Journal of Management Studies 45:4 June 2008 0022-2380 Trustworthiness, Risk, and the Transfer of Tacit and Explicit Knowledge Between Alliance Partners Manuel Becerra, Randi Lunnan and Lars Huemer Instituto de Empresa, Madrid; Norwegian School of Management; Norwegian School of Management  The transfer of knowledge in alliances entails risk to partners, whose willingness to accept it presumably relies on the trustworthiness that they perceive in their partners. We investigate the extent to which the perceptions of trustworthiness and the willingness to take risk determine the transfer of knowledge between alliance partners and their ultimate impact on alliance success. The results show that the transfer of tacit versus explicit knowledge have very different trust and risk profiles. Whereas explicit knowledge is closely associated with the firm’s willingness to take risk, tacit knowledge is intimately related to high trustworthiness. The results support the important role of trust and the transfer of tacit knowledge on the success of learning alliances. INTRODUCTION Learning from a partner has been recognized as one of the key reasons why alliances may be formed (Doz, 1996; Dussauge et al., 2000; Gulati, 1999; Hamel, 1991; Inkpen, 1998). For instance, Teece and Pisano (1994) argue that alliances can be regarded as mechanisms for firms to learn from each other, which helps them recognize dysfunctional routines and blindspots. Despite the benefits of learning, knowledge transfers entail some level of risk because the partners in an alliance cannot fully control the use that the other partner may make of the transferred knowledge. Allies may actually become strong competitors as a result of the newly developed competencies through the alliance (Inkpen and Beamish, 1997; Hladik, 1988; Khanna et al., 1998). Referring to this unique aspect of learning alliances, Khanna et al. (1998) warn that any alliance will contain both private and common scopes, which create tension in the trade-off between sharing and protecting firm knowledge while maintaining a careful balance between competition and cooperation (Parkhe, 1993). Thus, the transfer of knowledge between alliance partners may be beneficial for them, but it also puts them in a situation of potential vulnerability to the actions of the other partner. Address for reprints: Manuel Becerra, Accenture Chair in Strategic Management, Instituto de Empresa, Maria de Molina 12, 28006 Madrid, Spain (Manuel.Becerra@ie.edu). © Blackwell Publishing Ltd 2008. Published by Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. 692 M. Becerra et al. In these circumstances of potential vulnerability in learning alliances, the perceptions of trustworthiness of the other side become essential for the partners to be willing to take risks (Mayer et al., 1995) and to engage in a possible exchange of knowledge with their allies. For this knowledge transfer to happen, an adequate level of trust in their partner seems necessary. However, we still do not know much about how the perceptions of trustworthiness of the other party affect the transfer of knowledge between alliance partners, the risks that they generate, and their ultimate impact on alliance success through the facilitation of knowledge transfers. Our study builds on the existing literature on knowledge exchange in alliances (Inkpen and Dinur, 1998; Simonin, 2004) as well as research on trust (Mayer et al., 1995; Rousseau et al., 1998). More specifically, we will look into the differences between tacit versus explicit knowledge transfers with regard to their risk profile and their associated requirements for trustworthiness, as well as their effect on alliance success. Thus, our paper investigates the following two research questions: To what extent do tacit versus explicit knowledge exchanges in alliances differ with regard to the perceptions of trustworthiness between partners and their actual willingness to take risks? How do trust, risk, and tacit vs. explicit knowledge transfers contribute to the success of the alliance? The categorization of knowledge into tacit or explicit has become a cornerstone in the literature on learning and knowledge management. Our research questions are aimed at pushing forward our understanding of how these two types of knowledge flows between alliance partners differ with respect to their associated levels of trust and risk. We will also study its implications for the performance of the alliance. This analysis is especially relevant for the successful management of knowledge transfer in alliances, in particular with respect to the levels of trust and risk that are required for different types of knowledge to be actually transferred between alliance partners. After formulating a set of hypotheses, we will test them in a sample of 65 alliances in Norway for which knowledge exchange was particularly important. The results confirm that indeed the transfer of tacit versus explicit knowledge has very different profiles for risk, overall trustworthiness, and its different dimensions (Mayer et al., 1995). While tacit knowledge transfer is intimately related to trustworthiness, the transfer of explicit knowledge requires a greater willingness to accept risks. The results also show that the transfer of tacit knowledge mediates the positive effect of trust on the success of the alliance; in contrast, the explicit knowledge transfers and the willingness to accept risks associated with them have a negligible impact on alliance success. TRUST AND THE TRANSFER OF