ABSORPTIVE CAPACITY (ACAP) The impact of governance mechanisms on the exploitative and explorative capabilities of firms engaging in knowledge sharing alliances.

(1)

1

ABSORPTIVE CAPACITY (ACAP)

The impact of governance mechanisms on the exploitative and

explorative capabilities of firms engaging in knowledge sharing

alliances.

MSc MARKETING

ANDREA ZANGIACOMI

University of Groningen Faculty of Economics and Business

7th_{February, 2016}

First supervisor: Dr. J. Hans Berger Second supervisor: Dr. Yannick Joye

Paterswoldseweg 271-2 9728 AE Groningen (+31)6 57992312

(2)

2

Abstract

This study centers on the impact of governance mechanisms on PACAP and RACAP from a resource-based view, focusing on inter-firm learning, exchange of knowledge, and value creation by contributing to extending the existing research on the broad topic of inter-firm ACAP.

In order to reach the objectives of this research, the survey utilized a questionnaire of 166 matched-pair relationships between buyers and suppliers.

Researchers have examined the determinants of ACAP from a social perspective, focusing on the governance mechanisms that determine the interchange of power and collaborative spirit among the allies.

The paper specifically aims to provide new insights into the importance of relational norms and contracting governance. In fact this study reveals that governance mechanisms, relational norms and contracting can be used as substitutes from the buyers, in relevance to the dimension of PACAP and RACAP.

The role of the duration of the relationships has been assessed by using it as a moderator in the relation between the governance mechanisms and the dimensions PACAP and RACAP.

Consequently, the significant effects of the two sub-dimensions of ACAP on the explorative and exploitative learning performances have been examined.

(3)

3

1. Introduction and context setting

Today’s world is turbulent, rapidly evolving, and firms are constantly searching for novel ways to generate a competitive edge. Knowledge is essential to ensure the development and prosperity of a company over time. In fact, organizations embark on innovative processes starting from the information available, combining and increasing it in new and creative ways. It is important that companies do not focus solely on the knowledge developed within the firm, through research and development. Otherwise, the risk is to start only incremental innovation processes, rather than those of radical type (Cohen and Levinthal, 1990; Kats and Allen, 1982). Looking out to the external environment, observing the dynamics of the market and its changes, recording the moves of competitors and studying the new knowledge available in their context, are all sources of stimulus for organizational activities and they allow to overcome the limitations of the internal knowledge. Only by integrating internal and external information, knowledge of an enterprise can be completed and, in this way, promote innovation processes (Cassiman and Veugelers, 2006). In fact, firms that successfully develop those resources have higher chances than their competitors to realize competitive advantages, create as well as accept innovations and perform in rapidly changing markets (Barney, 1991)

The capability that allows organizations to identify the worthiness of new knowledge in the external environment, useful to its processes, to acquire, assimilate, transform and combine them with the information already available and, finally, to use them for commercial purposes is defined as absorptive capacity “a set of organizational routines and processes by which firms acquire, assimilate, transform, and exploit knowledge to produce a dynamic organizational capability” (Cohen and Levinthal, 1990; Zahra and George, 2002).

1.1. ACAP – Absorptive Capacity

(6)

6 Tacit knowledge is difficult to imitate, intangible, and highly protected by firms who have the resources and competencies to acquire it (Wiklund and Shepherd, 2003; Barney, 1991). Firms thus enter into contracts in order to ensure the transfer of knowledge and especially to support the early stages, which are considered the most vulnerable moment of cooperation and the most risky phases for a successful exchange of knowledge. Therefore the effectiveness of inter-firm knowledge transfer is attributed to numerous factors such as relational and contractual attributes (Minbaeva, 2007). The governance mechanisms determine the synergy of participants and whether the firm will be able to manage the knowledge (Fowler and Reisenwitz, 2013).The coexistence of these relational and contractual norms (determinants of ACAP) is inter-related, the ineffectiveness of any characteristic would occasion a less effective and less successful knowledge acquisition. As firms continue to incorporate more difficult and complex processes into practice in order to acquire, assimilate, transform and exploit the knowledge there is a need for close theoretical explanations regarding the relative influence of each related characteristic and their combined effect on the absorptive capacity.

1.2. PACAP & RACAP

Zahra and George (2002) have grouped ACAP into two categories: PACAP (potential absorptive capacity) and RACAP (realized absorptive capacity).

PACAP, potential absorptive capacity, includes a set of routines and strategic processes by which organizations acquire and assimilate new external knowledge (Zahra and George, 2002). The potential absorptive capacity indicates whether a company is receptive to new information, as it relates to the mode of acquisition, analysis, interpretation and understanding (Zahra and George, 2002 ). The PACAP, unlike the RACAP, is a process that is developed especially at a personal and an individual level (Cepeda-Carrion et al., 2010) . In fact, the driving forces of this capability are reflections, intuition and interpretations that entrepreneurs develop around the new knowledge (Cepeda-Carrion et al., 2010).

(7)

7 PACAP and RACAP are two complementary skills and are coexisting, their union resulting in the absorptive capacity. At the same time, these two dimensions differ because the potential absorptive capacity (PACAP) concerns the acquisition and assimilation of information, while the realized absorptive capacity (RACAP) focuses on how the knowledge is then transformed and exploited (Zahra and George, 2002). As largely established in his research, March (1991) determined these two aspects: “Exploration includes things captured by terms such as search, variation, risk taking, experimentation, play, flexibility, discovery, innovation. Exploitation includes such things as refinement, choice, production, efficiency, selection, implementation, execution”. As different and distinct concepts, they imply different strategies and structures (Newey and Zahra, 2009). While the PACAP requires change, flexibility and creativity, RACAP needs order, and stability control (Cepeda and Carrion, Cegarra and Jimenez, 2010).

Although they are distinct concepts, there is an equal importance. For example, the information acquired and assimilated, if unused, does not lead to any benefit for the organization. As reported in his study, Berger (2015) assesses that firms that over-emphasizing exploration will suffer the costs of experimentation without the benefit of commercialization, while firms accentuating their efforts more on the exploitation may find themselves trapped in “suboptimal stable equilibrium” (March, 1991).

