Master Thesis MSc. BA Strategic Innovation Management
Do resources explain the probability of
becoming an R&D coordinator? Investigating
the Bio-based Industries Joint Undertaking
January, 20th 2020 Mehmet Guenes
University of Groningen Faculty of Economics and Business
S3889653
Supervisor: dr. I. (Isabel) Estrada Vaquero Co-assessor: dr. K.J. (Killian) McCarthy Abstract:
R&D consortia are inter-organizational collaborations led by coordinators and supported by participants. However, scientists have limitedly investigated whether a firm's resource richness is more likely to lead to becoming an R&D coordinator. A binary logistic regression analysis is conducted on a sample of 346 participants and 29 coordinators belonging to profit organizations that started to operate in 2014 in the European Union consortia Bio-based Industries Joint Undertaking. The empirical findings show that resource-richness increases the likelihood of becoming an R&D coordinator and that this relationship is enhanced by the presence of prior accumulated coordinator experience. These results indicate that resource-abundant firms are more likely to attract participants, receive positive evaluations from policymakers and can allocate more resources for the coordination.
Keywords: R&D consortium, coordinator, participant, resources, coordinator experience, leader-member-exchange theory, collaboration, network
2
Table of Content
1. Introduction ... 3
2. Literature Review ... 5
2.1 Leader-Member-Exchange Theory ... 5
2.2 From Leader-Member-Exchange Theory towards Alliance and Network Theory ... 6
2.3 The objective and procedure of an R&D consortium ... 8
3. Hypotheses Building ... 10 3.1 Resource-richness ... 11 3.2 Coordinator Experience ... 13 4. Methodology ... 15 4.1 Industry Selection ... 15 4.2 Data collection ... 16 4.3 Measurements of variables ... 18 4.4 Dependent variable ... 20 4.5 Independent variable ... 20 4.6 Moderating variable ... 21 4.7 Control Variables ... 22 4.8 Method of Analysis ... 24 5. Results ... 26
5.1 Descriptive statistics and correlations ... 26
5.2 Binary logistic regression ... 32
5.3 Robustness checks ... 33
6. Discussion & Conclusion ... 34
6.1 Theoretical implications ... 34
6.2 Managerial implications ... 36
6.3 Limitations and future research... 37
6.4 Conclusion ... 38
7. References ... 40
3
1. Introduction
Economic processes are constantly changing over time and require ever newer innovations, either to keep up with the foreign competition or to meet the changing demands of consumers. However, research to transform economic processes can be very costly for companies, which is why they do not make the necessary investments. Therefore, policymakers can subsidize companies within an R&D consortium to give them financial incentives to stimulate their R&D spendings (Clausen, 2009; Corey, 1996; Irwin & Klenow, 1996). Companies apply together as an independent consortium for a tendered project and present their solution approaches. A consortium consists of a coordinator and participants who jointly research the project. Firms wishing to become a coordinator must first succeed in attracting sufficient and qualified participants for the project. Moreover, potential coordinators must not only convince possible partners to join its consortium but also outpace other potential coordinators from persuading their intended partners. Afterward, a selection committee must evaluate the application, as they try to select companies as coordinators that have the necessary capacity to make the project a success. Based on these processes, a firm being able to become a coordinator has gone through these stages (Borch & Solesvik, 2016; Mothe, 1999; Schiavoni & Simoni, 2011; Schiavoni & Simoni, 2016).
4
Schiavone & Simoni, 2016). Moreover, policymakers were able in previous consortia to assess the skills and capabilities of the reapplying firm (Sakakibara, 2000; Schiavone & Simoni, 2016; Schiavone & Simoni, 2011). However, scholars have limitedly investigated whether resource-rich firms are more likely to be selected as coordinators and if coordinator experience affects this relationship. Furthermore, management theoretical frameworks lack explanatory power to determine the differences between a coordinator and participants.
Alliance and network theory argue that firms enter collaborative agreements to access complementary resources, market opportunities and increase their competitive position (Gulati, Norhia & Zaheer, 2000; Dhanaraj & Parkhe, 2006; Olk & Young, 1997). However, these theoretical frameworks fail to identify what characteristics determine a leader and a member. One theoretical approach to determine what characteristics coordinators and participants possess is the Leader-Member-Exchange (LMX) theory. This theory, which stems originally from psychology literature, highlights the formal and informal exchanges of resources between a leader and a member but mainly focuses on the exchange itself rather than mainly classifying role characteristics (Dansereau, Graen & Haga, 1975; Graen & Scandura, 1987; Graen & Uhl-Bien, 1995; Wilson, Sin & Conlon, 2010). Derived from the exchanges between both parties and the qualities they possess, the LMX theory will be combined in a novel approach with the alliance and network theory from management literature. This unique composition of different literature streams will aid in explaining the role of resources for a firm to become an R&D coordinator. Based on the identified literature gap, this paper aims to answer the research question as follows:
Research Question: Does resource-richness increase the likelihood of being a coordinator and how is coordinator experience affecting this relationship?
resource-5
richness increases the likelihood of being a coordinator and that this likelihood is increased if a firm possesses prior coordinator experience. These results contribute to the literature by indicating that resources facilitate a competitive advantage to prevail (Barajas & Huergo, 2007; Barney, 1991; Barney, 2000; Gerringer, 1991; Shah & Swaminathan, 2008; Teece, 1986) and that experience has a positive moderating effect through prior accumulated knowledge (Quelin, 2000; Sakakibara, 2000; Schiavone & Simoni, 2016; Schiavone & Simoni, 2016). To be more precise, this study indicates that resources have a role in partner attraction and are well regarded by institutional decision-makers. The addition of coordinator experience has advantages as it simplifies the processes of becoming a coordinator.
The structure of this research paper consists of the following sections: First, the characteristics of coordinators and participants will be discussed and then the setting of R&D consortia explained. Afterward, the hypotheses will be developed through elucidating the hypotheses on how resource-richness increases the likelihood of being a coordinator and how coordinator experience is affecting this relationship. Subsequently, the methodology sector will highlight the measurement methods and is then followed by the presentation of the results. Lastly, theoretical and managerial implications will be discussed and limitations and future research directions elaborated.
2. Literature Review 2.1 Leader-Member-Exchange Theory
6
multiple members as well in which the leader is closer to high-quality members than low-quality members, with whom it shares more resources since qualitative members are more useful (Graen & Uhl-Bien, 1995). Nevertheless, both the leader and the members relationship are interdependent because they rely on each others capabilities to increase their own performance (Bauer & Green, 1996). In sum, a good leader possesses the necessary resources that are considered desirable by members and helps his members reach their full potential (Graen & Uhl-Bien, 1995; Sparrowe & Liden, 1997; Wilson, Sin & Conlon, 2010).
2.2 From Leader-Member-Exchange Theory towards Alliance and Network Theory
7
Liden, Sparrowe & Wayne, 1997). However, experienced leaders also seek members who complement their work since they do not prefer to have a large group of members who duplicate their work (Phillips & Bedeian, 1994). This behavior is observed in alliance theory as well in which firms prefer to collaborate with moderate diverse partners since they provide the most novel insights into technologies and capabilities compared to too distant or too similar partners which hamper the recombinative potential (Noseleit & de Faria, 2013; Sampson, 2007; Subramanian & Soh, 2017).
