-
1
First Draft. Please do not quote without permission.
Comments are most welcome.
Hybrids within Hybrids:
A Challenging Organizational Arrangement in
the Airline Industry
Claude Ménard and Gézia Damergy
Centre d’Economie de la Sorbonne,
Université de Paris (Panthéon-Sorbonne)
EMNET CONFERENCE, AGadir (Morocco),
November 21-23 2013
-
2
Hybrids within Hybrids:
A Challenging Organizational Arrangement in the Airline
Industry
Claude Ménard and Gézia Damergy
Centre d’Economie de la Sorbonne,
Université de Paris (Panthéon-Sorbonne)
1. Introduction.
Strategic alliances on international markets are now a dominant
feature in the airline
industry. Restrictions on cross border mergers, fed by national
pride embedded in tight
regulation, and the need to offer a seamless service “from
anywhere to anywhere” with
efficient worldwide connections have pushed major airlines to
take advantage of the
deregulation movement of the 1980s-1990s to build networks in
order to achieve significant
economies of scope and density. The three international global
alliances nowadays provide
about 60 % of air transport capacities.
In what follows, we refer to these alliances as “hybrid” modes
of organization, in line with
recent developments in organization theory inspired by
Williamson (1996).1 By hybrids, we
mean organizational arrangements between two or more economic
entities that put
together significant decision rights and, in numerous cases
property rights, while remaining
legally and economically distinct, each entity keeping control
over key assets in last resort.
The intensity of shared rights largely determines the type of
hybrid at stake and the
modalities of governance chosen. Although there is already an
abundant literature available
on airline alliances, not much is available on these issues of
allocation of rights and the
associated problems of internal organization and governance
1 For a detailed review and discussion of this literature, see
Ménard (2012).
-
3
Recent evolution in the airline industry has made the picture of
these hybrid arrangements
with their overlapping rights even more complex. Indeed, subsets
of members within the
same alliance are now developing specific agreements, with much
tighter coordination and
the acceptance of a complex set of rules to pilot the venture.
The agreement between Air
France-KLM/Alitalia/Delta implemented over the recent period
provides a vivid illustration.
As a result we have hybrid forms blooming within existing
hybrids, with distinctive
characteristics that make them apart from the global alliance of
which they are nevertheless
central players.
These layers of intricate arrangements raise important problems
with respect to the stability
of the alliance and the governance of the new structure. Our
paper focuses on two key
issues in that respect. First, these “hybrids within hybrids”
are getting close to full
integration when it comes to strategic decision rights although
parties to the arrangement
remain almost entirely separate when it comes to property
rights. Why is it so? Are
institutional constraints the only explanation or are there also
issues of performance that
would favor these arrangements? Second, what mode of governance
can overcome the
tensions resulting from the sharp distinction between strategic
decision rights that are
pooled while control over key assets remain in the hands of the
different entities?
Our paper develops elements of answers to these complex
questions. Section 2 reviews part
of the literature on strategic alliances, particularly in the
airline industry, in relation to the
literature on hybrids, thus providing the basis of our
theoretical framework. Section 3
examines the different steps in the development of hybrid
arrangements in the airline
industry, from tactic alliances to strategic alliances to the
more recent organizational
ventures, and the associated differences in the allocation of
rights. Section 4 explores more
systematically the core properties of the ‘hybrid within hybrid’
mode of organization though
a review of the pioneering venture developed by Air
France/KLM/Alitalia/Delta since 2008.
Section 5 discusses more specifically the complex governance
structure implemented to
monitor this venture and draws some lessons about the issues at
stake, particularly the
sustainability of such arrangements. Section 6 concludes with
lessons for organization theory
drawn from this analysis.
2. Strategic alliances as hybrids: where do we stand?
-
4
Economists were initially reluctant to take into account
strategic alliances and other
‘hybrids’, perceived as non-standard modes of organization that
did not fit well in their
“production function” approach to the firm.
Consequently, most early contributions came from managerial
sciences and sociology or
from economists working on the frontier of these disciplines.
Strategic alliances were
perceived as specific forms of inter-firm arrangements. For
example, Parkhe (1991: 581)
defined alliances as « a relatively enduring inter-firm
cooperative arrangement, involving
flows and linkages that utilize resources and/or governance
structures from autonomous
organizations, for the joint accomplishment of individual
goals». Similarly, two early surveys
looked at strategic alliances as one form among the more general
network arrangements
(Grandori and Soda, 1995) and as arrangements that escaped
well-established theories and
that were approximated through messy analytical frameworks
(Oliver and Ebers, 1998). One
of the very first books that looked at strategic alliances as a
class of their own deserving
specific analysis was provided by Nooteboom (1999), who also
devoted an entire chapter to
governance issues.
2.1: Theoretical background.
Progressively, three leading approaches emerged that still
dominate the landscape with
respect to alliances. A pioneering perspective was offered by
the so-called ‘resource-based
view’, initiated by Wernerfelt (1984), which developed a
distinction between resources,
which are tradable and non specific to the firm so that they can
be acquired through markets
or inter-firm arrangements; and capabilities, which are specific
to a firm so that to trade or
share them would challenge the firm’s identity and even its
survival (Amit and Schoemaker,
1993). However, the main focus of Resources Based View is about
how a firm secures its
capabilities and creates barriers to imitation rather than
providing an explanation to
strategic alliances and other cooperative agreements that would
go beyond means to
circumvent scarcity of resources.
More influential among economists trying to understand the
nature of alliances has been
agency theory. Indeed, when confronted to problems of
controlling and monitoring some of
their resources, firms may have an incentive to outsource to
partners while securing the
provisions of goods or services at stake through credible
contractual agreements (Lafontaine
-
5
and Slade, 2012: 962 sq.). However, the emphasis in that
perspective is less on the existence
of strategic alliances or other alternative modes of
organization than on the incentive for
firms to deal with partners to overcome internal problems.
Hence, the theory focuses on the
firm, not on alternative arrangements.
In that respect, it is the transaction costs perspective
inspired by Williamson (1991) that has
prevailed in the development of a positive analysis of
inter-firm agreements such as
alliances, franchise, joint ventures, and so forth.2 A recent
contribution (Baker, Gibbons and
Murphy, 2009) proposes a distinction between different forms of
alliances based on the
allocation of rights and outcome, complementing Transaction Cost
Economics with the
introduction in the analysis of alliances of a fruitful
distinction between property rights and
decision rights. Our theoretical framework largely builds on
these contributions.
2.2 : A short review of the literature on air alliances.
Before turning to our model, it is important to take note of
several contributions more
specifically focused on airline alliances, a topic of intensive
research over the last decade or
so.
Several authors paid special attention to the concept of
strategic alliances. Li (2000) made an
important contribution to the identification of different forms
of strategic alliances and the
evolution of their modes of cooperation analyzed from an
industrial organization
perspective. Morrish and Hamilton (2002, p. 401) qualified these
modes of organization as
“any collaborative arrangement between two or more carriers
involving joint operations
with the declared intention of improving competitiveness and
thereby enhancing overall
performance”.
In their effort to circumvent the concept, authors also
exhibited factors that could motivate
this choice. Button et al. (1998) emphasized that alliances
generate synergies by combining
resources and expertise, opening the door to functional
complementarities (the most
illustrative being code sharing and frequent flyer programs that
gave way to the so-called
‘dog-bone’ networks), while allowing members to have easier
access to important financial
2 For a survey of the dominance of the TCE approach to franchise
systems, see Lafontaine and Slade (2007); and
Ménard (2012) for the more general hybrid arrangements.
-
6
resources. Thanks to these arrangements, firms otherwise rivals
learned to share goals in
order to improve performance and reach higher level of
profitability. Other arguments focus
on uncertainty and complexity. According to Luo (2007) alliances
are an effective way for
members to reduce costs but also uncertainty by sharing risks.
