Substitution or Complementarity: Influence of Industry–University–Research-Institute Cooperation Governance Mechanism on Knowledge Transfer—An Empirical Analysis from China

Abstract

China’s economic growth is transforming from being traditional and factor-driven to being innovation-driven. How to improve the independent innovation ability and build the sustainable competitiveness of enterprises through knowledge transfer in industry–university–research-institute (IUR) cooperation has become an urgent problem to be solved. The obstacles to knowledge transfer in an IUR cooperation include internal and external cooperation risk factors. Improving the governance mechanism of an IUR cooperation, reducing cooperation risks, and promoting knowledge transfer are effective means for overcoming such obstacles in the short term. Interorganizational governance mechanisms include contractual governance and relationship governance. Obvious differences exist in the target functions of the different types of governance mechanisms, and their effects on knowledge transfer also differ. Based on a questionnaire survey on the IUR cooperation innovation of 364 enterprises, this study discusses the impact of the contractual governance mechanism and relationship governance mechanism on knowledge transfer in the IUR cooperation. Different types of contractual governance mechanisms have different effects on knowledge transfer. The contractual coordination mechanism significantly promotes explicit knowledge transfer in an IUR cooperation but has no significant impact on tacit knowledge transfer, whereas the contractual control mechanism significantly promotes explicit knowledge transfer but hinders tacit knowledge transfer. Meanwhile, the relationship governance mechanism has a significant positive impact on explicit knowledge transfer and tacit knowledge transfer. The joint use of the contractual coordination mechanism and relationship governance mechanism can significantly promote explicit knowledge transfer and tacit knowledge transfer, whereas the joint use of the contractual control mechanism and relationship governance mechanism can significantly weaken the two types of knowledge transfer. This study provides not only a theoretical explanation for the dispute over the “complementarity” or “substitution” relationship between the contractual governance mechanism and relationship governance mechanism in knowledge transfer in an IUR cooperation, thereby enriching relationship governance theory and knowledge management theory, but also a reference to the government, enterprises, and universities/scientific institutions participating in an IUR cooperation.

1. Introduction

China’s R&D investment reached CNY 2786.4 billion in 2021, thereby ranking second in the world, with an R&D intensity of 2.44%. The country’s scientific and technological innovation ability improved continuously, enabling China to enter the third echelon of innovative countries in the world. Industry–university–research-institute (IUR) cooperation plays a crucial role in attaining such achievements. According to statistics, at present, more than 80% of China’s major technological innovation projects are completed via IUR cooperation. In the era of the knowledge economy, IUR cooperation and innovation can integrate the advantages of all parties involved, create new knowledge, and develop new technologies. IUR cooperation is not only a magic weapon with which enterprises can improve their technological capacity and gain a sustainable competitive advantage but also an effective means for the government to enhance industrial competitiveness and promote regional economic development [1,2]. IUR cooperation is an important and special form of strategic alliance and the main channel through which enterprises and universities/scientific institutions can obtain complementary resources and share technology and knowledge [3].

While affirming the achievements of IUR cooperation innovation, we must be aware of the existence of various problems that must be solved. The most prominent problem is the poor sustainability and low conversion rate of an IUR cooperation. The conversion rate of scientific and technological achievements is less than 20% in China, which is far lower than the proportion of 40% in developed European and American countries. The main reasons for this low percentage include unstable and unsustainable partnerships, mismatched partners, and failure to give full play to the synergy of the resources and capabilities of all parties involved. In addition, cooperation methods, such as the “short and fast” technique, are not conducive to knowledge sharing and new knowledge creation. Owing to the differences in functional objectives, on the one hand, IUR cooperation organizations are prone to cognitive obstacles in communication and coordination. On the other hand, such organizations face a high moral hazard and adverse selection risks. In the absence of an effective governance mechanism, such obstacles and risks render IUR cooperation an empty statement and hinder enterprises from realizing effective knowledge learning and technical ability improvement. Typical alliance governance mechanisms include contractual governance mechanisms and relationship governance mechanisms, which are interrelated but contradictory. The literature shows that contractual governance and relationship governance have important and complex effects on alliance knowledge transfer and learning [4,5,6]. How to use contractual governance and relationship governance mechanisms scientifically to overcome the endogenous obstacles and risks of an IUR alliance and promote the sustainability of knowledge transfer from universities/scientific institutions to enterprises is of considerable strategic significance for developing enterprises’ independent innovation capability.

As interrelated but different cooperation governance mechanisms, the joint use of contractual governance and relationship governance will inevitably have an interactive effect on interorganizational cooperation. Regarding the results of such an interaction, two contradictory views exist in the literature: substitution and complementarity. The first view states that an alternative relationship exists between contractual governance and relationship governance. Relationship governance means that the expected opportunistic risk is small, which can reduce dependence on contractual governance, whereas stringent contractual governance may damage relationships among organizations [4,7]. When relationship governance among alliance members is weak, strengthening the contractual governance mechanism is necessary to prevent cooperation opportunistic behaviors, whereas when the relationship among cooperation members is satisfactory and opportunistic risks are low, the function of contractual governance may be weakened. Relationship governance and contractual governance are two types of alternative governance mechanisms [8,9]. The second view is that contractual governance and relationship governance have a symbiotic and complementary relationship. Contractual governance clearly stipulates the responsibilities and obligations of the parties involved in an alliance. In addition, contractual governance increases cooperation transparency and promotes learning among partners to develop relationship governance [7,10]. When changes and contradictions in the cooperation exist, relationship governance can compensate for the shortcomings of contractual governance by strengthening bilateral relations and overcoming the rigidity of contracts. In addition, a complete contract can strengthen the trust relationship among organizations by reducing transaction risks [9,11].

For the relationship between contractual governance and relationship governance, some scholars have proposed a third view: the relationship between “substitution” and “complementarity” is not either/or but can coexist simultaneously. The relationship between contractual governance and relationship governance depends on the motivation for signing a contract [12,13]. First, if a contract is not made for the purpose of ensuring strict legal protection, then contractual governance and relationship governance will complement each other and have a complementary relationship. Second, if a trust relationship exists among the cooperation organizations, then they will not stipulate detailed protection clauses in the contract. In this scenario, contractual governance and relationship governance are alternatives. Finally, if distrust exists among the alliance members, then they must replace the trust mechanism with a strong contractual governance mechanism to prevent opportunistic behaviors.

To sum up, many inconsistent views exist on the direct and interactive effects of contractual governance and relationship governance on knowledge transfer. The reason for such differences is that the literature mainly takes the contractual governance mechanism and knowledge transfer as a one-dimensional concept and barely considers the complex relationship between different types of contractual governance mechanisms and different forms of knowledge transfer. We believe that the different types of contractual governance mechanisms are based on different functions and purposes, and the effects of alliance governance naturally differ, which will inevitably have different impacts on interorganizational knowledge transfer. In addition, knowledge transfer can be divided into explicit knowledge transfer and tacit knowledge transfer. The attributes of the two types of knowledge differ, and the influence mechanism of knowledge transfer also differs considerably. In view of this argument, in this study, we take IUR alliance as our research object and adopt the large sample empirical research method to further analyze the direct and interactive effects of different types of contractual governance mechanisms and relationship governance mechanisms on different types of knowledge transfer. Our study can provide not only a reasonable explanation for the contradictory views in studies on alliance governance but also an important supplement to theory of IUR cooperation. In addition, our study can provide valuable theoretical guidance to Chinese enterprises for improving their independent innovation capability through IUR cooperation.

2. Theoretical Background and Hypothesis Development

2.1. Alliance Governance Mechanism

An innovation alliance describes a long-term cooperation agreement between enterprises and partners based on the strategic goal of innovation. On the one hand, enterprises hope to obtain the maximum innovation value from the alliance. On the other hand, the difference of alliance partners can lead to contradictions and conflicts. An effective alliance governance mechanism can promote the innovation alliance, improve innovation capability, realize the synergy of resources, and enhance the knowledge transfer performance of the IUR cooperation [14]. The governance mechanism was originally used in political economy and government management research and simply understood as control, manipulation, and guidance. The essence of governance lies in the coordination of various relations in a system [15]. Based on transaction cost theory and social transaction theory, the literature mainly divides alliance governance into contractual governance and relationship governance [4,16,17]. Contractual governance forms rigid management means through legally effective agreements or contracts, whereas relationship governance is a flexible method based on alliance trust and capability [16].