8
The stability of the LMX relationship is closely linked to the perceived performance of the members and the reliability of the leader (Gerstner & Day, 1997). If they do not perceive that their efforts get positively influenced by their leaders, the quality of the exchange is decreasing due to lower satisfaction with the results (Fisk & Friesen, 2012; Volmer, Niessen, Spurk, Linz & Abele, 2011). The dependence on satisfaction with the overall performance is present in firm network theory as well such as that members are satisfied if their innovation efforts can be successfully appropriated (Dhanaraj & Parkhe, 2006). Olk & Young (1997) argue that firms tend to leave consortia if the network does not provide the necessary benefits that participants are aiming for such as a decrease in financial performance. Based on the low performance, they would rather invest their funds into alternative projects.
After considering the theory and combining it with management literature, it stands out that the following aspects are essential for an R&D leader: First, an R&D leader must possess the necessary resources that others need in order to collaborate. Second, the leader is the center of the group and is able to manage all relationships even though that some are distant. Third, a leader must be able to deliver consistent results over the whole period, which are in the interest of the participants.
2.3 The objective and procedure of an R&D consortium
9
the welfare of consumers, make domestic firms more competitive against foreign competitors and create valuable spill-overs to related firms that benefit from the output as well (Corey, 1996; Irwin & Klenow, 1996). The goal of R&D consortia is to develop key technologies, accelerate the creation of a novel market and lead to a greater economic impact rather than solely benefitting for the producing companies. Furthermore, it also serves as a collaborative network in which firms can reduce their costs, exploit economies of scale and diversify the risk through shared investments (Mothe, 1999).
10
Table 1: Consortium formation process
3. Hypotheses Building
Table 2 shows the conceptual model for which arguments will be introduced in the upcoming section.
11
3.1 Resource-richness
To receive the necessary funding from policymakers, firms desiring to be coordinators aim for the most competitive consortium against other potential consortia coordinators to be approved by policymakers. Therefore, a coordinator must be able to attract participants who are willing to join its planned consortium. However, the participants are in a dilemma because they cannot assess which consortia offers the most promising project with its leading coordinator, which is why they need to be convinced to join a certain consortium (Schiavone & Simoni, 2011; Schiavone, Simoni, 2016). Borch & Solesvik (2016) emphasize that R&D consortia require valuable partner attractiveness for firms to join a consortium such as the necessary resources and skills. Therefore, the first stage in becoming a coordinator requires partner attractiveness for participants to join a potential coordinators consortium. Partner attractiveness is defined by Shah & Swaminathan (2008) as follows: "Partner attractiveness is defined as the degree to which the initiating firm in a particular alliance project sees a partner as desirable, favorable, appealing, and valuable. As the attractiveness of a given partner increases, its likelihood of selection also increases” (p. 473). Based on this requirement to become an R&D leader, the coordinators must prove the existence of their superior capabilities to attract participants. Thus, they need to signal their advantages compared to other potential leaders and prove their innovative capabilities to win potential participants over to join them (Barajas & Huergo, 2007; Sakakibara, 2000).
12
2002). Market access is another crucial motivation for firms to enter an inter-organizational collaboration. This is especially helpful for companies with innovation expenses because a potential leader might already possess valuable knowledge of the interested participants targeted customer segments to aid in appropriating the innovation efforts (Mu & Lee, 2005; Wassmer, 2010). However, entering new markets may be restricted by limited resources that may be held by a potential coordinator that has secured these rare resources (Bierly & Gallagher, 2007; Koza & Lewin, 2000; Teece, 1986). Due to these entry restrictions, a potential coordinator can secure participants by providing them a competitive advantage through opening up a new market segment that otherwise could not have been developed by participants on their own (Barney, 1991; Park, Mezias & Song, 2004; Wassmer, 2010). After a coordinator has secured a small percentage of participants, the attraction of other participants is reinforcing itself as now more resources and capabilities are available (Aversa, Hervas-Drane & Evenou, 2019). Based on this, the coordinators require a large number of resources to attract participants to join their consortium to convince them of their capabilities and potential support for conducting R&D (Barajas & Huergo, 2007 Borch & Solesvik, 2016; Schiavone & Simoni, 2011; Schiavone, Simoni, 2016). Resource-Rich firms' organizational resources reflect the internalized capabilities within an organization such as physical assets, human capital and knowledge available (Barney, 2000). Based on these arguments, a resource-abundant firm is more likely to attract participants in the first stage of becoming a coordinator.
13
For policymakers, the amount of resources is one way to measure the likelihood of success for a project. The finance structure of a consortium depends partly on public money and partly on firms' investments since policymakers want to stimulate R&D spendings of the firms (Clausen, 2009; Irwin & Klenow, 1996). Resource-rich firms have the advantage of being able to spend a larger proportion of their resources in R&D projects compared to SMEs that are resource-poor (Lee & Cin, 2010). This, in turn, is the desired effect of policymakers to stimulate private R&D spendings which can be achieved more likely if it approves larger firms (Gonzalez & Pazo, 2008). Moreover, these spendings create greater opportunities for experimenting and are therefore associated with a more successful outcome of the consortium objective (Bizan, 2003). Furthermore, a potential coordinator with a large number of resources has secured relatively large participants as well and thus asserts itself with the consortium against other consortia in the selection stage. Another positive evaluation are the coordination capabilities in an inter-organizational collaboration because interacting with multiple partners and institutional authorities is resource consuming (Mothe & Quelin, 2001). Thus, a firm with a large number of resources is more likely to deal with governmental and bureaucratic barriers by allocating sufficient resources for these processes that emerge in the attraction phase and lasts till the end of the consortium. Therefore, resource-constrained firms are less likely to deal with these problems and actually submit a proposal (Barajas & Huergo, 2007; Schiavone & Simoni, 2016).
Collectively, resource-rich firms are more likely to be coordinators because they can attract partners to join their consortium and receive positive evaluations by policymakers:
H1: Resource-rich firms are more likely to become R&D consortia coordinators than resource-constrained firms.
3.2 Coordinator Experience
14
might be detrimental for the selection process (Chang, Chen & Lai, 2008). These experiences are internalized in the firm in the form of organizational routines and structures which help to facilitate the goal to become a coordinator (Hoang & Rothaermel, 2005). Another positive effect of coordinator experience is that an R&D consortium is one kind of accumulated international experience because it involves collaborative work with multiple nationalities (Nielsen, 2007). This international experience, in turn, not only helps in the search and attraction for partners but also increases the reach for additional partners beyond national borders (Mowery, 1998; Rothaermel & Deeds, 2006). Experience as a coordinator also has the advantage that a potential coordinator has established relationships with other participants as well, which eases the process of conviction (Sakakibara, 2002; Schiavone & Simoni, 2016). Furthermore, participants who are willing to join a consortium can access information about past performance and thus reduce their dilemma of finding the right leader. This, in turn, increases the attractiveness to join if interested participants can appropriate their innovation efforts (Olk & Young, 1997).
15
Through earlier information and statements from former partners, the selection committee can ascertain that the risk of cheating in a reapplying company is lower since it did not happen in the past (Sakakibara, 2000).
