Technological coopetive initiatives and firm performance in
international markets:
the case of global airline alliances
Master Thesis Strategic Innovation Management
February, 2017
Wouter Michiel Hansen
S2533758
w.m.hansen@student.rug.nl
Supervisor: I. Estrada
Co-assessor: J. Surroca
MSc BA Strategic Innovation Management
Faculty of Economics and Business
TABLE OF CONTENT
TABLE OF CONTENT... 2
1. INTRODUCTION... 3
2. THEORY AND HYPOTHESES... 6
2.1 Coopetition and TCIs ... 6
2.2 Conceptual framework ... 8
2.3 Hypotheses ... 9
2.3.1 Tactical alliance engagement and the creation of TCI scope... 9
2.3.2 TCI scope and firm performance ...10
2.3.3 Timing of entry ...11
3. METHODOLOGY...12
3.1 Research setting and sample ...13
3.2 Measures...14
3.3 Regression assumptions ...17
4. ANALYSIS AND RESULTS ...18
4.1 Tactical alliance engagement and TCI ...19
4.2 TCI scope and firm performance...20
4.3 Moderator: timing of entry...22
5. DISCUSSION...24
5.1 Theoretical and managerial implications ...25
5.2 Limitations and future research ...26
W. M. Hansen
Technological coopetive initiatives and firm performance in international
markets:
the case of global airline alliances
AbstractThis study examines the coopetition-performance relation in the context of alliance portfolios. After previous studies indicated mixed evidence on the coopetition-firm performance relation, this study divides the relation into measurable sub-relation. By focusing on these sub-relations it is possible to study which tactical alliance partner conditions are optimal for technological coopetitive initiative (TCI) development. Subsequently, by dividing firm performance into financial and market performance it is possible to analyze the effect of TCIs on the internal firm as well as the external consumer environment. Using a sample of tactical multilateral alliances in the airline industry (global airline alliance) in the period 2011 to 2014 this study is able to examine and empirically test the business literature. Results from the empirical tests suggest the importance of collaborations and engagement with tactical alliance partners to increase the TCI scope development. Furthermore, it was found that an increase in TCI positively relates to financial performance and, opposed to what the literature stated, the TCI has an insignificant relation with market performance. Besides, the moderating effect of timing of entry depicts a positive effect on the relation between TCI and financial performance, no significant effect was found for the moderating effect between TCI and market performance. This study contributes to the extant business literature by providing evidence for; (1) the optimal partner conditions in alliance portfolio to improve TCI; (2) the effect of TCI on firm performance; (3) the moderating role early entry in the tactical multilateral alliances.
1. INTRODUCTION
Verona and Ravasi, 2003). Since no firm has all the needed resources and capabilities to gain and sustain competitive advantage, firm performance is highly related to an increase in the ‘TCI scope ’. The TCI scope is the total number of commercial services offered by the focal firm that is able to meet the need on the consumer market. It is the extent to which the focal firm is able to acquire and combine value chain activities from the alliances and thus is the actual number of collaborative activities of the focal firm (Oxley and Sampson, 2004; Mascarenhas and Koza, 2008). As the industry is globalizing and competition intensifies, strategic management should focus on cost and risk sharing with partner firms through innovative processes and manage the collaborative value creation and appropriation strategies (Bengtsson and Kock, 2014; Teece, 2007). These compelling factors striving to a more technology and knowledge based economy stimulate the process of business model innovation facilitated by e-technologies to create synergistic gains (Das and Teng, 2000; Grant and Baden Fuller, 2004; Teece, 2010). By focusing on an increase in TCI scope it is possible to accelerate firm performance. Consequently, through alliances the focal firm is able to profit from innovation, namely by financial performance and market performance, which are measures of firm performance (Murray et al., 2005).
Although existing research on strategic alliances and TCI development is enlightening, several outstanding issues remain. Collaborations with competitors through alliances are increasing in popularity, even though the success rate is perceived to be rather low, only 30 to 50 percent (Lazzarini, 2007). Despite this low success rate, firms are still highly motivated to join strategic alliances as a way to survive in the highly competitive industries. Hence, an important question appears based on this phenomenon; do technological coopetive initiatives (TCIs) enhances firm performance? Moreover, whereas several studies focused on the relation between TCI and firm performance, mixed results were found and extant research can not provide a clear answer on the effects. While several studies reported a negative relationship between TCI and firm performance (Nieto and Santamaria, 2007; Bengtsson et al., 2010; Bengtsson and Kock, 2014; Gnuawali and Park, 2011) other studies found more intriguing arguments supporting the positive relationship between TCI and firm performance (Estrada et al., 2014; Park et al., 2014; Gnyawali et al., 2008; Wu, 2014).
alliance strategies for network expansion and firm positioning in the highly competitive industries and thus improve firm performance, which are of greatest significance for policy-makers. Therefore, the leading question within this study is: Under which conditions can the focal firm benefit most from technological coopetives initiatives in the case of alliacnes? To answer the research question the following sub-questions are developed: (1) What tactical alliance partner conditions are most beneficial in the creation of TCIs? (2) Does an increase in TCI scope improves firm performance, and more specific is there a difference between the effect on firm performance and market performance. (3) And finally, what is the role order of entry in the tactical alliance on this relation?
By answering these questions, it is possible to extent the theory on the optimal alliance partner conditions of the coopetition – performance relation. Hence, through this research it will be possible to replace the previous inconsistent findings and put them into one clear direction. Unique possibilities can appear when the right complex alliance partner conditions are found which will increase the TCI scope and consequently firm performance; (a) financial and (b) market. Although firms may outrival in technological invention, firms can still lack value creation and commercialization (Teece, 1986). These differing performance measures can create a more specific indication of the tension between value creation and appropriation and can be used to elaborated on the effects of an increased TCI scope on firm performance (Teece, 2007). Additionally, this relation will be empirically tested by the moderator timing of entry, to see if it is beneficial to be a founding partner or a late entrant. Consequently, this sheds lights on the comprehensive understandings about timing of entry in tactical multilateral alliances.
the alliance strategy the focal firm can create an efficient and effective alliance portfolio which consequently increases firm performance.
The paper is structured in the following sections. First, the theory will discuss coopetition through strategic alliances and subsequently elaborate on the mixed findings between coopetition and performance by focusing on the benefits and drawbacks of TCIs. Subsequently, the hypotheses are developed on the optimal conditions of multilateral alliance portfolio partners and the TCI scope and the relation between TCI scope and firm performance moderated by timing of entry. Then the methodology section is presented in which the method is discussed. This part is followed by the analysis and result section of the findings. Finally, the discussion focusses on the theoretical and managerial implications followed by the main limitations and suggestions on avenues for future research.