TACIT AND EXPLICIT KNOWLEDGE Alliances are often defined as a durable, voluntary business arrangement between two firms involving exchange, sharing, or co-development of products, technologies, and services (Gulati, 1998). They may be based on inter-firm relations regulated by a contract or involve some form of shared ownership, like a joint venture. The management literature views alliances as important vehicles for learning and knowledge acquisition and it regards this type of learning as an alternative to internal knowledge generation © Blackwell Publishing Ltd 2008 Transfer of Knowledge 693 within a company (Almeida et al., 2002). Learning in an alliance involves accessing and internalizing critical information, capabilities, or skills from a partner (Kale et al., 2000) and it represents a way to short circuit the learning process compared to other alternatives for developing new competencies (Hamel, 1991). The literature on knowledge management often makes a critical distinction between tacit and explicit knowledge, though other dimensions of knowledge have also been studied, such as complexity, ambiguity, and specificity (Simonin, 1999, 2004; Zander and Kogut, 1995). Tacit knowledge is defined as knowledge that is non-verbalized, intuitive, and unarticulated (Polanyi, 1967) and it has often been regarded as an important basis for competitive advantage from a resource-based view of the firm (Barney, 1991; Grant, 1996). In contrast, explicit knowledge can be coded and articulated, which makes it easier to transfer (Simonin, 1999). Both tacit and explicit knowledge transfers have been analysed in the context of alliances. For instance, Inkpen and Dinur (1998) studied the exchange of tacit and explicit knowledge within an American–Japanese joint venture. Tacit knowledge was associated with culture and philosophy about business rather than knowledge covering specific rules or guidelines, whereas explicit knowledge transfer referred to product designs and specific manufacturing process exchanges between the partners. Notwithstanding this important distinction between tacit and explicit knowledge transfer, in most cases both types of knowledge exchanges are likely to occur in alliances concurrently. Lei et al. (2001, p. 217) argue that ‘in practice, the distinction between explicit and tacit knowledge is not always very clear, since a large component of both organizational and technical knowledge are interwoven with the firm’s culture, product development routines, and human resource practices’. Thus, explicit and tacit knowledge are sometimes difficult to disentangle, though their transfer between partners usually takes place through different means and processes. While explicit knowledge is codified and transferred verbally, tacit knowledge is more embedded in social relations and transferred primarily through direct contact and observation of behaviour. For the successful exchange of knowledge to take place in cooperative relationships, it is presumably necessary to have a trusting relationship (Ring and Van de Ven, 1994). Highly effective collaborative relationships have been described as possessing relational capital (Kale et al., 2000), relational embeddedness (Dhanaraj et al., 2004; Kogut and Zander, 1992), and transparency (Hamel, 1991), which are typically associated with high trust between alliance partners (Krishnan et al., 2006; Lane et al., 2001). All of these partially overlapping characteristics of relationships, though they are not equivalent, are generally considered to have a very positive effect on the relationship, particularly in facilitating the transfer of knowledge between alliance partners. However, the flip side of trusting relationships and the exchange of knowledge within them is vulnerability and the potential abuse from a partner. When a firm is able to build a close and trusting relationship with another company, the risk potentially increases that its own core distinctive competencies may be transferred to its partner with potentially harmful consequences (Inkpen and Beamish, 1997; Luo, 2002). An attempt to learn something from a partner may then end up with having trained a new competitor (Hladik, 1988; Khanna et al., 1998). Within the literature on alliance management, this © Blackwell Publishing Ltd 2008 694 M. Becerra et al. potential vulnerability has been acknowledged, for instance, by Kale et al. (2000) as protection of proprietary assets and Simonin’s (2004) concept of partner protectiveness, when the partner may not behave in a desired manner (Das and Teng, 2001). Thus, we may expect some level of trust and risk to be both present in learning alliances. Earlier research has explored the positive role of trust in knowledge transfer in alliances (Badaracco, 1991; Inkpen, 1998; Kale et al., 2000; Lane et al., 2001) and the link between risk and knowledge transfer (Kale et al. 2000; Simonin, 2004), but researchers have not yet looked into the different characteristics of tacit and explicit knowledge with regard to their implications for trust and risk. To the extent that explicit and tacit knowledge are inherently different and they are transferred through different verbal and social processes, they may likely be associated with different levels of trust and risk. We still do not know how trust, risk, and different types of knowledge transfers are related to each other and how they affect alliance performance. In the next section we will try to fill this gap and formulate specific hypotheses about the effects of trust and risk