Taken together, coordinator experiences enable a potential coordinator to be even more effective in attracting companies, to be better evaluated by politicians and to simplify coordination through previous experience. Collectively, these arguments lead to the second hypothesis as follows:
H2: R&D coordinator experience positively moderates the likelihood for resource-rich firms to become coordinators
4. Methodology 4.1 Industry Selection
This research paper examines the R&D consortia of the European Union in the H2020 project, which has been implemented in 2014. Under the H2020 umbrella, the “Bio-based Industries Joint Undertaking” (BBIJU) public-private partnership dealt specifically with the challenges of climate change, ending the dependency on fossil products, and the creation of organic products. All these measures serve to create a sustainable bioeconomy within the European Union and its partners (European Union, 2014). The setting in the BBIJU is that potential coordinators assemble a group of potential participants and work with each other towards a solution to a problem. Afterward, they submit a proposal to a board of experts who need to evaluate the proposal and its likelihood of success. If it is being accepted by the board of experts, the coordinator and the participants must sign an agreement to adhere to the rules given by the BBIJU that ensures a certain law regulatory within the consortium. Afterward, the coordinator and its participants start their collaborative research efforts as an independent consortium under the umbrella of the BBIJU (BBI, 2019).
16
participants to join its consortium. Secondly, the ultimate decision whether the proposal is being approved or not is made by the board of experts that allocates funds to the projects. Based on this, a coordinator has proven its capabilities by convincing participants and the board of experts.
In sum, the BBIJU provides an adequate setting for this study because the sample consists of the following characteristics: Firstly, the sample has a diverse industrial spectrum with firms from the agricultural, chemical, manufacturing and consumer products industries engaging in Bio-innovations. Secondly, the sample ranges from small to very large firms which further enriches the diversity. Collectively, these two conditions provide a satisfactory justification for the statistical procedure.
4.2 Data collection
The data sources were uniquely merged from several existing databases, which will be further elaborated on in the later stage. The BBIJU consortia set was collected from The Community Research and Development Information Service (CORDIS). Secondary sources were collected on Orbis, COMPUSTAT, Zephyr and The World Bank for the firm- and country-specific data. Lastly, consortia experience data were collected on ResearchRanking.
17
non-profit organizations. However, within the BBIJU, non-profit organizations were also assigned by policymakers as coordinators and/or participants in the consortia. Nevertheless, non-profit organizations were eliminated because analyzing non-profit organizations require different assumptions than market competition. The data of the CORDIS database included the company name, project acronym, project description, project budget, requested budget, location, and the role of the company within a consortium (either coordinator in a project or participant in a project).
Secondly, secondary sources such as COMPUSTAT, Orbis and The World Bank (The World Bank, 2019) were used to collect firm- and country-specific data (e.g. firm size, country, firm age, operating industry, GDP of country, etc.). Since the data sample consists of small and medium-sized enterprises as well, additional data for each company was obtained individually from the company websites to collect more data. However, these data were not in line with the focal period of 2013 which is why for some firms the last available date of 2019 had to be chosen which was displayed on their websites.
Thirdly, every firm was checked for its R&D consortia experience under the European Union framework by collecting data from EuropeanResearchRanking. The EuropeanResearchRanking is a non-profit public database aiming for creating a transparent insight into the research funding of the European Commission which collects data on R&D consortia projects implemented by the European Union (ResearchRanking, 2019). From this source, I have collected data on R&D consortia experience for each firm individually ranging from the year 2000 to 2013, since 2014 was the beginning of the BBIJU under the H2020 project and replaced previous subsidy programs.
18
After following these steps, 375 companies remained in the sample of which 346 were listed as participants and 29 as coordinators operating from 2014 to 2019.
4.3 Measurements of variables
19
Table 3: Summary of measurements
Variables Definition Mean Std. Dev.
Roles
The roles that firms take in the consortia. Coordinators received a "1" and participants received
a "0".
0,0775401 0,2678048
Firm Age
The start year of the BBIJU in 2014 was subtracted by the establishment year of a firm. A logarithm was
undertaken
3,041311 1,009391
Industry Classification
A dummy variable that was calculated on NACE Rev. 2-digit code in which firms were classified as high or
low technology sectors
0,3502674 0,4776927
GDP Nominal GDP of each country of origin of a company
from the year 2013. A logartihm was undertaken 27,61007 1,290895
Project Budget Project Budget for each consortia that was assigned
by policy makers. A logarithm was undertaken 15,72363 0,8807216
MNC A binary variable in which MNC firms were coded as
"1" 0,2700535 0,4445819
EU Contribution per national share
Share is calculated as the EU member contributions on the Gross National Income of each country of
origin of a firm. Non-European Union firms were assigned as "0"
8,135749 6,877477
Experience as
Participant Sum of all previous participations as participant 3,3171123 18,41225 Multiple
Participations
A binary variable in which firms that participated in
other consortia simultaneously were coded as "1". 0,155496 0,3628636 Firm Size Number of employees. A logarithm was undertaken 4,779876 2,717586 Recent Coordinator
Experience
Sum of all coordinator experience in the last three
20
4.4 Dependent variable
Since CORDIS provided a list with consortia operating under the BBIJU, the
coordinator and participant roles were already defined by European officials under the "Role" section. This section for every firm included the character string "Participant" or
"Coordinator". A "Coordinator" character string was coded to a "1" which reflects, in
probability measures, a 100% allocation as taking the role as a coordinator. In contrast to the coordinators, the "Participant" strings were coded as "0", as they do not take part in the leading role. Based on this coding approach, a setting was created which defined a Yes or No scenario in terms of the question of whether a firm is a coordinator. This Yes or No scenario is defined in mathematics as a binary or dichotomous variable, in which firms can only take one form: In this setting, it is either the role as a coordinator (1) or the role as a participant (0). Resulting from this approach, the dependent variable Coordinator was created because only firms with an assigned "1" hold the characteristic of being a coordinator while all others remain "0" (Lemmeshow & Hosmer, 1982). This binary approach was done similarly by Barajas & Huergo (2007) in assessing whether a consortium proposal was accepted or rejected by policymakers (0 as rejected and 1 as accepted). However, some firms in the BBIJU sample take up both roles as coordinators and participants. This, in turn, is skewing the distribution towards being a participant. For example, a firm with a large number of resources is taking once a coordinator role and three times a participant role. Therefore, if a coordinator has participant roles in other consortia as well, the additional participant roles were eliminated because a firm has proven its superiority in becoming a coordinator, which is the aim of this study to test if the resources-richness explains this. Nevertheless, this process will be thoroughly inspected later in the control section.