“From a self-selection
perspective, joint ventures are more likely to be chosen than
other entry modes in situations
with high uncertainty, meaning that environmental volatility
plays a role on the selection of
the joint venture mode” (Luo, 2007, p. 40). Globalization is
also viewed as an important force
having pushed towards alliances. According to Ireland, Hitt and
Vaidyanath (2002), in a
globalized environment no single firm has all resources needed
to gain and sustain a
competitive advantage on the long run. Bilotkach and Huschelrath
(2012) substantiate the
point, emphasizing that in a global market, the level of demand
on most city-pair markets is
not sufficient to sustain regular non-stop services, so that
forming international global
alliances is an important way for airlines to facilitate
inter-connections for consumers. This is
close to the argument made by Nielsen who pointed out that as
market complexity increases
“inter-firm collaboration becomes a crucial component in
building a competitive advantage”
(Nielsen, 2010: 682).
However, the underlying element to all these views is that
alliances improve performance. Is
it so obvious? Porter (1990) submitted that alliances are a tool
for extending or reinforcing
competitive advantage, but rarely a sustainable means for
creating it. Morrish and Hamilton
(2002) went a step further. In an extensive review of alliances
that developed in the period
1986-2000, they suggested that despite the increase of load
factors and productivity, there
is no convincing evidence that alliances have had a positive
impact on the competitive
position and profitability of members, the increased frequency
of flights having been
counter-balanced by decreasing fares. At about the same time,
Oum et al. (2000) published
one of the most important empirical analyses of the impact of
alliances on airlines’
performance. Airline alliances are portrayed as a strategic
management choice, designed to
enhance competitive capabilities in business rivalry. The
authors identified two main
advantages in joining an alliance: improved operational
productivity and improved
competitive position through market power. Other important
contributions are from Barla
and Constantatos (2006), who showed that profits might be
greater under strategic alliance
than under full merger; and from Brueckner and Whalen (2000) and
Bilotkach (2005) who
-
7
examined the impact of alliances on airfares as well as the
effect of competition among
alliances, and concluded that alliances benefit passengers,
thanks to the complementary
nature of the products and the removal of double
marginalization. More recently, Brueckner
and Picard (2012) showed how antitrust immunity limits the
potential anti-competitive
effects of cooperation in alliances by reducing incentives to
collude among international
alliances.
However, notwithstanding their richness, most analyses fall
short so far of looking at the
governance issue that is so central to the sustainability of
alliances. It is so, with a few
exception (see De Man, Roijakkers and De Graauw, 2010) because
these analyses rely so
much on a standard industrial organization approach, with
inter-firm agreements explored
essentially through their impact on market structures.
2.3 Our Analytical Framework.
Without ignoring the significance of the issues raised in this
market-oriented literature, we
would like to propose a theoretical framework that we derive
from transaction cost
economics, enriched by the relational contract approach (Baker
et al., 2008). This framework
intends to make room for the problems of governance on which so
much of the success or
failure of the redistribution of rights among parties to an
alliance depends. In the context of
alliances, we understand governance as the process through which
rights and assets are
allocated and transactions are established, implemented, and
monitored through specifically
designed mechanisms. In line with the concept developed by
Williamson (1996: chap. 4),
strategic alliances can be understood as forms of hybrid
arrangements, in which two or more
partners pool strategic decision rights as well as some property
rights, while simultaneously
keeping distinct ownership over key assets, so that there is the
need for specific devices to
coordinate their joint activities and arbitrate the allocation
of payoffs.
Let us consider a simplified case with only two firms, 1 and 2,
and four assets {A,a; B,b}. A
and B are assets essential to the core activity of 1 and 2,
respectively, and remain operated
within their respective boundaries, while a and b are assets
valuable if and only if used
jointly. Each firm holds full decision rights DA and DB, while
rights da and db require
coordinated decision making since linked to the joint usage of a
and b. The resulting payoffs,
associated to property rights at hands, are therefore ПA, ПB,
πa, and πb, with the last two
-
8
generated if and only if the corresponding assets are jointly
used (profits are zero
otherwise). The basic hypothesis underlying our analysis of
hybrids is that in order to benefit
from the joint usage of a and b, the partners have to develop
specific devices to coordinate
actions related to these assets, while they often remain
competitors on other sets of action.
As discussed below, the nature of these devices and the
extension of their power depend on
the type of decisions, and possibly assets, that are shared.
Figure 1 synthesizes the complex allocation of rights and assets
that characterize hybrids:
black arrows suggest the assets and rights that rely on joint
coordination, while the
horizontal arrow suggests the possibility that partners
simultaneously compete. However,
the figure does not provide indications on the exact nature of
the coordinating devices
implemented by partners, an issue to be discussed below.
FIGURE 1: THE COMPLEX ALLOCATION OF RIGHTS AND ASSETS IN
HYBRIDS
Source: Ménard 2009
Important problems of coordination and implementation of
decisions immediately arise
from such arrangements. The nature and magnitude of these
problems precisely determine
the choice of governance mechanisms … and the possibility of
their misalignment with tasks
to be done. Indeed, a key issue in strategic alliances (as in
other forms of hybrids) relates to
FIRM 1 A D
A
ПA
FIRM 2 Β DΒ
ПB
STRATEGIC COORDINATION
=> {a, b, d
a, d
b, π
a, π
b}
-
9
the matching between the transactional hazards that partners are
facing and the allocation
and monitoring of rights among them.
There are several sources of transactional hazards that can open
room for opportunistic
behavior, threatening the stability or even the existence of
such alliances. First, we assumed
above that part of the decision to go hybrid comes from the fact
that some assets have no
value (or a negligible one) if they are not exploited jointly.
It means that bilateral
dependence develops and that a party cannot presume that her
partners are in symmetric
position, so that incentives to behave strategically may occur.
Second, measurement
problems often emerge, due to the overlapping of rights. This is
typically so when partners
invest jointly in a new technology, or in marketing, or in R
& D projects. Third, conditions
that prevailed at the time the alliance was negotiated and
accepted may change, creating
unexpected reallocation of costs or profits, with the risk that
one or several partners will not
pass the adequate information, or will not transfer information
on time. Fourth, the
agreement may not have defined property rights properly, either
because doing so would be
too costly (contracts are most of the time incomplete) or
because the contract was
purposely left open to facilitate adjustments. Still, events may
require more substantial
adjustments than initially expected, as the financial crisis of
2008 illustrates, thus imposing
drastic changes to the alliance. Fifth, the institutional
environment may not provide
adequate guarantees, as when some partners to an alliance are
from a country that does not
properly secure property rights, so that deviants could hardly
be sanctioned.
2.4 A governance issue.
Potentially exposed to these hazards, hybrids must find a mode
of governance that gives
partners some control over the resulting risks of opportunism or
at least some control over
the behavior of partners. In that context, governance refers to
the devices implemented to
monitor the allocation of decision rights as well as property
rights, while simultaneously
minimizing the transaction costs generated by these actions. The
allocation of decision rights
determines the degree of centralization/decentralization of
devices in charge of coordinating
the joint use of assets of type a and b. The allocation of
property rights determines the
expected benefits from joint use of assets and the degree of
control transferred by partners
to the coordinating device.
-
10
Three stylized modes of governance can be adopted to monitor
this allocation of rights.3
When parties share very few rights, particularly rights over
specific assets, but still can
expect value added from some joint activities, they tend to rely
on a limited coordination,
developing information-based networks. There is no need in such
cases to develop well-
specified rules regarding rent-sharing and to establish complex,
detailed contracts.
Nevertheless, hazards can come out of asymmetric information,
giving parties incentives to
endorse devices that facilitate unbiased transfer of
information, e.g., shared software.
Therefore, some forms of control are needed, particularly with
respect to the quality of
information shared. As we will see in the next section, this
type of arrangement corresponds
quite well to the tactic alliances initially developed in the
airline industry.
When expected quasi-rent from joint activities becomes more
significant but requires
sharing important rights, parties become increasingly aware of
the risks of potential
opportunism. Pooling important decision rights and, in some
situations, limited specific
assets requires more intense coordination and forms of control.
Contracts become more
complex, adjustments more difficult to monitor, and the need to
punish deviants more
significant. Cooperation in this context requires mechanisms of
governance with power to
constrain and sanction. Typically, partners tend to rely on
third party enforcers: clauses will
refer to mediators designed as such in the agreement, to
external arbitrators with
procedures well defined, or even to courts in charge of filling
contractual blanks, regulating
adjustments, solving disputes when parties are not able to
overcome directly their
disagreement but still want to maintain cooperation, and even
punishing deviants without
challenging the possibility of shared activities. The shift from
tactic to strategic alliances
corresponds to that phase (see Section 3).