in the transfer of tacit and explicit knowledge between alliance partners, as well as the final impact of trust, risk, and knowledge transfers on alliance performance. HYPOTHESES First, we will analyse the basic link between a partner’s trustworthiness and the willingness to accept vulnerability by the other partner drawing from the model suggested by Mayer et al. (1995). These authors define trust as ‘the willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor and control that other party’ (Mayer et al., 1995, p. 712). This conceptualization of trust has been adopted by Rousseau et al. (1998), who also make a similar distinction between the willingness to accept vulnerability, i.e. trust, and the evaluation of the partner’s intentions or behaviour (i.e. perceptions of trustworthiness) on which trust is based upon. Mayer et al. (1995) argued that the willingness to accept vulnerability to the actions of another party results from the evaluations that the ‘trustor’ makes about the trustworthiness of the ‘trustee’ with respect to several specific dimensions. Building on the trust literature, they identify three main dimensions of trustworthiness: integrity (i.e. the overall moral character and ethical behaviour of the partner or trustee), benevolence (i.e. the positive vs. egocentric orientation of the trustee in dealing specifically with the trustor), and ability (i.e. general competence and expertise of the trustee). Other trust researchers have identified different elements of trustworthiness; Butler (1991), for example, finds ten dimensions including also consistency and promise fulfilment. Similarly, Cummings and Bromiley (1996) developed a scale based on three dimensions of trustworthy behaviour: keeping commitments, negotiating honestly, and avoiding taking excessive advantage of partner organizations. For these authors, perceived trustworthiness comprises the cognitive, intentional, and emotional evaluations of whether the other party can and will fulfil expectations. In this paper we use Mayer et al.’s (1995) conceptualization of trustworthiness and its three dimensions of integrity, benevolence, and ability. We do not make specific hypotheses for each dimension, because we would expect the same effect for each dimension as for the overall scale of trustworthiness. © Blackwell Publishing Ltd 2008 Transfer of Knowledge 695 However, we will also analyse empirically each dimension separately to explore the extent to which the three dimensions of trustworthiness may not have the same effects. Regardless of the differences in the conceptualization of trust among researchers and the actual dimensions of trustworthiness that can be measured, there seems to be substantial agreement that greater perceptions of trustworthiness should enhance the willingness to accept vulnerability in dealing with another party (Rousseau et al., 1998). This general link between trustworthiness and willingness to take risks should also be observable in the context of knowledge exchange between alliance partners. In the alliance context, relational risk is concerned with the probability and consequences of a partner firm not committing itself to the alliance in the expected manner (Das and Teng, 2001). It is for this type of relational risk, as opposed to environmental risk, that trust is particularly important, and it has positive consequences (Krishnan et al., 2006). The risk that the partner may behave opportunistically is an important threat for alliances, including the partner’s potentially damaging use of the newly acquired knowledge, whose consequences need to be taken into account by alliance partners (Heide and Miner, 1992). Drawing from the trust literature and consistent with Mayer et al. (1995), we will argue that only when the perceptions of partner trustworthiness are sufficiently high, will partners be willing to accept the risks that may exist in collaborating with an alliance partner. Only if the firm believes that its potential partner has sufficient ability, overall integrity, and benevolent attitude towards the firm, would it be reasonable for the firm to accept the vulnerability associated with the alliance. Otherwise if the firm considers that the potential partner is not trustworthy, it may be a better option to avoid a risky alliance and to look for another partner with greater trustworthiness. Consequently, we expect a positive relation between trustworthiness and the willingness to take risk. Hypothesis 1: Greater perception of a partner’s trustworthiness is associated with greater willingness to take risks in the alliance. Earlier research has shown that social relations characterized by high trust facilitate the transfer of knowledge in organizational settings, particularly for the transfer of tacit knowledge (Hansen, 1999). From a social learning theory perspective, intensive social interaction facilitates the development and transfer of tacit knowledge (Brown and Duguid, 2001). While explicit knowledge can be effectively transferred through written documents without the need for direct contact and attachment between partners, tacit knowledge relies primarily on direct observation and a close relationship between individuals. In this type of close social relationship, trust is also likely to exist. Inkpen and Dinur (1998) claim that the transfer of tacit knowledge is more likely to occur among individuals in a high trust context. They propose that the transfer of tacit knowledge will occur through individuals rather than collectives; furthermore, initiatives