Scholars often divide the dimensions of contractual governance into contractual coordination and contractual control from the perspective of root cause differences [11,18]. Contractual coordination is based on human cognitive limitations. Deviations in understanding caused by the different cognitive limitations of members on the same objective subject must be reduced through formal contractual determination and coordination procedures to ensure the efficient transfer of innovation alliance knowledge [11]. Meanwhile, contractual control is based on differences in the interest motivation of members. Opportunistic behaviors driven by interest require a formal contract to clarify the control methods and punishment measures [11]. Influenced by various factors, such as education, intellectual status, and work experience, people’s cognition of the objective world has inherent limitations. This cognitive limitation can lead easily to understanding deviations and communication obstacles in interorganizational cooperation. Determining the effective coordination procedures and methods in advance through formal contracts, overcoming members’ cooperation obstacles, reducing the cooperation complexity, and improving the alliance efficiency are necessary, which are contractual coordination mechanisms. Typical contractual coordination mechanisms include clarifying the expected cooperation objectives, division of the responsibilities of the members, technology and resource investment, and the project reporting system in the form of contracts [4,10,19].

The contractual control mechanism is rooted in the differences in the interest motivation of alliance members [19]. As rational economic individuals driven by different interest motives, alliance members may have opportunistic behavioral tendencies. Therefore, formulating effective control and punishment measures in advance in the form of contracts for potential opportunistic behaviors is necessary. This function is known as the contractual control mechanism, which mainly includes the inspection and supervision of partners, intellectual property protection, the unilateral termination of contractual agreements, dispute arbitration penalties for breach of contract, and legal proceedings [4,11,12].

With regard to the constituent dimensions of relationship governance, scholars have proposed a variety of division methods from different perspectives [8,20]. Among them, the trust dimension is the most frequently mentioned. Therefore, in this study, we select trust as the subdivision dimension of relationship governance. Trust refers to the creation of an atmosphere of mutual trust for the strong supervision and promotion of knowledge transfer by improving transfer willingness and reducing transfer difficulties [21]. The literature mainly defines trust from two perspectives, the first of which is positive expectations. Trust is an individual’s degree of confidence in others and willingness to act based on others’ words, behaviors, and decisions [22,23]. According to Coote et al. (2003), trust means that one party has confidence in the honesty, reliability, and integrity of another party [24]. Hall et al. (2004) defined trust as the belief that others will not engage in opportunistic and self-interest behaviors [25]. The second perspective is willingness to be vulnerable. Trust is the willingness to accept the risks brought about by the type and degree of dependence on a given relationship [26,27]. Williams (2001) and Baer et al. (2018) defined trust as people’s willingness to rely on others’ behaviors, which involve opportunistic risks [28,29]. Trust exists in many levels, including interpersonal trust and interorganizational trust. In this study, we take the organization members of an IUR cooperation as the research object. Therefore, we refer to trust as interorganizational trust and define it as a partner’s willingness to accept its vulnerability from its willingness to fulfill its commitment to the other organization and to not take advantage of its positive expectations of the other party’s weaknesses.

Based on this definition, we divide the governance mechanism in an IUR alliance into contractual governance and relationship governance and further divide the former into two dimensions, namely contractual coordination and contractual control, and the latter into the trust dimension.

2.2. Effect of Contractual Governance Mechanism on Knowledge Transfer

2.2.1. Influence of Contractual Coordination on Knowledge Transfer

Obvious differences exist between universities/scientific institutions and enterprises in value objectives and knowledge structures. Universities/scientific institutions pursue the frontiers and universality of knowledge and advocate the academic atmosphere of free exploration, whereas enterprises focus on the market benefits of technology and emphasize the low cost and rapid effect of technology development [30]. Such differences make both parties prone to cognitive bias, communication barriers, and behavioral disorders in cooperation. The key to reduce performance risks in an IUR cooperation is to formulate an effective contractual coordination mechanism in advance. The contractual coordination mechanism plays different roles in the transfer of explicit and tacit knowledge in an IUR cooperation. Explicit knowledge transfer mainly involves tangible objects as carriers, such as project reports, product descriptions, technical documents, databases, and computer programs. In addition, the content and value of explicit knowledge are easy to measure and determine by coding. Therefore, explicit knowledge transfer is often the result of an economic exchange. As the embodiment of the economic exchange relationship, contractual coordination is closely related to explicit knowledge transfer.

From the perspective of knowledge and technology input–output, a complete contractual coordination mechanism generally describes the expected output objectives of the cooperation, such as the content of the technical achievement, product performance indicators, and appearance quality. Such technical objectives are formed based on the existing knowledge base reserves, technical prediction insights, and market development experiences of the alliance members, which belong to the important content of the knowledge of the member organizations. The more detailed the technical objectives of a contract, the richer the knowledge transferred, and the more the opportunities of an enterprise to obtain explicit knowledge. In addition, to ensure the smooth realization of the goals of an IUR cooperation, creating explicit agreements in advance on the technical resource investment of the alliance members, especially universities/scientific institutions, is necessary, including patent intellectual property rights, technical development materials, and process design schemes that should be shared, and the R&D effort that should be exerted [30]. Such a contractual coordination mechanism can establish an explicit knowledge-sharing platform based on economic transaction relations for all the partners and become an important condition for enterprises to obtain coded knowledge and research institutions.

From the perspective of knowledge exchange, a complete contractual coordination mechanism includes reporting systems for the progress of cooperation projects and timely written notices regarding behaviors that deviate from the original objectives [31,32] to build a formal explicit knowledge exchange network for the operation of an IUR cooperation. A complex contractual coordination mechanism generally stipulates the specific technology required to perform a cooperative task and requires the task executor to communicate and report the information content, use, and implementation effect of the technology regularly or irregularly through formal communication channels such as meetings, emails, and documents [4,33] to provide the enterprise with an opportunity to acquire and absorb a large amount of explicit knowledge. Barthélemy and Quélin (2006) pointed out that a relatively complete contractual coordination mechanism typically requires alliance members to standardize the phased performance indicators they achieved [34]. In essence, this performance indicator is also an exchange of knowledge symbols, which increases the degree of the explicit knowledge exchange when the partners perform relevant tasks [35,36]. In addition, when one partner encounters an unexpected condition in the project implementation process and must readjust the original target, technical route, and design scheme, it shall send a corresponding written notice to the other party in time. This type of written notice typically contains new technical objectives and problem solutions, which are part of the partners’ knowledge resources, and can provide an effective means for the enterprise to acquire explicit knowledge from the other party. Thus, a complete contractual coordination mechanism can create an effective explicit knowledge exchange network for the members of an IUR alliance.

In contrast to explicit knowledge transfer, tacit knowledge transfer is typically carried out in the form of face-to-face communication, stories, metaphors, and so on. Tacit knowledge is not easy to express, and the boundaries are difficult to define; thus, it lacks a knowledge value evaluation basis. Therefore, partners are unable to formulate clear written statements and conduct accurate valuations of the input of this type of knowledge in the form of contracts [37,38]. In addition, effectively supervising and controlling tacit knowledge transfer with the contractual coordination mechanism are difficult. According to Adler (2001), formal contracts have little effect on knowledge acquisition and sharing among organizations [39] because the most valuable knowledge that an organization can acquire is tacit knowledge [40]. Based on this argument, we present the following assumptions:

H1a: 

The contractual coordination mechanism has a significant positive effect on the explicit knowledge transfer performance of an IUR alliance.

H1b: 

The contractual coordination mechanism has no significant effect on the tacit knowledge transfer performance of an IUR alliance.

2.2.2. Influence of Contractual Control on Knowledge Transfer

In an IUR cooperation, enterprises and universities/scientific institutions have different interest demands. Enterprises pursue their economic interests, whereas universities/scientific institutions value their “academic reputation.” However, as rational economic entities, universities/scientific institutions cannot ignore the economic benefits of cooperation. An IUR cooperation may have opportunistic risks owing to differences in interests [41,42,43]. To control the relationship risks within an acceptable range, enterprises must effectively restrict opportunistic behaviors through the contractual control mechanism. The contractual control mechanism can reduce opportunistic behavior risks in two ways. First, contractual control can increase the cost of enterprises’ self-interest behaviors, change the payment structure, and reduce the probability of the game result of opportunistic behaviors [44]. Second, contractual control clearly stipulates the allowable or impermissible scope of the behavior of all the parties involved, improves the transparency of the cooperative relationship, and is conducive to reducing the possibility of opportunistic behaviors [32]. In reality, the contractual control mechanism is embodied in protective contractual terms mainly involving audit and supervision rights [19], intellectual property protection [12], confidentiality agreements [31], dispute arbitration [31], and legal litigation [19]. In an IUR cooperation, the contractual control mechanism can indirectly affect the knowledge transfer effect of the cooperation by restricting opportunistic behaviors.