There is a point, though, when these external enforcers cannot
do the job adequately or can
expose partners to too high transaction costs. When expected
benefits require sharing
strategic rights and depend on substantial joint investments
and/or joint usage, risks in
relying on third party become dissuasive: external enforcers may
not have the required
expertise or may depend on the support of highly priced experts,
their decision can be highly
unpredictable and at risk of being disruptive for the
relationship, and so forth. Partners then
3 Details on this typology can be found in Ménard (2012, section
5)
-
11
have an incentive to build a joint strategic center to which
essential decision rights and even
some key assets can be transferred, control over parties to the
agreement can be delegated,
and capacity to punish deviation may even be allowed. This
strategic center can take
different forms, although it almost always relies on a specific
entity, as in joint ventures. As
developed below, an original mode of governance of this type has
been implemented in the
venture between Air France-KLM, Alitalia, and Delta.
Because these modes of governance are alternative ways to
monitor and control hybrids,
there is a trade-off. Our basic assumption, inspired by a
transaction cost approach, is that
parties have an incentive to select devices that can
appropriately match the allocation of
rights while facing the transactional hazards this allocation
generates. The more rights are
shared, the more parties have incentives to switch from
information-based networks to the
building of a strategic center. We now turn to the exploration
of this evolution in the airline
industry.
3. From Tactic Alliances to Strategic Global Alliances to Joint
Ventures
Beside being now a leading force in the airline industry, H
alliances have deeply evolved with
respect to the nature of rights shared and their allocation
among parties. In what follows, we
briefly assess the importance of these alliances and, after a
short reminder of their origin
and the factors that gave the impulse to their development, we
review the arrangements
that marked their evolution and opened the way to the emergence
of their most recent
form, which we identify as ‘hybrids within hybrids’.
3.1: How empires were born.
Alliances between airlines are now a dominant feature of the
industry. The three leading
alliances represented in 2011 almost 60 % of all commercial air
traffic (Star Alliance: 24,8 %;
SkyTeam: 19,6 %; Oneworld: 15,1 %), with a total of 1547 million
passengers transported
that year4. The number of their active members grew from a
handful in the early 1990s to 58
4Star Alliance was founded by Air Canada, Lufthansa, SAS, Thai
Airways, and United Airlines in 1997; Oneworld
was founded by American Airlines, British Airways, Canadian
Airlines, Cathay Pacific and Qantas in 1999; and
Sky Team was founded by AeroMexico, Air France, Delta Air Lines,
and Korean Air in 2000. Qualiflyer, created
by Austrian Airlines and Swissair in 1992, was dissolved after
Swissair went bankrupt in 2002.
-
12
at the end of 2011 (membership of
Oneworld: 12 airlines). Members of the alliances are listed in
table 1 below.
TABLE 1: MEMBERS OF THE THREE GLOBAL ALLIANCES.
Although a few members have
Spanair), each alliance is still on strategy of expansion,
targeting new members in specific
parts of the world. For example,
East member of a global alliance. However, the boom
‘open skies’ policies is now almost over and has left way to a
more guarded ap
focused on the provision of well
for more efficient modes of organization, particularly through
the development of varieties
of joint ventures among subset of members within existing
alli
The power of the Big Three is well illustrated by some basic
data.
$445billion of total revenues, with shares relatively
proportiona
Alliance: USD 201.8 bn; SkyTeam
yielded collective net profits of
data with respect to these alliances and their power.
TABLE 2: MAIN CHARACTERISTICS OF THREE LEADING ALLIANCES
membership of Star Alliance: 27 airlines; Sky
Members of the alliances are listed in table 1 below.
LE 1: MEMBERS OF THE THREE GLOBAL ALLIANCES.
Source: Authors’ compilation
have disappeared or have been absorbed (e.g.,
), each alliance is still on strategy of expansion, targeting
new members in specific
rts of the world. For example, Saudia recently joined Sky Team,
becoming the first Middle
East member of a global alliance. However, the boom that
followed the implementation of
‘open skies’ policies is now almost over and has left way to a
more guarded ap
focused on the provision of well-tailored capacities,
consolidated revenues, … and the search
for more efficient modes of organization, particularly through
the development of varieties
of joint ventures among subset of members within existing
alliances.
is well illustrated by some basic data. In 2011, the
$445billion of total revenues, with shares relatively
proportional to their traffic (
SkyTeam: USD 137.1 bn; Oneworld: USD 106.6 bn
collective net profits of USD 5.5 bn. Table 2 summarizes some of
the most illustrative
data with respect to these alliances and their power.
: MAIN CHARACTERISTICS OF THREE LEADING ALLIANCES
Sky Team: 19 airlines;
Members of the alliances are listed in table 1 below.
LE 1: MEMBERS OF THE THREE GLOBAL ALLIANCES.
disappeared or have been absorbed (e.g., BMI, Malév, or
), each alliance is still on strategy of expansion, targeting
new members in specific
, becoming the first Middle
that followed the implementation of
‘open skies’ policies is now almost over and has left way to a
more guarded approach,
tailored capacities, consolidated revenues, … and the search
for more efficient modes of organization, particularly through
the development of varieties
2011, they generated
to their traffic (Star
bn). That year, they
some of the most illustrative
: MAIN CHARACTERISTICS OF THREE LEADING ALLIANCES
-
13
Star Alliance Skyteam Oneworld
Year of formation 1997 2000 1999
Members (airlines) 27 19 12
Passengers (millions) 449 436 300
Employees 448 926 436 000 299 967
Daily departures 21 900 15 465 8 837
Countries served 194 187 157
Total revenues
(in USD billions )
* data for 2011
201,8* 137,1* 106,6*
Source: Compilation by the authors.5
* data for 2011
It took these alliances almost 25 years to reach that level of
development. According to Oum
and Yu (1998), the first international alliance linked Air
Florida with British Island in 1986,
although Li (2000) suggests that alliances emerge even earlier,
as with the alliance between
Avico and Iberia, established in 1948, and some may go as far as
1919.
The early alliances were essentially bilateral, corresponding to
tactical choices, with some
shared flights, schedule coordination, maintenance consortia
and/or joint ground/catering
services. These bilateral tactical alliances are still used by
airlines to address specific
deficiencies in their networks, but most carriers providing
international services now prefer
to join a global alliance to take advantage of their worldwide
networks. As emphasized by Li
(2000: 65) “The general consensus is that if an airline does not
join an alliance it will be
operating under a serious handicap.”
5 Based on the website of each alliance as of April 2013, with
the exception of the item ‘Total Revenues’ based
on Airline Business, September 2012
-
14
However, it is only in the mid-1990 that broad multilateral
alliances took off. With the
development of computerized reservation systems, early alliances
evolved towards more
complex arrangements, with code-sharing agreements, joint
Frequent Flyer Programs,
shared ground handling, and so forth. One of the earliest
alliance of the new generation was
the ‘Global Excellence Alliance’ between Singapore Airlines,
Delta Airlines and Swissair; it
started in 1989 and ended in 1997. The main impulse likely came
from the deregulation of
air traffic, giving incentives for airlines to go further than
bilateral agreements. As noted by
Button (2008: 61), “Open Skies air service have not only removed
restrictions governing
rates and fares, market entry, and the ways revenues are
allocated, but have also permitted
the emergence of various forms of business alliances”. It is
noticeable in that respect that
the three leading global alliances were all created between 1997
and 2000, and that “ [they]
are no longer mere loose arrangements between a few airlines to
share flight codes and
cross-sell tickets. Rather, they are aiming at virtual mergers
to bypass national rules
governing foreign ownership of domestic carriers and provisions
on cabotage” (Li, 2000: 66).
3.2: Evolution
Notwithstanding their short life span, these alliances went
through quite spectacular
organizational transformations. To better understand this
evolution, it is useful to keep in
mind the different ingredients that can enter in an alliance,
the mix of which determines
different hybrid forms. Li (2000:67) listed fourteen such
ingredients that we can find in an
airline alliance:
BOX 1: INGREDIENTS THAT CAN BE FOUND IN AN AIRLINE ALLIANCE
C01 – Equity: Taking up of shares in another carrier.
C02 – Joint Frequent Flyer Program.