of tacit knowledge transfer from the collective often fail. It is in the personal networks between trusted individuals that tacit knowledge is gained and effectively transferred. In contrast, trust is not really important for the exchange of explicit knowledge, which can be codified and therefore transferred independently of specific individuals. The transfer of explicit knowledge is thus less dependent on close relationships based on trust. © Blackwell Publishing Ltd 2008 696 M. Becerra et al. There is already abundant empirical evidence in the alliances literature that is consistent with the idea that the transfer of tacit knowledge is more likely to take place in a context of high trust (Inkpen, 1998; Lubatkin et al., 2001). While explicit knowledge transfer may also be enhanced by the presence of trust, it does not require a high trust relationship to be effectively transferred. In their study of transfer of knowledge from a foreign partner to an international joint venture, Dhanaraj et al. (2004) found that relational embeddedness, which is usually related to high trust, had a much stronger influence on the transfer of tacit than explicit knowledge. They showed that tacit knowledge is more closely connected to trust because it requires more intensity in the transfer process than explicit knowledge, which does not need the same socialization. In summary, we will argue that high trustworthiness is more critical when knowledge is tacit, and consequently transferable only through direct contact, than when knowledge is explicit. In the case of explicit knowledge, it can be effectively transferred independently of the individuals involved in the exchange and the level of trust between them. Hypothesis 2: The transfer of tacit knowledge is more strongly associated with perceptions of a partner’s trustworthiness than the transfer of explicit knowledge between alliance partners. As we have claimed above, because of its codifiable nature, explicit knowledge needs less effort in communication, teaching, and interactive processes to be successfully transferred. This codifiability makes it more transparent, less sticky, and essentially easier to transfer at greater speed (Szulanski and Capetta, 2003). In contrast, tacit knowledge is much more difficult to identify, evaluate, and absorb since the transfer process is plagued with inherent complexity and possible misunderstandings between the parties involved in the process. The lower ambiguity and specificity of explicit knowledge should make it more risky to transfer than tacit knowledge, which has inherent difficulties for replication. The more tacit the knowledge is, the harder it is for competitors to assess the true value of this knowledge because it is more context-specific (Chakravarthy et al., 2003; Peteraf, 1993). Therefore, we could expect firms to heavily guard all their explicit knowledge and to consider very risky any possible transfer of such type of knowledge, while they may see less risk in showing their more highly-specific tacit knowledge. Consequently, it seems reasonable to believe that it is far more dangerous for a firm to contribute explicit knowledge to an alliance, since this type of knowledge can quickly spread with great ease of replicability. Explicit knowledge has greater probability to leak outside the alliance partners and be perfectly absorbed and utilized by third parties. Sharing explicit knowledge may be considered to place partners at greater risk than the transfer of tacit knowledge, which can only spread slowly through close interaction and is likely to require highly-specific assets and complementary skills. Thus, we will argue that the transfer of explicit knowledge may be perceived as inherently riskier by alliance partners and, as a result, sharing explicit knowledge should require greater willingness to accept risk than the transfer of tacit knowledge. © Blackwell Publishing Ltd 2008 Transfer of Knowledge 697 Hypothesis 3: The transfer of explicit knowledge requires greater willingness to accept risk than the transfer of tacit knowledge between alliance partners. The possible benefits of knowledge transfers in alliances are well established in the literature (Dussauge et al., 2000; Inkpen, 1998). For instance, in one of their earlier studies of Hungarian joint ventures, Lyles and Salk (1996) showed that knowledge acquisition was positively related to performance. Thus, we may expect a positive relationship between knowledge acquisition and alliance success, at least for alliances whose major goal is learning. However, there is some controversy about the relative importance of the transfer of tacit versus explicit knowledge for alliance success. Given the presumed greater relevance of tacit knowledge for competitive advantage in the resource-based view (Barney, 1991; Grant, 1996), it may be argued that the transfer of tacit knowledge should be more critical than explicit knowledge for the performance of an alliance. However, Dhanaraj et al. (2004) found a positive relation between explicit knowledge transfer and performance and failed to find a direct relationship between tacit knowledge transfer and performance. They explained this somewhat surprising finding by arguing that it may be due to explicit knowledge’s relative low cost of transfer, clarity, and actionability of routines. It is possible that the transfer of explicit knowledge may be easier to detect and evaluate and therefore deemed more important, though its possession could have a smaller effect on the long-term