The distinction of these two categories, PACAP and RACAP, is very important to assess and it is also necessary to understand their link to the concept of dynamic capabilities, defined as “the firm’s ability to integrate, build and reconfigure internal and external competencies to address rapidly changing environments” (Teece, Pisano, and Shuen, 1997) and to consequently recognize their specific impact on the enterprise’s competitive advantage. Finally, the division between PACAP and RACAP helps to explain how some companies are more efficient than others in activating the absorptive capacity and where the inefficiencies creep in scope. This helps to understand why some businesses fail as a result of environmental changes, while others thrive on equal terms (March, 1991).

As both PACAP and RACAP concepts will be crucial in this thesis, their definitions and key points are again shortly summarized below.

PACAP (potential absorptive capacity) comprises the capabilities to acquire and assimilate

(8)

8 knowledge. Assimilation, on the other hand, refers to the structures set in place to allow for the analysis, processing, interpretation, and comprehension of the newly acquires external knowledge (Zahra & George, 2002).

RACAP (realized absorptive capacity) are the knowledge transformation and exploitation

capabilities of a firm. It consists of two components, transformation, and exploitation. Transformation deals with the refinement and development of the routines that enable the newly acquired knowledge from external sources and the existing in-house knowledge to be assimilated. Exploitation is the ability of this acquired knowledge to be utilized in the firm’s operations (Zahra & George, 2002).

1.3. Problem Statement

As described before, tacit knowledge is intuitive, unarticulated, difficult to acquire and to exploit, and can only be acquired through practice and observed while being applied (Wiklund and Shepherd, 2003; Barney, 1991). The acquisition of this kind of knowledge is subject to time compression diseconomies that make it very hard to accelerate its learning no matter the resources invested to acquire them (Lin, 2005). Moreover, it involves a lot of transaction costs. These asset specific costs are undertaken in support of the knowledge acquisition quest (Williamson, 1985). They include dedicated human, physical, and site assets borrowed from other endeavors in the organization. With the amount of time and assets invested to acquire this knowledge and considering that the information has to be used timely to ensure a competitive advantage, many corporations want assurance that the alliances for knowledge transfer will generate results. Firms face a tradeoff between the opportunity to acquire knowledge and the potential opportunism of the other allies (Inkpen, 2000). They thus seek ways to safeguard themselves, while at the same time ensuring they maintain a conducive working relationship with the other partners.

(9)

9 attributes.

The research of Berger (2015) is one of the few papers, which studies the multidimensionality of ACAP, while focusing on the inter-firm level. His study takes into account ACAP’s four dimensions, instead of using measures as the number of patents or R&D investments. This will enable to analyze with accurateness. Furthermore, dyadic data was collected from buyer and supplier in order to examine the perception of their relationship from both their sides, since literature proposes that each partner might experience different perceptions of the relationship (Oosterhuis et al., 2013; Ambrose et al., 2010). The current research has been inspired by these mentioned pioneer approaches. However, this research extends Berger’s study (2015) by assessing how both, contracting and relational norms, help to enhance PACAP and RACAP when also the duration of the relationship as a moderator is taken into account. Moreover, the work of Berger (2015) will be extended additionally by considering whether the governance mechanisms can be seen as substitutes or complements.

In the following, reasons why this paper contributes to the existing literature will be listed. Firstly, many studies have shown that the interplay between contracting and relational norms from a TCE prospective influences the performance between the partners, opportunism and satisfaction (e.g., Poppo and Zenger, 2002; Cannon et al., 2000; Li et al., 2010; Yang et al., 2012). This study centers on the impact of contractual and relational governance mechanisms on PACAP and RACAP from a resource-based view, focusing on inter-firm learning, exchange of knowledge, and value creation by contributing to extending the existing research on the broad topic of inter-firm ACAP. By doing so, this research also extends the study of Jansen et al. (2005), which examines the effect of contractual governance on PACAP and RACAP, but not the effect of relational governance on PACAP and RACAP.

(10)

10 and how they influence the organizational learnings from a resource-based view. Thus, the present study aims at adding clarity to the topic of the interplay of governance mechanisms and furthermore focuses on knowledge acquisition, the overall effectiveness of the learning experience, and assesses the role of contractual and relational governance as narrow links.

Thirdly, several studies have tried to assess the role of the duration of the relationships (Gounaris and Venetis, 2002; Claro et al., 2003; Kumar et al., 2003) but results have shown opposite and disharmonious results without giving determined findings. Fink et al. (2008) put worthy considerations towards the timing of the relationship as an important indicator to assess the relationship between buyer and supplier. They have mentioned in their study that “the factors that influence customer–supplier relationship outcomes vary with time and more studies need to be conducted to further enhance our understanding of time as a variable”. In addition various researchers have found that a direct effect of the duration of the relationship exists on knowledge transfer at inter-firm level (Squire, Brown & Cousins, 2009). However they were not able to find prove that the moderation effect of the duration of the relationship between partners can increase the strength of the governance mechanisms on PACAP and RACAP. According to Ring & Van de Ven (1992) a longer lasting relationship can be caused by shared norms between partners and that they create the foundation to improve ACAP at an inter-firm level. Thus, the duration of the relationship might strengthen two of the variables that were distinguished by Berger (2015) and which enhance the absorptive capacity: relational norms and contracting. Therefore, the duration of the relationship is used as a moderator in this analysis.

To sum up, this research will extend the study of Berger (2015) and will contribute to the existing literature. Therefore, the following research questions seem to be of particular interest.

1.4. Research Questions

1. How do inter-firm governance mechanisms affect the overall effectiveness of the collaboration exercise:

(11)

11 2. What is the magnitude of the duration of the relationship and how does this effect affect

the interplay of the two inter-firm governance mechanisms on PACAP and RACAP?

1.5. Conceptual Model

As it will be explained in chapter 2, the expected relationships being investigated are summarized in the following conceptual model. The main concept is that the contractual and relational norms have a narrow interaction. The contractual and relational norms have a bearing on both RACAP and PACAP. The relationship duration acts as a moderator for the model, and it will be assessed if it enhances the effect of the governance mechanisms. In addition, PACAP has a bearing on explorative performance, while RACAP has a bearing on exploitative learning performance. The following chapter will further elaborate on the variables and the links between them.