C03 – Code-sharing: A marketing arrangement between airlines
allowing them to sell seats on each other’s
flights under their own designator code. In the case of
connecting flight of two or more code-sharing
carriers, the whole flight is displayed as a single carrier
service on a computer reservation system.
C04 – Block seat and space arrangements: A contractual
arrangement between two airlines, whereby a
specified number of passenger seats or cargo space are allocated
between more than two points on
a carrier’s route for a given period of time. The space swap
agreement (usually as a side agreement
-
15
of a code-share package) is also included here.
C05 – Joint marketing: usually involving setting up of joint
sales office.
C06 – Joint purchasing: An arrangement where partners in the
agreement buy things in bulk to enjoy
economies of scale. Examples include utensils, engine parts and
etc.
C07 – Ground support: Including ground handling, engineering,
maintenance and cargo.
C08 – Airport facilities sharing arrangements: These facilities
include passenger lounge, hangars, slots and
other equipment in the airport.
C09 – Flight scheduling: Joint planning of flight schedule among
carriers to minimize the transit time.
C10 – Pool agreements (revenues and costs): Agreement between
two or more airlines to combine capacity
operated. Tickets for the agreed sectors are used
interchangeably without endorsement. Aspects
include revenue pooling, revenue sharing and cost sharing.
C11 – Franchise/royalty: The use of another airline’s
color/arrangements.
C12 – Information technology co-operation: Linking of IT systems
together.
C13 – Joint operations: Including joint services/flights and
business ventures. A scheduled air service operated
by an airline between points on its own behalf and that of
another airline by sharing the costs and
revenue on an agreed basis. Each carrier assigns its own airline
code and flight number, which is
both, displayed as a joint service.
C14 – Commercial agreement: Arrangement from one airline to
another for use of certain traffic rights. This
arises because there is no reciprocity in the exchange or usage.
For example, Country A may allow a
carrier from Country B to operate to Country A even when its own
carrier has no plans to operate to
Country B. To facilitate this, a carrier from country B may
agree to a commercial agreement with a
carrier from Country A.
Source: Li (2000: 67)
The number and type of ingredients entering in an alliance does
not only define the variety
of observable arrangements, it also corresponds to different
steps in the development of
alliances in the airline industry. In that respect, we can
identify three major phases, with
uneven degree of achievements depending on the alliance under
review. These phases
correspond to switches in the typology of modes of governance
summarized in section 2.3.
Phase 1: From Information-based network to shared decision
rights.
Alliances were initially tactical agreements that did not
implied shared property rights and
that pooled a very limited set of decision rights, such as the
possibility to buy a ticket that
would allow connection with a flight from the partner carrier
(e.g. block seats, the C04
ingredient in Li’s list). These agreements typically involved
two carriers and covered a limited
number of routes, the main objective being to facilitate
correspondence for travelers having
to use both networks. The coordination of Frequent Flyer
Programs became a natural
extension of these agreements.
-
16
It is obvious from these examples that tactical alliances were
created and implemented to
overcome two complementary difficulties: (1) to address
loopholes in the respective
networks of the partners, thus allowing their passengers to
reach a wider set of destinations;
(2) to circumvent restrictive regulations (e.g., connecting
routes not opened to one partner).
These arrangements remained very close to standard commercial
agreements, that is, to
market relationships based on some shared services, while
property rights remained entirely
distinct. However, shared information and coordination over some
decision rights pushed
towards better interconnected networks and more cooperation.
Phase 2: Extended coordination: the development of strategic
alliances.
Tactical alliances remain part of the picture in the airline
industry. There are still numerous
bilateral agreements on limited issues. However, the need for
more intense coordination
pushed towards new forms of cooperation. In a changing
environment, new features
developed that are no longer limited to bilateral agreements or
shared flight codes and
cross-sell tickets and that has been identified as “strategic
alliances” or “global alliances”
(which is a variation on strategic alliances).
First, expected benefits from economies of scale, scope, and
density extend progressively
beyond flights’ interconnections. Examples of increased
coordination over important
decision rights include joint marketing, airport slot sharing,
check-in and baggage handling
systems, joint purchase and repair of spare parts or fuel,
shared use of hubs, the joint
development of technical and training procedures, as well as
facilitated access to
financial/capital resources (Holloway 2003; Kleyman et al.
2004). Indeed, in the airline
industry, scale economies come out of the extension of the
network and the related increase
in the number of passengers carried. Thanks to cost reduction
through the development of
joint activities (ground handling, shared lounge etc.) and
through the possibility of market
presence in areas that a single company could hardly serve
directly, alliances can boost
benefits of each partner (Gudmundsson and Rhoades 2001: 210).
Alliances can also take
advantage of scope economies, giving members the possibility to
offer customers a ‘global
seamless network’ that minimizes their travel time, increases
potential connections, and
extends benefits from frequent flyer programs (Tretheway and
Oum, 1992). For airlines, the
-
17
possibility to offer an extended network also facilitate
coordinated schedules, increased
capacities, and the offer of a variety of services at lower
prices.
Second, there is an institutional dimension that plays a key
role in the development of
strategic alliances in the airline industry. The propensity
towards better coordination for the
reasons mentioned above has been amplified by the deregulation
of the industry initiated in
the US in the late 1970s and extended progressively to other
countries in the 1980s and
1990s. Deregulation intended to open skies to competing
airlines. However, major
constraints governing foreign ownership of domestic carriers and
the possibilities of
“cabotage” prevent airlines to freely serve international
markets, making them dependent
on bilateral agreements politically negotiated among countries.
As emphasized by Kasper
and Lee (2009, p. 2):
« Airlines are precluded by a host of laws and regulations in
the United States and
abroad from acquiring control of foreign airlines and from
carrying domestic traffic in
other countries. These barriers to investment and market entry
effectively preclude
any single carrier from building a global network using its own
network resources and
aircraft and thus pose a particular problem for the airline
industry, which has become
an increasingly network---based business since deregulation. As
a result, the only way
a U.S. airline can provide convenient connections, common
service standards, lounge
access, and frequent flyer credits for its customers traveling
internationally to points
that the U.S. carrier cannot itself serve is by forming an
alliance with a foreign carrier.
Likewise, foreign carriers must rely on alliances with U.S.
carriers to provide their
customers access to routes that the foreign carrier cannot serve
for legal or economic
reasons».
In that context, strategic alliances also became arrangements
facilitating bypass of restrictive
national rules.
Phase 3: Building a strategic center.
The next step in the industry was the development of global
alliances structured by strategic
center. Iatrou and Alamdari (2005: 127-128) define global
alliance as “airlines participating in
-
18
a commercial relationship or joint-venture, where a joint and
commonly identifiable product
is marketed under a single commercial name or brand”. The
strategic alliances created at the
end of the 1990s rapidly consolidated in to the three dominant
alliances already mentioned
which, nowadays, the basic choice offered to non-member
airlines. However, parallel to
their consolidation global alliances became increasingly
submitted to inter-alliance
competition. In such a context, economic theory suggests that
the absence of shared
property rights as well as the maintenance of autonomous
decision rights should threaten
the stability of these arrangements. Parkhe (1993: 794) already
emphasized that strategic
alliances are “voluntary inter-firm cooperative agreements,
often characterized by inherent
instability arising from uncertainty regarding a partner’s
future behavior and the absence of
higher authority to ensure compliance.” This did not happen in
the airline industry, at least
so far. Beside the peculiar cases of Qualifyier, which was
dissolved, and of KLM-Northwest
which splitted when KLM merged with Air France and joined Sky
Team , there are very few
cases of airlines switching alliances.6
Actually, two strategies have been developed to prevent
instability. First, there is the
classical solution of mergers and acquisitions. With one
exception (the acquisition of
Canadian Airlines by Air Canada), the main mergers and
acquisitions over the last period
have been among members of the same alliance (Air France-KLM,
Delta-Northwest,
Continental-United, British Airways-Iberia). However,
competition authorities have been
very reluctant to authorize mergers and acquisitions, fearing
the potential negative impact
on consumers’ surplus, particularly on routes that would fall
under monopoly power.
Hence, a second ongoing solution developed with the pioneering
experiment initiated by Air
France/KLM-Delta-Alitalia in 2008: building a hard core alliance
within the global alliance.