success of each firm. Lane et al. (2001) also studied the relationship among trust, learning, and performance. They found that trust is very limitedly connected to learning, but strongly related to performance. Furthermore, despite the high correlation between tacit and explicit knowledge, they found a negative relationship between tacit learning and performance, but a positive correlation between explicit learning and performance. Based on these studies, we may conclude that the empirical evidence shows that the transfer of tacit and explicit knowledge transfer takes place through different processes and may have different effects on performance. We believe that different alliances created primarily for learning purposes may have different goals about the type of knowledge that they seek. In general, both tacit and explicit knowledge transfers may be presumed to have a positive effect on performance, as knowledge acquisition is a critical goal in learning alliances. However, either tacit or explicit knowledge transfer may be more or less closely associated with alliance performance, depending on the goals of specific alliances. For alliances created primarily for the sake of sharing information, explicit knowledge and willingness to take risk should be more closely associated with the success of the alliance. In contrast, trust and tacit knowledge are likely to be more relevant for alliances created for more complex mutual learning and development of new competencies, which require close social interaction. In both cases, the transfer of knowledge in learning alliances has an important mediating role between relational context and alliance success. On the one hand, trust needs to exist for the transfer of tacit knowledge to take place, which ultimately should be positively associated with alliance success for those partners that seek this type of knowledge. Referring specifically to knowledge transfer in alliances, Kale et al. (2000, © Blackwell Publishing Ltd 2008 698 M. Becerra et al. p. 217) showed that ‘relational capital based on mutual trust and interaction at the individual level between alliance partners creates a basis for learning and know-how transfer across the exchange interface’. Thus, trust acts as a behavioural lubricant and enables partners to be more flexible, open, and tolerant in the cooperation process (Madhok, 1995), which should result in better performance. On the other hand, the transfer of explicit knowledge should play an analogous mediating role in the effect of willingness to take risks on alliance success. As discussed above, the empirical literature shows that the transfer of explicit knowledge has a positive effect on alliance performance (Dhanaraj et al., 2004; Lane et al., 2001). However, without the willingness to take risks, the transfer of explicit knowledge would be less likely to occur. Therefore, we can formulate a different process for the effective transfer of explicit knowledge. In this case, greater willing to take risks facilitates the transfer of explicit knowledge, which in turn may result in better performance for the alliance. Hypothesis 4a: The transfer of tacit knowledge between alliance partners mediates the effect of partner trustworthiness on the success of the alliance. Hypothesis 4b: The transfer of explicit knowledge between alliance partners mediates the effect of the willingness to take risks on the success of the alliance. METHODOLOGY To test these hypotheses about trust, risk, and knowledge transfer between alliance partners, we surveyed 155 firms from a variety of industries in Norway, primarily printing, construction, fishing, and the public sector. These firms had voluntarily participated in a Research Project on Networks and Organizations funded by the Norwegian Research Council (NRC), for which improving information technology and sharing knowledge among participants were some of the key project goals. First, we telephoned each company and talked directly to the CEO in almost all cases. Getting access to the CEO was not difficult because of the small size of the companies and the firm’s involvement in a larger government-financed project. In a few cases, the CEO referred us to a different person who was formally in charge of managing the firm’s alliances. We then mailed the questionnaire to the CEO or to this referred person that we had talked to. Two weeks later, we made a reminder phone call to follow up on the survey. Though we relied on a single informant per firm, we believe that the questions that we wanted to address about trust, risk, and the transfer of knowledge with alliance partners should be answered by the person that constitutes the actual link with the external partner, who was the CEO in most cases. We sent out a questionnaire to the 155 firms involved in this project and received 65 completed and usable questionnaires. Though they were not formally required to participate in our study, the official support that our project obtained from the NRC and the direct contact with the CEO probably persuaded many of them to take part in our study, which resulted in a relatively high response rate of 42 per cent. From the firms in the sample, 70 per cent of the firms had fewer than 50 employees, only 3 per © Blackwell Publishing Ltd 2008 Transfer of Knowledge 699 cent had more than 1000 employees, and 80 per cent of the alliances reported by the firms had existed for less than five years. Each company was asked a set of questions about the closest partners (up to five alliances) in the alliances