Conceptual model

(12)

12

2. Theoretical Framework

2.1. Governance mechanism: Contractual Norms

Formal contracts represent promises or obligations to perform particular actions in the future (Macneil, 1978). It is also referred to as formalization, contracting, or formal protection mechanisms. Specifically, Jansen et al. (2005) define it as the extent to which institutions, processes, rules, and effective communication are defined in written contracts or formal documents. It also refers to the explicit rules of the knowledge exchange endeavor (Ghosh and John, 2005). These rules are especially necessary during the early stages of cooperation as they reduce both cooperating partners’ worries about opportunism and risks since it establishes a certain amount of transparency in the exchange relationship (Noordhoff et al., 2011; Wathne and Heide, 2000). It may be relaxed later when the degree of collaboration and ability to transfer knowledge of the partners are increased between the partners.

Firms feel the need to protect themselves for various reasons, majorly for the restriction of ease of access to proprietary knowledge (Hau & Evangelista, 2007). It is an increased trend that firms in a buyer-supplier alliance are competitors in the market (Hamel et al., 1989), unlike in the case of a joint venture. It is for example common practice in the electronic industry (Hamel et al., 1989). If some information provides a sustainable advantage for some firms, other firms in the inter-firm relationship may exhibit opportunistic tendencies. It may also be that some alliance partners maintain an effective ability to transfer knowledge for selfish reasons.

(13)

13 2.1.1. Impact on PACAP

The presence of absorptive capacity is not only a determinant of the firm as a whole but it also has an important impact on each single unit of the organization (Cohen and Levinthal, 1990). Therefore, it is important to understand how contracts impact PACAP.

Weick (1979) argued that if firms limit themselves and thus their inside workers by activating contracts and stipulating rules, in turn will limit the workers’ behavior and their field of expertise and consequently the potential explorative acquisition. Furthermore, Jansen et al. (2005) mentioned that high levels of contracting forms can reduce the level of the firm’s commitment and decrease the willpower to explore and assimilate new external knowledge. Specifically, contracts can obstacle the exchange of knowledge at unit’s organizational level.

Jansen et al. (2005) found that in some cases formalizing PACAP through contracts does not cause its decrease. Furthermore, contracting allows for more effective acquisition and assimilation of new external knowledge (Adler and Borys, 1996).

Consequently, Jansen et al. (2006) stated that a unit’s exploratory innovation performance does not decrease as a result of contractual protections, thus acquisition and assimilation might not be affected.

For enterprises, low levels of contracting do not appear as detrimental to PACAP and exploratory as expected. The posited reason is that low levels of protections allow employees to carry out their functions better than when there are no protections (Adler and Borys, 1996). In fact, contracting might actually improve the relationship and ultimately firm’s PACAP. From a coordination point of view, contractual formalizations might show commitment to the relationship instead of a lack of trust (Malhotra and Lumineau, 2011). Zollo and Winter (2002) added that contractual formalizations might actually contribute to the exploratory orientation of a firm by empowering the knowledge in form of a contract.

(14)

14

In the light of the above, it is believed that the following hypothesis could be generated at inter-firm level:

Hypothesis 1a-b:

Contractual governance will have a positive relationship with PACAP (H1a), that levels off at higher levels of contractual governance (H1b).

2.1.2. Impact on RACAP

According to research conducted by Jansen et al. (2006), it has been empirically affirmed that RACAP is positively influenced by formalization. Formalizations enable firms in the alliance to discover new opportunities for transformation of the knowledge derived by external environment. In particular, contractual formalizations might make it easier for firms to retrieve the existing knowledge in a timely manner. When used with knowledge derived internally, the comprehension of the exchanged knowledge is amplified (ibid). Contractual formalizations can bring transparency to the relation by removing any incidences of opportunism, and with it the elimination of obstacles to the sharing of knowledge (Carson et al., 2006). Finally, when knowledge has already been internally stored, formalizations lead to easier exploitation and simpler and faster implementation (Jansen, 2006).

Based on the above, it is hypothesized:

Hypothesis 1c:

Contractual formalizations will have a positive effect on RACAP (H1c).

2.2 Governance mechanism: Relational Norms

Relational governances are the implicit rules of exchange and are comprised of relational norms and a trust mechanism. These norms cannot be created; they have to be developed over time. The better the relational norms, the more the openness and transparency between partners. Gupta (1987) defines this openness as the degree of informality of the relationship between alliance managers to promote the spontaneous and open exchange of ideas.

(15)

15

relational norms. Trust can enhance cooperation and evolves over time (Mayer et al., 1995; McAllister, 1995). Besides enhancing cooperation and openness, trust also lessens inter-firm transaction costs (Dyer and Chu 2003). Indeed, when firms perceive high level of trust in the cooperating partner they are inclined to become less strict. On the opposite, when firms detect less level of trust, they are more inclined to inspect and adopt a stricter behavior by enforcing the openness with more barriers, which lead to monitor each other’s behavior (Selnes and Sallis, 2003) and a tendency to not sharing the knowledge for fear of opportunistic behavior. Furthermore, thanks to high level of trust, firms get also easier access to useful referrals, hence leading to the improvement of ACAP (Uzzi, 1996).

Thus, both scopes of ACAP will benefit from the increase of trust over time. Indeed, potential absorptive capacity (PACAP) will be improved as a result of the firms opening up and being more willing to share their accumulated knowledge. Realized absorptive capacity (RACAP) will be positively impacted because trust will lead to an easier way of sharing knowledge. Trust is especially crucial in sharing inexplicit and uncertain knowledge, for which formalizations seem to not be suited (Makhija and Ganesh, 1997; Cannon et al., 2000; Uzzi and Lancaster, 2003; Inkpen and Curall, 2004).

However, at higher levels of relational ties and norms, ACAP is impacted upon negatively. According to Selnes and Sallis (2003), this point forms the “hidden costs” point. This is due to the fact that the negative effects are not easily discernible and recognizable by the firms in the alliance. Firms might avoid any information deemed negatively by the other party (Selnes and Sallis, 2003). Emotions kick in and firms begin avoiding difficult but potentially significant information for fear of rupturing the relationship (Selnes and Sallis, 2003). Another argument is that it might also occur that in a group thinking between partners, as firms become more acquainted with each other, creativity and the incentives for innovation are undermined due to the facts that partners are used to work together and do not rethink their ways of working (Selnes and Sallis, 2003; Noordhoff et al., 2011), leading to a reduced level of PACAP and RACAP

Finally, it is hypothesized that:

Hypothesis 2a-b:

(16)

16

Hypothesis 2c-d:

Relational governance will have a positive relationship with RACAP (H2c), that levels off at higher levels of relational governance (H2d).