This arrangement, which is often described as a joint venture,
can better be understood as ‘a
hybrid within a hybrid’. 7 It is based on an agreement
characterized by a substantial increase
6 Canadian Airlines joined Star Alliance after its acquisition
by Air Canada in 2000; Mexicana left Star Alliance in
2004 then joined Oneworld a few years later; and Continental
switched from Sky Team to Star Alliance in 2009
after it merged with United Airlines. Ansett and Varing are
different in that they left Star Alliance because they
went bankrupt. 7 According to Bilotkach and Hüschelrath (2010,
page 3) “the theoretical rationales for forming joint ventures
rather than entering into regular contracts includes transaction
costs savings, strategic behavior, and
capitalizing on the organizational knowledge. The former of the
three is well in line with the traditional
-
19
in shared decision rights and the building of strategic
coordination devices that bring these
arrangements very close to integration and require special
agreements from antitrust
authorities. The next sections come back to the nature and
governance of
of organization.
To sum up, arrangements in the airline industry have evolved
within 30 years from tactic
alliances that were and remain very close to standard commercial
agreements of the market
type, to strategic alliances based on multi
without shared property rights, to global alliances within which
a subset of partners develop
quasi-integrated forms of coordination with respect to decision
rights while property rights
remain distinct. Figure 2 summarizes these various forms
according to the degree of integration among partners to an
agreement
FIGURE 2: A COMPLEX COMBINATION OF VARIOUS ARRANGEMENTS
4. Quasi-joint ventures within strategic alliances.
Stiglerian boundaries of the firm argument; t
latter views joint ventures as a means by which the firms learn
or retain their capabilities
in shared decision rights and the building of strategic
coordination devices that bring these
arrangements very close to integration and require special
agreements from antitrust
authorities. The next sections come back to the nature and
governance of
To sum up, arrangements in the airline industry have evolved
within 30 years from tactic
alliances that were and remain very close to standard commercial
agreements of the market
type, to strategic alliances based on multilateral inter-firm
agreements (usually contracts)
without shared property rights, to global alliances within which
a subset of partners develop
integrated forms of coordination with respect to decision rights
while property rights
summarizes these various forms that developed over time,
posited
according to the degree of integration among partners to an
agreement.
FIGURE 2: A COMPLEX COMBINATION OF VARIOUS ARRANGEMENTS
joint ventures within strategic alliances.
Stiglerian boundaries of the firm argument; the second relates
to longer-term profit maximization; and the
latter views joint ventures as a means by which the firms learn
or retain their capabilities
in shared decision rights and the building of strategic
coordination devices that bring these
arrangements very close to integration and require special
agreements from antitrust
authorities. The next sections come back to the nature and
governance of these new modes
To sum up, arrangements in the airline industry have evolved
within 30 years from tactic
alliances that were and remain very close to standard commercial
agreements of the market
firm agreements (usually contracts)
without shared property rights, to global alliances within which
a subset of partners develop
integrated forms of coordination with respect to decision rights
while property rights
that developed over time, posited
FIGURE 2: A COMPLEX COMBINATION OF VARIOUS ARRANGEMENTS
term profit maximization; and the
latter views joint ventures as a means by which the firms learn
or retain their capabilities ».
-
20
We now turn to the analysis of the last development in airline
alliances, the building of a
hard core within existing strategic alliances. However, it
should be noted upfront that what
we actually observe, with respect to figure 2, is not the
disappearance of initial forms such as
tactical alliances but rather a complex overlapping of various
forms of hybrid arrangements.
Indeed, the three leading global alliances nowadays typically
involve numerous partners with
various degree of cooperation, with forms quite closely to the
‘pure’ hybrid arrangements
identified in section 2.
Nevertheless, confronted to the increasing competition that
characterize a globalized
market, with competition switching from firm versus firm to
alliance versus alliance, and
facing institutional constraints that limit their capacity to
integrate or even coordinate,
airlines had an incentive to find new modes of organization.
This pressure was amplified by
the crisis that started in 2008. Operating in a pro-cyclical
industry, the major airlines quickly
understood that they had to reduce their capacity, the frequency
of flights, etc., and to
reorganize their network configuration in order to cut costs and
adapt offer to the changing
demand. These difficulties were accentuated by inter-firm
competition, including among
members of the same alliance, on the most important markets.
Before 2008, some key
players were already looking at means to circumvent this
“destructive competition.”8. They
put high on their agenda the creation of ‘joint ventures’ that
would coordinate much more
closely substantial decision rights. The existing literature
does not provide a clear definition
and delineation of joint ventures as distinct from strategic
alliances. Jay Barney (2002)
proposed to subdivise alliances into non-equity alliances,
equity alliances, and joint ventures.
Bilotkach and Hüschelrath (2011: ) went a step further, noting
that: « The key difference
between the latter and the former two is that only a joint
venture leads to the creation of a
legally independent new corporation in which the parent
companies hold shares. Following
this delineation, airline alliances must typically be
categorized as ‘non-equity alliances’
rather than ‘joint ventures’ ».
The organizational arrangement under review does not fit well
within these definitions. On
the one hand, some members of alliances have developed joint
ventures, usually with
8 For the time being, Oneworld is unambiguously dominated by
American Airlines and British Airways; Star
Alliance by Lufthansa and United airlines; and Sky Team by Air
France-KLM and Delta-Northwest.
-
21
outsiders to the alliance. For example, All Nippon Airways
(ANA), a member of Star Alliance,
got involved in a joint venture with the low cost Air Asia. On
the other hand, more puzzling
for organization theory and with respect to the stability of
contractual agreements is the
development among selected members of a broader alliance of
ventures run through very
tight coordination of decision rights, although property rights
remain entirely distinct. In that
respect they differ from joint ventures, although they do
ressemble joint ventures if one
looks at the specific activities on which they share almost all
decision rights, so that they
could be labelled ‘quasi-joint ventures’. Examples are provided
by the Air France-KLM / Delta
/ Alitalia or Air Canada / Lufthansa / United Airlines
agreements on transatlantic routes.
Before looking at the expected benefits and difficulties of such
arrangements, the
institutional issue must be clarified. Indeed, these
arrangements go against the regulatory
constraints imposed by several national antitrust authorities,
so that there implementation
requires a derogatory regime. Nowadays, the three global
alliances have been able to get
antitrust immunity (ATI) for their North Atlantic routes (see
the graph below).
GRAPH 1: HYBRIDS INSIDE HYBRIDS: THE CASE OF SKY TEAM
-
22
4.1: Getting antitrust immunity
Indeed, building a tight network among a small number of firms
and for a limited number of
routes often overlapping represents a major source of concern
for competition authorities
since it involves price setting, with potentially negative
impact on consumers’ surplus,
particularly through price setting. Consequently, the
possibility to create the new type of
arrangement described above required approval from competition
authorities of the
countries in which participating carriers are based and/or
operate.
In that respect, the US Department of Transportation (DOT) is a
key player. Getting ‘antitrust
immunity” from DOT requires submitting the proposed arrangements
to standards quite
close to those applied to mergers. Two criteria are of
particular significance in that respect:
(1) would the proposed alliance “substantially reduce or
eliminate competition”? and (2)
What conditions should be imposed (particularly ‘carve outs’
solutions) to cure potential
-
23
competitive problems? Referring to these criteria, the US DOT
took a positive stance
towards alliances as early as 1999. As stated in a report from
that year:
« [N]o airline, however strong, is able to efficiently provide
service with its own
aircraft and crew to every destination its customers require.
Using the transatlantic
market as an example, there are several hundred cities in the
U.S. and also in Europe
that will never have the benefit of non-stop service.
[Alliances] are the only practical
way to provide better service to thousands of passengers in long
distance, low-density
international markets. (….) Multinational alliances have fueled
enormous increases in
connecting traffic, ...newly stimulated traffic accounts for a
large proportion of alliance
growth” (US DOT, 1999: 2)
Based on this analysis, US DOT developed a position favorable to
the delivery of ‘Antitrust
Immunity’, conditional to the demonstration by applicants that:
(a) the proposed
arrangements would allow significant public benefits; and (2)
these benefits would not
materialize if ‘Antitrust Immunity’ was not granted. The impact
on prices is of course a major
concern in that perspective.