4.5 Independent variable
21
to smaller firms that need to access external linkages while knowledge is already embedded in the human diversity within a large firm (Huggins & Johnston, 2010). Furthermore, the number of employees within a firm is related to total R&D expenditure and increases the efficiency for collaborations such as in an R&D consortium (Fritsch & Meschede, 2001). Duysters, Heimeriks, Lokshin, Meijer & Sabidussi (2012) highlight that firm size is an indicator that larger firms are more likely to find a partner and manage an alliance due to the availability of sufficient resources. This is highly important because it justifies that a potential coordinator can find suitable partners and can manage an R&D consortium. In addition, firm size is an indicator of how much bargaining power a company has in R&D consortia, as larger companies have more room to maneuver due to the number of resources they have. This has a positive effect on the fact that a larger company can create more attractiveness and thus create dependencies for smaller companies (Gardet & Mothe, 2012). Measuring the firm size is thus an effective approach to test the number of overall resources within a firm's boundary. Since the built-up H1 argues that resource-rich firms are more likely to be selected as coordinators, firm size aims to investigate the relationship between the likelihood of being selected as a coordinator in terms of resource-richness in contrast to participants. To control for the skewness of the distribution, a logarithm was undertaken because the number of employees within the sample ranges from very small firms with less than 10 employees to firms with more than 10.000 employees.
4.6 Moderating variable
22
reside within the employees as tacit knowledge. Thus, recent alliance experience for up to three years is linked to managers still being employed within the firm. Based on this, recent experience as a coordinator reflects the current leading experience on which a firm has built up recently a certain degree of managing skills and reputation.
4.7 Control Variables
Firm Age was chosen as the first control variable measured as the year 2014 minus the official registration date of the company. It is important to control because it reflects if younger companies are more likely to be chosen as coordinators or more mature ones. Younger firms are more likely to focus on riskier R&D to enter successfully a market while older companies rather focus on process innovations to refine their existing competencies (Coad, Segarra & Teruel, 2016). Firm age also reflects the performance of a company and the experience accumulated throughout the existence of a firm (Galbreath & Galvin, 2008). Furthermore, firm age controls for the reputational effect because older companies enjoy a higher degree of credibility and reputation than younger ones. Moreover, reputation is one important resource that a firm possesses since potential partners may prefer to access it to increase its reputational resource (Flanagan & O'shaughnessy, 2005). To control for skewness, a logarithm was undertaken to receive a normal distribution because Firm Age ranges from very young companies with less than five years to very mature companies with a history of more than 100 years.
knowledge-23
intensive sector, the observation was coded as a "1" while medium- and low-technology industries were coded as a "0" to create a dummy variable (Eurostat, 2008).
GDP of a country, measured as the nominal GDP value, was adopted in Euros (€) from The World Bank (2020) for each firm origin country in 2013 that is taking part in the BBIJU. This control variable checks for institutional support and the number of resources available within a country. Furthermore, it reflects the market size within a country and firms enjoy higher revenues due to a larger purchasing power (Jadhav, 2012). Thus, this measure controls for coordinators located in high GDP countries who may enjoy larger benefits due to the strong economic performance and support. To receive normal distributions and reduce skewness, a logarithm was undertaken to account for the great differences among the countries since Germany's nominal GDP value exceeds massively the nominal GDP value of Cyprus for example.
Project Budget, measured as the allocated financial subsidy to an independent consortium, is the fourth control variable that checks the overall financial power and size within the group. This financial grant was adopted from CORDIS itself as a project-specific control variable. However, a logarithm was undertaken since the budgets varied greatly in their size.
MNC is the fifth control variable that checks for the degree of internationalization of a company. A company was identified as an MNC if it possessed a wholly-owned subsidiary in at least one foreign country (Mayrhofer & Prange, 2015). The information was taken from Zephyr and after a firm was identified as an MNC, it was coded as a binary variable. MNCs were assigned as "1" and non-MNCs as "0". Checking the degree of internationalization is important because companies with subsidiaries located elsewhere than in the domestic country enjoy access to a large number of heterogeneous resources that are not available within the domestic country. Furthermore, these companies benefit from foreign market access and gain competences from experiences from abroad as well which are transferred to the overall MNC (de Faria & Sofka, 2010; Sofka, Shehu & de Faria, 2014).
24
based on the Gross National Income of each country. This is important to measure because countries such as Germany and France contribute a larger share (and in absolute terms) of their public available resources into financing the European Union which in turn finances the European consortia BBIJU (Cipriani, 2014; European Parliament, 2017; European Union, 2014). Thus, they might expect a higher reciprocal benefit by their national companies being chosen as coordinators. Based on this, a national share contribution per country was added to control for this effect. However, some firms in the sample are not part of the European Union such as the United States and Turkey, wherefore, their country of origin firms received a “0” since they do not contribute financially to the maintenance of the European Union and do not delegate politicians to the European Parliament.
Experience as Participant is the seventh control variable measured as the total sum of experience as participant accumulated since the existence of the European Union ranging from 2000 up until the year 2013, one year before the start of the BBIJU. This allows us to determine whether coordinators have previous experience as participants and are therefore more familiar with consortium procedures such as bureaucratic tasks, collaborating with several partners and at the same time accumulated somewhat of coordination skills through observations in the participant role (Schiavone & Simoni, 2011; Schiavone & Simoni, 2016). Nevertheless, some firms do not possess any experience as participants, wherefore they received a "0" since it is their first time participating in a European consortium.
Multiple participation within the BBIJU is the eighth control variable that determines whether coordinators can be and or manage other consortia simultaneously. In particular, it controls if coordinators have distinctive capabilities to coordinate their consortia and simultaneously coordinate and/or participate in others as well (Hazir, 2013). It is measured as a binary variable in which I have assigned a "1" for firms participating more than once, either as coordinator or participants, and a "0" for firms that are only present once. This effect also controls for the measures that were taken for the operationalization of the independent variable.
4.8 Method of Analysis
25
two opposite groups, coordinators and participants, need to be opposed to each other. A binary logistic regression is suitable for this analysis since the assumption in the model is that the outcome must be binary, either "1" as being a coordinator or "0" as not being a coordinator. The standard binary logistic regression takes the form as follows (Hosmer, Lemeshow & Sturdivant, 2013):
E(Y|x) = β0 + β1 x
With a binary outcome, the conditional mean (being a coordinator), must be greater than or equal to zero and less than or equal to one. However, since a moderator is included, the binary logistic regression takes the form as follows (Hosmer, Lemeshow & Sturdivant, 2013):
(E(Y|x) = β0 + β1 x1 + β2 x2 + β3 x1x2
Based on this mathematical equation, the model for the main analysis will take the form as follows:
E(Y|x) = β0 + Control Variables + β1 Firm Size + β2 Recent Coordinator Experience
+ β3 Firm Size * Recent Coordinator Experience
The model will be introduced firstly with control variables that will check for other explanations and is then followed by the introduction of the independent variable firm size. Subsequently, the moderate variable will follow and afterward interact with the independent variable.
26
OR = eβi
Descriptive statistics will highlight the mean, standard deviations, and correlations among the variables to ensure that the independent variables remain independent from the influence of other variables that may bias the analytical model. Furthermore, a variance inflation factor will highlight if multicollinearity exists among the variables.
Moreover, a t-test will be conducted to show if a significant difference exists between the coordinators and the participants regarding their firm size and recently accumulated experience as coordinators. A t-test is determining whether the means from two groups differ significantly from each other. This approach is important to undertake because if the Null-Hypothesis can be rejected, the statistical difference among the coordinators and the participants is given.
5. Results
In the following section, the statistical results will be presented. The first step includes descriptive statistics, t-tests and the correlations among the variables that are going to be commented on. Furthermore, a variance inflation factor analysis will be carried out to detect potential multicollinearity. The second step includes the binary logistic regression in which the hypotheses are tested. Subsequently, robustness checks will follow to determine the reliability and validity of the results.
