Prior research has shown that the level of communication has an on the learning outcomes In R&D alliances and, subsequently, on performance. The level of communication depends on Issues such as the systems that enhance knowledge sharing and the factors that deter knowledge flows, such as tactless of knowledge and (deliberate) barriers set up to Limit knowledge transfer.
Innovations and value can only be created through sufficiently open knowledge sharing. But on the other hand, too open knowledge sharing might outrun the value gained by innovations if the knowledge lost would leave the knowledge sharing firm vulnerable. Yet. There is still a lack of understanding on which factors are responsible of the level of communication and to what extent. In particular, it has often been forgotten that HARM-related mechanisms are not only functional within firms, but also between firms.
Thus, in this study, the HARM-related mechanisms that contribute to both of hose areas are examined by using empirical quantitative data gathered from 83 Finnish R&D intensive firms. Theory and Hypotheses Strong HARM system for knowledge sharing has been extensively researched in different disciplines, but especially in the inter-firm context (Doze and Hammer, 1997; Mowers et al. , 1 996;Sahara and George, 2002) and in many cases, knowledge sharing has been discussed in terms of innovation (e. G. , Hooch and Trot, 1999; Jaunted, 2005).
Knowledge sharing between firms has often been thought of Including firms that specialize in speedy and efficient knowledge creation and transfer (Gout and Gander, 1992). Danger and Parke (2006) discuss knowledge mobility as the ease of knowledge sharing and transfer In collaborative networks. Without knowledge sharing, the potential residing In the varying knowledge bases of partners cannot be realized, which would diminish the Innovativeness and success of the collaborative endeavourer (Danger and Parke, 2006).
Thus, knowledge sharing should not be Limited by bureaucracy or too tight formal arrangements, but formal structures should support efficient knowledge exchange (see, e. G. , Miller and Savannah, 2006: O'Neill and Day, 2007: Ridge, 2005 on organizational factors affecting knowledge haring). Cohen and Leviathan (1990) argue that the ability to exploit external knowledge resources is a critical component of innovative capabilities, and that in individuals play an important role.
While the processes that lead to the creation of new knowledge are quite well tackled, the ways to govern employees in aim to support knowledge sharing are not so well understood. Considering that the employees of a firm are the ones that hold a lot of relevant knowledge (although some of it can also be found in the routines, practices, technologies, and documents f the firm), it seems quite natural that the HARM system and practices should be a vehicle in adjusting the levels of knowledge flows.
If such practices are very confusing, the employees of a firm cannot be sure what information is free to be discussed in collaboration, or how and when it can be disseminated. In such situations, they may deter knowledge flows ? Just to be on the safe side, or be unmotivated to share knowledge (Gang©, 2009). Doze and Hammer (1998) have noted that, especially in the initial phases, the context of the collaboration seldom encourages cooperation. The people involved will likely find themselves in an unfamiliar situation in which they have no clear frame (Kelly et al. , 2002).
In inter-firm collaboration, it has been noted that firms that feel their background knowledge is properly protected by contracts, 'Pros or non-disclosure agreements may be more willing to share such knowledge with a partner. In similar vein, when employees are not sure what can be shared in collaboration, they may be reluctant to do so. However, knowledge sharing is as important internally to achieve innovation as it is in inter-firm collaboration. Employees' motivation is another important issue affecting knowledge sharing propensity both internally and in inter-firm collaboration.
Motivation can be enhanced by a strong HARM system and clear messages or diminished by unclear messages and a lack of leadership (Gang©, 2009; Ridge, 2005). The process of finding order, forming collective patterns of interpretation, and creating a frame of reference is critical to the early stages of the collaboration because it reduces ambiguity, legitimates collaboration, and facilitates knowledge sharing (Kelly et al. 2002). Thus, clarity in the conduct of knowledge sharing is necessary for suitable levels of knowledge sharing.
Kelley attribution theory suggests that people form attributions about cause-effect relationships in situations depending on the degree of distinctiveness (which means that the target is highly observable), consistency (target behaves the same way in different situations), and consensus (other targets share their views on the situation) (Kelley, 1967). Applying this to HARM systems, Bowen and Castoff (2004) note that in strong HARM systems, messages regarding what is appropriate behavior are communicated (via HARM practices) to employees in an unambiguous, consistent, and consensual way.