2. THEORY AND HYPOTHESES
2.1 Coopetition and TCIsAlliances in general can contain anything from arm’s length contracts to joint ventures, and function as a form of collaborating with competitors: coopetition (Bouncken et al., 2016). Hence, the consensus about coopetition is on the inter-firm relation between competing firms that cooperate in the achievement of mutual benefits through technological coopetive initiatives in the pursuit of creating a competitive advantage (Bouncken et al., 2016; Park et al., 2014; Tsai, 2002). Based on the work of Yoshine and Rangan (1995) and Bouncken et al. (2016) it is possible to create a more detailed and specific condition to define coopetition through alliance: (1) There should be a set of agreed upon collective goals, but firms should remain independent after alliance formation. (2) Alliance benefits, performance, and control, including tasks, should be shared. (3) All partners should continuously contribute to the strategic areas and goals of the alliance. Besides, coopetition exists on several levels and can include suppliers, consumers and competitor relations on the bilateral or the multilateral level. Bilateral agreements contain relations with one single alliance partner and thus only involving two firms, whereas multilateral agreements are a cooperative arrangement involving at least three partner firms (Lavie et al. 2007; Li et al., 2012). Within this study the alliance resides on the horizontal relation between several competing firms that formed a tactical multilateral partnership wihtin the same business industry in the pursuit of creating TCIs (Bengtsson and Kock, 2014).
increase in technological complexities, firms are internally restricted to innovate on a continuous basis. Consequently, firms reach out to external sources of knowledge through alliances since competitors often have the most significant and valuable resources and capabilities because they face comparable business challenges (Elyas, 2013; Lavie, 2007; Park et al., 2014). Hence, this suggest that alliances help to enhance collaborative innovation in the form of TCIs. Following the literature, TCIs can be explained as a collaborative innovation through technology and information based systems that enhances efficiency and effectiveness in the creation of a shared competitive advantage (Barua et al., 2004; Laukkanen et al., 2007; Nigro, 2016).
As a motive of this study, previous research indicated mixed findings concerning the coopetition – firm performance relation. Therefore, the following sub-paragraphs will elaborate further on these mixed findings.
2.1.1 Benefits of coopetition and TCIs
Several studies have demonstrated positive effects of an increase in coopetition on TCI development and firm performance. This stream of research indicates more efficient and effective new processes that focus on improving firm performance (Nigro, 2016). Hence, with implementing information-technology systems it is possible for alliances to execute collaborative products and processes that provide integrated solutions to consumer needs and an increase in economic returns (Ku and Fan, 2009). For example, joint R&D can change products and processes dynamically to not only create value but also capture value from innovation (Pisano and Teece, 2007). Also, using collaborative information-technology based structures can help the firm to improve lead-time, customer service, financial control and standardization. This is made possible by accessing new and complementing information on resources, markets, and technologies (Gulati et al., 2000). Hence, TCI is the driving factor behind business model innovation and is a result of the growth and development of internet and e-commerce, the emerging knowledge economies, outsourcing and offshoring of business
activities, and restructuring service systems (Lahiri and Narayanan, 2013; Teece, 2010).
Consequently, innovation in business models through TCI changes the way firms create profit nowadays (Teece, 2010). Since innovation is the essential strategy in which firms and alliances can create value and appropriate their competitive position (Brown and Eisenhardt, 1997; Ahuja and Katila, 2001).
2.1.2 Drawbacks of coopetition and TCIs
can harm further development of TCIs. For example, coopetition creates the need for a critical balance between common alliance benefits and focal firm benefits (Park et al., 2014). More fundamentally, rigidities and difficulties can arise from these contradictory demands that exist between the different alliance partners. Also, to realize joint TCI benefits competing firms should engage in close interaction with each other, which can be difficult to perform or it can create risks and costs related to trust and knowledge protection (Gnyawali and Park; 2011). This since the developing process of joint value creation emphasizes on close interaction to recombine knowledge to create synergistic gains through TCIs (Gnyawali and Park; 2011). Furthermore, firms can lack the ability to appropriate value from the opportunities that are created through coopetition. In order to realize these TCI benefits, firms should overcome certain jeopardies of knowledge spillovers which can trigger cost of protection against partner opportunism and behavioral uncertainty (Nieto and Santamaria, 2007; Bengtsson et al., 2010; Bengtsson and Kock, 2014). Thus, when a new coopetive business model is established, firms can experience problems with a strategy of value creation, value delivery and capturing mechanisms which is noteworthy for firm performance (Teece, 2010). More specific, research suggests (Park et al., 2014) that tensions in coopetition can result in rivalry, conflicts and uncertainties within the alliance that reduce stability and effectiveness and decrease TCI development.
2.2 Conceptual framework
The conceptual model explains all the relations between the hypotheses. First, the positive relation between an increase in engagement with the tactical alliance partners is displayed. Hereafter, the positive relation between an increase in TCI scope and an increase in firm performance is depicted in which financial and market performance are used as variables. This is followed by the moderating variable of timing of entry in which early entrance has a positive effect on the relation between TCI scope and firm performance.
Tactical alliance
engagement TCI scope
2.3Hypotheses
2.3.1 Tactical alliance engagement and the creation of TCI scope
The tactical multilateral alliances focus on the creation of technological coopetive initiatives (TCIs). Increasing the TCI scope is a way how firms should deal with market uncertainties by creating competitive advantages and reducing risks and secure market survival. Using the work of Schumpeter (1934), innovation includes combining and recombining new and existing resources. Besides an increase and intensification of partner collaboration it provides improved availability of external knowledge to the focal firm. The multilateral alliances involve multiple partners in the development process, which functions as a heterogeneous knowledge pool (Li et al., 2014). These new TCI based business models define how the alliance can combine consumer needs and technological trajectories to create and deliver consumer value (Teece, 2010).
In this study it is expected that participating within these safe and close tactical multilateral alliance provides a conductive business environment that helps to maximize the collaborative innovative potential (Peltokorpi, 2014). For example, the increase in engagement with the alliance results in the emerge of improved and more efficient inter-organizational processes that are the resource to be innovative. Hence, an increase in the alliance collaboration positively influences firm performance through knowledge recombination, knowledge spillovers and learning structures. Also, more recent research shows that an increase in the intensification of alliance participation results in an increase in the total number of new services offered by the firm as well as an increase in the number of innovations (Baum et al., 2000; Rothaermel and Deeds, 2004; Goncalves and Goncalves, 2008; George et al., 2001),
Moreover, the increase in engagement results in more resources, increased market power, risk sharing, scale advantages and joint R&D that increase and stimulate the development of TCIs and consequently results in a broader TCI scope (Bouncken et al., 2016; Song, 1998; Goh and Uncles, 2003) Thus, an increase in multilateral tactical alliances engagement results in an increased and more diverse knowledge pool which consequently relates positively to TCI development (Lavie et al., 2007; Li et al., 2014).
collaboration and reduce hazards. When the firms apply these different and costly governance structures, the willingness to share knowledge decreases (Sampson, 2007).