2.3. Duration of the relationship as moderator

Several researchers pointed out that relational characteristics like the duration of the relationship can influence the relationship between partners (Gounaris and Venetis, 2002; Sweeney and Webb, 2007; Yen and Barnes, 2011).

According to Ford (1980), the nature of a buyer-supplier relationship changes over time and may impact several aspects such as uncertainty and relational-distance. The author explains further, that these negative aspects for a relationship might decrease in a mature relationship and that for this reason buyer and supplier should try to rely on the duration of the relationship, in order to avoid unpleasant circumstances. Furthermore, being in a long-term relationship with a partner may result in the creation of substantial barriers to enter in connection with other potential future partners, due to mainly two factors: inertia and avoiding to jeopardize the business by changing the partner (Yong-Ki Lee et al., 2015). Related to this research paper, these mentioned aspects can be seen as positive effects on the transfer of knowledge through formalization. Contracts can contribute to hinder partners to end a relationship at an early stage. If partners cannot rescind contracts, the relationship is “forced” to exist and over time might satisfy both sides.

(17)

17 According to Gounaris and Venetis (2002), the duration of the relationship moderates the service attributes and trust by enhancing friendliness and empathy. The authors have also mentioned that in a mature relationship between buyer-supplier, each part puts more efforts in short-term outcomes, whereas partners in new alliances emphasize long-term quality. Furthermore, another research that reported a significant moderating effect of the duration of the relationship has been developed by Sweeney and Webb (2007). The authors highlighted the significant moderating effect of the duration of the relationship on three types of relationship benefits (functional, psychological and social) in the manufacturing industry. Especially, social and psychological benefits have been found to play an important role in the experienced and mature relationships. However, these findings are contradicted by other authors, who reported results (Gounaris and Venetis, 2002; Sabiote and Roman, 2009), leading to inconclusive findings on the moderating role of the duration of the relationship. Kohtamaki and Bourlakis (2012) found no significant role of the duration of the relationship as a moderator, especially regarding the socials aspects. In addition, also Claro et al. (2003) found no significant effect of the duration of the relationship on the relational governance behaviors such as joint planning and joint problem solving.

However, our research follows the logical reasoning that long relationship between firms enhance trust and openness, thus the effect of relational norms on the processes to acquire, assimilate, transform and exploit will be improved and will lead the firms to improve the transfer of external knowledge. Furthermore, the current study tries to put clarity by adding further evidence to the literature by examining the role of the duration of the relationship in the B2B industry at inter-firm level. In this research, the role of both governance mechanisms on PACAP and RACAP are retained to be greater when the duration of the relationship will be taken into consideration rather than when it will be not taken into consideration.

In summary, this review of the literature suggests that the duration of the relationship may play an important role in an inter-firm relationship aspect. Therefore, the following hypothesis is constructed.

Hypotheses 3a-d:

(18)

18

2.4 The interplay between Contractual and Relational Governance

In this research, the study on contractual vs. relational governance has been inspired by examples from numerous theoretical reviews (e.g. Poppo and Zenger, 2002; Li et al., 2010; Cannon et al., 2000). These theoretical reviews show that the interplay between contracting and relational norms can be viewed from the ‘substitute view’ as well as from the ‘complementarity view’. Poppo and Zenger (2002) report that governance mechanisms, specially contracting and relational norms act as complements. They reason that contracting may lead both partners to build a relational governance between them (Cannon et al., 2000; Poppo and Zenger, 2002). In addition, Cao et al. (2015) put worthy considerations towards the complementary interplay between governance mechanisms. They found in their research that when contracting and relational norms act as complements a result could be the improvement of the satisfaction and enterprise performances. In particular, the authors mention when these governance mechanisms are used as complements they also decrease the level of opportunism on the TCE view. Additionally, Bastl et al. (2012) point out that when both mechanisms are used as complements relational norm can advance the development of new contractual mindset and simplify the contracting processes.

Previous researches are mainly focused on the TCE perspective. This study however is based on the outcomes of RBV, which is the reason why in the following the ‘substitutes view’ is described based on RBV.

(19)

19 on sharing experience considering the integrity-processes on the improvement of dimensions of ACAP: PACAP and RACAP. They should be mentioned in the frameworks of this research as alternative substitutes. Thus, this research is designed to discover the interaction in governance mechanisms in two different perspectives, as they may be used in governing functioning as relocated substitutes towards each other (Huber et al., 2013; Lui and Ngo, 2004; Li et al., 2010).

Hypothesis 4:

Both governance mechanisms, contracting and relational, function as substitutes to enhance PACAP and RACAP (H4).

2.5. Learning performance

Many researchers have tried to study and analyze the concept of absorptive capacity (e.g. Cohen et al., 1990; Zahra et al., 2002) due to its correlation with innovation, performance and competitive advantage. The different studies are aimed, firstly, at trying to understand this connection, and secondly, at providing the tools to businesses in order to develop a skill that will enable them to achieve these results. The literature has mainly focused on the relationship between ACAP and innovation, identified by Cohen and Levinthal (1990) and Zahra and George (2002).

(20)

20 efficiency when this process occurs. The differences in terms of access to knowledge and learning skills impact innovation and performance (Tsai, 2001).

2.5.1. Interlink of PACAP with Explorative Learning Performance

Explorative learning suggests that the company advances from its existing technology base and encompasses information that is both new and tacit. In order to implement the explorative learning, firms might base their activities on search, variation, risk taking, experimentation, play, flexibility, discovery, or innovation (March, 1991). As mentioned in the previous section, the company might often lack in-house knowledge that forces it to team up with other organizations that are more knowledgeable.

Connecting to the external environment, observing the dynamics of potential partners and the possibility of new knowledge available in the business (Keil, 2002), are all abilities that create incentives to overcome the limitations of the internal knowledge. Often the external environment offers knowledge way different from the existing knowledge of the firm (Katila, 2002).

According to Weick (1979), putting efforts and focusing only on restricted areas will reduce the explorative learning. Firms should act wisely, by only integrating internal and external information, so that knowledge of an enterprise can be completed and, in this way, promote innovation processes (Cassiman and Veugelers, 2006).

(21)

21

coalitions is exceeded (Duysters and de Man, 2003). It would be of high interest to assess the potential impact of PACAP and determine if it interacts positively with the explorative learning performance.