in which they participated. To avoid problems of autocorrelation of observations, we selected the first relationship they identified, except for the few situations where some firms described an alliance to another firm in the sample. In those rare cases, we selected the second tie that they identified. We followed the recommendations from Churchill (1979) to design and pretest the survey. Eleven pilot interviews were conducted early on to check the relevance of our topic and develop some preliminary ideas about the research questions and the variables to include in the study. Then, we developed our measurement instrument and pretested the questionnaire on two colleagues and two CEOs from the sample, who suggested some minor changes to the questionnaire. All independent and dependent variables, except the two control variables (firm size and alliance years), were measured on multiitem scales. The items based on English studies were translated to Norwegian and then translated back to English, so that the non-Norwegian researchers could check for content changes. All items were measured on a seven-item Likert scale (strongly disagree–strongly agree). The reliability of all scales was always above the threshold of 0.70 recommended by Nunnally (1978). The Appendix includes the items for all of the scales. The main variables in the study are further discussed below. Transfer of Tacit and Explicit Knowledge Because of the lack of a widely accepted instrument to measure tacit and explicit knowledge transfers, we developed several items based on earlier studies and our pilot interviews. From the initial set of items, we selected six items, three for tacit and three for explicit knowledge transfer. The items for explicit knowledge transfer deal with the exchange of information and knowledge about products, industry trends, and key industry participants (i.e. customers, suppliers, and competitors), which the partners must presumably verbalize before it can be transferred. On the other hand, the tacit knowledge items are concerned with shared activities, observation of behaviour, and direct contact, which are associated with more complex ways to interact and acquire tacit knowledge from the partner. Exploratory factor analysis of the six items revealed two clear factors with eigenvalues greater than one (2.31 and 1.74), which jointly account for 67 per cent of the total variance. All items had at least 0.65 loading coefficients and no cross-loadings reached 0.40. However, one of the items for explicit knowledge (‘We exchange knowledge about customers, suppliers, and competitors with this company’) loaded on the factor with the other three items for tacit knowledge. Apparently, the managers distinguish between the transfer of any kind of information (i.e. explicit knowledge about products and industry trends) on the one side and the exchange of knowledge and direct learning from the partner on the other side (i.e. tacit knowledge). Thus, we moved this item to its appropriate scale. The reliability (Cronbach alpha) of the final two scales was 0.76 for explicit knowledge (two items) and 0.77 for tacit knowledge transfer (four items). © Blackwell Publishing Ltd 2008 700 M. Becerra et al. Willingness to Accept Risk Three items were used to measure the extent to which the firm tried to minimize risks when dealing with the alliance partner. It is important to note the distinction between the willingness to assume risk and the actual risk-taking behaviour, which has been analysed in the literature (Mayer et al., 1995; Sitkin and Pablo, 1992). These items do not measure the actual assumption of risk, which in our study is reflected in the transfer of knowledge. Instead, this variable captures the firm’s general intentions to minimize risk, defend themselves, and avoid sharing risky projects with the alliance partner. The scale obtained a reliability of 0.75. Because the arguments in the hypotheses section were developed following the literature on the willingness to accept risk (rather than the minimization of risk, which the items measure), we reversed the initial coding of the items to facilitate the interpretation of this variable. Evaluations of Trustworthiness Many different dimensions of trustworthiness have been proposed in the past. The three dimensions identified by Mayer et al. (1995) from their analysis of the trust literature parsimoniously capture the key aspects of the expectations that one person may have about others’ intentions and behaviour. As we mentioned earlier in the paper, these three dimensions of trustworthiness are integrity, benevolence, and ability. The items to measure each dimension were taken from Mayer and Davis (1999) and adapted to the alliance context of our study. The person (i.e. the trustor) that managed the relationship with alliance partners was asked his/her opinion about its alliance partner (i.e. the trustee), with regard to the partner’s integrity (i.e. the overall moral character and ethical behaviour of the partner), benevolence (i.e. the positive vs. egocentric orientation of the partner in dealing specifically with his/her firm), and ability (i.e. general competence and expertise of the partner). The 12 items were aggregated to compute an overall measure of trustworthiness, which obtained a reliability of 0.91. To further explore the effect of the three dimensions of trustworthiness, the analysis was also conducted for each dimension separately, so that we could see how each dimension uniquely affects the willingness to take risks