Based on these facts, an increase in engagement with the tactical alliance relates positively to an increase in the TCI scope, resulting in the following hypotheses:
H1: The higher the focal firm’s engagement in the tactical alliance the higher the TCI scope
2.3.2 TCI scope and firm performance
The main focus of strategic alliances is on reducing costs and driving revenues by the means of creating a collaborative competitive advantage (Lavie, 2007). Combining these collaborative competitive advantages will be the source of the TCI scope development, which is a mean by which the focal firm can comply with the increasing market demands (Teece, 2010). To illustrate, an increase in engagement with the tactical alliance partners increases the number and availability of collaborative competitive advantages which will increase the TCI scope of the focal firm. The TCI scope is the total number of commercial services offered by the focal firm that is able to meet the needs of the consumer market. It is the extent to which the focal firm is able to acquire and combine value chain activities from the alliances and thus is the number of collaborative activities of the firm (Oxley and Sampson, 2004; Mascarenhas and Koza, 2008). This since this is the extent to which the firm is able to combine value chain activities of the alliance to broaden and increase the commercial services offered (Carpenter and Fredrickson, 2001; Oxley and Sampson, 2004). Consequently, when the focal firm is able to increase its services offered it is possible increase firm performance.
improves the flexibility to do business and enables the firm to match the constantly changing environment (Belderbos et al., 2004). By more efficiently incorporating innovation these collaborative technologies create a future-focused IT solution in the high-performance industries. Besides, seamless integration by technology, IT infrastructures can make data available throughout the whole alliance and improve informed decision making. Consequently, these powerful applications result in technological architectural frameworks that create a highly integrated business model to maintain the collaborative competitive market position and improve the companies’ financials (Sarkar et al., 2001;Hoang and Rothaermel, 2005).
By improving both the company financials and the consumer experiences an increase in the TCI scope relates positively to financial and market performance. Thus, the following hypotheses were developed:
H2a: There is a positive relation between an increase in TCI scope and an increase in financial
performance.
H2b: There is a positive relation between an increase in TCI scope and an increase in market
performance.
2.3.3 Timing of entry
The early mover advantages of the founding partners are derived from the path dependency in alliance evolution, efficient management and other advantages. The founding partners have the possibility to influence the evolution of the strategic alliance and influence the TCI development process (Hite and Hesterly, 2001; Lavie, 2007). For instance, advanced industry standards and technological specifications can help to ensure efficient operations (Lavie, 2007). The founding partners have a central position within the strategic alliance, additional they have a different alliance focus based on their rich information flows and internal coalitions. These benefits of early entry allow for optimizing strategies and selecting the right new partners that are complementary for further TCI scope development (Zollo et al., 2002). Besides, when focusing on firm performance the founding partners do outperform late entrant firms since they have better developed performance goals, processes and strategies (Su et al., 2009). Subsequently, early engagement in the alliance improves processes, increases market power, improves risk sharing, resource complementarity and scale advantages (Hitt et al., 2000).
Given these circumstances, when focusing on timing of entry, this study sheds light onto the important early stages of alliance formation and specifically the positive moderating effect of early entrance in the tactical alliance on the relation between TCI scope development and firm performance (Jinju et al., 2016). Consequently, the following hypotheses were developed:
H3a: Early entry into the multilateral alliance has a positive moderating effect on the positive
relation between TCI scope development and improved financial performance.
H3b: Early entry into the multilateral alliance has a positive moderating effect on the positive
relation between TCI scope development and improved market performance.
3. METHODOLOGY
3.1 Research setting and sample
The research setting of this study is the airline industry from which the data is retrieved and empirically tested. To test the hypotheses a comprehensive pooled panel dataset (2011-2014) was used on a sample of airlines in the global airline industry. The airline industry provides an ideal setting for testing the hypotheses since this business field is highly characterized by extreme forms of competitions and demand response time. The data used within this study is obtained from multiple
databases (airline route, airline business and SDC), top airline rankings, annual reports and
corporate websites.1First, the relevant records and data was compiled for the period of 2011 – 2014,
which created a longitudinal dataset with multiple lags per variable.
By performing empirical tests, it is possible to determine the causality and create a trend analyses. When using a pooled panel measure, it is possible to test time series for data points in t and
t+3 and preform a robustness test for data points in t+1 and t+2. A disadvantage of this pooled panel
data testing is data loss during the statistical test. Hence, this data measures frequency over time and contains multiple time periods for the same firm. Since the data used in this study contains data points that are quite closely related and linear in time the test can still give useful indications and the data loss is covered by estimating a robustness test on these points.
In order to ensure a complete coverage, the database has been completed and corrected by global airline reports and market indexes. In total this created insights in the three global airline alliances and developed a sample of 60 airlines. The sample consists of airlines that are member of one of the three global airline alliances. Due to this membership the sample consist only out of firms that have their information publicly available, which in this study is a perquisite to collect and complete the dataset. Therefore, with this sample it is possible to create a sufficient indication and comparison of one of the most competitive international markets. The change in the international business environment is especially evident in the global airline industry in which competitive advantages can be originated from being a member of an airline alliance (Lazzarini, 2007). The airline industry has a major part in the world economy since air transport has a broad impact on the development of world trade and tourism but also the role in the global challenge of reducing carbon emissions (Tugores-Garcia, 2012). Passengers- and air-transport are operated by commercial airlines that comprise the challenging industry which is characterized by low profits and high volatility in returns. Important is the appearance of alliances which combine the largest airline carriers in the 1The data within this study is retrieved from the following websites as well as all the corporate global alliance websites.
(n.d.). Retrieved December 21, 2016, fromhttp://www.airlinequality.com/
(n.d.). Retrieved December 21, 2016, fromhttp://www.skytraxresearch.com
world and are rephrased as the global airline alliances (table 1): Star Alliance, oneworld, and SkyTeam.