In light of the above, it is hypothesized:

Hypothesis 5:

PACAP will have a positive impact on the explorative learning performance of the firm (H5).

2.5.2. Interlink of RACAP with Explorative Learning Performance

Exploitative learning is the refinement of existing technologies. Exploitation activities include refinement, choice, production, efficiency, selection, implementation, and execution (March, 1991). It is less tacit and uncertain since the company has much of the requisite information in-house (Levinthal and March, 1993). Companies team up with partners to share R&D costs and risks, obtain complementary know-how, and avoid the duplication of effort.

Since the difficulty is defined concisely, alternative solutions are known, and the allies have a fairly good comprehension of tasks, the opportunities for opportunism are high. Moreover, the partners in the alliance have the required knowledge in-house and thus the alliance is only one of many alternative ways of strengthening their competitive advantage. Finally, the partners are competitors (Hamel, 1989) and thus the level of trust is low (Hansen et al., 2001). Realized absorptive capacity might provide instant effects on exploitation of learning and thus for example on financial performance (Uotila et al., 2009).

Since the knowledge is mostly already acquired, firms mainly need to implement specific information, this process might determine a faster exchange of know-how and thus better performances.

Therefore, following hypothesis will be tested:

Hypothesis 6:

(22)

22

3. Methodology

3.1 Introduction to ACAP and its Measurement

The literature dealing with the issues of ACAP has focused mainly on the definition of the concept and its key dimensions (Cohen et al., 1990). Several authors, as Jansen et al. (2005) and Jimene et al. (2011) have also tried to find the units of measurement to verify the presence or absence of ACAP within an organization and its staff. Being able to measure this capability is useful for two main reasons. Firstly, to better investigate the correlation between absorptive capacity and business results, and secondly to be able to compare a company to its competitors. Despite the fact that absorptive capacity has been recognized as a dynamic and multidimensional skill (e.g. Thuc Anh et al., 2006), most of the studies in the literature provide indicators that treat the concept as one-dimensional. Moreover, these indicators often refer to research and development.

(23)

23 same way.

Research and development is referred to the knowledge management in terms of information flow. This way, the leading companies can work out on the innovation and discovering the new up-to-date solutions when combined together. These factors stand closely linked between R&D and ACAP actions. However, the limitation in research and development is specific to what happens within the organization, which do not consider the knowledge that is developed externally and that a company can acquire.

Khoja and Maranville (2010), starting from the definition of Lubatkin and Lane (1998) and Tsai (2001) that the absorptive capacity is "the ability to acquire and assimilate new knowledge obtained from external sources", measure the ACAP only through the indicators acquisition and “absorption”. The focus of these authors is primarily on human capital, which is why the measures proposed by them are primarily to evaluate the actions of the staff.

Lane et al. (2006) suggest that the absorptive capacity should be explored empirically in contexts other than those of R&D, using a metric that captures each of the dimensions of ACAP and to take account the specific context in which the company operates. The criticism that these authors move towards indicators addressed here, relates to the fact that research and development, with reference to both its output and input, is the absorptive capacity as a static resource and not as a capacity or a process.

In order to exceed these limits described above, a multidimensional indicator is needed, that is actually able to grasp every aspect and process activated from ACAP. This research has utilized existing scales already developed in previous researches, which are presented in the next sections.

3.2 Research Design

The object of analysis is to test the effect of both governance mechanisms. Contracting and relational norms are hypothesized to affect the two sub-dimension of ACAP: PACAP (potential absorptive capacity) and RACAP (realized absorptive capacity) (Zahra and George, 2002). These four effects will be also assessed when moderated by the duration of the relationship.

(24)

24 between contracting and relational norms and further in order to assess whether they can be used as substitutes as hypothesized.

Consequently, PACAP and RACAP are hypothesized to influence positively the explorative and exploitative learning performance.

3.2.1.

Sample and Data Collection

In the frameworks of this research, the data has been collected by Dr. J. Berger, professor of the faculty of Economics and Business of the University of Groningen. The questionnaire has been developed by Berger (2015) to assess the inter-firm absorptive capacity and its effect on ambidexterity. The current study has utilized data collected from 166 matched-pair relationships between buyers and suppliers. The interviews have been organized by contacting the buyer-respondents who were located in the Netherlands and asked to contact their suppliers located all over the world.

The insights of the questionnaire provide relevant and sensitive information on aspects driving the business of the firms, thus the questionnaire was distributed to participants to collect the data of buyers and suppliers in a confidential way. The distribution was performed physically or via e-mail.

In order to assess the engagement towards the firm (Campbell, 1955) and in order to fill out the survey, it has been necessary to identify the employees who play a central role in the customer relationship with partner firms. In the period between June 2011 and April 2013, among many other things, they have been asked questions about their current position and the current number of years they worked in their positions.

In relation to the object of our analysis, the informants have responded to the questionnaire that includes questions related to contractual governance as well as relational norms. The participants have also answered to questions related to explorative and exploitative capabilities (PACAP and RACAP) as well as explorative and exploitative learning performances. Thus, the database contains measures that can be utilized to test the suggested hypotheses presented in the conceptual model.

(25)

25 advantage through technology innovations to boost up business. In fact, it is important to highlight that this questionnaire includes 332 enterprises in total to represent three industries: pharmaceutical (28.9%), automobile (40.8%) and semiconductors (13.3%).

3.2.2. Operationalization

In relation to our research, the survey is delivered in 10 sections to represent six variables as previously presented in the conceptual model. Following the study of Berger (2015), the conceptual model was measured with multi-items that were developed and assessed also in previous researches. All items were tested on a Likert scale, anchored by 1 (strongly disagree) to 7 (strongly agree) and from 1 (to no extent) to 7 (to a great extent) (Berger, 2015; Jap, 1999). First of all however, in the present study the presence of first- and second-order constructs calls for a clarification of these two terms before the conceptual model can be tested. “A first-order construct is a latent construct that has observed variables as indicators, whereas second-order constructs have other latent (first-order) constructs as their indicators” (Berger, 2015). Related to the conceptual model in this study: PACAP, RACAP as well as relational norms are second-order constructs. Therefore, they consist of compositions of a series of latent constructs.