The codification of explicit knowledge gives the knowledge activities of the members of an IUR alliance a contractual supervision and control basis. When the contractual control mechanism is weak, a contract will not clearly define the scope of the explicit knowledge exchange among and legal behaviors of the alliance members. A formal, detailed, and clear contractual guidance framework for determining the explicit knowledge that can be shared and the behaviors that fall within the legal scope is lacking. This contractual ambiguity can easily induce the excessive knowledge acquisition and improper knowledge utilization of alliance members. At the same time, a low degree of contractual control can lead to a lack of effective supervision on partners’ behaviors, which can easily lead to a lack of credit. One party takes advantage of the information disadvantage of the other party not fully observing its behavior, goes beyond the reasonable boundaries, and engages in opportunistic behaviors, thereby resulting in the loss of the other party’s knowledge assets [45]. Once universities/scientific institutions perceive the high risk of intellectual property rights in the cooperation, they will reduce the potential risks by limiting knowledge sharing, thereby resulting in the poor effect of the explicit knowledge transfer. In addition, a low degree of contractual control makes sanctions for the opportunistic behaviors of alliance members insufficient. Meager penalties for a breach of contract, loose contractual termination, and legal litigation conditions will reduce the cost of the opportunistic behaviors of partners and increase the risks of a cooperative relationship. Therefore, low-level contractual control can easily induce opportunistic behaviors in a cooperation through fuzzy transactions, ineffective monitoring, and ineffective sanctions. To avoid unilateral and disproportionate loss in their core knowledge and skills, alliance members may deliberately hide valuable knowledge and abilities [46]. As a written contract can clearly define only the sharing, use, licensing, and transfer of explicit knowledge, the contract mainly has a direct control over explicit knowledge-related behaviors.

If the contractual control degree in an IUR cooperation alliance is moderate, then the contractual agreements on the knowledge that can and cannot be shared in the cooperation process and boundaries of legitimate behaviors are clear. In addition, detailed punishment provisions and legal litigation provisions are stated for breaches of contract. Moderate contractual control can serve as an effective deterrent to partners’ “loopholes” and knowledge theft or disclosure. Once knowledge infringers violate the provisions in the contract, they will face high punishment costs, thereby making their opportunistic behaviors unprofitable. Thus, such a contractual control mechanism can eliminate infringers’ profit motivation at the source and form an effective binding force [47]. An appropriate contractual control level sends a signal to the cooperation members that the potential intellectual property risks are controllable. Therefore, it is conducive to reduce the uncertainties of a transaction, enhance communication and understanding among allies, and stimulate members’ behavioral motivation to engage in explicit knowledge sharing and exchange.

However, if the contractual control degree in an IUR cooperation is too strong, provisions on the supervision and punishment of alliance members may be highly complex and detailed, which can easily create an atmosphere of mutual suspicion among the IUR cooperation members and may be regarded by the partners as a manifestation of distrust. When a relationship of suspicion and distrust exists among the members of an IUR cooperation, they may take noncooperative action, such as deliberately retaining their valuable explicit knowledge or treating the knowledge transfer between the organizations passively. According to Keller et al. (2021) [20], excessive contractual control will lead to a “moral hazard” (partners will not try their best to transfer knowledge) or “adverse selection” (partners will not demonstrate their ability). Such potential risks will affect the willingness of both parties to transfer explicit knowledge to a certain extent.

Based on the above analysis, we propose the following hypothesis:

H2a: 

The contractual control mechanism has a significant inverted U-shaped effect on the explicit knowledge transfer performance of an IUR cooperation. That is, the explicit knowledge transferred from universities/scientific institutions to enterprises first increases then decreases with the enhancement of the contractual control mechanism.

From the above analysis, it can be seen that the codification of explicit knowledge enables a formal contract to control related behaviors directly. However, tacit knowledge is fuzzy and difficult to encode and express. At the same time, tacit knowledge is rooted in the human brain and difficult to detect. Specifying the sharing, use, and transfer of tacit knowledge to written forms in formal contracts is difficult; thus, directly restricting related behaviors is likewise difficult [48]. Owing to such difficulties, contractual control enhancement will not promote tacit knowledge transfer [23]. By contrast, with the continuous strengthening of contractual controls, the sense of distrust and suspicion of both parties may increase gradually. Ghoshal and Moran (1996) argued that taking rational control measures means that no mutual trust exists between partners [49], so using external authoritative monitoring forces to ensure the legitimacy of allies’ behaviors is necessary. In other words, excessive contractual control hinders the development of satisfactory relations between partners [8]. Tacit knowledge sharing is based mainly on the principle of voluntariness and satisfactory interpersonal relationships, which are often manifested as altruistic knowledge behaviors. Exceedingly strict contractual control can lead to a decline in the quality of the relationship between partners, which may cause the two parties to reduce or stop their tacit knowledge sharing and transfer. Based on this idea, we present the following assumption:

H2b: 

The contractual control mechanism has a significant negative effect on the tacit knowledge transfer performance of an IUR alliance.

2.3. Effect of Relationship Governance Mechanism on Knowledge Transfer

In the literature on interorganizational cooperation, trust, as an important factor affecting cooperation success, has attracted the attention of scholars. Trust means that one partner has positive expectations of the honesty, reliability, and integrity of the other party; believes that the other party will not engage in opportunistic and self-interest behaviors; and is willing to be vulnerable [24,25]. Trust is a subjective psychological state held in a specific social environment and a series of behavioral manifestations arising from the environment. In an IUR cooperation, trust is an important prerequisite for interorganizational knowledge transfer. The mechanism is reflected mainly in two aspects, that is, to improve willingness for knowledge transfer and to reduce knowledge transfer difficulties [50].

First, from the perspective of knowledge transfer willingness, when trust is high among cooperative members, one partner has a positive expectation that the other party will consciously uphold the contract without taking advantage of the other party’s weaknesses to engage in opportunistic behaviors. At the same time, risk of loss of self-worth is perceived to be reduced accordingly. This positive psychological state will make the knowledge sender willing to bear a certain degree of controllable relationship risks, reduce the control and defense behaviors of the other party, and encourage the adoption of an open attitude to transfer valuable knowledge. Trust can enhance the knowledge transfer willingness of IUR partners by reducing perceived opportunistic risks and improving future potential cooperation benefits, which is consistent with the findings of Inkpen and Tsang (2005) [51]. In addition, trust can improve partners’ willingness to send and receive knowledge. Second, from the perspective of knowledge transfer difficulties, trust can reduce cooperative members’ anxiety about opportunistic behaviors [48,52], thereby urging them to increase the frequency and means of communication and improve the breadth and depth of communication to objectively expand the amount of information that can be transmitted and reduce knowledge transfer difficulties and transaction costs [30,53]. According to Zaheer, Mcevily, and Perrone (1998) and Majuri (2022), trust reduces people’s requirements for short-term interests [54,55], makes them start from long-term interests, and creates a satisfactory information communication environment to directly or indirectly reduce knowledge transfer difficulties.

In summary, the trust between IUR partners can have a positive effect on enterprises’ knowledge acquisition by improving knowledge transfer willingness and reducing knowledge transfer difficulties. This view was confirmed by many studies. For example, through an empirical survey, Maurer (2010) found that a high degree of trust is conducive to enhancing motivation for knowledge sharing and acquisition between enterprises and external organizations [56].

On the basis of clarifying the relationship between trust and knowledge transfer, knowledge transfer was further divided into explicit knowledge transfer and tacit knowledge transfer. Trust has different degrees of influence on the two types of knowledge transfer. Tacit knowledge is based on specific situations, rooted in the human brain, difficult to express clearly, and based on experience. Moreover, disseminating tacit knowledge through coded written forms is difficult and depends mainly on face-to-face communication [55,57,58]. Such knowledge constitutes sticky knowledge that is difficult to transfer in an IUR cooperation [4]. Holste and Fields (2010) summarized several factors that hinder the transfer of tacit knowledge based on previous studies, including willingness to share and use tacit knowledge, awareness of the tacit knowledge of others [59], the difficulty of expressing tacit knowledge, and the constraints of the situational transplantation of tacit knowledge. The key to overcome such obstacles is to build trust [60]. Tacit knowledge is highly personalized knowledge based on personal cognition and experience and is the most valuable type of knowledge. Understanding the breadth and depth of this type of knowledge is difficult, except for the knowledge owner. Therefore, the sharing and transfer of this type of knowledge depend completely on the principle of individual voluntariness. When a high degree of trust exists among cooperative members, one partner will no longer worry about the loss of valuable knowledge and have high expectations of future cooperation benefits, hoping to establish a long-term cooperative relationship. To show its sincerity, a party will respond quickly and positively to the other party’s tacit knowledge sharing requirements or take the initiative to share its unique tacit knowledge with the other party to demonstrate its subjective willingness to share. In addition, trust can objectively help improve tacit knowledge transfer efficiency. Trust fosters the intimate social relationship between partners [61], and the close, informal communication brought about by this intimate relationship is an important tacit knowledge transfer method. First, frequent private contact enables cooperative members to gain a comprehensive understanding of one another’s knowledge structure, especially highly personalized experience and knowhow, which is conducive to identifying partners’ valuable tacit knowledge. Second, close, social communication can enable both parties to establish and develop common experiences, values, mental models, and cognitive structures [62]; enhance the understanding of each other’s tacit knowledge embedding situation; and overcome the situation transplantation obstacles of tacit knowledge transfer. Moreover, informal interaction includes oral communication, body language, facial expression, onsite observation, and other personalized ways, which are suitable for the transmission of tacit knowledge. Furthermore, trust can promote the objective transfer of tacit knowledge by enhancing social and emotional ties between partners.

