Mouyrin Renaud Ph.D. Student CEROG IAE Aix en Provence Financed - - PowerPoint PPT Presentation

Does Multipoint Competition amongst European Airlines Lead to a Reduction of Competitive Pressure?

Mouyrin Renaud

Ph.D. Student CEROG IAE Aix en Provence Financed and Supported by Eurocontrol Experimental Center Strategy Unit

Agenda

• Theoretical Background
• Purpose of the study
• The Research Model & Hypotheses
• Methodology

– Data – Measures

• Preliminary results provided by an alternative

model

• Next steps

Theoretical background: What is Multipoint Competition?

• Situation in which firms meet simultaneously the

same rivals in more than one market

• As each City Pair connection can be considered

as discrete market…

• …airlines probably have multimarket contacts

with competitors

• One major element of the competitive interaction

between 2 airlines that may influence their competitive behavior towards each other.

Theoretical background: The Mutual Forbearance Hypothesis

• Theory: Familiarity

Multimarket Competitive Contacts Intensity Deterrence

• Multiple contacts give the option to respond to an attack

by a rival not only in the challenged market

• Airlines Competitive moves are linked across markets
• Specter of retaliation reduces the motivation to act

aggressively

Example in the US airlines industry (1989)

• AW entered in H. with low introductory

fares.

• CA responded by lowering prices from P.
• AW stopped immediately its low fares in H.
• AW and CA entered into a code share

agreement

Phoenix Houston America’s West hub Continental Airline’s hub

Theoretical background: Past Researches

• Empirical studies in the US airline Industry
• Multimarket contacts lead to higher averages fares

(Evans and Kessides, 1994; Gimeno, 1999)

• Multimarket contacts in rival’s strategic markets (High

Market share, High Market dependence, City Pair connectivity in the airline network) are more effective than contacts in rival’s

secondary markets. (Gimeno, 1999)

• Airline’s Market Entry in and Market Exit from rivals

markets is unlikely when the 2 airlines are strong multipoint competitors (respectively to avoid competitive escalation and to maintain a deterrence mechanism) (Baum and Korn, 1996 and 1999)

Purpose of he study: Why testing Mutual Forbearance Hypothesis in the European Airlines Industry?

• Data source geographical transposition, from the

US airlines industry to the European One.

• We have not yet reviewed an empirical study on

the Multipoint Competition effects on the European industry

• Is the competitive and the institutional

frameworks of the European industry is adapted for airlines to achieve Mutual Forbearance Strategy?

• Understand and highlight one potential source
• f European airlines industry competitive

distortion.

The Model Research

Multipoint Competition Airline competitive commitment change in rival’s territories Airline exit from rival’s territories Airline Entry in rival’s territories Multimarket Contacts Airlines competitive Interactions. Competitive Moves.

? Period t Period t+1

Multimarket Contacts and the dynamics of competitive interactions

• Decreased aggressiveness that follows

recognition of mutual dependence and potential for multimarket retaliation

• Recognition of mutual dependence and

potential for multimarket retaliation Suppressed by Suppressed by HIGH

• Incentive to establish a foothold to signal

ability to counterattack

• Aggressive interactions among airlines that

meet in one or only a few market routes

• Aggressive interactions among airlines that

meet in one or only a few routes Stimulated by Stimulated by LOW Market exit rate / decrease in flight frequency and seat capacity Market entry rate / increase in flight frequency and seat capacity

Multimarket contact

Effects of multimarket contacts on rates of market entry and exit and change in flight frequency (Baum and Korn, 1999)

Multimarket Contacts and the dynamics of competitive interactions

Likelihood of Market entry in and exit from rival’s territories

Multimarket contacts

Hypothesis

• H1: A firm’s rate of entry into a competitor’s

markets is related in an inverted manner to the level of multimarket contacts with the competitor.

• H2: A firm’s rate of exit from a competitor’s

markets is related in an inverted manner to the level of multimarket contacts with the competitor.

• H3: The likelihood of a firm competitive

commitment change into a competitor’s markets is related in an inverted manner to the level of multimarket contacts with the competitor.

Methodology: Data and sample description

• Longitudinal and cross-market study. OAG data Base

from 2002 to 2006.