Table 1 - Global airline alliances
Global airline alliances Star Alliance (1997) (n27) oneworld (1998) (n13) SkyTeam (2000) (n20)
ADRIA Airberlin Aeroflot
AEGEAN American Airlines AerolineasArgentinas
AIR CANADA British Airways AeroMexico
AIR CHINA Cathay Pacific AirEuropa
AIR INDIA Finnair Airfrance
AIR NEW ZEALAND Iberia Alitalia
ANA JAL Japan Airlines China Airlines
ASIANA AIRLINES TAM/LAN/LATAM China Eastern
Austrian Malaysia Airlines China Southern
EVA AIR Qantas CSA Czech Airlines
AVIANCA Qatar Airways Delta
Brussels Airlines Royal Jordanian Garuda Indonesia
Copa Airlines S7 Airlines Globus, LLC. Kenya Airways
CROATIA AIRLINES KLM
Egyptair Korean Air
Ethiopian MEA
LOT POLISH AIRLINES Saudia
Lufthansa Tarom
SAS Vietnam Airlines
Shenzen Airlines Xiamen Air
SINGAPORE AIRLINES South African Airways SWISS TAP PORTUGAL THAI TURKISH Airlines United 3.2 Measures
Tactical alliance engagement
and is conducted as a relative presence of partners of the global airline alliance in the focal firm’s alliance portfolio. (tactical alliance partners’ / alliance portfolio) And, (b) the actual number of tactical (global airline) alliance partners of the focal firm. The difference appears in the fact that the relative presence explains the importance of engagement with partners from the tactical alliance, whereas the real number of partners is the exact number of agreements. Since the focal firm is not required to have alliances with all partners from the tactical alliance every firm has a different alliance portfolio consisting of partners from the tactical alliance and other partners. By creating two separate measures it is possible to test how an increase in the relative presence of partners of the global airline alliance in the focal firm’s alliance portfolio and an increase in the actual number of global airline alliance partners increase the TCI scope. Hence, by doing this it is possible to develop insights in the optimal condition of the global airline alliance in the pursuit of TCI scope development.
Technological Coopetive Initiative scope
The TCI scope is a measure of the number of initiatives and activities of the focal firm that is a result of the technological coopetive collaboration created by the tactical alliance as well as other alliance partners. This collaborative created competitive advantage within the airline industry is called a code-sharing agreement. This contains the possibility where one airline’s designator code is used on flights that are operated by an alliance partner airline in 2011 (Gayle, 2008). In more detail, this variable is based on the code-shared (a) routes and (b) destinations of the focal firm as a result of the TCIs and function as a collaborative competitive advantage. The code-shared routes and destinations are the result of the collaboration of the focal firm in technology and services with alliance partners that are both from individual alliance and the tactical alliance. These numbers are retrieved from
routesonline which contains the code-shared routes of 2011 and are counted manually per airline.
For the focal firm an increase in TCI supports the increase of firm performance as it results in improved possibilities to serve the markets. Consequently, therefore the variable will be tested as a combined measure called the TCI scope. Furthermore, the level of analysis is the focal firm its code-shared routes and destinations together per year. This can contain any figure depending on the size and collaboration intensity of the focal firm. Through this measure it is possible to compare the focal firm with the other firms in the tactical (global airline) alliance and see how engagement with tactical alliance partners relates to the TCI scope. Hence, this is interval data.
Firm performance
Within this study firm performance is obtained reflected through financial performance and market performance in the period of 2011 - 2014. By selecting two types of performance measures it is possible to cover financial as well as market based indicators. This to overcome the limitations of measuring performance only from the company’s view. Rather market performance enhances insights of the consumer experiences and thus firm performance now focuses on the multiple sides of performance. Financial performance is indicated as a proxy of revenue and is depicted from the airlines annual reports that are made publicly available. Consumer satisfaction is a form of market
performance and will be measured by the the Skytrax quality indicator (1-5 scale). The Skytrax
indicator is an independent quality indicator that exists since 1999 in the esteem to provide clarity on the global airline quality in the respect of the global airline passengers. Based on around 20 million customer surveys with respondents from over 104 nationalities, airlines are judged on 41 key performance and quality indicators.2Based on those two indicators of firm performance it is possible
to empirically test the literature. Hence, both variables are interval data.
Moderator: timing of entry
The variable timing of entry of the focal firm in the tactical alliance has a moderating effect on the relation between TCI scope in 2011 and firm performance in 2011 – 2014 in which early entrance has a positive moderating effect due to increased efficiency and effectiveness (Jinju et al., 2016;
Tugores-Garcia, 2012). To measure timing of entry, a dummy variable is created were (1) is a measure of early entrance and (0) stands for late entrant firms. Within this study we use the alliance members of the first two years as a measure for early entrant firms and late entrant firms are the firms that joined the alliance after two years. By testing the effects of early and late entrance on the TCI scope (innovation) – performance relation it is possible to explain the effect of timing of entry into the alliance and help firms to enrich their alliance strategy in the context of innovation and firm performance.
Controls
included to make the outcome more reliable. Relying on the number of countries, the daily departures and the number of annual passengers it is possible to control for any size related influences on performance in the period of 2011 – 2014. Based on previous work (Jinju et al., 2016;
Tugores-Garcia, 2012) there are indications that these control variables can have a positive influence on the performance factors as an increase in these control variables can relate to efficient and effective business practices. The control variables used can entail circumstances that can influence the assumed relation between the two variables and can increase the possibility of a positive relation. Hence, these variables function as interval data as they are direct measures of size.
Table 2 – Measures
Variables Tactical alliance engagement
TCI scope Firm performance
Timing of entry
Controls
Measures Relative presence of tactical alliance partners in the alliance portfolio Code-shared routes Financial performance (Revenue) Founding firms (early entrants) - Countries active - Daily departures - Annual passengers Actual number of tactual alliance partners Code-shared destinations Market performance (SkyTrax quality indicator) Late entrants 3.3 Regression assumptions
The regression assumptions will first be tested to make sure that the dataset can be used. First of all, the regression assumptions state that there need to be two variables of metric scale, so ratio or interval data. Also, as a rule of thumb for the sample size, regression analysis requires at least twenty cases per independent variable to do an analysis. As the basic regression assumptions are met, then the following regression assumptions should be checked and tested. Finally, if all assumptions are met the statistical test can be performed.
autocorrelation, in which scores around the 2 indicate no autocorrelation. Fifth, the last assumption test for homoscedasticity, which illustrates consistence and equality of the error terms, performing the test homoscedasticity should be rejected and a hetroscedastic dataset is needed.
For the dataset a log-transformation was used to met the normality assumption. When implementing the log-transformation on one variable, the same has to be done to all variables to compare the differences between the variables.