Compared with tacit knowledge, the influence of trust on explicit knowledge is weaker. Explicit knowledge has the characteristics of observability, expressibility, and measurability; thus, it can be used as a commodity for economic exchange. For example, explicit knowledge such as patents, software, and design drawings can be classified as tradable knowledge commodities. At the same time, because explicit knowledge can be coded, the responsibilities, obligations, and risks of both parties in the knowledge transaction are clearly stipulated in their contract, which determines that the sharing of explicit knowledge is often based on an economic relationship and contractual responsibility. Li, Poppo, and Zhou (2010) pointed out that most transactions take explicit knowledge transfer as the premise for establishing a reliable and effective economic exchange relationship [4]. Although trust can encourage cooperative members to share explicit knowledge beyond that indicated in the provisions of the contract, it is not a necessary condition for such knowledge transfer. In addition, from the perspective of knowledge transfer channels, explicit knowledge is transmitted mainly through formal and coded means, and the informal, intimate communication brought about by trust is not necessary [63,64].

Based on the above discussion, we present the following assumptions:

H3a: 

The relationship governance mechanism has a significant positive effect on the explicit knowledge transfer performance of an IUR alliance.

H3b: 

The relationship governance mechanism has a significant positive effect on the tacit knowledge transfer performance of an IUR alliance.

2.4. Interactive Effect of Contractual Governance Mechanism and Relationship Governance Mechanism on Knowledge Transfer

2.4.1. Interaction between Contractual Coordination and Relationship Governance

Contractual governance and relationship governance are not completely independent factors but are interrelated. When both mechanisms exist at the same time, they will have a certain impact on knowledge transfer in an IUR cooperation. Several scholars showed that contractual governance can strengthen relationship governance in transactions, and relationship governance is conducive to improving the content of contractual governance and cooperation stability. A complementary relationship exists between contractual governance and relationship governance [9,10]. Regarding knowledge transfer in an IUR cooperation, we believe that the interaction between the contractual governance mechanism and relationship governance mechanism differs.

From the perspective of contractual coordination, it is believed that a perfect contractual coordination mechanism can enhance relationships among organizations to promote cooperative knowledge transfer. Owing to their lack of trust, enterprises and universities/scientific institutions are unaware of each other’s technical capabilities; thus, giving full play to the synergy of technological innovation is difficult. Without the intervention of the contractual coordination mechanism, alliance members will be unable to establish formal and informal knowledge exchange networks, thereby resulting in poor information and a loose situation of their affairs, which will hinder the transfer of knowledge from universities/scientific institutions to enterprises [9]. With the help of external authorities, a complete contractual coordination mechanism will urge alliance members to build a formal knowledge communication network to deepen their understanding of one another continuously in the process of formal communication, gradually reach a consensus on issues and stable relations [9], and further develop formal communication into informal communication based on personal friendship to continuously strengthen relationship governance and promote knowledge transfer in the IUR alliance.

Meanwhile, the relationship governance mechanism is an important condition for the improvement of the contractual coordination mechanism. A perfect contractual coordination mechanism is often accompanied by a satisfactory relationship governance mechanism, which can jointly promote knowledge transfer in an IUR alliance. The conclusion of detailed and comprehensive contractual coordination terms not only requires alliance members’ subjective need to design, conceive, and formulate contractual coordination terms but also requires them to have a satisfactory contractual design ability. When the governance level in an IUR alliance is low, the cooperative relationship among the members is relatively loose, and the level of cooperation is low. It needs only to formulate simple and rough contractual coordination terms to meet the needs of the cooperation. With the gradual improvement of the governance mechanism of an IUR cooperative relationship, the cooperative relationship among the alliance members will become tightknit, and the breadth and depth of the cooperation content will expand, which can be a strong motivation for building a complex alliance value creation system [7]. Therefore, increased comprehensive, complex, and detailed contractual coordination terms are necessary to adapt to it. In addition, under a perfect relationship governance mechanism, enterprises can deepen their understanding of their resources and capabilities through close interactions with academic and research institutions and gradually gain a tacit understanding of technical cooperation behaviors. With the continuous deepening of cognitive experiences and tacit understanding of behaviors, they will become fixed in the form of formal contracts to supplement and improve the content of the original contracts in terms of contractual coordination mechanisms such as resource supply and distribution, the tasks and responsibilities of all the parties involved, and project communication mechanisms to establish an effective platform for knowledge transfer.

The simultaneous use of contractual coordination and relationship governance will complement each other and produce an expansion effect of one plus one greater than two on explicit and tacit knowledge transfer in an IUR alliance. Based on the above analysis, we present the following assumptions:

H4a: 

The contractual coordination mechanism and relationship governance mechanism exist synchronously and can strengthen the explicit knowledge transfer performance of an IUR alliance.

H4b: 

The contractual coordination mechanism and relationship governance mechanism exist synchronously and can strengthen the tacit knowledge transfer performance of an IUR alliance.

2.4.2. Interaction between Contractual Control and Relationship Governance

In contrast to the contractual coordination mechanism, the contractual control mechanism prevents opportunistic behaviors in a cooperation and promotes knowledge transfer through rigid means, such as monitoring and sanctions. The relationship governance mechanism can eliminate the opportunistic behavior tendencies of partners and improve knowledge transfer willingness through soft means, such as satisfactory relationship capital and mutual benefits. When two organizations interact, mutual transfer risks will be reduced.

On the one hand, the relationship governance mechanism will reduce the dependence of IUR cooperation partners on contractual control [65]. When the relationship governance mechanism is imperfect, enterprises will likely have doubts about whether their partners illegally possess and steal knowledge, fail to fulfill their due obligations of knowledge transfer, or free ride in the technology R&D. To prevent the loss of valuable knowledge, cooperative members may adopt a strict contractual control mechanism, force partners to share and transfer knowledge according to the contract, and control the potential risks caused by the lack of a relationship governance mechanism through strong external binding on opportunistic behaviors. By contrast, if the relationship governance mechanism is complete, enterprises’ expectations of the opportunistic behaviors of their partners will be reduced considerably. Moreover, partners will not engage in illegal knowledge possession or theft. At the same time, based on the principle of mutual benefit, sincere cooperation and moral behaviors will arise. In an IUR cooperation with a perfect relationship governance mechanism, enterprises and universities/scientific institutions will no longer fear opportunistic behaviors [52], gain a certain psychological tacit understanding of their legal behavior boundaries, and engage in cooperation in an orderly manner. At this time, detailed contractual control terms are redundant. Therefore, as an efficient and low-cost self-implementation mechanism, trust can reduce the necessity of contractual monitoring and serve as an effective substitute for the contractual control mechanism [66], thereby weakening the impact of contractual control on knowledge transfer.

On the other hand, overly strict contractual control can easily hinder the development of relationship governance. The original intention of the contractual control mechanism is to restrain potential opportunistic behavior risks, and its completeness increases as risk perception increases. If the IUR cooperation pays too much attention to the role of contractual control, then it will send a negative signal to the partners. That is, it will believe that the partners will have a tendency to engage in opportunistic behaviors and likely use every opportunity to illegally erode the knowledge assets in the cooperation; thus, it will require strong supervision and punishment measures to control it. This pessimistic expectation of partners’ behaviors may destroy the foundation of the interorganizational relationship governance mechanism and make partners feel mistrusted [67]. In response, the other party will also defend itself against the violation of the cooperative party by strengthening its contractual control means and reducing knowledge exchange and sharing, which will further worsen the relationship governance and lead the cooperative members into a vicious cycle of mutual suspicion and revenge [68,69]. Therefore, the excessive use of the contractual control mechanism will indicate and trigger a sense of mistrust among the partners, which can easily harm the existing trust and further inhibit knowledge exchange and cooperative innovation among the members [12,70,71].