4.2: Expected benefits and impact on prices
This paper is not about issues of fares and welfare but about
the characteristics of the new
arrangements and the mode of governance implemented to make them
efficient. However,
some expected consequences of these arrangements on prices must
be mentioned because
they explain several restrictions imposed under the ‘antitrust
immunity’, largely as the result
of tensions between the US DOT and the US DOJ. Indeed, the US
Department of Justice
expressed serious doubts about the benefits to be expected from
this derogation to general
rules as well as about the governance embedded in the new
arrangements. 9
Similar concerns can be observed in the position adopted by the
European Commission in
2011, as assessed by IATA (2011). According to this comment,
which is of course a bit biased
in favor of alliances, “[H]orizontal cooperation agreements can
lead to substantial economic
benefits in particular if they combine complementary activities,
skills or assets. Horizontal
cooperation can be a mean to share risk, save costs, increase
investments, pool know-how,
9 See US Department of Justice, 2009.
-
24
and enhance product quality. On the other hand, horizontal
cooperation agreements may
lead to competition problems …” (IATA 2011: 5).10 The expected
benefits of the (quasi) joint
ventures under examination can come out from (among other
factors) . They « , These
competition problems are mainly be related to pricing strategies
and their impact on fares.
On the one hand, “for interline trips, cooperation in fare
setting under anti-trust immunity
eliminates a type of horizontal double marginalization. Instead
of non-cooperatively setting
the components of an interline fare for travel across two
networks, the two carriers under
the antitrust immunity jointly set the entire fare reducing
their two separate “markups” to
one”, which would benefit passengers. (Brueckner and Proost
2010: 657). On the other
hand, antitrust immunity could well introduce anti-competitive
effects, particularly on
overlapping non-stop hub-to-hub routes.
There are few studies already available on these effects (see
Brueckner 2001 and US DOJ
2009). In their recent paper, Brueckner and Proost (2010)
provide an analysis of both effects
and show that the net impact on price and welfare is ambiguous.
Making a distinction
between inter-hub and interline markets,11 Barla and
Constantanos (2006: 412) already
noticed that: « In both cases, cooperation leads to an
internalisation of each firm’s impact on
the other firm’s profits. In the inter-hub market, this results
in less competition and thus
higher prices. In the interline markets however, cooperation
leads to lower prices, since
airlines’ demands are complementary. The empirical findings […]
confirm that cooperation
significantly reduces interline fares while the evidence on fare
increases in inter-hub markets
is less conclusive ».
The risks of a negative impact on consumers’ welfare,
particularly with respect to inter-hub
markets may explain why competition authorities have imposed
‘carve-outs’ conditions12
10
According to the same comment, expected benefits can come out of
demand-side – relating to the creation
of new or improved services through expanded networks or
seamless service, or supply-side – essentially the
ability to produce the same services at lower cost taking
advantage of traffic densities, improved utilization of
capacity and lower transaction costs. Potential demand-side
benefits include the elimination of double
marginalization, expansion of route networks, expansion of
flight frequency, and improved ‘online’ service
options. Supply-side efficiencies include cost reductions
through economies of traffic density, cost reductions
through coordination of second-degree competition parameters
(sharing of facilities), and cost reductions
through coordination of first degree competition parameters
(pricing and yield management, capacity ) » (IATA
2011: 6) 11
Interline trips mean that passengers have to cross the networks
of an alliance partners. 12
A carve-out prohibits collaboration in hub-to-hub fare setting,
while allowing cooperation in other markets.
-
25
while delivering antitrust immunity, restrictions that have had
ambiguous results according
to Brueckner and Proost (2010)13 .
4.3: Property rights issue: The « metal neutral » concept.
Beside their ambiguous impact on fares, another puzzling aspect
of the new arrangements is
that they share many characteristics of a full joint venture or
even resemble a merger when
it comes to strategic decision rights (on specific line,
partners share revenue, costs, and
profits –more on this below), while they do not pool property
rights.
Part of the explanation might be institutional. Antitrust
Immunity allows parties to the
arrangement to share information, pricing, capacity and
frequency, as well as route
strategies, scheduling, pricing, distribution, frequent flyer
programs, and so forth on on
specific routes. However, immunity does not allow merger or
acquisition. The resulting
arrangement can be qualified as a ‘restricted joint venture’, as
opposed to a full joint
venture in which significant assets would be jointly owned.
Indeed, participating carriers engage in “comprehensive revenue
sharing on international
routes, so that a partners' revenue from a passenger is
independent of which airline actually
provides the service. This arrangement leads to what is known as
a “metal neutral” alliance
structure, in the sense that the identity of the “metal” (the
aircraft) involved in the service is
irrelevant to individual airline revenue », (Brueckner and
Proost, 2010: 658). At the same
time, rights over the physical assets, particularly the
aircrafts, as well as over human assets
(e.g. labor contracts, management of human resources) remain
strictly separated. In a word,
it is merger-like in that on routes specified by the immunity,
partners no longer compete
with each other; but it does not share with mergers or
acquisitions the main characteristic
which is the transfer of property rights. The arguments
developed by airlines to obtain this
agreement from competition authorities have been summarized as
follows:
“This ‘metal neutrality’ is significant in that it maximizes the
opportunity for pro-
competitive efficiency gains from density economies. There is
now substantial evidence
that existing alliance relationships have led to significant
consumer benefit for
13
In this paper the two authors refer to the idea that “carve-outs
are beneficial when the alliance does not
involve full integration of the partners’ operations on the
hub-to-hub route, its key point is that carve-out may
be harmful when imposed on a joint-venture alliance. A JV
alliance involves full exploitation of economies of
traffic density on the hub-to-hub route, and a carve-out
prevents the realization of these benefits”, Brueckner
and Proost 2010: 657.
-
26
passengers on interlining trips, both in terms of improved
service and lower fares.
Concerns about anti-competitive effects arising, for passengers
flying hub to hub
itineraries, where cooperating airlines’ services overlap,
should be lessened by the
potential for metal-neutral JVs to generate pro-competitive
efficiencies » (IATA 2011:
1).
Not all competition authorities nor all economists have been
fully convinced! Nevertheless,
several such arrangements have now been accepted, so that we
have these multiple forms
of alliances as well as multiple layers within specific
alliances, as described in section 3.
5. Hybrids within hybrids: the governance issue.
Beside the sharp separation between decision rights and property
rights, another puzzling
issue in the new type of arrangement developed within existing
airline alliances concerns the
combination of characteristics that should challenge the
governance and stability of these
alliances. First, the new venture is an arrangement among a
small subset of partners,
essentially the most powerful ones, among tens of members (Air
Canada/Lufthansa/United
Airlines in Star Alliance; Delta/Air-France-KLM/Alitalia in Sky
Team; and British
Airlines/Iberia/American Airlines in Oneworld). Second, the new
ventures exist because they
benefit from Antitrust Immunity, although strictly limited to
certain routes. Third, these
arrangements pool most strategic decisions on these routes while
they remain competitors
on these routes as well as on the global network. In what
follows, we take advantage of
access to privileged information on the pioneer venture linking
Air France-KLM / Delta-
Northwest / Alitalia to explore characteristics that may explain
the success and stability of
these arrangements.
5.1: A short historical reminder.
The transatlantic venture between Air-France, KLM, Delta,
Northwest and Alitalia is the
result of a long process of mutual commitments. This process
began in 1997 with the
signature between Northwest and KLM of the first ever joint
venture agreement in the air
transport industry. At that time, the motivation behind that
move was largely due to the
restrictions imposed by US laws on foreign ownership of national
carriers.
-
27
Not long after that, in 1999, Air France and Delta started
building a global alliance, Sky Team,
officially created in June 2000. A few years later, in 2004, Air
France and KLM got the
authorization from the European Union to merge, a movement
followed not long after by
the acquisition of Northwest by Delta Airlines (in 2008). During
that period, the Air France-
KLM group and their American counterpart and nevertheless rival,
Delta, developed an
increasingly diversified commercial and technical
cooperation.