and the transfer of tacit and explicit knowledge. Each sub-scale obtained a Cronbach alpha of at least 0.80. Alliance Success Three items formed a scale that subjectively measured the degree to which the alliance could be considered successful by the key informant. As alliances are entangled in other firm activities, alliance performance can be difficult to discern from the performance of the firm, and they are often considered to be successful in the long term (Anderson, 1990); consequently objective, short-term measures are not useful. Evaluations based on multiple indicators, including an assessment of financial performance, are found to be reliable proxies for alliance performance (Anderson and Narus, 1990). Thus, one item assesses the net results of the alliance in the previous year, whereas the other two items provide a subjective evaluation of the alliance success and its contribution to growth. The scale achieved a reliability of 0.73. © Blackwell Publishing Ltd 2008 Transfer of Knowledge 701 Control Variables The empirical analysis included two control variables: firm size (in number of employees), and alliance years (i.e. years since the alliance was created). Firm size may indicate the resource scope and slack of an organization that may influence the ability to access resources. Young alliances are generally perceived as more risky ( Yan and Gray, 1994). Because there was substantial skewness in the distributions of the size of the firms in the study and also in the number of years since they alliances were created, we used their logarithm values in the analysis. Construct Validity and Common Method Variance All the instruments used in the survey revealed a sufficient degree of convergent validity, as reflected by Cronbach alphas reported above. However, using perceptual data coming from a survey raises the threat of common method variance. To see the extent to which this could be a problem, we tested the discriminant validity of the scales through exploratory and confirmatory factor analysis (Podsakoff et al., 2003). First, we did a Harman’s one factor test (Podsakoff et al., 2003), which resulted in a very poor fit (chi-squared = 541 (p < 0.001); CFI = 0.60). Consequently, we could discard the idea that all variables, including trustworthiness and tacit knowledge transfer, were just one evaluative image in the minds of the respondents. More than one construct clearly emerged from the data. Second, exploratory factor analysis shows that there are indeed five factors with eigenvalues greater than 1 in the full set of 24 items that cover the five variables (trust, risk, tacit knowledge, explicit knowledge, and alliance success). The items primarily load on the expected factors after varimax rotation was conducted (principal component analysis that explains 67 per cent of the total variance). In conclusion, even though we are measuring perceptions of individuals with one survey, there are several distinct and interconnected variables in the minds of these managers. The relationships among these cognitive constructs constitute the focus of his study, including partner trustworthiness, willingness to take risk, tacit and explicit knowledge transfers, and alliance success. Given the goals of this paper and the results from the convergent and discriminant validity of the variables, we concluded that the potential problems of common method variance were not significant. RESULTS Hierarchical regression analysis was used to test the hypotheses. Descriptive statistics for all the variables and their zero-order correlations are shown in Table I. Several results from the correlation table should be noted. There is significant correlation among the variables that could be expected, for instance, between trustworthiness and willingness to take risks, between trustworthiness and tacit knowledge transfer, between willingness to take risk and explicit knowledge transfer, between explicit and tacit knowledge, and also between alliance years and trustworthiness. Only one set of correlations seems to be sufficiently high to suggest problems of multicollinearity: the correlations among the three dimensions of trustworthiness (up to © Blackwell Publishing Ltd 2008 702 © Blackwell Publishing Ltd 2008 Table I. Descriptive statistics and correlations table Mean Explicit knowledge Tacit knowledge Willingness to take risks Alliance success Trustworthiness Integrity Benevolence Ability Firm size Alliance years 4.44 4.93 4.15 4.75 5.16 5.33 4.61 5.53 2.76 1.11 1.41 1.27 1.40 1.18 0.95 1.12 1.10 1.04 2.04 0.86 Notes: N = 65. ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed). Pearson correlations 1 2 3 4 5 6 7 8 9 1 0.43** 0.57** 0.18 0.22 0.30* 0.05 0.27 -0.06 0.19 1 0.14 0.51** 0.56** 0.43** 0.43** 0.57** 0.03 0.22 1 -0.05 0.32* 0.41** 0.18 0.22 -0.10 0.15 1 0.48** 0.36** 0.43** 0.45** -0.05 0.14 1 0.89** 0.84** 0.88 -0.05 0.34* 1 0.59** 0.71** -0.10 0.39** 1 0.56** -0.06 0.09 1 0.02 0.34** 1 0.18 M. Becerra et al. 1. 2. 3. 4. 5. 6. 7. 8. 6. 7. Std. dev. Transfer of Knowledge 703 Table II. Regression analysis of the willingness to take risks Dependent variable: Willingness to take risks Independent variables Model 1 Model 2 Model 3 Alliance years Firm size Overall trustworthiness Integrity Benevolence Ability 0.16 -0.14 0.04 -0.10 0.33* -0.04 -0.05 R2 Change in R2 R2 change F test 0.56** -0.08ns -0.07ns 0.039 1.07 0.136 0.097 5.85* 0.213 0.174 3.69* Notes: Standardized coefficients reported. ** Significant at the 0.01 level (2-tailed). * Significant at the 0.05 level (2-tailed). ns This coefficient was not significant when the other two dimensions of trustworthiness were dropped from the analysis. 