5.1 Descriptive statistics and correlations
27
Table 4: Descriptive statistics
Variables Obs Mean Std. Dev. Min Max
Roles 374 0,0775401 0,2678048 0 1 Firm Age 374 3,041311 1,009391 0 5,484797 Industry Classification 374 0,3502674 0,4776927 0 1 GDP 374 27,61007 1,290895 21,71093 30,4515 Project Budget 374 15,72363 0,8807216 12,51379 17,66427 MNC 374 0,2700535 0,4445819 0 1
EU Contribution per national
share 374 8,135749 6,877477 0 21,3
Experience as Participant 374 3,3171123 18,41225 0 326
Multiple Participations 374 0,155496 0,3628636 0 1
Firm Size 374 4,779876 2,717586 0 12,19877
In Table 5, the descriptive statistics highlight the differences between the two opposite groups. Regarding the means of Firm Size and the moderating effect of Recent Coordinator Experience, it is observable by group mean comparison that the coordinator group consists of firms with a greater Firm Size and Recent Coordinator Experience. To see if these variables differ statistically compared to the participant group, a t-test by mean comparison will be conducted to highlight whether this observation is also significant.
28
29
Table 5: Descriptive statistics for participants and coordinators
Variables Participants
Group (0) Obs Mean Std. Dev. Min Max
Firm Age 345 3,024448 0,996638 0 5,484797 Industry Classification 345 0,342029 0,4750783 0 1 GDP 345 27,61731 1,28598 21,71093 30,4515 Project Budget 345 15,70638 0,852317 13,52696 17,66427 MNC 345 0,2521739 0,4348913 0 1 EU Contribution per national share 345 8,23687 6,943214 0 21,3 Experience as Participant 345 1,994203 6,371263 0 69 Multiple Participations 345 0,119186 0,3244793 0 1 Firm Size 345 4,632854 2,650878 0 11,95118 Recent Coordinator Experience 345 0,0637681 0,495905 0 8 Variables Coordinator
Group (1) Obs Mean Std. Dev. Min Max
30
Table 6: Pearson’s correlation matrix
Table 7: Variance Inflation Factors
Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (1) Role 1 (2) Firm Age 0,0574 1 (3) Industry Classification 0,059 -0,3676 1 (4) GDP -0,0198 0,0812 -0,03 1 (5) Project Budget 0,0684 0,0273 -0,1585 0,0765 1 (6) MNC 0,1407 0,42 0,2424 0,0571 0,0838 1
(7) EU Contribution per national
share -0,0518 0,0259 0,0264 0,7157 0,1045 -0,0067 1 (8) Experience as Participant 0,2207 0,1569 0,0541 0,0087 0,0535 0,147 -0,012 1 (9) Multiple Participations 0,3451 0,0536 0,0873 -0,0803 -0,0929 0,1745 -0,01204 0,1979 1 (10) Firm Size 0,1869 0,4945 -0,3259 0,1188 0,1182 0,6881 0,0047 0,2528 0,1724 1
(11) Recent Coordinator Experience 0,1673 0,0816 0,1109 0,0691 -0,0034 0,042 0,0266 0,5454 0,1548 0,1135 1
Variables VIF 1/VIF
Firm Age 1,46 0,685
Industry Classification 1,286 0,777
GDP 2,125 0,471
Project Budget 1,063 0,941
MNC 1,957 0,511
EU Contribution per national share 2,111 0,474
Experience as Participant 1,542 0,648
Multiple Participations 1,118 0,894
Firm Size 2,314 0,432
31
Table 8: Binary logistic regression
Model 1 Model 2 Model 3 Model 4
Control Variables Firm Age 1,076 0,984 0,976 0,975 (0,264) (0,242) (0,242) (0,244) Industry Classification 2,152 2,538* 2,478* 2,695* (1,077) (1,309) (1,287) (1,41) GDP 1,026 0,946 0,93 0,91 (0,227) (0,215) (0,214) (0,211) Project Budget 1,507* 1,509* 1,502* 1,46* (0,355) (0,36) (0,359) (0,362) MNC 1,869 0,858 0,868 0,835 (0,957) (0,579) (0,589) (0,599) EU Contribution per national share 0,985 0,994 0,994 0,993
(0,044) (0,044) (0,045) (0,046) Experience as Participant 1,014 1,006 1,002 0,956 (0,016) (0,012) (0,012) (0,029) Multiple Participations 8,341*** 8,254*** 8,073*** 8,044*** (3,797) (3,81) (3,727) (3,825) Independent Variable Firm Size 1,239* 1,253* 1,246* (0,148) (0,15) (0,154) Moderator Variable
Recent Coordinator Experience 1,253 0,282
(0,36) (0,274)
Interaction Effect Firm Size X Recent Coordinator
Experience 1,47* (0,323) Number of observations 374 374 374 374 LR chi2 40,85 43,98 44,46 48,15 Prob > chi2 0,000 0,000 0,000 0,000 Pseudo R2 0,2003 0,2156 0,218 0,2360 Log likelihood -81,571 -80,007 -79,766 -77,921
32
5.2 Binary logistic regression
The results of the binary logistic regression are presented in Table 8. In a first step, only control variables are included in Model 1, following with both the control and the independent variables in Model 2, before adding the moderator variable in Model 3. Lastly, the interaction term was included in Model 4. The overall likelihood-ratio chi-square increased steadily from 40.85 to 48.15, which indicates that the added variables increase the overall explanatory power of the model.
Hypothesis 1 argues that resource-rich firms are more likely to be coordinators, which is tested in Model 2 by adding the independent variable Firm Size. The odds ratio highlights a weakly positive significant relationship (OR=1.239, p < 0.1), meaning that Firm Size affects the odds of being a coordinator. Therefore, H1 can be supported.
Hypothesis 2 suggests that Recent Coordinator Experience positively moderates the relationship between resource-richness and the likelihood of being a coordinator. To verify this suggestion, the moderator variable was included in Model 3 and interacted with the independent variable Firm Size in Model 4. The odds ratio for the interaction term is weakly positive significant (OR=1.47, p < 0.1). This finding highlights that Recent Coordinator Experience has a positive moderating effect on the odds of being a coordinator. Therefore, H2 can be supported.
33
5.3 Robustness checks
As for the robustness check, one potential bias needs to be checked with an alternative test that is given due to the constellation of the two groups: Since the amount of coordinators is 29 and the remaining 345 observations are participants, the logistic model might suffer from small-sample bias due to the large difference in the in-group observations. These low amounts of 30 coordinator observations are called "rare events" in contrast to the 346 participants. Based on this, the maximum likelihood estimate in logistic regressions might be biased due to rare events. To check the robustness of the analysis, a penalized likelihood method is required, which corrects the problem of rare events in logistic regressions. Compared to a logistic regression, in which the maximum likelihood estimate tends to infinity while calculating, the penalized likelihood method converges the maximum likelihood estimates to finite calculations (Leitgoeb, 2013). However, one difference remains compared to the main analysis since the output in the penalized likelihood method is not estimated as odds ratios, which is why the listed results must be interpreted as coefficients. In Appendix B, the robustness check with a penalized likelihood method is attached that controls for the rare events of coordinators. The listed results in the table support the robustness of H1 since Firm Size remains significant and further throughout all models (p < 0.1) at the same significance level. Contrary to the main analysis, the moderating effect of Recent Coordinator Experience does not yield significant results, wherefore the robustness of the moderator needs to be questioned.