These actors linked to strong HARM systems are next described in more detail. Distinctiveness builds on four dimensions. First, the visibility of HARM defines the degree to which HARM practices are salient and readily observable. This attribute allows employees to perceive and understand the meaning of HARM practices. Second, understandability refers to the way in which employees categorize the information gathered from each HARM practice. Third, the legitimacy of authority is related to the perceived power of HARM in the organization. Finally, relevance is present when HARM practices are linked to an important goal.
Consistency, in turn, is a sum of three attributes. Instrumentality refers to the perceived cause-effect relationship between degree to which expected behaviors are rewarded by the HARM system. Consistent HARM messages refer to the degree to which HARM practices express similar and consistent messages in terms of time and throughout the organization. Finally, consensus consists of agreement among principal HARM decision makers (the degree to which employees perceive agreement among HARM top decision makers) and fairness, namely, the perceived Justice in the use of HARM practices.
All of these areas would be in line in order to produce wanted knowledge sharing outcomes. Clear guidelines given by the managers influence the tendency of employees to communicate with partner organizations. In a firm with an open culture, communication is encouraged and, thus, distinctiveness, consistency, and consensus should improve the level of communication between the company and its R;D partner as well. Thus, the following is hypothesized: HARM for knowledge protection Not only the confusing HARM systems within firms, but also other HARM-related mechanisms may come into play when regarding barriers to knowledge flows.
Intellectual property protection and the protection of knowledge have been extensively studied when it comes to mechanisms such as intellectual property rights (APRs; e. G. , Cohen et al. , 2000; Davis, 2004; Heartfelt et al. , 2006). However, not that many studies have dealt with the role that HARM has in knowledge protection (for exceptions, see, e. G. , Heralding-Lausanne and Phenylalanine,2007; Hannah, 2005; Vaughn et al. , 1997). Nevertheless, the management of human resources matters in collaborative R;D ? again, because the firm's employees are the ones that communicate with the partner firm's employees.
The employees at the interface of two collaborating firms can be seen as both gatekeepers and recipients of partner know-how (see also Cheeseburger, 2003, on employee identification of and access to external 'P). This also means that the employees of the partner will act in similar roles. The gatekeeper role of employees may thus become quite notable. With suitable HARM practices, like giving clear instructions regarding intellectual property protection, enhancing employees' commitment to the firm, or paying attention to hiring and firing practices (e. G. Reminding leaving employees of their obligations to pep the former employer's trade secrets to themselves) several problems can be avoided. As Vaughn et al. (1997, 112) note, "control over human resources is a primary means for protecting intellectual capital", and such HARM practices that increase the loyalty of employees may be very relevant. Immobility of human resources and commitment created through HARM mechanisms protect the firm's core knowledge and can, thus, be used to define HARM protection (Heralding-Lausanne and Phenylalanine, 2007). HARM practices can be augmented with the means of labor legislation.
Long-term employment contracts (with sanctions attached to early aperture), non-competition contracts preventing working for competitors or in a competing business for a certain time after the employment relationship ends (only eligible for a very cogent reason and with certain limitations on time, for instance, as enacted by the law), or non-disclosure agreements can be utilized to keep employees Lausanne and Phenylalanine (2007) consider the legal loyalty responsibility of employees and drafting of non-disclosure agreements together with contracts used for preventing key personnel from moving from one company to another as the means of labor legislation to control knowledge losses related to employees. Labor legislation, with the institutional and quite strict air attached to it, may be slightly controversial (see, e. G. , Hannah, 2005, on the different sides related to secrecy required from employees), but it can also be an efficient way of slowing down knowledge flows.
Further, both of these knowledge protection mechanisms can be accompanied with tackiness of knowledge, its embeddings in personnel, routines, and practices. The type of knowledge has an effect on how easily it is transferred to a partner. Certain knowledge to be used commonly in collaborative R;D activities s represented in documents and other explicit forms what Gout and Gander (1992) call "information". The other type of knowledge is what they call "know-how' or tacit knowledge (Plainly, 1966). Tacit knowledge is sticky, complex, and difficult to codify (Nelson and Winter, 1982) as it is embedded in the individuals of an organization (He-man and Nickering,2004). Typically, it is the tacit knowledge that is of the most valuable kind for individual firms and the collaborative endeavourer.
Thus, from the knowledge protection point of view, companies may not put too much time and effort n transforming tacit knowledge to codified or explicit forms, even if that would be beneficial for knowledge sharing. Further, it is very likely that certain tacit knowledge is simply not transformable (see, e. G. , Nelson and Winter, 1982). Gander and Gout (1995) identify five constructs that measure the degree to which a capability can be easily communicated and understood and, thus, how well tackiness acts as a protection mechanism and protects the firm's core knowledge. These constructs are testability, product absorbability, complexity, justifiability, and system dependency Gander and Gout, 1995).