In summary, the contractual control mechanism and relationship governance of an IUR alliance are mainly manifested in the substitution effect. The simultaneous use of the two mechanisms will weaken their role in knowledge transfer in the alliance. Accordingly, we present the following assumptions:

H5a: 

The synchronous existence of the contractual control mechanism and relationship governance mechanism will weaken the explicit knowledge transfer performance of an IUR alliance.

H5b: 

The synchronous existence of the contractual control mechanism and relationship governance mechanism will weaken the tacit knowledge transfer performance of an IUR alliance.

3. Methodology

3.1. Sample Selection and Data

To test the theoretical hypotheses, we used the questionnaire survey method to obtain empirical data. As enterprises are important to the transformation of the innovation achievements of an IUR cooperation, and their innovation output can effectively represent the innovation performance of the IUR cooperation, we selected manufacturing enterprises engaging in IUR cooperation in the Yangtze River Delta as our research object. Manufacturing enterprises are the mainstay in China’s economic development but are facing the severe challenges of industrial upgrading and structural adjustment. Thus, achieving sustainable development through innovation is urgent. In addition, the Yangtze River Delta is one of the regions with the most intensive manufacturing industry and the most developed economy in China. The region is also home to numerous universities/scientific institutions with a strong scientific research thrust and has a satisfactory environment for IUR cooperation. Therefore, our sample selection has universal practical significance for research on IUR cooperative innovation [72].

This research has conducted a lot of searching, consulting, sorting and summarizing on the relevant research on the topics of contract governance, relationship governance and knowledge transfer in the existing literature, and has deeply understood and grasped the essence and meaning of the key concepts. On this basis, we further collect, analyze and compare the existing measurement scales of relevant variables. Considering that researchers and managers have different cognitive structures and language habits, managers are often confused by the academic terms commonly used by researchers. In order to reduce the academic nature of the questionnaire and make it easy for managers to understand, it is necessary to fully communicate with managers before the survey and further modify the questionnaire. Next, we selected six manufacturing enterprises engaging in IUR cooperation in Hangzhou to participate in the questionnaire survey. We revised the questionnaire based on the feedback results and relevant expert opinions. To ensure the validity and reliability of the variable measurements, we selected the maturity scale used in the literature. The questionnaire design process followed standard procedures and methods. We rearranged and placed the scale items in different subject modules to prevent the participants from guessing the hypothetical logic or potential constructs to reduce common method bias [73]. At the same time, the questionnaire reduced the impact of systematic deviation by emphasizing anonymity and confidentiality, avoiding sensitive problems, and using neutral words and reverse items.

We distributed a total of 900 questionnaires and recovered 532. Next, we excluded 168 questionnaires that did not meet the requirements, such as having missing information, inconsistent answers, identical questionnaire options, and too many “uncertain” opinions. Finally, we obtained 364 valid questionnaires, with an effective recovery rate of 40.44%. The descriptive statistics of the sample enterprises are shown in

Table 1. Sample distribution characteristics.

Variable	Category	Count	Percentage	Variable	Category	Count	Percentage
enterprise scale	≤20 employees	6	1.648%	ownership	state owned	92	25.275%
	20~300	135	37.088%		private enterprises	215	59.066%
	300~1000	171	46.978%		foreign funded	57	15.659%
	1000~5000	40	10.989%	industry	electronic/communication equipment	46	12.637%
	≥5000	12	3.297%		biomedical	30	8.242%
enterprise age	≤1 year	7	1.923%		chemical	41	11.264%
	1~3	38	10.440%		automobile/transportation equipment	32	8.791%
	3~5	151	41.484%		electrical/machinery manufacturing	58	15.934%
	≥5 years	168	46.154%		textile and clothing	35	9.615%
R&D intensity	≤1%	43	11.813%		IT service	47	12.912%
	1~2%	139	38.187%		instrumentation	26	7.143%
	2~3%	165	45.330%		food/beverage/cigarette	31	8.516%
	≥3%	17	4.670%		others	18	4.945%

. From the perspective of enterprise ownership, among the enterprises, 92 were state-owned, accounting for 25.275% of the total; 215 were private enterprises, accounting for 59.066% of the total; and 57 were foreign funded, accounting for 15.659% of the total sample. From the perspective of enterprise scale, 6 enterprises had 20 employees or fewer, accounting for 1.648% of the total; 135 enterprises had 20–300 employees, accounting for 37.088% of the total; 171 enterprises had 300–1000 employees, accounting for 46.978% of the total; 40 enterprises had 1000–5000 employees, accounting for 10.989% of the total; and 12 enterprises had more than 5000 employees, accounting for 3.297% of the total sample. From the perspective of enterprise age, 7 enterprises were less than 1 year, accounting for 1.923% of the total; 38 enterprises were between the ages of 1 and 3 years, accounting for 10.440% of the total; 151 enterprises were between the ages of 3 and 5 years, accounting for 41.484% of the total; and 168 enterprises were older than 5 years, accounting for 46.154% of the total sample. From the perspective of the average R&D intensity in the past 3 years, 43 enterprises were below 1%, accounting for 11.813% of the total; 139 enterprises were between 1% and 2%, accounting for 38.187% of the total; 165 enterprises were between 2% and 3%, accounting for 45.330% of the total; and 17 enterprises were above 3%, accounting for 4.670% of the total sample. From the perspective of industry categories, 46 enterprises belonged to the electronic/communication equipment industry, accounting for 12.637% of the total; 30 enterprises belonged to the biomedical industry, accounting for 8.242% of the total; 41 enterprises belonged to the chemical industry, accounting for 11.264% of the total; 32 enterprises belonged to the automobile/transportation equipment industry, accounting for 8.791% of the total; 58 enterprises belonged to the electrical/machinery manufacturing industry, accounting for 15.934% of the total; 35 enterprises belonged to the textile and clothing industry, accounting for 9.615% of the total; 47 enterprises belonged to the IT service industry, accounting for 12.912% of the total; 26 enterprises belonged to the instrumentation industry, accounting for 7.143% of the total; 31 enterprises belonged to the food/beverage/cigarette industry, accounting for 8.516% of the total; and 18 enterprises belonged to other industries, accounting for 4.945% of the total sample.

3.2. Variable Measurement

According to the research hypotheses and theoretical model, the variables to be measured included contractual coordination mechanism, contractual control mechanism, relationship governance mechanism, explicit knowledge transfer, and tacit knowledge transfer. Quantifying the variables with objective indicators is difficult; thus, we used the subjective scoring method of Likert scales for the evaluation, ranging from 1 (“highly disagree” or “not very detailed”) to 7 (“highly agree” or “very detailed”). A score of 3 indicated neutrality.

3.2.1. Dependent Variables

According to the characteristics of knowledge, our explained variable, namely knowledge transfer, can be divided into explicit knowledge transfer and tacit knowledge transfer. The main differences between explicit knowledge and tacit knowledge are their expression forms and transfer channels. Such differences are researchers’ basic basis for measuring the two types of knowledge transfer. From the perspective of expression forms, explicit knowledge generally exists in the form of formalization and writing, whereas tacit knowledge is expressed mainly as experience, intuitions, and skills. Many studies designed knowledge transfer measurement terms according to such differences [4,74]. We believe that the essential differences between explicit knowledge and tacit knowledge lie in their knowledge characteristics and forms of expression, which lead to differences in their communication channels. Therefore, the design of the scale items of the two types of knowledge transfer according to their expression forms should effectively reflect the characteristics of the two types of knowledge. Our research objects were universities/scientific institutions and enterprises. The type of knowledge transferred between the entities was typically technical knowledge, which is similar to the research of Sherwood and Covin (2008) [74]. At the same time, the scale items were also based on the knowledge expression forms. Therefore, we mainly used the scale of Sherwood and Covin (2008) [74] and referred to the research of Dhanaraj et al. (2004) and Li, Poppo, and Zhou (2010) [4], combined with the enterprise interviews, to formulate the 10-scale items of knowledge transfer [75]. The items included “We successfully learned and mastered written technical description knowledge from our partners” and “We successfully learned and mastered unwritten technical fault resolution experience from our partners.” Among the items, five were for explicit knowledge transfer, with a Cronbach’s α coefficient of 0.911, and five were for tacit knowledge transfer, with a Cronbach’s α coefficient of 0.909.