• Scheduled, Regular and Direct flight within the European

Union (25) + Norway and Switzerland: Homogeneous Institutional framework

• 3963 City Pair (markets); 119 carriers; 10506 unique

Competitor-dyad/year (pairs of airlines by year)

Methodology: Measure of Independent variables

• Multimarket contactsijt between 2 airlines=

Sum Mit [Cimt x (Dimt x Djmt)]+ Sum Mjt[Cjmt x (Dimt x Djmt)] Mit + Mjt

Sets of Markets served by firms i and j at time t Centrality of the route m to the route networks of airlines i and j at time t (airlines routes’ strategic importance) Indicator variables set equal to 1 if airlines i and j are active in route m at time t and 0 otherwise

Examples

0.02 12 34.04 285 Air France 38.44 232 Iberia W2002 0.02 12 38.44 232 Iberia 34.04 285 Air France W2002 0.11 12 39.72 150 Alitalia 100.00 12 Volare Airlines W2002 0.11 12 100.00 12 Volare Airlines 39.72 150 Alitalia W2002 0.17 12 98.11 106 KLM 85.71 20 Transavia Airlines W2002 0.17 12 85.71 20 Transavia Airlines 98.11 106 KLM W2002 0.32 16 80.21 24 Blue1 98.57 60 Finnair W2002 0.32 16 98.57 60 Finnair 80.21 24 Blue1 W2002 0.44 18 74.07 24 Aegean Airlines 44.44 35 Olympic Airlines W2002 0.44 18 44.44 35 Olympic Airlines 74.07 24 Aegean Airlines W2002 Weighted multipoin t contacts Net Multipoint Contacts Average Airline j's centrality of routes in contacts with i Airline j set of markets Airline j Average Airline i's centrality of routes in contacts with j Airline i set

• f

Markets Airline i PERIO D

Reliability Statistics .717 3 Cronbach's Alpha N of Items

Measures: Dependant variables

• Airline i Entry in rival’s j markets

– The seasonally number of entry by airline i into each of its j competitor’s route

• Airline i Exit from rival’s j markets

– The seasonally number of exit by airline i from each of its j competitor’s route

• Airline i change in competitive commitment in rival’s j markets

– Average Airline i Change in flight frequency – Average Airline i Change in seat capacity in each of its j competitor’s route – Average Airline i Change in market share

• Data Reduction Process through Principal Component Analysis

Factorial Score Computation for each observation

Other Control Variables

• Relative size (size j / size i)
• Age of Airline i and j
• Airline i’s route dominance over competitor j
• Airline i’s entries into competitor j’s route (previous year)
• Airline i’s average capacity (seats) route
• Airline i’s average route density (number of competitors)

Preliminary Results

• Not yet achieved empirical test of the research

model presented.

• Previous Model based on a market unit analysis

(Airline route) not at firm level (interaction between airlines), provided exploratory and complementary results.

15.48

Airlines Maintaining in markets

10.86

Airlines Exiting Markets

11.78

Airlines Entering new markets

15.13

All Airlines Average Multimarket Contacts with each City Pair Competitors

Variable Selection

Preliminary results: Entry Rate and Multimarket Contacts

100,00 80,00 60,00 40,00 20,00 0,00

Entry_rate Correlations 1

• .239**

.000 1437 1437

• .239**

1 .000 1437 1437 Pearson Correlation

• Sig. (2-tailed)

N Pearson Correlation

• Sig. (2-tailed)

N Entry rate Multimarket contact Entry Rate multimarke t contact Correlation is significant at the 0.01 level (2-tailed). **.

Preliminary results: Airlines’ competitive commitment variance decreases while MC increase.

2,50000 0,00000

• 2,50000

Airlines_change_comp_commitment

Multimarket contacts

Preliminary results conclusions

• The choice of markets to enter in and to exit

from is partly determined by the level of multimarket contacts an airline has with its competitors.

• The magnitude of airline’s changes in flight

frequency, seats capacity and in market share, is also determined by multimarket contacts it has with its competitors.

• Stability of an airline competitive behavior is

more likely when it has a high level of multimarket contacts with its competitors

Next Steps

• The preliminary analysis allows us to pursue

investigations on European Airlines Mutual Forbearance Strategies but not to (in)validate hypotheses.

• As Multipoint Competition is a property of the

relationship between two airlines…

• …we need to determine a fully model,

integrating all competitor-dyadic level

• bservations (with control variables), to be

statistically tested through Poisson and Binomial regression analysis, which is actually running…