4. ANALYSIS AND RESULTS
Table 3 reports the correlation matrix and the descriptive statistics on the global airline alliance sample, it depicts an indication of the mean, standard deviation, minimum and maximum, and the Pearson correlation of the variables within the dataset. The average number of the actual number of alliance partners is 17,3 partners with a minimum of 3 and a maximum of 31. The average relative number of engagement of global airline alliance partners is 49,7% with the lowest score of 11.1% and the highest score of 85,7%. Financial performance is averaged by a revenue of 9,1 and market performance has a 3,6 on the scale from 1 to 5. Furthermore, the average number of annual passengers is 29 million. The daily departures are averaged on 763,3 and the countries in which the airlines are active are 39,2 on average.
Moreover, table 3 reports high correlations between several variables. Hence, the variables that depict medium to high correlation are only the control variables and have a logical explanation for the correlation. Besides, those there are no other variables that indicate higher correlation than 0,6. Thus, this is a sign that there is no multicollinearity among the key variables. High correlation indicates there there is a linear relation between several variables in which the values range between 0 to 1. Hence, values of 0 indicate no relationship whereas 1 is an indication of perfect correlation. Also, when the correlation coefficient sign is negative it indicates a negative and inverse relation between the variables. Problems can occur when multiple independent variables are highly correlated. But, since there can be certain sources and causes related to the appearance of correlation this can be a logical explanation.
Table 3 - Correlation matrix and descriptive statistics
* correlation is significant at the 0.05 level ** correlation is significant at the 0.1 level
4.1 Tactical alliance engagement and TCI
In table 4 the results of the regression analysis of the hypotheses 1 are depicted, consisting of several models. The variable is tested with (model 5 - 7) and without (model 2-4) the control variables as well as separate tests of the measures of the independent variable and a model with only control variables (model 1). Within all the models the dependent variable is TCI scope. This hypothesis focusses on the direct effects of tactical alliance partner engagement on the TCI scope. Further, to control for unobserved effects the number of annual passengers, the number of daily departures, and the countries in which the airline is active are used.
Before the statistical tests, a number of checks concerning the regression assumptions were conducted. Due to the high correlation in table 3 for certain variables it was needed to perform a Log-transformation since the normality assumption was not met. Besides, due to the indications of table 3 the VIF factors and the related tolerance levels were checked. Table 4, model 1 – 7 all had VIF factors for all variables between 1 and 2,607 and tolerance levels >.422 which do not indicate any statistical problems.
Model 5 depicts the empirical results by examining the measure relative presence of global airline alliance partners in the focal firm’s alliance portfolio and the relation to the TCI scope. According to model 5, there is an indication of the predicting strength and significance of the
Mean S. D. Min Max 1 2 3 4 5 6 7 8 9
variables, the R-Square of the model is .659. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 65,9% with a positive and significant coefficient (p<0.01). Model 6 depicts the empirical results by examining the measure of the actual number of tactual alliance partners and the relation to the TCI scope. According to model 6, there is an indication of the predicting strength and significance of the variables, the R-Square of the model is .733. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 73,3% with a positive and significant coefficient (p<.001). This empirical result confirms the stated hypothesis 1
When combining all measures in model 7, the empirical results do depict a significant outcome that is able to accept model 7 and confirms the relations stated in hypotheses 1: The higher
the focal firm’s engagement in the tactical alliance the higher the TCI scope. According to model
7, there is an indication of the predicting strength and significance of the variables, the R-Square of the model is .757. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 75,7% with a positive and significant coefficient (p<.001).
Table 4 - Regression H1
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
4.2 TCI scope and firm performance
Table 5 depicts the empirical results for hypotheses 2a and 2b, in which the relation between TCI scope and firm performance is tested by separating firm performance into (a) financial performance and (b) market performance. This study empirically tests the direct effects of the independent variable in 2011 on the dependent variables in t (2011) and t+3 (2014). The t+1 (2012) and t+2 (2013) will function as a robustness test.
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
Relative presence GALs
partners -,155(,288) -,877***(,231) ,018 (,356) -,429**(,170)
Actual number GALs
Before the statistical tests, a number of checks concerning the regression assumptions were conducted. Due to the high correlation in table 3 for certain variable it was needed to perform a Log-transformation since the normality assumption was not met. Moreover, due to the indications of table 3 the VIF factors and the related tolerance levels were checked. Table 5, model 8 – 19 all had VIF factors for all variables between 1 and 3,148 and tolerance levels >.318 which do not indicate any statistical problems.
TCI scope and financial performance
Table 5 depicts the empirical results of the TCI scope and financial performance in t and t+3. According to the theory the following hypothesis was developed: H2a: There is a positive relation
between an increase in TCI scope and an increase in financial performance. Based on model 10 (t)
and 13 (t+3) it is possible to accept the hypothesis 2a. According to model 10 and 13, there is an indication of the predicting strength and significance of the variables, the R-Square of the models are .621 and .651. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 62,1% and 65,1% with a positive and significant coefficient (p<.001). In appendix 2 we can find the robustness test for t+1 and t+2 (2012 and 2013) from which the empirical results indicate similar figures.
Table 5 - Regression H2a
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
TCI scope and market performance
Table 6 depicts the empirical results of the TCI scope and market performance in t and t+3. According to the theory the following hypothesis was developed: H2b: There is a positive relation
Variables Model 8t Model 9t Model 10t Model 11t+3 Model 12t+3 Model 13t+3
between an increase in TCI scope and an increase in market performance. Based on model 16 (t)
and 19 (t+3) it is possible to reject the hypothesis 2b. According to model 16 and 19, there is an indication of the predicting strength and significance of the variables, the R-Square of the model are .021 and .046. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 2,1% and 4,6% with a positive but insignificant coefficient model 16 (t) .165 and model 19 (t+3) .160. In appendix 3 we can find the robustness test for t+1 and t+2 (2012 and 2013) from which the empirical results indicate similar figures.
Table 6 - Regression H2b
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
4.3 Moderator: timing of entry
Timing of entry functions as the moderating variable on the stated relation in hypothesis 2 between an increase in TCI scope and the positive effect on firm performance, in which early entrance by founding firms has a more positive effect than late entrance. To make the variables suitable for empirical testing, the independent and the moderating variable will be centralized. By using the mean centering technique, the moderating variable and the independent variable were combined with an interaction term to create a new independent variable that will be used to re-examine hypotheses 2a and 2b in table 7 and 8 (MacKinnon, 2011).