Secondly, it is important to point out the presence of reflective and formative constructs in the research scales. Jarvis, MacKenzie and Podsakoff (2003) inform about the differences between these two scales. Regarding reflective constructs, covariance among the indicators is caused by the variation in the underlying latent factor. Regarding formative latent constructs, the indicators cause changes to the underlying construct (Fornell and Bookstein, 1982).

In addition, it is necessary to mention that several indicators have been recoded in order to be sure that the answers were well considered from the participants.

(26)

26 1) Contractual governance are formal documents and are developed to foster the relationship between partners and to facilitate the explicit norms to be exchanged (Ghosh and John, 2005). These measures were developed by Deshpande and Zaltman (1982), Buvik and Reve (2002), Cannon and Perreault (1999) and Jansen et al. (2006). Contractual governance was measured by the questions 14.1 – 14.6 of the questionnaire of Berger (2015).

2) Relational governance is explicable implicitly in the exchange, being captured under relationally significant norms and trust measures. Heide and John (1992) assess relational norms in three sub-dimensions: flexibility, exchange of information and solidarity that are reflective first-order constructs. Scales from existing research were used to measure relational norms: questions 15.1 – 15.3 measure flexibility, questions 15.4 – 15.7 measure information exchange and questions 15.8 – 15.10 of the questionnaire of Berger (2015) measure solidarity norms. As to test for PACAP and RACAP, the three first-order constructs will be assessed by validity and reliability tests and if satisfying results emerge from these tests, the three first-order constructs will be merged in a second-order construct in order to test the inner-model.

3) PACAP and RACAP are equally determined in two sub-dimensions: assimilation and acquisition capabilities (PACAP) and transformation and exploitation capabilities (RACAP) (Zahra and George, 2002). The measures were gathered from the studies of Camisón and Forés (2010) and Szulanski (1996).

In Berger’s (2015) questionnaire PACAP stands for the capabilities of a firm to recognize, acquire and assimilate knowledge. Questions 1.1 – 1.5 are used to measure acquisition (formative scale). Questions from 2.1 – 2.5 are used to measure assimilation (reflective scale).

RACAP was measured through questions about transformation and exploitation. The questions 3.1 – 4.4 are used to measure transformation (formative scale) and questions 5.1 – 5.5 are used to measure exploitation (reflective scale).

(27)

27 4) Explorative learning performance is the need to cultivate procedures and alliance management skills that result in an increased innovative output for the affiliating company. Concerning this variable Lane et al. (2001) developed measures. Questions 8.1 – 8.5 (formative scale) measure explorative learning performance. All items were tested on a Likert scale, anchored by 1 (to no extent) to 7 (great extent) (Berger, 2015).

5) Selnes and Sallis (2003) developed the measures of exploitative learning performance in order to assess the joint creation of new expertise, with a special focus to test the ability of a firm to improve the efficiency and effectiveness when sharing the knowledge with the partner. Questions 9.1 – 9.7 (formative scale) measure exploitative learning performance. All items were tested on a Likert scale, anchored by 1 (strongly disagree) to 7 (strongly agree).

6) Relationship duration is referred to the number of years that buyers and supplier have been engaged in cooperative work. It will be assessed by the question 0.1 and is expressed in the number of years.

An overview of all the different items and scales used to measure the variables of the model is presented in the appendix.

3.2.3. Plan Of Analysis

The analysis of the survey will be measured and processed through SmartPLS 3, which is based on the variance Structural Equation Model (SEM) that integrates factor analysis and multiple regression analysis. The aim of the SEM is to minimize the residual variance within the explained variables (Hair et al., 2011; Berger, 2015).

Partial Least Squares (PLS) stands for the statistical technique to combine factor analysis and the consequent analysis. This type of technique is considered the most appropriate in processing the first- and second-order constructs along with the formative and reflective scaling (Chin, 1998; Echambadi et al., 2006). For this reason the data used for this research is analyzed by using Partial Least Squares (PLS).

(28)

28 (e.g. Relational norm: flexibility, solidarity and information exchange) and through finally linking dependent and independent variables the graphic model has been created.

An assessment of the measures of the model has been performed. Therefore, the reliability of the survey is tested using Cronbach’s Alpha and Composite Reliability for the reflective scales (Berger, 2015). Additionally, p-values and VIF have been used to test for convergent validity of the formative scales (Berger, 2015).

Subsequently, it was possible to start with the analysis of the research model and the testing of the framed hypotheses. The first part of the analysis consists of focusing only on the direct effects of the 6 variables. The effects were assessed through the calculation of beta coefficients, which inform about the influence that an independent variable possesses on a dependent variable. Corresponding p-values are used to assess the level of significance of the measured influences. Bootstrapping was used regarding the treatment of the intervals of confidence.

In the second part of the analysis the effects of the moderator and their interactions with the independent variables have been investigated.

4. Results

The fourth chapter presents the results of the analyses. Chapter 3 has already provided a general description of the survey and its respondents as well as the data gathered, the operationalization, and the plan of analysis. Thus, the results of the reliability analysis examining the internal consistency of the scales will be represented. Moreover, the main output of the factor analysis will be mentioned. Subsequently, the data gathered by the analysis will be presented. Finally, the hypotheses validation will show which of the hypotheses formulated in the second chapter are supported and which are rejected.

4.1 Data Reliability and Validity

(29)

29

Given the fact that all the items used in this research are based on an antecedent study and were previously validated, it is assumed that all the scales and measured variables are reliable. However, the reliability analysis has been performed, in order to check the internal consistency of the scales.

Moreover, a factor analysis has been performed to examine how the factors solution would change and to double-check the reliability of the scales. In order to test the two different scales,

formative and reflective scales, two different approaches for their construct and convergent validity is needed (Hair, Black, Babin & Anderson, 2010).

Reflective scales were measured through Cronbach’s Alpha and also through composite reliability to double-check the reliable measure of the Cronbach’s Alpha (Henseler et al., 2009). The loadings confirmed the convergent validity by scoring more than 0.5 with a p-value under the level of 0.05. According to Coltman et al. (2008) and Fornell and Larcker (1981) AVE has been used to assess the discriminant validity that resulted over 0.5.