34
Surprisingly, an inverted U-shape relationship can be found, because Firm Size as the regular term is positive significant at the beginning and the squared term of Firm Size is negative significant at the end (Haans, Pieters & He, 2016). This observation indicates that Firm Size has a positive significant effect on the odds of being a coordinator up until a tipping point has been reached and afterward has a negative significant effect on the odds of being a coordinator. Despite the difference between the main analysis and the robustness check, we can still assure certain credibility for the positive effect of Firm Size on the likelihood of being a coordinator. Regarding the moderating effect of recent coordinator experience on Firm Size, it is evident that the interaction term increases the odds of being a coordinator. However, it is also noticeable that Recent Coordinator Experience also influences the likelihood of being a coordinator as an independent effect. Nevertheless, it is still assuredly that recent coordinator experience moderates the relationship between firm size and the likelihood of being a coordinator.
6. Discussion & Conclusion 6.1 Theoretical implications
35
previous consortia, regardless of the inter-organizational objective. However, these results are weakly supported and only have a small effect, which is why they require further inspection.
First of all, coordinators must be able to win over potential participants to join its consortium in an attraction process by highlighting the existence of their superior capabilities and resource complementarity (Barajas & Huergo, 2007; Borch & Solesvik, 2016; Schiavone & Simoni, 2011; Schiavone, Simoni, 2016). The study findings are in agreement with Gerringer (1991) and Katila, Rosenberger & Eisenhardt (2008), who identified resources as a crucial element for firms to attract partners because it highlights their internalized capabilities in forms of physical assets, human capital and knowledge residing within the company (Barney, 2000). Furthermore, the coordinators' resource-richness indicates a broader range of possible complementarity among the smaller participants' heterogeneous resource bases, which, if combined, can create highly desired synergistic effects (Nielsen, 2002; Subramanian & Soh, 2017; Wassmer, 2010). Therefore, coordinators are more likely to be resource-rich because it attracts the highly needed participants to form a consortium and submit a proposal. This finding contributes to the literature as resources being a signaling method of capabilities that reside within the coordinator's firm boundaries.
36
The management of an inter-organizational collaboration require coordinative capabilities which are more likely to be found in resource-rich firms (Barajas & Huergo, 2007: Quélin, 2000). Since coordinators are centered within consortia and interact with all participants and governmental institutions simultaneously, they face greater coordination costs than participants do (Barajas & Huergo, 2007; Schiavone & Simoni, 2011; Schiavone, Simoni, 2016). The study's findings complement these insights from literature by highlighting that coordinators are more likely to be resource-rich firms because they can allocate more of their internal resources to manage several inter-organizational relationships at the same time. Another result of this research complements this view because coordinators are more likely to manage several participations simultaneously in different consortia. This result is also in line with Barajas & Huergo (2007) who identified that larger firms are more likely to possess the necessary capabilities, such as human and financial resources, to effectively tackle coordination burden.
Since in this study coordinators have been observed retrospectively and their roles have already been clearly defined, it becomes clear that a coordinator has gone through three different phases and that resource-richness plays a role in mastering them successfully. The availability of resources facilitates the search for partners, is positively evaluated by politicians and offers the necessary coordination capacities in an R&D consortium.
Further insight in this research paper is that the effectiveness of resources is strengthened by the presence of recent coordinator experience. This effect takes place because firms are more experienced in attracting reliable partners in the first stage, policymakers have been able to get an idea of the leadership role in the past in the second stage, and experienced firms have already established routines to deal with the coordination burdens.
6.2 Managerial implications
37
capabilities. In addition to that, resources are an advantage when dealing with coordination tasks. Furthermore, if a firm has been a coordinator recently, regardless of the type of consortium, it is easier for them to be able to become a coordinator once again through their accumulated experience. Therefore, they should reapply as soon as possible. Potential participants, on the other hand, should join consortia whose coordinator has ample resources, as these proposals are more likely to be accepted. This selection approach makes it much easier to find consortia that are more likely to be accepted by institutional authorities.
6.3 Limitations and future research
This study has several limitations that need to be addressed but at the same time creates new opportunities for future research. First, the study focused on companies that carried out Bio-innovations, which is why generalizability is limited. Even though the sample contained firms from all major industries with different firm sizes, it may be possible that results for firms conducting non-Bio-innovations differ. Furthermore, this study focused solely on the European Union framework and did not include consortia from other nations. Future research needs to address other consortia types and national frameworks as well and clarify whether these results remain similar or vary from each other.
38
would also offer insight into how much politics are involved in the R&D consortia since national governments from the EU participate alongside firms (Barajas & Huergo, 2007). Another bias that emerged through the elimination process is that the merging of the subsidiaries into their MNCs excluded subsidiary specific data. It might be possible that specific subsidiary resources or the location qualify a company as the leader that remained unobserved in this study. Therefore, future research needs to address this limitation by looking at geographical proxies and its special features.
Third, the analysis itself focused on the number of resources and consortia experience but not on which resources and types of consortia experience qualify a company as an R&D coordinator. It might be possible that not resource-richness per se increases the likelihood but rather particular resources such as tangible and intangible assets. Due to the unavailability of data, it was not possible to test for specific resources and other types of consortia experience. Future research needs to investigate the individual resource categories and classify consortia experience. This would also clarify the weak significance and inconsistencies in the results section.
6.4 Conclusion
39
40
7. References
Aversa, P., Hervas-Drane, A., & Evenou, M. (2019). Business model responses to digital piracy. California Management Review, 61(2), 30-58.
Barajas, A., & Huergo, E. (2007). International R&D cooperation within the EU Framework Programme: The case of Spanish firms. In Paper contributed for the 2007 Conference on Corporate R&D (CONCORD), Seville, Spain.
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of management, 17(1), 99-120.
Barney, J. B. (2000). Firm resources and sustained competitive advantage. In Economics Meets Sociology in Strategic Management (pp. 203-227). Emerald Group Publishing Limited.
Bauer, T. N., & Green, S. G. (1996). Development of leader-member exchange: A longitudinal test. Academy of management journal, 39(6), 1538-1567.
Belderbos, R., Gilsing, V., Lokshin, B., Carree, M., & Sastre, J. F. (2018). The antecedents of new R&D collaborations with different partner types: On the dynamics of past R&D
collaboration and innovative performance. Long Range Planning, 51(2), 285-302.
Bierly III, P. E., & Gallagher, S. (2007). Explaining alliance partner selection: fit, trust and strategic expediency. Long range planning, 40(2), 134-153.
Bio-based Industries Joint Undertaking (2019). About BBI JU derived from:
https://www.bbi-europe.eu/about/about-bbi Retrieved at: 12. December 2019
Bizan, O. (2003). The determinants of success of R&D projects: evidence from American–Israeli research alliances. Research Policy, 32(9), 1619-1640.