TCI scope * founding firms (early entrants)
According to this method it is possible to modify the strength of the relation between the dependent and independent variable, using the moderating variable timing of entry it is possible to create insights in the importance of the conditions of entry, early or late entry in the tactical alliance. Based
on the theory the following hypothesis will be empirically tested: H3a: Early timing of entry into the
multilateral alliance has a positive moderating effect on the positive relation between TCI scope development and improved financial performance, and H3b: Early timing of entry into the multilateral alliance has a positive moderating effect on the positive relation between TCI scope development and improved market performance. Table 7 and 8 (model 20 – 31) indicate the
hypotheses empirical results, besides the VIF factor scores were all between 1 and 3,373 with tolerance levels >.418, indicating no statistical measuring problems.
Table 7 depicts the empirical results of hypothesis 3a and tested the moderating effect of early entry on the relation between TCI scope and financial performance. According to table 7, there is an indication of the predicting strength and significance of the variables, the R-Square of the model is (t) .606 and (t+3) .645. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 60,6% and 64,5% with a positive and significant coefficient (p<0.01). This empirical result confirms the stated hypothesis 3a. This indicates a more positive relation between TCI scope and financial performance incase of early entrance in the tactical alliance. Besides the same indications were found in the robustness test in appendix 4.
Table 7 - Regression timing of entry H3a
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
Table 8 depicts the empirical results of hypothesis 3b and tested the moderating effect of timing of entry on the relation between TCI scope and market performance. According to table 8, there is no indication of and significance of the variables, the R-Square of the model is (t) .020 and (t+3) .046. This specifies the relation between the X-variable and it explanatory power of the change in Y, and is 2,0% and 4,6% with a positive but insignificant coefficient (p<0.01). This empirical result rejects
Variables Model 20t Model 21t Model 22t Model 23t+3 Model 24t+3 Model 25t+3 TCI scope * Early entry ,556*** (,067) ,286 (,110) ,531*** (,062) ,238 (,098)
the stated hypothesis 3b. Besides the same indications were found in the robustness test in appendix 5.
Table 8 - Regression timing of entry H3b
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
Note: In order to check a moderating effect, you have to have a model containing [control variables + dependent variables+ interaction term]. I did run this regression and obstained the same results than in model 25, but did not include it in the manuscript.
Also: In light of the current findings, H3 does not seem really supported.
5. DISCUSSION
The main findings of the suggest that an increase in engagement with tactical alliance partners has a positive relation with an increase in TCI scope. This for both the relative presence of global airline alliance partners in the focal firms’ alliance portfolio, as for the actual number of global airline alliance partners. Also, the TCI scope relates positively to financial performance, which relation is even stronger when the moderator (early) timing of entry is taken into account. However, the impact of TCI scope on market performance and the moderating effect of timing of entry indicated both insignificant results.
Where in the literature mixed findings made suggestions for further research, this study was able to enhance a deeper understanding on the subject and developed the importance of alliance portfolio partner conditions in which the tactical alliance partners relate positive to the creation of technological coopetive initiatives (TCIs). More specific, based on this relation it is possible accept the positive relation between TCI scope and financial performance which relation also was suggested in the business literature. Unfortunately, in the relation between TCI scope and market performance, empirical results depict opposite outcomes than predicted in the business literature. Therefore, it is possible to conclude that TCI scope is rather focused on efficiency and effectiveness the internal firm rather than the external consumer market.
Variables Model 26 t Model 27 t Model 28 t Model 29 t+3 Model 30 t+3 Model 31 t+3 TCI scope * Early entry ,023 (,018) ,000 (,032) ,016 (,016) -,019 (,028)
Passengers ,119 (,064) ,119 (,067) ,110 (,057) ,119 (,058) Daily departures -,203 (,253) -,202 (,299) -,205 (,222) -,109 (,261) Countries active ,128 (,154) ,129 (,174) ,163 (,135) ,211 (,152) Constant ,590*** (,103) ,533*** (,012) ,590*** (,130) ,597*** (,090) ,545*** (,010) ,549*** (114) Observations 60 60 60 60 60 60 Adjusted Rsquare ,037 ,012 ,020 ,055 ,001 ,046 F 1,765 1,704 1,300 2,146 1,066 1,719
Furthermore, the business literature suggested a positive effect of timing of entry on the TCI scope – performance relation. Hence, there are several important benefits and incentives that let founding firms have high paybacks compared to late entrant firms. The literature suggests that founding alliance partners have higher cumulative experiences, efficiency, and effectiveness. Thus, timing of entry has a positive moderating effect on the positive relation between TCI scope and financial performance.
5.1 Theoretical and managerial implications
Based on the insights of previous research this study makes a number of contributions and extends the understandings of the coopetition-performance relation both empirically as well as conceptually. First, while most business studies focus on the alliance portfolio or even collaborations in general, this study contributes by focusing on the multilateral alliance portfolio partner conditions and the coopetition in the context of firm performance. By focusing on the tactical alliance partners which took part in a multilateral or in this study the global airline alliance it is possible elaborate and empirically proof the importance of engagement with tactical multilateral alliance partners to survive in the highly competitive international markets. By doing this, we are able to create a proxy that is able to help firms with strategizing and managing their alliance portfolio. Following the alliance formation and innovation literature, this study found proof for the importance of the relative as well as the actual number of tactical multilateral alliance partners as an important condition to increase the TCI scope and improve firm performance.
Furthermore, by splitting firm performance this study provides insides on the effect of TCI scope, this by measuring financial performance as well as focusing on the consumer experiences through market performance. By doing this, it was possible to find empirical proof for the positive relation between TCI scope and financial performance. Despite the positive findings, the empirical findings did not found proof for the existence and relevance in the relation between TCI scope and market performance, which indicates that coopetition is focused on internal firm improvements rather than an increase in consumer experiences. This depicts the importance of TCI as a tool to survive in highly competitive markets by improving the focal firms financials through efficiency and effectiveness of internal processes as a result of information and technology based systems.
performance relation. Thus, for the focal firm being a founding partner in a coopetition driven alliance would enhance an even better financial performance.
Lastly, from a practical point of view, the study goal was to contribute to the managerial awareness of inter-firm coopetition through alliances in the highly competitive international markets. It is important to consider the conditions under which the TCI – performance relation is most effective as it highly relates to the internal processes and financial performance of the firm.
5.2 Limitations and future research
This study is subject to several limitations, consequently those limitations provide interesting directions for future research. Within this study the focus is on conditions that have a positive effect on the coopetition-performance relation. This in the aim to provide direction in the previous found mixed results. Hence, through studying this subject it is only possible to focus on limited conditions and perspectives related to the subject as it is only proven that the positive relation is significant. Therefore, it is believed that future studies should focus on others aspects and characteristics of tactical multilateral alliance partnership or try to reject any negative relations that exists within the coopetition – performance relation. Also, future scholars can elaborate on this study and combine it with new insights. For example, differences between firms and alliance characteristics like levels of interaction and intensity would create interesting insights.
