.
Australian Aviation (2016) TAAAF Calls on Government to Privatise Airservices,
accessed 1 August 2016 at http://australianaviation.com.au/2016/04/
taaaf-calls-on-government-to-privatise-airservices/.
Averch, H. and L.L. Johnson (1962) Behavior of the Firm Under Regulatory
Constraint, The American Economic Review, 52, 1052–69.
Brown, D., T. Berry, S. Welman and E.J. Spear (2014), CAA International
Structures, McLean, VA: MITRE.
Button, K. and J. Drexler (2006) Are Measures of Air-Misses a Useful Guide
to Air Transport Safety Policy?, Journal of Air Transport Management, 12(4),
168–74.
Button, K. and R. Neiva (2013) Single European Sky and the Functional Airspace
Block: Will They Improve Economic Efficiency?, Journal of Air Transport
Management, 33, 73–80.
Button, K. and R. Neiva (2014a) Economic Efficiency of European Air Traffic
Control Systems, Journal of Transport Economics and Policy, 48, 65–80.
Button, K. and R. Neiva (2014b) European Air Navigation Services Industry
Regulatory Reform and Averch–Johnson Effects, International Journal of
Transport Economics, 41, 33–50.
Civil Air Navigation Services Organisation (2015), Global Air Navigation Services
Performance Report 2015 – 2010–2014 ANSP Performance Results – The ANSP
View, Montreal, QC, Canada: Civil Air Navigation Services Organisation.
Dempsey, P.S., R. Janda, Y. Nyampong, J. Saba and J. Wilson (2005), The McGill
Report on Governance of Commercialized Air Navigation Services, Montreal,
McGill University Center for Research on Air & Space Law.
Federal Aviation Administration (2016a) Airport and Airway Trust Fund (AATF)
Fact Sheet, accessed 1 August 2016 at https://www.faa.gov/about/budget/aatf/
media/AATF_Fact_Sheet.pdf.
Federal Aviation Administration (2016b) FAA Aerospace Forecasts, accessed
1 August 2016 at https://www.faa.gov/data_research/aviation/aerospace_fore
casts/.
Helios Economics and Policy Services (2006) Cost Benchmarking of NATS Relative
to Selected Air Navigation Service Providers – Phase 2 Report, Farnborough and
London: Helios Economics and Policy Services and Civil Aviation Authority –
Economic Regulation Group.
Majumdar, A. (1995) Commercializing and Restructuring Air Traffic Control:
A Review of the Experience and Issues Involved, Journal of Air Transport
Management, 2(2), 111–22.
mbs Ottawa Inc. (2006) Air Traffic Control Commercialization Policy: Has it Been
Effective?, Ottawa, ON: mbs Ottawa Inc.
National Civil Aviation Review Commission (1997), Avoiding Aviation Gridlock
and Reducing the Accident Rate – A Consensus for Change, Washington, DC:
National Civil Aviation Review Commission.
National Partnership for Reinventing Government (2001) A Brief History of Vice
President Al Gore’s National Partnership for Reinventing Government During
the Administration of President Bill Clinton 1993–2001, accessed 1 August 2016
at http://govinfo.library.unt.edu/npr/whoweare/historyofnpr.html.
Neiva, R. (2015) Institutional Reform of Air Navigation Service Providers: A
Historical and Economic Perspective, Cheltenham, UK and Northampton, MA,
USA: Edward Elgar Publishing.
NERA Economic Consulting (2005) Cost Benchmarking of Air Navigation Service
Providers: A Stochastic Frontier Analysis, London: NERA Economic Consulting.
Nolan, M.S. (2010) Fundamentals of Air Traffic Control, 5th edn, Boston, MA:
Cengage Learning.
Office of the Inspector General (2014) ADS-B Benefits are Limited Due to a Lack
of Advanced Capabilities and Delays in User Equipage, Washington, DC: US
Department of Transportation.
Oster, C.V. and J.S. Strong (2007) Managing the Skies: Public Policy, Organization
and Financing of Air Traffic Management, Aldershot: Ashgate.
Standfuß, T., M. Schultz and M. Kreuz (2015) Multidimensional Benchmarking
of Air Navigation Service Providers, Air Transport and Operations Symposium
2015, Delft.
Train, K. (1991) Optimal Regulation: the Economic Theory of Natural Monopoly,
Cambridge, MA: MIT Press.
United States Government Accountability Office (2003) National Airspace System:
Reauthorizing FAA Provides Opportunities and Options to Address Challenges,
Washington, DC: United States Government Accountability Office.
United States Government Accountability Office (2005) Air Traffic Control:
Characteristics and Performance of Selected International Air Navigation Service
379
fare differentials for US–international Gudmundsson, S.V. 244, 246, 248, 250,
routes 7–8, 18–23 262
fares studies
airline liberalization 7–8, 10, 12–24 Herfindahl–Hirschman concentration
empty cores 294 index (HHI)
low fares, as strategy to reduce Canadian 40
market instability 296–7 explanation 255–6
oligopolization of markets 247–8, US domestic 258–9
250, 252–4, 256 US international 258, 260–61
in theories of market instability The Hindu 99, 102–3, 106
292–3 Hooper, P. 93–6, 115, 121, 266
Federal Aviation Administration horizontal aviation agreements 98,
(FAA) 167–8, 181
attempts at reform 370–72 horizontal relationships 309–12
comparison with NAV CANADA hub-and-spoke (HS) systems
359, 368, 372–3 airport capacity utilization/capacity
history, governance and funding constraints 276–82
366–8, 372 background 266–8
Hub Classification scheme 225, hub classification scheme 225, 227
241–2 traffic concentration 268–9, 270–76,
International Aviation Safety 280–81
Assessment programme 168–70
Ireland’s categorization 79 India
Modernization and Reform Act 236 Air Corporations Act 96, 121
NextGen 369–70, 372 Air Deccan 97, 100–101, 115–16,
space-based satellite navigation 373, 122, 124
375 Air India 92, 96–8, 115–16, 121,
fifth freedom 9, 36–7, 41, 45, 48, 63, 123–4, 126, 128–32, 134, 307,
81, 141, 188 337
Findlay, C. 59–60, 62, 115, 121, 123 Air Taxi system 94–5, 97, 121, 124
Finger, M. 206, 340, 343, 356 air traffic management 105–7
foreign ownerships Aircraft Act 103–4
eliminated limitations in Colombia Airports Authority of India (AAI)
175 102–4, 106–7
raised limit in China 118–20 aviation market post-deregulation
restrictions 100–102
Canada 38, 43–4 civil aviation industry
efficiency challenge 23 evolution 92–3, 112, 121–13, 124
post-Brexit scenario 151–2 literature review 115–16
setting limits 172 comparison with China
Forsyth, P. 52–7, 60, 162, 206 efficiency performances of airlines
Fu, X. 24, 114, 118 128–32, 134–5
full-service network carriers (FSNCs) overview 123–8
274–7, 281 policy environments 132–5
Director General of Civil Aviation
Gelhausen, M.C. 268, 277 (DGCA) 92, 96, 104, 106–7,
Gillen, D. 308, 337 126
Goetz, A. 76–9, 85, 266 East West Airlines 94–6, 121, 124
Goldstein, A. 115, 121, 123 environmental impact 107–8
greenfield airports 103–5, 107 foreign ownership limits 44