41
Borgatti, S. P., & Halgin, D. S. (2011). On network theory. Organization science, 22(5), 1168-1181.
Chang, S. C., Chen, S. S., & Lai, J. H. (2008). The effect of alliance experience and intellectual capital on the value creation of international strategic alliances. Omega, 36(2), 298-316.
Cipriani, G. (2014). Financing the EU Budget: Moving forward or backwards? CEPS Paperback.
Coad, A., Segarra, A., & Teruel, M. (2016). Innovation and firm growth: Does firm age play a role?. Research Policy, 45(2), 387-400.
Clausen, T. H. (2009). Do subsidies have positive impacts on R&D and innovation activities at the firm level?. Structural change and economic dynamics, 20(4), 239-253.
Corey, E. R. (1996). Technology fountainheads: the management challenge of R&D consortia. Harvard Business School Press.
Cummings, J. L., & Holmberg, S. R. (2012). Best-fit alliance partners: the use of critical success factors in a comprehensive partner selection process. Long Range Planning, 45(2-3), 136-159. Dansereau Jr, F., Graen, G., & Haga, W. J. (1975). A vertical dyad linkage approach to
leadership within formal organizations: A longitudinal investigation of the role making process. Organizational behavior and human performance, 13(1), 46-78.
Dhanaraj, C., & Beamish, P. W. (2003). A resource‐ based approach to the study of export performance. Journal of small business management, 41(3), 242-261.
Dhanaraj, C., & Parkhe, A. (2006). Orchestrating innovation networks. Academy of management review, 31(3), 659-669.
Donnenwerth, G. V., & Foa, U. G. (1974). Effect of resource class on retaliation to injustice in interpersonal exchange. Journal of Personality and Social Psychology, 29(6), 785.
42
Draulans, J., Deman, A. P., & Volberda, H. W. (2003). Building alliance capability::
Management techniques for superior alliance performance. Long range planning, 36(2), 151-166.
Duysters, G., Heimeriks, K. H., Lokshin, B., Meijer, E., & Sabidussi, A. (2012). Do firms learn to manage alliance portfolio diversity? The diversity‐ performance relationship and the
moderating effects of experience and capability. European Management Review, 9(3), 139-152. Dyer, J. H., & Nobeoka, K. (2000). Creating and managing a high‐ performance knowledge‐ sharing network: the Toyota case. Strategic management journal, 21(3), 345-367.
Etemad, H., Wright, R. W., & Dana, L. P. (2001). Symbiotic international business networks: collaboration between small and large firms. Thunderbird International Business Review, 43(4), 481-499.
European Commission, (2014). H2020 online manual. Derived from:
https://ec.europa.eu/research/participants/docs/h2020-funding-guide/user-account-and-roles/roles-and-access-rights_en.htm#RolesAssigned Retrieved at: 19. October 2019
European Parliament (2017). The EU budget at a glance - European Parliament - Europa Derived from: https://www.europarl.europa.eu/external/html/budgetataglance/default_en.html Retrieved at: 06. January 2020
European Union, (2014). BBI Joint Undertaking Derived from:
https://europa.eu/european-union/about-eu/agencies/bbi_en Retrieved at: 19. October 2019
Fisk, G. M., & Friesen, J. P. (2012). Perceptions of leader emotion regulation and LMX as predictors of followers' job satisfaction and organizational citizenship behaviors. The Leadership Quarterly, 23(1), 1-12.
Flanagan, D. J., & O’shaughnessy, K. C. (2005). The effect of layoffs on firm reputation. Journal of management, 31(3), 445-463.
43
Galbreath, J., & Galvin, P. (2008). Firm factors, industry structure and performance variation: New empirical evidence to a classic debate. Journal of business research, 61(2), 109-117. Gardet, E., & Mothe, C. (2012). SME dependence and coordination in innovation networks. Journal of Small Business and Enterprise Development, 19(2), 263-280.
Geringer, J. M. (1991). Strategic determinants of partner selection criteria in international joint ventures. Journal of international business studies, 22(1), 41-62.
Gerstner, C. R., & Day, D. V. (1997). Meta-Analytic review of leader–member exchange theory: Correlates and construct issues. Journal of applied psychology, 82(6), 827.
González, X., & Pazó, C. (2008). Do public subsidies stimulate private R&D spending?. Research Policy, 37(3), 371-389.
Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The leadership quarterly, 6(2), 219-247.
Graen, G. B., & Scandura, T. A. (1987). Toward a psychology of dyadic organizing. Research in organizational behavior.
Gulati, R., Nohria, N., & Zaheer, A. (2000). Strategic networks. Strategic management journal, 21(3), 203-215.
Haans, R. F., Pieters, C., & He, Z. L. (2016). Thinking about U: Theorizing and testing U‐ and inverted U‐ shaped relationships in strategy research. Strategic Management Journal, 37(7), 1177-1195.
Hazir, C. S. (2013). Multilateral R&D collaboration: an ERGM application on biotechnology. In The geography of networks and R&D collaborations (pp. 221-237). Springer, Cham.
Hoang, H., & Rothaermel, F. T. (2005). The effect of general and partner-specific alliance experience on joint R&D project performance. Academy of Management Journal, 48(2), 332-345.
44
Huggins, R., & Johnston, A. (2010). Knowledge flow and inter-firm networks: The influence of network resources, spatial proximity and firm size. Entrepreneurship & regional development, 22(5), 457-484.
Hung, D. K. M., Ansari, M. A., & Aafaqi, R. (2004). Fairness of human resource management practices, leader-member exchange and organizational commitment. Asian Academy of
Management Journal, 9(1), 99-120.
Inkpen, A. C. (2005). Learning through alliances: General Motors and NUMMI. California Management Review, 47(4), 114-136.I
Irwin, D. A., & Klenow, P. J. (1996). High-tech R&D subsidies Estimating the effects of Sematech. Journal of International Economics, 40(3-4), 323-344.
Jadhav, P. (2012). Determinants of foreign direct investment in BRICS economies: Analysis of economic, institutional and political factor. Procedia-Social and Behavioral Sciences, 37, 5-14. Katila, R., Rosenberger, J. D., & Eisenhardt, K. M. (2008). Swimming with sharks: Technology ventures, defense mechanisms and corporate relationships. Administrative Science Quarterly, 53(2), 295-332.
Koza, M., & Lewin, A. (2000). Managing partnerships and strategic alliances: raising the odds of success. European Management Journal, 18(2), 146-151.
Laursen, K., & Salter, A. (2004). Searching high and low: what types of firms use universities as a source of innovation?. Research policy, 33(8), 1201-1215.
Lee, E. Y., & Cin, B. C. (2010). The effect of risk-sharing government subsidy on corporate R&D investment: Empirical evidence from Korea. Technological Forecasting and Social Change, 77(6), 881-890.
45
Liden, R. C., Sparrowe, R. T., & Wayne, S. J. (1997). Leader-member exchange theory: The past and potential for the future. Research in personnel and human resources management, 15, 47-120.
Lorenzen, M. (2001). Ties, trust, and trade: Elements of a theory of coordination in industrial clusters. International Studies of Management & Organization, 31(4), 14-34.