3.2.2. Independent Variables

The first independent variable was contractual governance mechanism, which included two dimensions: contractual coordination mechanism and contractual control mechanism. Two main methods were used in the literature to measure contractual governance mechanism. First, various contractual terms listed in advance are scored in the form of “yes” or “no”; then, the scores are summed up, that is, contractual governance mechanism is quantified as $\sum Cont\_g{m}_i$, where $Cont\_g{m}_i=1$ indicates that an i contractual clause exists, and $Cont\_g{m}_i=0$ indicates that such a clause does not exist. Lui and Ngo [44], Lumineau [7], Reuer and Ariño [31], and Lumineau and Henderson [11] adopted the aforementioned method. Second, with a Likert scale, we scored and measured the details of the various contractual terms to reflect the complexity and strength of contractual governance mechanism. Owing to China’s imperfect legal environment, the general contractual template is often used in an IUR cooperation. On the surface, all the terms were complete, but the actual content was empty, which may be in vain during implementation. If the first measurement method is used, then reflecting the actual situation by judging the strength of contractual governance mechanism only by whether specific terms exist would be difficult. Therefore, the second method is more in line with the reality of China’s IUR cooperation [4,76,77,78].

Based on the literature, Lumineau and Henderson [11] clearly presented measurement items for contractual control mechanism and contractual coordination mechanism. The measurement items of contractual control mechanism included audit and supervision rights, protection clauses, third-party control and supervision, punishment clauses, and termination and dissolution clauses. Meanwhile, the measurement items of contractual coordination mechanism included the allocation of functions and responsibilities, renewal duration and conditions, the operating conditions of the task redistribution of participants, strategic coordination, and dispute resolution provisions. On the one hand, we drew on lessons from the literature. On the other hand, combined with China’s special situation, we referred to the technology contractual law of the People’s Republic of China and made localization adjustments in the content and wording based on enterprise discussions and expert consultations. Finally, we formed the measurement scale for contractual coordination mechanism and contractual control mechanism, which included 14 items such as “The contract signed between us and our partners specifies the division of the tasks and responsibilities of both parties” and “The contract signed between us and our partners specifies the information and technology confidentiality agreement.” Among the items, seven were for contractual coordination mechanism, with a Cronbach’s α coefficient of 0.874, and seven were for contractual control mechanism, with a Cronbach’s α coefficient of 0.882.

Another independent variable was relationship governance mechanism, which used trust as the proxy variable. Trust in an IUR cooperation can be classified as interorganizational trust, and its measurement method for organizations differs from that for individuals. Although the literature on trust measurement is very rich, most studies used interpersonal trust as the research object, and research on interorganizational trust is lacking. In designing the trust variable, on the one hand, we should pay attention to drawing lessons from the literature on interorganizational trust. On the other hand, we should adapt to the Chinese situation as much as possible, such as referring to the Chinese literature. According to the above principles, for our trust scale design, we mainly drew on lessons from Zaheer et al. [16] and Lui and Ngo [44]. According to our definition, trust refers to the willingness and behavior of one partner to fulfill its commitment to the other organization, not take advantage of its positive expectations of the other’s weaknesses, and to accept its own vulnerability. Trust mainly reflects the confidence of one party on the reliability and loyalty of the other party and emphasizes the goodwill dimension. Therefore, our use of this definition to learn from the goodwill trust measurement indicators of Lui and Ngo [44] was suitable. The scale included five items, such as “We believe that our partners sincerely care about our interests” and “We believe that our partners will try their best to help us when we encounter difficulties,” with a Cronbach’s α coefficient of 0.889.

3.2.3. Control Variables

To ensure the reliability of the conclusions, we selected a series of characteristic variables as our control variables. Li, Poppo, and Zhou [4] pointed out that the larger the enterprise size and the older the enterprise, the stronger the learning ability. Large enterprises and enterprises with a long history have abundant resources for learning and innovation and thus can absorb external knowledge more effectively than small enterprises [79]. A commonly used enterprise scale index in the literature is the number of enterprise employees [4,79,80,81,82]. Based on the literature, we used the logarithm of the number of employees to represent the enterprise scale and the number of years of the establishment of the enterprise to represent the enterprise age. The cooperation experience between enterprises and universities/scientific institutions is also an important factor affecting knowledge transfer. On the one hand, previous cooperation experience can enable both parties to have a certain understanding of each other’s knowledge and technical structure to build an effective learning foundation and improve the effect of the enterprise’s knowledge transfer. On the other hand, previous cooperation experience can help deepen mutual trust, increase knowledge sharing, and promote knowledge transfer. Kotabe et al. [83] confirmed that a positive relationship exists between cooperation experience and knowledge sharing. Extending the research of Li, Poppo, and Zhou [4], we used previous cooperation relationship between the enterprises and universities/scientific institutions to measure cooperation experience. In addition, we selected industry as a control variable, including 10 industry categories, that is, the electronic/communication equipment industry, biomedicine industry, chemical industry, automobile/transportation equipment industry, electrical/mechanical manufacturing industry, textile and garment industry, IT service industry, instrument industry, and the food/beverage/cigarette industry and other industries [72].

3.3. Model Setting

Based on the above analysis, in order to test the impact of different governance mechanisms of IUR alliance on knowledge transfer, we set the following model:

$$\begin{array}{cc}\hfill Kno{w}_{TransE}=& {\beta }_0+{\beta }_{1}Contrac{t}_{coord}+{\beta }_{2}Contrac{t}_{contr}+{\beta }_{3}Relation\hfill \\ & +{\beta }_{4}Contrac{t}_{contr}{}^2+{\beta }_{5}Relation\ast Contrac{t}_{coord}\hfill \\ & +{\beta }_{6}Relation\ast Contrac{t}_{contr}+{\beta }_{i}Control+ϵ\hfill \end{array}$$

(1)

$$\begin{array}{cc}\hfill Kno{w}_{TransT}=& {\beta }_0+{\beta }_{1}Contrac{t}_{coord}+{\beta }_{2}Contrac{t}_{contr}+{\beta }_{3}Relation\hfill \\ & +{\beta }_{4}Relation\ast Contrac{t}_{coord}+{\beta }_{5}Relation\ast Contrac{t}_{contr}\hfill \\ & +{\beta }_{i}Control+ϵ\hfill \end{array}$$

(2)

where $Kno{w}_{TransE}$ and $Kno{w}_{TransT}$ refer to explicit knowledge transfer and tacit knowledge transfer, $Contrac{t}_{coord}$ and $Contrac{t}_{contr}$ refer to contractual coordination mechanism and contractual control mechanism, and $Relation$ is the relationship governance mechanism.

4. Results

4.1. Descriptive Statistics and Correlation Analysis

The calculation results of the descriptive statistics (mean and standard deviation) of the main variables and Pearson correlation coefficients are presented in

Table 2. Descriptive statistics and Pearson correlation coefficients.

Variable	Mean	S.D.	1	2	3	4	5	6	7
1. Explicit knowledge transfer	5.499	1.084	(0.532)
2. Tacit knowledge transfer	5.542	1.103	0.302 *	(0.603)
3. Contractual coordination	5.024	1.124	0.286 **	0.093	(0.524)
4. Contractual control	4.987	1.104	0.216 **	−0.271 **	0.178 **	(0.517)
5. Relationship governance	5.131	0.973	0.155 **	0.318 **	0.236 **	−0.374 **	(0.587)
6. Enterprise scale	3.854	0.905	0.103	0.068	0.153	0.101	0.120
7. Enterprise years	7.946	4.558	0.132	0.077	0.031	0.063	0.021	0.033
8. Cooperation experience	3.510	1.719	0.089	0.112	0.136	0.066	0.154 *	0.012	0.043

Note: ** p < 0.05, * p < 0.10. AVE in brackets.

. The results show that the maximum absolute value of the Pearson correlation coefficient between relationship governance mechanism and contractual control mechanism is 0.374, which is far less than 0.75. Thus, we can basically rule out the possibility of multicollinearity. From the correlation significance of the main variables, we can see that a significant correlation exists between contractual coordination mechanism, contractual control mechanism, and relationship governance mechanism and explicit knowledge transfer and tacit knowledge transfer, thereby indicating a possible interaction. To further eliminate the problem of multicollinearity, we calculate the variance inflation factor (VIF). The results indicate that the VIF values of the variables are between 1 and 10, thereby ensuring that the interference of multicollinearity can be ignored. At the same time, the preliminary results reveal that the theoretical model and assumptions are reasonable, and the internal mechanism between the variables can be tested further.