For the formative scales, since the correlation amongst the indicator might have a lower score (Rossiter, 2002), it has been necessary to perform a different procedure than the FA and the respective Cronbach’s Alpha. To indicate the validity of the construct, weights have been assessed as well as the threshold that VIF (variance inflation factors) should be below the level of 3.3 (Diamantopoulus and Siguaw, 2006). It is important to highlight that deleting one variable, might lead to a different outcome of the construct. Also important to mention is, that according to Coleman (2008) there is no perfect procedure to determine whether the construct is valid or not. Below first the formative and consequently the reflective approach will be described in detail.

4.1.1. Reflective scales

(30)

30 In this relation, the Cronbach’s Alpha needs to be checked, if an item is deleted from the scale, because this might lead to a changed coefficient alpha. However, the Cronbach’s Alpha in this scale of this study with more than one item was appropriately provided for all the variables.

The results of the reflective scales are presented in Table 1, which shows that the alpha scores are above 0.7. It proves that the indicators for all first-order constructs are appropriated and none of them created confusion for the respondents.

The measurements of the loadings can be resulted from reliability and validity analysis in the evaluation for the overall construction. Hair, Black, Babin & Anderson (2010) mentioned in their researches, that a loading should be above 0.5 with a p-value lower than 0.05. The present study therefore, also presents in Table 1, that all items are sufficiently loading, as all overpassed the threshold of 0.5.

TABLE 1

Construct reliability and convergent validity of reflective scales. Buyer and Supplier

Outer Loading Crombach’s alpha Composite Reliability

(31)

31 Flexibility 1 0.81 Flexibility 2 0.84 Flexibility 3 0.77 Information Exchange 0.78 0.86 Information Exchange 1 0.74 Information Exchange 2 0.78 Information Exchange 3 0.77 Information Exchange 4 0.81 Solidarity 0.83 0.89 Solidarity 1 0.88 0.9020 Solidarity 2 0.90 Solidarity 3 0.80 4.1.2. Formative scales

According to Johnson et al. (1996) and Hulland (1999) formative variables are representing different dimensions of the construct, thus they are not correlating among the other variables as the reflective scales.

Although traditional assessments of convergent and discriminant validity are irrelevant for formative measures and validity is largely based on content validity (Tsang, 2002), there are some measurements that could be provided in this research and also the evidences of the appropriateness of the formative scales.

Furthermore, the convergent validity is measured by checking the p-values and VIF’s of outer weights as calculated from PLS software. The VIF score in reflective scales should not overpass the cut-off point of 10.0. Controversially, as mentioned before, the VIF value should not represent a score higher than 3.3 when it is used to process formative scales (Diamantopoulos & Winklhofer, 2001). The lower the value of the VIF, the higher the appropriateness. Based on inappropriate measures, ‘Acquisition 2’, ‘Transformation 2’ and ‘Transformation 3’ were excluded. However, given the characteristics of a formative construct, excluding measurement items could potentially influence the content of the construct, because all items are related to different dimensions of the construct.

(32)

32 As discussed from Berger (2015) the possibility of the exclusion of the first three items (‘Acquisition 2’, ‘Transformation 2’ and ‘Transformation 3’) does not affect the theoretical domain of the two constructs, which are still captured adequately by the remaining items. As mentioned above, the same is assumed for the exclusion of Explorative L.P 1. However, Exploitative L.P 3 has been held, since it presents the only ‘quality’ performance outcome measure. Thus, it is important to be integrated in the construct. Finally, the remaining other constructs and indicators show appropriateness and the scores are supporting the convergent validity of the formative scales as presented in the table below (Table 2).

TABLE 2

Convergent validity of formative scales. Buyer and Supplier

Outer Weights p-value VIF

(33)

33

4.2. Results of the structural model

The validation of the hypotheses presented in our research, has been calculated through Smart PLS3. The software has been set differently from the default measures: the number of iterations for PLS algorithm was set on 1000 while the number of bootstrap subsample was set on 5000 instead of 300. Bootstrapping has been set based on the percentile of the p-level of significance of 0.1 and 0.05.

To test the hypothesis 4, whether contracting and relational norms can be used as substitutes to enhance RACAP and PACAP, also Smart PLS3 was used for calculations.

Furthermore, Poppo and Zenger (2002) mentioned a method, called “primary support” for complementarity, which enables to test, if complementarity occurs. When relational norms have a positive effect on contracting and vice versa, they are considered complements. However if the effect between those two governance mechanisms is negative, they are considered to be substitutes.

In our research study we have tested (H4), as outlined in Table 3, whether relational norms have a negative effect on contracting and if contracting has a negative effect on relational norms. Our results outline the following relationship: the relation between contracting and relational norms is negative regarding the “buyer-only” database. In the “supplier-only” database the relationship between contracting and relational norms shows no significant effect between the two governance mechanisms.

TABLE 3

Direct effects between contracting and relational norms

Buyer Supplier

(34)

34

Table 4 presents the results that are based on the ‘buyer-only’ database, while table 5 presents the results that are based on the ‘supplier-only’ database. As it can be seen in those two tables, the results of the structural model are supporting most of the hypothesis stated above in chapter 2. In the following the results regarding the ‘buyer-only’ database will be presented, followed by the results regarding the ‘supplier-only’ database.

Considering “only-buyer” database, contracting has a significant positive effect on PACAP (H1a: β= 0.22 and p-value= 0.00) and there are not evidences for an inverted U-shaped relationship (H1b: β= 0.04 and p-value= 0.43). As expected, a positive direct effect between contracting and RACAP (H1c: β= 0.20 and p-value= 0.00) has been found.

Regarding the relationship between relational norms and PACAP a direct positive effect (H2a: β= 0.57 and p-value= 0.00) can be reported, there are also evidences to the expectations of an inverted U-shaped relationship (H2b: β= -0.09 and p-value= 0.02). Relational norms have a positive effect on RACAP (H2c: β= 0.86 and p-value= 0.00) and evidence has been found for an inverted U-shaped relationship (H2d: β= -0.11 and p-value= 0.02).

Regarding, PACAP is positively related to explorative learning performance (H5: β= 0.34 and p-value= 0.00) and RACAP significantly supports the relation with exploitative learning performance (H6: β= 0.66, p-value= 0.00).

Finally, the moderation effect of duration of the relationship (DR) positively influences the effect of contracting on PACAP (H3a: β= 0.15 and p-value= 0.00) as well the effect of contracting on RACAP (H3b: β= 0.09 and p-value= 0.07). While the effect of the moderator, the duration of the relationship, has not significant effect between relational norms and PACAP (H3c: β= -0.01 and p-value= 0.86) and relational norms and RACAP (H3d: β= -0.04 and p-value= 0.63).