In the context of the ‘open sky’ negotiations between Europe and
the USA, this evolution
found an achievement with the grant of ‘Antitrust Immunity’ for
transatlantic routes by both
the European Commission and the US DOT, allowing Air France-KLM
and the newborn Delta-
Northwest group to implement a venture that officially began
operating in 2008. This special
arrangement, joined by Alitalia in 2010, after Air France took
over 25 % of its shares, was
created to monitor their transatlantic routes. The result was
the emergence of a hybrid
arrangement within the existing hybrid, Sky Team alliance, which
at the time involved
already ten airlines. 14
We now turn to the analysis of this venture, which created a
hybrid form within the already
existing hybrid arrangement that is the Sky Team alliance.
5.2: A contractual venture
The very first quasi-joint venture in the airline industry,
between KLM and Northwest, lasted
for almost twenty years (1989-2007). It went through a series of
conflicts, adaptations and
modifications in its governance structures that provided a
unique source of lessons for the
venture that progressively emerged between Air France-KLM and
Delta (with which
Northwest later merged). Actually, the formal structure of the
venture under review is
deeply embedded in this pioneering experience.15
Like most inter-firm agreements with significant shared rights,
the new arrangement is
based on a contract. However, this contract has some unusual
characteristics that make it a
bit like a Russian doll, with sub-contracts within the core
contract! Indeed, the European
partners, Air France-KLM-Alitalia are linked through specific
contracts, part of which is
14
Aeroflot, Aeromexico, Air France-KLM, Alitalia, China Southern
Airlines, Continental Airlines, Czech Airlines,
Delta Airline-Northwest, Korean Air. Continental left this
alliance a few months later (June 2008) to join United
Airlines (with which Continental will merge in 2012). 15
For an examination of the KLM-Northwest model, see De Man,
Roijakkers, de Graauw (2010)
-
28
embedded in their contract behind their North Atlantic venture
with Delta-Northwest. As a
result, we have formally a two side arrangement, with the Delta
group on the one hand and
the Air France-KLM-Alitalia partners speaking as if they would
form a unique entity when it
comes to the venture on the other hand.
Formally this contract is relatively short (a bit over 150
pages) and quite dense. Although it
is strictly confidential, prohibiting parties to reveal publicly
any information about profits
and costs associated to the venture, information obtained
through extensive interviews
reveal two fundamental dimensions to the contract: (1) a
financial agreement; and (2) broad
indications about the governance structure, mainly focused on
the entities responsible for
piloting the agreement and the decision-making procedure to be
followed. This second
aspect, which is the one on which we focus, is complemented by
Codes of Conduct jointly
agreed upon, finalized by July 31th, 2009, approved by the
Steering Committee (more on this
below) and incorporated as part of the Agreement. The table
below summarizes some key
specifications of this agreement.
-
29
TABLE 3: KEY CHARACTERISTICS OF THE CONTRACT
GOAL Sharing revenue, costs and profits on the routes defining
the
scope of the contract (Bundle 1)
LEGAL FORM A joint venture (defined in a 150 pages contract)
TERMS OF THE
CONTRACT
The agreement is an « evergreen » : it will run indefinitely
and
the first possibility for a partner to notify termination of
the
agreement is after 10 years, and requires a three year
notice
period.
FINANCIAL
COMPONENT
The key clause is a 50–50 profit sharing rule between
European and US partners (AF/KLM/Alitalia & Delta).
SCOPE OF THE
VENTURE
Bundle 1, which identifies transatlantic routes
DECISION-MAKING Consensus (no unilateral action is allowed for
the routes agreed
upon in the contract
Discretionary Decision are taken into account in the
contract:
it is specified that they shall be listed in the codes of
conduct,
allowing unilateral actions (to make the contract flexible)
if in accordance with the Codes of Conduct, and after prior
written notice to Co-Chairmen of the relevant committee in
case of disagreements or conflicts.
GOVERNANCE CEOs’ committee
Steering committee 11 working groups
COMMUNICATION Viewed as central to the arrangement and organized
along two
channels.
Formal Channels Informal Channels Multiple points of contacts
are
defined in the governance
procedures
(JV’s weekly meetings,
steering committee, …)
-Personal relationships at
Board level and inside
the companies - Social events to build personal relationships
(even
between the CEOs)
Source: Authors.
-
30
5.3. A complex financial agreement
The financial part of the contract is of course essential to the
definition and stability of
the venture, particularly since it concerns only a subset of the
decisions each partner
has to make. Indeed, the members of the venture cooperate on a
bundle of
transatlantic routes (“trunks routes”) on which they operate
according to a ‘metal
neutral’ agreement: joint operations, competitive fares,
harmonized ground and in-
flight services, and shared revenues and costs characterize all
related routes, without
consideration for who owns the aircraft. At the end of 2012, the
agreement is relevant
for 27 points of entry in North America and 33 points of entry
in Europe, with 7 major
connecting hubs involved (Amsterdam, Atlanta, Detroit,
Minneapolis, New York-JFK,
Paris-CDG and Rome Fiumicino). All in all, this venture operates
250 transatlantic flights
per day with 144 aircrafts, and it represents 26% of the total
passenger capacity
between Europe and North America, generating revenues of USD 11
billion.
Although confidential, we have been able to collect information
about the main financial
characteristics of the agreement. First, there is a fifty–fifty
profit/cost sharing rule
between the European partners (AF/KLM/Alitalia) and the American
partner (Delta) on
all the routes (« trunks ») specified in the ventures’ scope.
Second, the profits (or losses)
are allocated by reference to a ‘Base Line’ determined by the
situation in 2008, that is:
the revenues before the venture was formally created. All
additional revenues (or
losses) on top of this base line are shared equally between
AF-KLM/Alitalia and Delta.
Third, 25% of the revenues generated beyond the bundle routes
and resulting from the
connections made possible by the agreement are transferred to
the other partner (US or
European). Fourth, once a year the total sum of the incremental
revenues (and costs) is
formally estimated (for trunk routes as well as for routes
connected to trunk routes) and
a « check » is delivered to one or the other partner (Air
France-KLM/Alitalia or Delta).
Fifth, the venture has its own accounting system. For example,
not all costs are
integrated in the costs charged to the venture. Also, some costs
are adjusted (e.g.
wages/charges for Pilots and/or stewards; fuel cost without
hedging). On the other hand
some important costs are referenced in well defined tables (e.g.
lease of aircrafts).
-
31
These contractual clauses frame the domain of action of the
partners to the venture.
However, they leave open the modalities of the implementation of
the contract as well as
the way partners are going to deal with the numerous unexpected
events that characterize
the airline industry. In other terms, the contract is
substantially INCOMPLETE. It defines a
framework within which partners operate. Hence the importance of
the governance
mechanisms adopted.
5.4: The Governance structure adopted in the agreement: building
a strategic center
Although property rights of Air France-KLM on the one hand, and
Delta-Northwest on the
other hand remain distinct,16 the new venture has indeed
implemented a specific structure
to monitor their joint activities and make joint decisions for
all relevant routes and their
associated services. This structure is hierarchical, close to
what we find in integrated firms,
but relies on procedures that work bottom-up. In other terms,
for a specific segment of their
activities, partners behave ‘as if’ they were an integrated
company (which they are not) and
combine a highly centralized decision-making process with a
highly decentralized
management of the joint activities.
There are three main components in this governance structure.
First, the ‘CEO committee’,
with the CEO of Delta and one CEO representing the European
partners is in charge of
strategic decisions and meets at least once a year. Second, a
steering committee of a
minimum of 10 members (5 from the European consortium, 5 from
Delta) is in charge of
monitoring and implementing major decisions with respect to the
organization of the
alliance, its marketing and sales activities, finance,
coordination of the network, and
operations on the relevant routes. This steering committee meets
at least once every three
months and is usually prepared on the European side by an
informal meeting since the three
participants have to contribute to the steering committee as one
single voice! Third, and
most importantly for the actual coordination of the venture, the
2008 agreement created
eleven working groups.17 Each working group is co-chaired by a
representative of Delta and a
representative of the European partners and report to the
steering committee. Working
16
It must be remembered that Air France and KLM merged in 2004,
Delta and Northwest in 2008, and that Air
France bought 25 % of the shares of Alitalia in 2010. 17
The groups are: Network, Revenue Price Management, Sales and
Distribution, Product, Loyalty, Operations,
Information Technology, Finance, Cargo, Advertising and Brand,
Corporate Communications.