4.2. Hypothesis Testing

We conduct hierarchical regression analysis to test the research hypotheses empirically. First, we construct a basic model containing only the control variables to test the impact of the control variables, namely enterprise scale, enterprise age, cooperation experience, and industry, on knowledge transfer. Second, we introduce the independent variables based on the basic model to test the influence of the control variables and independent variables on knowledge transfer. Finally, we introduce the interaction term of the independent variables to test the impact of the interaction term of relationship governance mechanism and contractual governance mechanism on knowledge transfer. To avoid the problem of multicollinearity between the variables in the regression equation, we centralize all the variables involving quadratic terms and interaction effects.

First, we test the influence mechanism of explicit knowledge transfer, and the detailed results are shown in

Table 3. Results of hierarchical regression analysis (dependent variable: explicit knowledge transfer).

Variable	Model 1	Model 2	Model 3
Enterprise scale	−0.125 * (0.050)	−0.129 * (0.047)	−0.121 * (0.045)
Enterprise years	−0.079 (0.004)	−0.079 (0.004)	−0.086 (0.005)
Cooperation experience	0.133 * (0.025)	0.121 * (0.024)	0.121 * (0.022)
Industry dummy	control	control	control
Relationship governance		0.145 * (0.078)	0.076 (0.064)
Contractual coordination		0.196 *** (0.063)	0.184 *** (0.060)
Contractual control		0.226 *** (0.066)	0.232 *** (0.069)
Contractual control 2		0.037 (0.054)	−0.066 (0.054)
Relationship governance × Contractual coordination			0.327 *** (0.078)
Relationship governance × Contractual control			0.313 *** (0.073)
R2	0.271	0.373	0.414
Adj R2	0.254	0.358	0.406
ΔR2	—	0.102	0.041
F-value	35.390 ***	22.069 ***	19.103 ***

Note: *** p < 0.01, * p < 0.10. Contractual control. ² is the secondary term of the variable "contractual control".

. We can see from the results that the F-values of the three models are significant at the 0.010 level, thereby indicating that each regression equation is valid. From the perspective of the model explanatory power, the R² of each model increases compared with that of the previous basic model, and the R² growth of each model is 0.102 and 0.041, thereby indicating that the latter model has a higher explanatory power on the changes in the dependent variables compared with the previous model. In addition, the VIF of each latent variable is between 1.347 and 1.981, which is far lower than the judgment standard of 10, thereby indicating that no multicollinearity exists between the latent variables.

Model 1 tests the impact of enterprise scale, enterprise age, and cooperation experience on explicit knowledge transfer. The results show that a negative relationship exists between enterprise scale and explicit knowledge transfer ($\beta =-0.125, p<0.100$), thereby indicating that the larger the enterprise scale, the lower the dependence on external knowledge, and the less explicit the knowledge transferred. Cooperative experience has a significant positive impact on explicit knowledge transfer ($\beta =0.133, p<0.100$), and the longer the cooperation experience, the more explicit the knowledge obtained by the enterprises from the universities/scientific institutions.

To test H1a, H2a, and H3a, we add the independent variables to model 2, such as relationship governance mechanism, contractual coordination mechanism, contractual control mechanism, and the quadratic term of contractual control mechanism, on the basis of the control variables. The regression results indicate that contractual coordination mechanism has a significant positive impact on explicit knowledge transfer ($\beta =0.196, p<0.010$), thereby indicating that the greater the strength of the contractual coordination mechanism, the more explicit the knowledge obtained by the enterprises. Thus, H1a is supported. Contractual control mechanism has a significant positive impact on explicit knowledge transfer ($\beta =0.226, p<0.010$), but no significant correlation exists between the quadratic term and explicit knowledge transfer ($\beta =0.037, n.s.$), thereby indicating that the stronger the contractual control mechanism, the better the explicit knowledge transfer effect of the enterprise. The relationship between the two variables is an inverted U-shape. Thus, H2a is unsupported. In addition, the positive impact of relationship governance mechanism on explicit knowledge transfer is significant at the 0.1 level ($\beta =0.145, p<0.100$), and its influence coefficient and significance level are lower than those of contractual coordination mechanism and contractual control mechanism, thereby indicating that the higher the trust level between enterprises and universities/scientific institutions, the easier to obtain explicit knowledge. However, the influence degree of relationship governance mechanism is weaker than that of contractual governance mechanism. Thus, H3a is supported by the statistical test.

On the basis of model 2, we further add the interaction terms of relationship governance mechanism and contractual coordination mechanism and relationship governance mechanism and contractual control mechanism to model 3 to verify H4a and H5a. We can see from Table 2 that the interaction term of relationship governance mechanism and contractual coordination mechanism has a positive impact on explicit knowledge transfer ($\beta =0.327, p<0.010$), thereby indicating that the joint use of relationship governance mechanism and contractual coordination mechanism will promote explicit knowledge transfer, and a complementary relationship exists between them. Thus, H4a is empirically supported. By contrast, the interaction term of relationship governance mechanism and contractual control mechanism has a significant negative impact on explicit knowledge transfer ($\beta =0.313, p<0.010$), which means that when relationship governance mechanism and contractual control mechanism exist at the same time, the effect of explicit knowledge transfer will decrease, and a substitution relationship will exist between them; thus, H5a is supported.

In the empirical test of the impact mechanism of tacit knowledge transfer, we conduct hierarchical regression analysis, similar to that for explicit knowledge transfer. The regression analysis results of the impact mechanism of tacit knowledge transfer are presented in

Table 4. Results of hierarchical regression analysis (dependent variable: tacit knowledge transfer).

Variable	Model 4	Model 5	Model 6
Enterprise scale	−0.114 * (0.052)	−0.133 * (0.047)	−0.097 * (0.044)
Enterprise years	−0.103 (0.005)	−0.093 (0.005)	−0.088 (0.007)
Cooperation experience	0.118 * (0.024)	0.117 * ( 0.021)	0.109 * (0.020)
Industry dummy	control	control	control
Relationship governance		0.246 *** (0.077)	0.231 *** (0.064)
Contractual coordination		0.084 (0.064)	−0.026 (0.061)
Contractual control		−0.278 *** (0.065)	−0.261 *** (0.062)
Relationship governance × Contractual coordination			0.247 *** (0.075)
Relationship governance × Contractual control			0.253 *** (0.071)
R2	0.261	0.339	0.416
Adj R2	0.244	0.327	0.403
ΔR2	—	0.078	0.077
F-value	21.343 ***	27.108 ***	29.823 ***

Note: *** p < 0.01, * p < 0.10.

. We can see from Table 4 that the F-values of the three models of tacit knowledge transfer are significant at the 0.01 level, thereby indicating that all the regression equations are valid. From the perspective of the model explanatory power, the R² of each model increases compared with that of the previous basic model, and the R² growth of each model is 0.078 and 0.077, thereby indicating that the latter model has a higher explanatory power on the changes in the dependent variables than the previous model. In addition, the VIF of each latent variable is between 1.127 and 1.886, which is far lower than the judgment standard of 10, thereby indicating that no multicollinearity exists between the latent variables.

Model 4 first tests the influence of the control variables, such as enterprise scale and cooperation experience, on tacit knowledge transfer. From the regression results in Table 4, we can see that enterprise scale has a negative impact on tacit knowledge transfer at the 0.1 level ($\beta =-0.114, p<0.100$), thereby indicating that the larger the scale of the enterprise, the less the acquisition of external tacit knowledge. In addition, cooperation experience has a significant positive effect on tacit knowledge transfer ($\beta =0.118, p<0.100$), thereby indicating that previous cooperation experience is conducive to tacit knowledge transfer between enterprises and universities/scientific institutions.

According to the regression analysis results in Table 4, contractual coordination mechanism has no significant impact on tacit knowledge transfer ($\beta =0.084,n.s.$), and a negative correlation exists between contractual control mechanism and tacit knowledge transfer ($\beta =-0.278, p<0.010$), thereby indicating that the stronger the contractual control mechanism of the IUR cooperation, the more unfavorable the acquisition of tacit knowledge to the enterprise. Thus, H2b is empirically supported. In addition, the positive impact of relationship governance mechanism on tacit knowledge transfer is significant ($\beta =0.246, p<0.010$), thereby indicating that a high level of trust between enterprises and universities/scientific institutions can significantly promote the transfer of tacit knowledge. Thus, H3b passes the statistical test.