Mayrhofer, U., & Prange, C. (2015). Multinational Corporations (MNC s) and Enterprises (MNE s). Wiley encyclopedia of management, 1-5.
Mothe, C. (1999). Creating new resources through European R&D partnerships. Technology Analysis & Strategic Management, 11(1), 31-43.
Mothe, C., & Quélin, B. (2000). Creating competencies through collaboration:: The case of EUREKA R&D consortia. European Management Journal, 18(6), 590-604.
Mothe, C., & Quelin, B. V. (2001). Resource creation and partnership in R&D consortia. The Journal of High Technology Management Research, 12(1), 113-138.
Mowery, D. C. (1998). The changing structure of the US national innovation system:
implications for international conflict and cooperation in R&D policy. Research Policy, 27(6), 639-654.
Mu, Q., & Lee, K. (2005). Knowledge diffusion, market segmentation and technological catch-up: The case of the telecommunication industry in China. Research policy, 34(6), 759-783. Nielsen, B. B. (2002). Synergies in strategic alliances: motivation and outcomes of
complementary and synergistic knowledge networks. Journal of Knowledge Management Practice, 3(2), 1-15.
Nielsen, B. B. (2007). Determining international strategic alliance performance: A multidimensional approach. International Business Review, 16(3), 337-361.
46
Olk, P., & Young, C. (1997). Why members stay in or leave an R&D consortium: Performance and conditions of membership as determinants of continuity. Strategic Management Journal, 18(11), 855-877.
Park, N. K., Mezias, J. M., & Song, J. (2004). A resource-based view of strategic alliances and firm value in the electronic marketplace. Journal of Management, 30(1), 7-27.
Phillips, A. S., & Bedeian, A. G. (1994). Leader-follower exchange quality: The role of personal and interpersonal attributes. Academy of Management Journal, 37(4), 990-1001.
ResearchRanking (2019). European Research Ranking - About. Derived from:
http://www.researchranking.org/index.php?action=about Retrieved at: 10. December 2019
Rothaermel, F. T., & Deeds, D. L. (2006). Alliance type, alliance experience and alliance management capability in high-technology ventures. Journal of business venturing, 21(4), 429-460.
Rubin de Celis, J. C., & Lipinski, J. (2007). R&D alliances and the effect of experience on innovation: A focus on the semiconductor industry. Journal of Leadership & Organizational Studies, 14(1), 26-37.
Sakakibara, M. (2002). Formation of R&D consortia: Industry and company effects. Strategic Management Journal, 23(11), 1033-1050.
Sampson, R. C. (2007). R&D alliances and firm performance: The impact of technological diversity and alliance organization on innovation. Academy of management journal, 50(2), 364-386.
Sampson, R. C. (2005). Experience effects and collaborative returns in R&D alliances. Strategic Management Journal, 26(11), 1009-1031.
47
Shah, R. H., & Swaminathan, V. (2008). Factors influencing partner selection in strategic alliances: The moderating role of alliance context. Strategic Management Journal, 29(5), 471-494.
Schiavone, F., & Simoni, M. (2011). An experience-based view of co-opetition in R&D networks. European Journal of Innovation Management, 14(2), 136-154.
Schiavone, F., & Simoni, M. (2016). Prior experience and co-opetition in r&d programs. Journal of the Knowledge Economy, 7(3), 819-835.
Sofka, W., Shehu, E., & de Faria, P. (2014). Multinational subsidiary knowledge protection—Do mandates and clusters matter?. Research Policy, 43(8), 1320-1333.
Sparrowe, R. T., & Liden, R. C. (1997). Process and structure in leader-member exchange. Academy of management Review, 22(2), 522-552.
Stine, R. A. (1995). Graphical interpretation of variance inflation factors. The American Statistician, 49(1), 53-56.
Subramanian, A. M., & Soh, P. H. (2017). Linking alliance portfolios to recombinant innovation: The combined effects of diversity and alliance experience. Long Range Planning, 50(5), 636-652.
Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research policy, 15(6), 285-305.
The Community Research and Development Information Service (Cordis): Derived from:
https://cordis.europa.eu/about/en Retrieved at: 12. December 2019
The World Bank (2020); GDP 2013: Retrieved from: https://data.worldbank.org/ Retrieved at: 09. January 2020
Tian, Q., Ma, J., Liang, J., Kwok, R. C., & Liu, O. (2005). An organizational decision support system for effective R&D project selection. Decision support systems, 39(3), 403-413.
48
Volmer, J., Niessen, C., Spurk, D., Linz, A., & Abele, A. E. (2011). Reciprocal relationships between leader–member exchange (LMX) and job satisfaction: A cross‐ lagged analysis. Applied Psychology, 60(4), 522-545.
Wassmer, U. (2010). Alliance portfolios: A review and research agenda. Journal of management, 36(1), 141-171.
Wilson, K. S., Sin, H. P., & Conlon, D. E. (2010). What about the leader in leader-member exchange? The impact of resource exchanges and substitutability on the leader. Academy of Management Review, 35(3), 358-372.
49
Appendix A – T-test by mean group comparison
Group (by Firm Size) Obs Mean Std. Err. Std. Dev. [95% Conf. Intervall]
Participants 345 4,632854 0,1427185 2,650878 4.352143 4.913565 Coordinators 29 6,528933 0,5450081 2,934959 5,412534 7.645331 combined 374 4,779876 0,140523 2,717586 4,503559 5.056193 diff -1,896079 0,5168646 13,52696 -2,912421 -0.879736 t-value -3,6684 degrees of freedom 372 Pr(T < t) 0,0001 Pr(|T| > |t|) 0,0003 Pr(T > t) 0,9999
Group (by Recent Coordinator
Exp.) Obs Mean Std. Err. Std. Dev. [95% Conf. Intervall]
50
Appendix B - Penalized log-likelihood method
Model 1 Model 2 Model 3 Model 4
Control Variables Firm Age 0,101 -0,012 -0,023 -0,026 (0,235) (0,239) (0,241) (0,242) Industry Classification 0,714 0,897* 0,875* 0,942* (0,481) (0,497) (0,5) (0,501) GDP 0,027 -0,062 -0,079 -0,099 (0,216) (0,222) (0,224) (0,225) Project Budget 0,407* 0,392* 0,382* 0,343 (0,227) (0,229) (0,23) (0,238) MNC 0,579 -0,176 -0,151 -0,141 (0,490) (0,637) (0,641) (0,688) EU Contribution per national share -0,013 -0,004 -0,004 -0,005
(0,043) (0,043) (0,043) (0,044) Experience as Participant 0,004 0,000 -0,004 (-0,043) (0,006) (0,006) (0,007) (0,028) Multiple Participations 2,077*** 2,019*** 1,978*** 1,944*** (0,434) (0,44) (0,438) (0,447) Independent Variable Firm Size 0,217* 0,221* 0,218* (0,113) (0,114) (0,154) Moderator Variable
Recent Coordinator Experience 0,291 -0,776
(0,206) (0,760)
Interaction Effect
Firm Size X Recent Coordinator Experience 0,296
(0,203)
Number of observations 374 374 374 374
Wald chi2 36 38,05 38,89 37,5
Prob > chi2 0,000 0,000 0,000 0,0001