Summing up the discussion, it seems that there are certain challenges in terms of utilizing HARM-related protection mechanisms the same way 'Pros or contracts might be used: 'Pros and contracts can quite easily be used to shift the emphasis from preventing others from accessing certain knowledge to enable its safe transfer. HARM related protection mechanisms, especially tackiness, have built in them the inherent difficulty of transferring knowledge. Also, as noted above, labor legislation is characterized by the need to protect the employer ? and their knowledge base. HARM practices related to protection are perhaps the easiest ones to gain control in this respect, but it is also evident that if such a protective approach is emphasized and if the means are really strong, knowledge flows are likely to diminish.
In line with the above stated, we expect to see following relationships to emerge: Framework to be examined Although we have formulated our hypotheses in a manner suggesting that clear HARM practices within a firm enable knowledge sharing in R&D collaboration, and that HARM-related means of protection limit it, we have also noticed that these issues are not straightforward. Thus, we formulate the following framework to assist us in the Discussion and Conclusions The aim of this study was to increase understanding on the effects that HARM related mechanisms may have on the levels of communication in R&D alliances and, thus, to further research on the roles and use of HARM-related mechanisms in knowledge sharing and knowledge protection. The findings suggest that the stronger the knowledge sharing related HARM mechanisms are, the higher the level of communication is. In particular, distinctiveness in the HARM system of an individual firm has influence on their knowledge sharing activity.
This can be explained by the fact that a distinctive HARM system allows the employees to observe and understand credible HARM messages which encourages the employees to share knowledge both within the firm and in inter-film collaboration. In R&D collaboration, the explicitness of the HARM system gives ground rules for employees to share knowledge with their partners. This is very important as the key persons in the interfaces might not dare to share all the needed knowledge ? especially if there are no supporting non-disclosure agreements, or if they otherwise were not sure whether knowledge haring is encouraged or not. On the other hand, no significant relationships exist between HARM-related protection mechanisms and the level of communication.
This may be because the effects of HARM protection mechanisms may be somewhat controversial (others increasing the tendency not to communicate with outside organizations, others diminishing it). In fact, although the relationships were not statistically significant, we can see that the protection mechanisms have mixed influences: some (tackiness and HARM protection) show negative and others (labor legislation) positive signs when the relationships are examined. This could be due to the fact that labor legislation can also be seen as producing ground rules and a safer environment for knowledge sharing in collaboration, whereas tackiness and HARM protection inherently diminish knowledge flows.
Also, regarding collaborative endeavourers, it may be that the protective side of HARM is deliberately downsized: the firms may rather rely on 'Pros or contracts for protection and try to make sure that HARM practices do not reduce the levels of communication. We believe that examining these issues with different ? especially larger ? datasets might prove to be more revealing. Considering both theoretical and empirical discussion presented in this paper, our study suggests that if a firm wishes to preserve its competitiveness, it needs to control knowledge transfer to its partners in collaborative R&D (or at least control how the knowledge can be used in the partnering firm).
After all, it has been noted in prior research that value capture is, in the end, individual activity ? even if value creation would be carried out Jointly. Our empirical evidence suggests, however, that HARM-related protection may not be the mechanisms that lower the level of communication. Nevertheless, at the value creation phase an adequate level of communication needs to be achieved in any case. Here, the internal HARM mechanisms may play a relevant role ?in particular such mechanisms that enable knowledge flows to take place. As a managerial implication, this study suggests that HARM mechanisms and a consistent HARM system should be introduced in all knowledge sharing within and between collaborating firms and, thus, enable circumstances for the creation of new knowledge.
Protection of core knowledge via HARM mechanisms should not be forgotten either, since they do not seem to diminish he level of communication in R&D collaboration. For firms of small size, mechanisms, such as contracts and 'Pros could even be out of reach because of limited resources. This study provides a starting point for future work, even if it has certain limitations. Even though the two-respondent data is a merit, the use of two separate survey forms also makes the sample relatively small. Also, data from Just one country is not enough to make wide-ranging generalizations. Further, the analyses could be made more detailed in the future. For instance, the regression analysis in 428 H. Oleander & P. Heralding-Likableness's 1 showed slight support for the R&D intensity control variable.
This suggests that industry sector issues could be studied further with a set of data that would allow clear distinctions between different industries. Further, we have suggested that the level of communication is related to innovative and firm performance, but we have not examined that relationship empirically.