Besides, one of the main concerns in testing a specific sample is the generalizability of the results for other industries. The degree of heterogeneity within the sample only accounts for airliners. Though the sample did not focus on budget airlines like RyanAir or fully government subsidized airlines like Etihad or Emirates. Hence, by not using these special cases it can also be concluded that the sample is more applicable for other industries since it focuses on the general players within the industry.
Hence, airline alliances are centralized alliances with formal hierarchical structures. To better understand formal hierarchical structures, the following explanations should be taken into account. A hierarchical structure is an organization that built upon a centralized authority relation with coordination achieved through vertical bureaucratic processes (Tsai, 2002). Since all three alliances have inter-organizational headquarters, this study copes with the question of their influence on airline performance.
Also, the measure of market performance does not give a clear indication and makes it hard to get empirically tested. Since market performance was measured by the use of the Skytrax rating, a good indication about the quality and consumer experiences was done but for further research other measures are suggested. The control variables within this research had mixed effects on the significance and the predicting effect of the independent variable. Hence, therefore future research is needed to add different control variables and measured that could change empirical results.
6. CONCLUSION
Under which conditions can the focal firm benefit most from technological coopetive initiatives? The findings within this study provide important implications for mangers and firms that are active in highly competitive markets. Based on the empirical study, it is possible to generate the right alliance portfolio strategy based on the listed optimal conditions for TCI development. Building on the business literature several sub-questions and hypotheses were empirically tested using a sample of the global airline alliances. Notwithstanding the mixed results in the literature concerning the coopetition-performance relation, this paper aimed to create new insights on this phenomenon and contribute to the coopetition-performance literature by rephrasing the ideal conditions for optimal TCI scope.
REFERENCES
Ahuja, G., & Katila, R. (2001). Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strategic management journal, 22(3)
Anderson, J., & van Wincoop, E. (2004). Trade Costs. NBER Working Papers, National Bureau of Economic Research, Inc.
Barua, A., Konana, P., Whinston, A. (2004). An empirical investigation of net-enabled business value. MIS quarterly, 28
Baum, J., Calabrese, T., & Silverman, S. (2000). Don’t go it alone: Alliance network configuration and startups’ performance in Canadian biotechnology. Strategic Management Journal, 21.
Belderbos, R., Carree, M., & Lokshin, B. (2004). Cooperative R&D and firm performance. Research Policy, 33
Bengtsson, M., & Kock, S. (2014). Coopetition—Quo vadis? Past accomplishments and future challenges. Industrial Marketing Management, 43
Bengtsson, M., Eriksson, J., & Wincent, J. (2010b). Co-opetition dynamics — An outline for further inquiry. Competitiveness Review 20
Bouncken, R. , Clauß, T., & Fredrich, V. (2016). Product innovation through coopetition in alliances: Singular or plural governance?. Industrial Marketing Management, 53,
Brown, S., & Eisenhardt, K. (1997). The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative science quarterly
Carpenter, M., and Fredrickson, J. 2001. ‘Top Management Teams, Global Strategic Posture, and the Moderating Role of Uncertainty,’ Academy of Management Journal, 44
Das, T., & Teng, B. (2000). A resource-based theory of strategic alliances. Journal of management, 26(1)
Elyas, S., & Mohamed, Y. (2013). Brand alliance, a strategy to enter new markets and a tool for positioning. Journal of Knowledge Management, Economics and Information Technology, 3
Estrada, I., Martin-Cruz, N., & Martin-Pérez, V. (2014). To cooperate or not to cooperate? The dilemma faced by inexperienced firms in R&D consortia. In T. Das, Strategic alliances for innovation and R&D. Information Age Publishing Inc.
Feldman, M. (2000). Organizational Routines as a Source of Continuous Change. Organization Science, 11(6)
Field, A. (2013) Discovering Statistics using SPSS. 4threvised edition. Sage publications LTD.
Gayle, P. G. (2008). An Empirical Analysis of the Competitive Effects of the Delta/Continental/Northwest Code‐Share Alliance. Journal of Law and Economics
George, G., Zahra, S. , Wheatley, K., & Khan, R. (2001). The effects of alliance portfolio characteristics and absorptive capacity on performance: A study of biotechnology firms. Journal of High Technology Management Research,.
Goh, K., & Uncles, M. (2003). The benefits of airline global alliances: an empirical assessment of the perceptions of business travelers. Transportation Research Part A: Policy and Practice,
Grant, R. , & Baden‐Fuller, C. (2004). A knowledge accessing theory of strategic alliances. Journal of management studies,
Gnyawali, D. , He, J., & Madhavan, R. (2008). Co-opetition: Promises and challenges. In C. Wenkel, Encyclopedia of 21st Century Management.
Hite, J. , & Hesterly, W. (2001). The evolution of firm networks: From emergence to early growth of the firm. Strategic management journal
Hitt, M., Dacin, M., Levitas, E., Arregle, J. & Borza, A. 2000, ―Partner selection in emerging and developing market contexts: resource-based and organizational learning perspectives‖, Academy of Management Journal,
Jansen, J.J.P., Van den Bosch, F.A.J., Volberda, H.W., 2006. Exploratory innovation, exploitative innovation, and performance: Effects of organizational antecedents and environmental moderators. Management Science,
Jinju, L., Namgyoo Kenny, P., (2016). Alliance Portfolio Complexity and Order-of-Entry Learning Effects in Int'l Alliance Formations. Academy Of Management
Hoang H, Rothaermel F. 2005. The effect of general and partner-specific alliance experience on joint R&D project performance. Academy of Management Journal
Kale P and Singh H. (2009) Managing strategic alliances: what do we know now, and where do we go from here? Academy of Management Perspectives
Ku, E., & Fan, Y. (2009). Knowledge sharing and customer relationship management in the travel service alliances. Total Quality Management
Lahiri, N., & Narayanan, S. (2013). Vertical integration, innovation, and alliance portfolio size: Implications for firm performance. Strategic Management Journal
Lavie, D., Lechner, H. Singh. 2007. The performance implications of timing of entry and involvement in multipartner alliances.
Laukkanen, T., Sinkkonen, S., Kivijärvi, M., & Laukkanen, P. (2007). Innovation resistance among mature consumers. Journal of Consumer Marketing
Lazzarini, S. G. (2007). The impact of membership in competing alliance constellations: Evidence on the operational performance of global airlines. Strategic Management Journal
Li, D., Eden, L., Hitt, M. A., Ireland, R. D., & Garrett, R. P. (2012). Governance in Multilateral R&D Alliances. Organization Science,
MacKinnon, D. P. (2011). Integrating mediators and moderators in research design. Research on Social Work Practice, 21(6), 675-681.