TABLE 4

Overview of the structural model and hypothesis in ‘buyer-only’ database

Buyer

Relations. Path-coefficient p-value Hypothesis Approved

Contracting PACAP 0.22 0.00*** H1a yes

-Quadratic PACAP 0.04 0.43 H1b no

Contracting RACAP 0.20 0.00*** H1c yes

(35)

35

-Quadratic PACAP -0.09 0.02** H2b yes

Relational Norms RACAP 0.86 0.00*** H2c yes

-Quadratic RACAP -0.11 0.02** H2d yes

PACAP Explorative L.P 0.34 0.00*** H5 yes

RACAP Exploitative L.P 0.66 0.00*** H6 yes

Contracting*DR PACAP 0.15 0.00*** H3a yes

Contracting*DR RACAP 0.09 0.07* H3b yes

Relational Norms*DR PACAP -0.01 0.86 H3c no

Relational Norms*DR RACAP -0.04 0.63 H3d no

Note: *p ≤ 0.1; **p ≤ 0.05; ***p ≤ 0.001

Regarding the ‘supplier-only’ database, contracting has no significant effect on neither PACAP nor RACAP. Consequently, both tested relationships are not representing the necessary scores to support our hypotheses (H1a and H1b). Contracting on PACAP (H1a) shows: β= -0.02 and p-value= 0.75 and no evidences have been found for an inverted U-shaped relationship (H1b: β= 0.06 and value=0.38), as well as for RACAP there is no significant effect: β= 0.04 and p-value= 0.96 (H1c).

The hypotheses H2a and H2c, which state that relational norms are positively related to PACAP and RACAP are supported, as both effects present a positive and significant effect regarding the supplier database (H2a: β= 0.36 and p-value= 0.00; H2c: β= 0.43 and p-value= 0.00). There are no evidences for an inverted U-shaped relationship for both PACAP (H2b: β= -0.04 and p-value= 0.58) and RACAP (H2d: β= -0.05 and p-value= 0.55).

PACAP is positively related to explorative learning performance (H5: β= 0.33 and p-value= 0.00), while RACAP significantly effects exploitative learning performance (H6: β = 0.53, p-value=0.00).

Finally, contracting has no significant effect on PACAP (H3a: β= 0.02 and p-value= 0.85) and RACAP (H3b: β= 0.04 and p-value= 0.62) when its effect is moderated by the duration of the relationship. Additionally, the moderation of the duration of the relationship has no significant effect on the relation between relational norms and PACAP (H3c: β= 0.02 and p-value= 0.85), while on RACAP the effect is significantly negative (H3d: β= -0.18 and p-value= 0.06).

(36)

36

Overview of the structural model and hypothesis in ‘supplier-only’ database

Supplier

Relations. Path-coefficient p-value Hypothesis Approv.

Contracting PACAP -0.02 0.75 H1a No

-Quadratic PACAP 0.06 0.38 H1b No

Contracting RACAP 0.04 0.96 H1c No

Relational Norms PACAP 0.36 0.00*** H2a Yes

-Quadratic PACAP -0.04 0.58 H2b No

Relational Norms RACAP 0.43 0.00*** H2c Yes

-Quadratic RACAP -0.05 0.55 H2d No

PACAP Explorative L.P 0.33 0.00*** H5 Yes

RACAP Exploitative L.P 0.53 0.00*** H6 Yes

Contracting*DR PACAP 0.02 0.85 H3a No

Contracting*DR RACAP 0.04 0.62 H3b No

Relational Norms*DR PACAP -0.05 0.58 H3c No

Relational Norms*DR RACAP -0.18 0.06* H3d No

Note: *p ≤ 0.1; **p ≤ 0.05; ***p ≤ 0.001

5. Discussion, Implications, Limitations and Conclusion

5.1. Discussion

This paper provides an examination of the role of governance mechanisms on the inter-firm relationship, in particular it focuses on the effect of contracting and relational norms on PACAP and RACAP and the role of the duration of the relationship considered as a moderator. In addition, this research examines whether both governance mechanisms can be used as substitutes at the inter-firm level.

The existing literature reported the positive effect of contractual governance on RACAP (Jansen et al., 2005). In this research however, this mentioned effect has only been found for the buyer perspective while the effect for the supplier perspective has not been found. These findings do not lead to a full comprehension of the role of contracting in the relationship between both partners. In addition, the effect of relational norms has been tested on both sub-dimensions PACAP and RACAP, which was not empirically assessed by prior research. The findings of this research are therefore exploratory and add new findings to the existing literature. In the following the results of each hypothesis will be discussed in detail.

(37)

37 al. (2005) no significant effect of contractual governance was found on PACAP for the supplier database, but controversially, a positive impact for the buyer point of view has been found. In contrast with the findings of Jansen et al. (2005) no significant effect of contracting on RACAP was found for the supplier perspective while there are evidences of the positive effect of contracting on RACAP for the buyer perspective. These results are not completely satisfying the expectation of this study and contradict prior research as done by Jansen et al. (2005). However, it needs to be considered, that this study is based on the relationship between buyer and supplier, while the study of Jansen et al. (2005) focuses on the unit level. The findings of this study underline that both partners might have different interpretations and different perceptions of contracting on PACAP and RACAP. For instance, buyers are giving more importance to the contractual governance in determining RACAP, while for the suppliers contracts seem not to have significant impact on RACAP. Several explanations can be given. First, according to Adler and Borys (1996), buyers that have based their relationship on contracting have defined functions and can feel safer than when there are no protections. This can be caused since innovation and exchange of knowledge could generate opportunism in the relationship between partners. From the supplier point of view, contractual governance might be less suitable and limit the workers’ behaviour, for example as the willpower in the relationship with the buyer might decrease. In fact, according to Wang et al. (2011) less strict and more flexible procedures should increase the performance of the inter-firm innovation activities. Second, since contracting is not effective when the exchanged knowledge between the partner is tacit (Cannon et al., 2000; Inkpen et al., 2004) and since this study has focused on mature relationships, most of the shared knowledge is tacit and that leads contracting to have less importance on RACAP.

ABSORPTIVE CAPACITY (ACAP) The impact of governance mechanisms on the exploitative and explorative capabilities of firms engaging in knowledge sharing alliances.