-
32
groups must formally meet at least four times a year but are
actually interacting much more
often, for some of them on a daily basis, as is the case for the
Network working group.
The decision process works essentially bottom up. The working
groups deal with all daily
problems as well as with specific issues defined by the Steering
Committee. When a decision
is required, the relevant working group examines the issue at
stake, arbitrates among
possible solutions, and prepares the implementation of the
decision. In most cases, the
approval by the Steering Committee is formal. This committee
mostly operates as an
arbitrator, overcoming diverging interests and making decisions
when conflicts develop
between working groups or, more likely, among them. It is very
noticeable that so far the
committee of CEOs has never been formally involved directly in
this bottom-up procedure,
although such a possibility is explicitly stipulated in the
contract. The CEOs are essentially
consulted on a relatively informal basis when important issues
are under discussion among
partners at the working group level. Nevertheless, the CEOs
clearly operate as a strategic
center, very much as a Board of Administrators would do in an
integrated firm.18
5.4: The key role of the working groups
Clearly the working groups play a central role in this
governance structure. Each group is led
jointly by two members, one appointed by Air
France-KLM-Alitalia, the other by Delta group.
The composition of each group is also symmetrical. Between the
so-called ‘joint venture
week’, that is, the formal meeting of the steering committee
which is planned four times a
year, the working groups monitor the activities of the venture
and make most decisions.
They are in constant contact, through conference calls, phone
calls, e-mails, and it is at their
level that most conflicts are solved. An indicator of their
significance is that before the
contract was signed in 2008, each group, provisory at that time,
had to write down its own
code of conduct. In other terms, they were considered from the
very beginning as teams
that had to create their own rules. The codes of conducts are
explicitly referenced in the
contract.
18
A recent decision about full flat seats illustrates well how the
governance works. Delta initially introduced at
the level of a working group a request to significantly improve
the quality of services, particularly through the
provision of full flat seats for business customers, in order to
compete with similar strategies developed by
competitors, e.g., British Airways and Lufthansa. Air France was
initially reluctant to make these costly changes.
The favorable recommendation of the different working groups
went all the way through the steering
committee to the CEOs and Air France finally accepted the
changes for its 2000 seats in business class (decision
made on June 21st
, 2012, and to be fully implemented in 2013-2014)
-
33
These groups walk on two legs, so to speak, with quite different
human resources required.
On the one hand there are experts, selected to participate to a
specific group because of
their technical knowledge (e.g., the network planner of Delta).
On the other hand, there are
so-called ‘generalists’, almost all of them directors of
different departments of the partners
(e;g., the alliance director of Air France), many of them having
played an active role in the
writing and initial implementation of the contract. These
generalists remain active in the
different working groups, and some are members of the Steering
Committee, so that they
can keep an eye on the various activities of the alliance and
signal problems early on.19 Since
both generalists and experts have been working together for
while, for many of them even
before the venture was formally implemented, they have developed
tight connections and a
‘team spirit’.
This ‘team’ approach, with its division of tasks among various
governing bodies, seems to
work very well. The governance implemented has effectively faced
numerous adaptations
required by a rapidly changing environment, it has solved
numerous conflicts without
noticeable clashes, and significant changes have been adopted
and implemented without
disruptions, for example after the 2008 crisis. The contract has
been renegotiated (five times
as of June 2012) to adapt to this changing business environment,
without any significant
crisis in the governance of the venture, which has remained
remarkably stable through all
these episodes.20
An important key for understanding this success is that the
contract provides a formal
framework substantially complemented by informal contacts.
Formal meetings are usually
paired with informal ones (e.g., the Board meetings of Air
France-KLM and Delta are held in
the same town, the same day, with an informal diner organized
among the CEOs). Other
events also provide opportunities to develop informal contacts
(e.g. the opening of the “S 4”
19
As emphasized by De Man, Roijakkers, and de Graauw (2010: 177).
“They act as a lubricant when different
interests need to be reconciled. (…) There are links between the
generalist groups and the specialists groups,
through personal unions. A balanced staffing ensures that
problems are identified early on. When they do
occur the right persons in the alliance can be found to deal
with them. This reduces the impact of internal
tensions in the alliance ». 20
Examples of this smooth adaptation are the entry of Alitalia in
the venture, or the inclusion of cargo freight in
the contract. Alitalia began to be closely associated to the
arrangement in April 2010, although its association
was formalized only later on, and the pact is planned to last at
least until the end of March 2022, with Alitalia
part of all the initiatives of the venture, including many that
have been decided long before it joined the
venture (e.g., the agreement on joint sales). As for cargo
freight, financial terms had to be substantially
renegotiated after the 2008 crisis. Again, the adaptation was
endorsed without any major problem.
-
34
satellite dedicated to Sky Team at Roissy airport, which created
an opportunity for an
informal meeting of the CEOs).
5.5: A sustainable agreement?
Of course, important questions remain to be answered regarding
the future of this hybrid
arrangement within the global alliance. Is it only a step
towards full integration? And how
will partners to the global strategic alliance who are not part
of this arrangement react in the
long run? It is difficult to provide a clear answer at this
point, since the venture is relatively
young, concerns only part of the activities (and therefore the
rights) of the parties involved,
and is developing in an industry submitted to rapid changes.
However, several factors
provide indications that the arrangement may remain stable,
without a formal integration,
for a significant period of time. Let us illustrate with a few
insights.
First, as already mentioned, parties to the agreement share
substantial decision rights on
part of their activities without sharing property rights, and it
is not clear that competition
authorities who delivered the antitrust immunity would accept
any move in that direction.
Nevertheless, partners have developed interdependence on
specific assets, which provides
incentives to keep the arrangement stable. For example, Air
France recently bought an
Airbus 380 to deserve NY for its own sake and that of its
partners. Would the venture
collapse, Air France could well end up with over-dimensioned
capacity and costly fixed costs
to support (pilots, stewards etc.). Symmetrically, Delta has
closed many of its European
agencies, leaving to Air France-KLM the task of selling its
flights. Would the venture be
terminated, Delta would have to rebuild a substantial sales
force. Other examples could be
provided, e.g., the joint development of a costly information
system, the sharing of strategic
information on costs and the efficiency and profitability of
different segments of the shared
network, etc. In other terms, we are typically in a situation of
interdependent firms with
respect to some specific assets while other strategic assets
remain distinct, which is typical
of hybrid arrangements and the source of their stability.
Second, this interdependence raises the risks of opportunistic
behavior fed by contractual
hazards. Transaction cost economics has pointed out that when
these hazards are paired
with important investments specific to the agreement, there are
strong forces pushing
towards integration. Clearly, the parties to the venture under
review have adopted
-
35
measures to face these difficulties. The mode of governance
specified in the contract, the
decision process it contains and the codes of conduct embedded
in the contract define ex-
ante a formal framework to shape the behavior of agents and
limit risks of opportunism.
Also, partners have substantially invested in measures oriented
towards building and
encouraging trust. The organization of working groups as teams
sharing joint procedures,
developing common habits overtime and learning to operate
cooperatively would involve
important switching costs, including psychological ones, would
the venture fail. The
significance of investments in resources that are complementary,
notwithstanding the fact
that property rights remain distinct, and the resulting
interdependence already mentioned
also favor stability of the contractual arrangement.
Third, we also know from the literature on hybrid arrangements
that important conditions of
stability rely on the mode of governance implemented and the
procedure of selection at
entry. We have already discussed the first issue, emphasizing
the important combination of
formal and relational mechanisms (see subsection 5.3), which
confirms what the literature
on relational contracts predicts (see Baker et al., 2008;
Malcolmson, 2012). Selection at
entry is another important issue. Parties to the arrangement
under review have been very
cautious in focusing on partners that had already developed a
long experience of
cooperation, including as founders of the global alliance, which
is the case for Air France and
Delta.
However, this selectivity may well create problems with parties
that are members of the
global alliance but not integrated in the venture. In a sense,
restrictions imposed by
regulators may contribute to the consolidation of the existing
venture, giving arguments to
its participants to not accept new members, or to do so very
selectively, in fear of
developing positions that would not be acceptable to competition
authorities. If it is so, it
means that the stability of the venture depends on a complex
combination of endogenous
factors, such as the mode of governance implemented, and of
exogenous factors coming out
of the regula