We introduce the interaction terms of relationship governance mechanism and contractual coordination mechanism and relationship governance mechanism and contractual control mechanism to test H4b and H5b. The results show that the interaction term of relationship governance mechanism and contractual coordination mechanism has a significant positive effect on tacit knowledge transfer ($\beta =0.247, p<0.010$), which means that the simultaneous existence of relationship governance mechanism and contractual coordination mechanism will promote the acquisition of tacit knowledge, and the two types of mechanisms have a complementary effect; thus, H4b is supported. In addition, the interaction term of relationship governance mechanism and contractual control mechanism has a significant negative impact on tacit knowledge transfer at the 0.010 level $(\beta =0.253, p<0.010$), thereby indicating that if relationship governance mechanism and contractual control mechanism are used together, then the effect of tacit knowledge transfer will be reduced, and a substitution effect will exist between them. Thus, H5b passes the empirical test.

5. Discussion, Implications, and Limitations

5.1. Conclusions and Theoretical Implications

In view of the contradictory views on contractual governance mechanism, relationship governance mechanism, and knowledge transfer in the literature, we deeply analyze the direct and interactive effects of the different dimensions of contractual governance mechanism and relationship governance mechanism on the two types of knowledge transfer in an IUR alliance from the perspective of a subdivision dimension using the large sample empirical research method. Our research conclusions are listed below.

(1) The contractual coordination mechanism, which is based on an economic transaction relationship, is conducive to explicit knowledge transfer in an IUR cooperation but has no significant impact on tacit knowledge transfer. Through the contractual coordination mechanism, IUR cooperation members can establish a formal knowledge sharing and exchange platform, which can provide an effective environment to enterprises to learn and absorb the knowledge of universities/scientific institutions. However, this type of contractual mechanism, which is based on written forms, is applicable only to the transfer of coded explicit knowledge and may not play a role in the transfer of tacit knowledge.

(2) The contractual control mechanism can accelerate explicit knowledge transfer in an IUR cooperation but will hinder tacit knowledge transfer. As an external authority, the higher the degree of contractual control, the stronger the deterrent to the members of the IUR cooperation, and the lower the potential risks of opportunistic behaviors. Under the guarantee of contractual control, from their confidence in the security of intellectual property rights, partners will be highly willing to share their knowledge and achievements. However, because the contractual control mechanism is limited to the protection of explicit knowledge, it can promote only the transfer of explicit knowledge and cannot increase the transfer of tacit knowledge. By contrast, strict contractual control will lead to suspicion and mistrust among partners, which is not conducive to the transfer of tacit knowledge.

(3) The relationship governance mechanism has a significant positive impact on explicit knowledge transfer and tacit knowledge transfer, but its impact on tacit knowledge transfer is greater than its impact on explicit knowledge transfer. According to social exchange theory, having a high degree of trust will enable enterprises to have positive expectations of the behaviors of partners and confidence in their own future benefits. Based on the principle of mutual benefit, partners will also engage in good faith cooperative behaviors, such as actively sharing their knowledge achievements with each other. In addition, trust can enable the cooperative parties to establish frequent and various means of interaction, which can objectively accelerate the transfer of explicit knowledge and tacit knowledge. However, the intimate relationship and informal communication channels developed by partners on the basis of trust are suitable for tacit knowledge transfer based on the principle of voluntariness but have little effect on explicit knowledge transfer characterized by economic transactions.

(4) The joint use of the contractual coordination mechanism and relationship governance mechanism increases the transfer of explicit knowledge and tacit knowledge. The contractual governance mechanism and relationship governance mechanism represent different governance mechanisms, and their simultaneous existence will have a certain interactive effect on knowledge transfer. A perfect contractual coordination mechanism can establish a formal interaction platform for IUR partners to gradually develop a common code of conduct and mutual trust in the formal interaction and further establish informal communication channels [9,10]. At the same time, a high level of trust will enable partners to understand each other’s technical capabilities and knowledge structure to improve the contractual coordination mechanism continuously. When the contractual coordination mechanism and relationship governance mechanism exist at the same time, their role can enhance the other, thereby making the effect of explicit knowledge transfer and tacit knowledge transfer obvious.

(5) The joint use of the contractual control mechanism and relationship governance mechanism reduces the transfer of explicit knowledge and tacit knowledge. A high degree of relationship governance will make it easy for IUR partners to have a satisfactory psychological tacit understanding based on positive expectations of their partner’s behaviors, thereby forming an effective self-implementation mechanism in the cooperation and making the external contractual control redundant. In addition, strict contractual control can easily be regarded as a manifestation of the distrust of the partners, leading them to respond with uncooperative defense behaviors, such as hiding their knowledge, not sharing, hindering knowledge transfer, and so on. Therefore, the simultaneous existence of the contractual control mechanism and relationship governance mechanism will reduce the interaction between them, which will have an adverse impact on the two types of knowledge transfer.

5.2. Practical Implications

According to the needs of different types of knowledge transfer in an IUR alliance, the appropriate contractual governance mechanism should be adopted. To successfully transfer explicit knowledge, such as patents, process drawings, and technical descriptions, from universities/scientific institutions, enterprises should improve the two mechanisms of contractual coordination and contractual control simultaneously, which can not only build an effective knowledge exchange platform but also strengthen the knowledge sharing motivation of allies to realize the smooth transfer of explicit knowledge. For the learning of tacit knowledge, such as R&D creativity, experimental experience, and technical knowhow, improving the contractual coordination mechanism is futile, and strengthening the contractual control mechanism will likely have a negative impact. Therefore, when facing the needs of different types of knowledge transfer, enterprises should carefully use contractual governance means, especially the contractual control mechanism, and choose the appropriate level. In addition, enterprises should not only ensure a certain degree of contractual completeness to facilitate the transfer of explicit knowledge but also not be too strict to avoid hindering the free exchange of tacit knowledge.

For the needs of different types of knowledge transfer in an IUR alliance, the appropriate relationship governance mechanism should also be chosen. If learning the tacit knowledge of universities/scientific institutions is the main goal of the enterprise, then relationship capital investment should be the strategic direction of the enterprise’s cooperation, such as spending time and money to create channels, environments, and opportunities for informal communication between the two sides; considering the interests and needs of the universities/scientific institutions in the cooperation; and encouraging the employees of both sides to develop private friendships. Such friendly behaviors are conducive to the cultivation and development of trust among organizations and create a favorable ground for tacit knowledge transfer. As the role of the relationship governance mechanism in promoting explicit knowledge transfer is relatively weak, if enterprises want to obtain explicit knowledge from researchers, then they should mainly improve their contractual governance mechanism, supplemented by the development of the relationship governance mechanism.

The contractual coordination mechanism and relationship governance mechanism reinforce and complement each other. In an IUR alliance with a low level of trust, enterprises can gradually enhance trust among the partners by formulating complex and detailed cooperation procedures and improving the contractual coordination mechanism. At the same time, in the process of developing trust, alliance members can continuously improve the contractual coordination function and jointly promote knowledge transfer in the alliance. Differing from the contractual coordination mechanism, the contractual control and relationship governance mechanisms have a weak substitution effect. To achieve optimal knowledge transfer performance, enterprises should pay attention to selecting the appropriate contractual control level, which can not only effectively inhibit opportunistic behavior risks and promote explicit knowledge transfer but also carefully avoid damaging the trust among organizations and hindering tacit knowledge transfer.

5.3. Limitations and Future Research Directions

After the theoretical analysis and empirical tests, we obtained some meaningful conclusions, which have certain theoretical value and practical significance. However, our study also has some limitations, which must be improved in future research. First, we regarded the relationship governance mechanism as a one-dimensional concept. However, some studies showed that the relationship governance mechanism is a multidimensional construct. Different relational governance mechanisms may have different effects on knowledge transfer, and their joint use with different contractual governance mechanisms may also have different interactive effects on knowledge transfer. Future research can consider further refining the relationship governance mechanism and discussing the impact of different types of relationship governance mechanisms on knowledge transfer and the effect of the interaction between different types of relationship governance mechanisms and different types of contractual governance mechanisms on knowledge transfer. Second, the IUR cooperative organizations we examined included only two participants. Many IUR alliances are composed of multiple participants, involving complex contractual governance mechanisms, relationship governance mechanisms, and knowledge transfer. Future research can be based on the IUR cooperation of multiple participants as the research object and further explore the contractual governance mechanism, relationship governance mechanism, and knowledge transfer relationship between multiple participants from the perspective of transaction cost theory, relationship theory, innovation network theory, and knowledge management theory.

Finally, we collected data from only 364 participants, and our sample size was relatively limited. Future research can consider increasing the sample size and collecting sample data within a large range to obtain universal research conclusions. Furthermore, we did not verify, analyze, and compare the sample data based on region, industry, and ownership type. Thus, future research can further refine the research conclusions from such aspects.