Mascarenhas, B., & Koza, M. P. (2008). Develop and nurture an international alliance capability. International Business Review
Murray, J. Y., Kotabe, M., & Zhou, J. N. (2005). Strategic alliance-based sourcing and market performance: Evidence from foreign firms operat- ing in China. Journal of International Business Studies
Nigro, G. (2016). The effect of early or late R&D inbound alliance on innovation. Journal of Business Research, 69
Nieto, M. , & Santamaria, L. (2007). The importance of diverse collaborative networks for the novelty of product innovation. Technovation
Pan, Y. & Jackson, R. (2008). Ethnic difference in the relationship between acute inflammation and serum ferritin in US adult males. Epidemiology and infection
Peltokorpi, V., & Hasu, M. (2014). How participative safety matters more in team innovation as team size increases. Journal of Business and Psychology
Pisano, G. P., & Teece, D. J. (2007). How to capture value from innovation: Shaping intellectual property and industry architecture. California Management Review
Rogerson, P. A. (2001). Statistical methods for geography. London: Sage.
Rothaermel, F. , & Deeds, D. (2004). Exploration and exploitation alliances in biotechnology: A system of new product development.
Sampson, R. C. (2007). R&D alliances and firm performance: The impact of technological diversity and alliance organization on innovation.
Sanders, W.G., and Carpenter, M.A. 1998. Internationalization and Firm Governance: The Roles of CEO Compensation, Top Team Composition, and Board Structure.
Sarkar, M.B., Echambadi, R. and Harrison, J.S., (2001) “Alliance entrepreneurship and firm market performance”
Schumpeter JA. 1934. The Theory of Economic Development. Harvard University Press: Cambridge, MA.
Song, X. M., Thieme, R. J., & Xie, J. (1998). The impact of cross‐functional joint involvement across product development stages: an exploratory study.
Stuart, T.E., (2000) “Interorganizational Alliances and the Performance of Firms: A Study of Growth and Innovation Rates in a High-Technology Industry”
Teece, D. J., Pisano, G. P., and Shuen, A. (1997) Dynamic Capabilities and Strategic Management. Strategic Management Journal.
Teece, D. J. (2010). Business models, business strategy and innovation. Long range planning, 43(2), 172-194.
Tsai, W. (2002). Social Structure of Coopetition within a Multiunit Organization Coordination, Competition, and Intraorganizational Knowledge Sharing.
Tugores-García, A. (2012). Analysis of global airline alliances as a strategy for international network development.
Verona, G., & Ravasi, D. (2003). Unbundling dynamic capabilities: An exploratory study of continuous product innovation. Industrial and Corporate Change.
Wu, J. (2014). Cooperation with competitors and product innovation: Moderating effects of technological capability and alliances with universities. Industrial Marketing
Management
APPENDICES
Appendix 2
Table 9 - Robustness H2a
Variables Model 31t+1 Model 32t+1 Model 33t+1 Model 34t+2 Model 35t+2 Model 36t+2
TCI scope 1,081*** (,123) ,575** (,202) 1,074*** (,121) ,568** (,199) Passengers ,085 (,238) -,051 (,217) ,092 (,221) -,042 (,213) Daily departures 4,629*** (,881) ,3,148** (,979) 4,573***(,867) 3,110*** (,963) Countries active ,792 (,536) ,113 (,559) ,791 (,528) ,120 (,549) Constant -1,385*** (,358) -1,888*** (,296) -2,012*** (,402) -1,350*** (,352) -1,858*** (,292) -1,970*** (,396) Observations 60 60 60 60 60 60 Adjusted Rsquare ,580 ,563 ,628 ,585 ,567 ,632 F 28,172 77,024 25,848 28,696 78,192 26,336
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
Appendix 3
Table10 - Robustness H2b
Variables Model 37t+1 Model 38t+1 Model 39t+1 Model 40t+2 Model 41t+2 Model 42t+2
TCI scope ,032 (,031) -,038 (,054) ,032 (,031) -,038 (,054) Passengers ,110 (,057) ,118 (,058) ,110 (,057) ,118 (,058) Daily departures -,205 (,222) -,109 (,263) -,205 (,222) -,109 (,263) Countries active ,163 (,135) ,208 (,150) ,163 (,135) ,208 (,150) Constant ,597*** (,090) ,469*** (,075) ,638*** (,108) ,597*** (,090) ,469*** (,075) ,638*** (,108) Observations 60 60 60 60 60 60 Adjusted Rsquare ,055 ,001 ,046 ,055 ,001 ,046 F 2,146 1,033 1,714 2,146 1,033 1,714
Appendix 4
Table 11 - Robustness timing of entry H3a
Variables Model 43t+1 Model 44t+1 Model 45t+1 Model 46t+2 Model 47t+2 Model 48t+2
Timing of entry ,539*** (,064) ,266 (,104) ,535*** (,063) ,262 (,103) Passengers ,085 (,238) -,042 (,220) ,092 (,221) -,033 (,217) Daily departures 4,629*** (,881) 3,315*** (,986) 4,573*** (,867) 3,279*** (,970) Countries active ,792 (,536) ,136 (,573) ,791 (,528) ,145 (,564) Constant -1,385*** (,358) ,684*** (,042) -,718*** (,430) -1,350*** (,352) ,696*** (,041) -,693 (,423) Observations 60 60 60 60 60 60 Adjusted Rsquare ,580 ,542 ,618 ,585 ,545 ,622 F 28,172 70,736 24,823 28,696 71,645 25,263
*** p<0.00 ** P<0.05 *P <0.1 (standard error in parentheses)
Appendix 5
Table 12 - Robustness timing of entry H3b
Variables Model 49t+1 Model 50t+1 Model 51t+1 Model 52t+2 Model 53t+2 Model 54t+2 Timing of entry ,023 (,018) ,000 (,032) ,016 (,016) -,019 (,028) Passengers ,119 (,064) ,119 (,067) ,110 (,057) ,119 (,058) Daily departures -,203 (,253) -,202 (,299) -,205 (,222) -,109 (,261) Countries active ,128 (,154) ,129 (,174) ,163 (,135) ,211 (,152) Constant ,590*** (,103) ,533*** (,012) (,130),590*** ,597*** (,090) ,545*** (,010) ,549*** (114) Observations 60 60 60 60 60 60 Adjusted Rsquare ,037 ,012 ,020 ,055 ,001 ,046 F 1,765 1,704 1,300 2,146 1,066 1,719