0.71) and also with its overall aggregate value (up to 0.89). We computed the variance inflation factors for all the variables; all of them were well below the level of 10 used to test for multicollinearity. However, when the three dimensions of trustworthiness were included in the same step, the overall increase in R2 was usually highly significant, but none of the three variables had a significant coefficient, which is clear evidence of some multicollinearity among this set of variables. To deal with this problem, we did the analysis for the overall measure of trustworthiness and for the three dimensions as a group, but also each dimension of trustworthiness was included separately without the other two and their significance is reported in the results. Table II investigates how trustworthiness perceptions between alliance partners are associated with the willingness to take risks within the alliance; more precisely, the effect of the trustor’s perceptions of the trustee’s trustworthiness on the willingness of the trustor to take risks in dealing with the trustee. First, the two control variables are included in Model 1. In the next model, we can see that indeed greater perceptions of trustworthiness are clearly associated with greater willingness to take risk, after controlling for alliance years and firm size. The results are significant for the overall measure of trustworthiness in Model 2 as well as for its different dimensions in Model 3. When the three dimensions are included in the last step in a hierarchical regression analysis, the increase in R2 from Model 1 to Model 3 is 0.174, which is a significant increase at the 5% confidence level. Thus, there is empirical evidence in favour of Hypothesis 1. It is interesting to see that integrity is the only dimension of trustworthiness with a significant impact on the willingness to take risks. The other two dimensions, ability and benevolence, are insignificant in this model, even when only one of them is left as the independent variable. It seems that the moral character and integrity of the partner © Blackwell Publishing Ltd 2008 704 M. Becerra et al. Table III. Regression analyses of the transfer of tacit and explicit knowledge Dependent variable: Tacit knowledge transfer Independent variables Model 4 Alliance years Firm size Overall trustworthiness Integrity Benevolence Ability Willingness to take risks 0.21 0.00 R2 Change in R2 R2 change F test 0.044 1.20 Model 5 Model 6 Model 7 0.03 0.06 0.55** 0.01 0.02 0.01 0.02 -0.02s 0.14s 0.51** -0.04s 0.14s 0.52** 0.05 0.355 0.311 7.85** 0.356 0.001 0.72 0.310 0.266 19.64** Dependent variable: Explicit knowledge transfer Independent variable Model 8 Model 9 Model 10 Model 11 Alliance years Firm size Overall trustworthiness Integrity Benevolence Ability Willingness to take risks 0.21 -0.09 0.15 -0.07 0.17 0.02 -0.06 0.05 -0.03 0.33ns -0.29ns 0.17ns -0.01ns -0.23ns 0.20ns 0.59** 0.145 0.101 1.91 0.422 0.277 23.09** R2 Change in R2 R2 change F test 0.046 1.27 0.070 0.024 1.35 Notes: Standardized coefficients reported. ** Significant at the 0.01 level (2-tailed). * Significant at the 0.05 level (2-tailed). s This coefficient was positive and significant when the other two dimensions of trustworthiness were dropped from the analysis. ns This coefficient was not significant when the other two dimensions of trustworthiness were dropped from the analysis. makes alliance members willing to accept risks within their relationship, rather than their perceptions about the partner’s ability or benevolence towards them. The analyses of the relative impact of trustworthiness perceptions and willingness to take risks on the transfer of tacit and explicit knowledge are shown in Table III. With regard to tacit knowledge, the control variables are included first in Model 4, then overall trustworthiness is included in Model 5 while its different dimensions substitute for it in Model 6, and finally willingness to take risks is added in Model 7. The results provide strong evidence that the transfer of tacit knowledge is significantly associated with higher © Blackwell Publishing Ltd 2008 Transfer of Knowledge 705 perceptions of trustworthiness (significant coefficient of 0.55 for overall trustworthiness and a significant increase in R2 of 0.311 for its three dimensions), though not with the willingness to take risk. These results support Hypothesis 2. With respect to the three dimensions of trustworthiness, the partner’s ability is the most important dimension of trustworthiness for the transfer of tacit knowledge, though the other two dimensions would also be significant if they were the only dimension left in the model. Thus, all the three dimensions of trustworthiness are associated with greater transfer of tacit knowledge, though the association is stronger with the partner’s perceived ability. Exactly the opposite situation takes place in the transfer of explicit knowledge. As we can see in Models 9, 10, and 11, overall trustworthiness and its dimensions (either independently or as a group) are not associated with the transfer of explicit knowledge. In this case, only the willingness to take risk determines the transfer of explicit knowledge, which is the last variable introduced in Model 11 (significant coefficient of 0.59). Therefore, there is also support for Hypothesis 3. Finally, the results provide mixed support for the last two hypotheses, as Table IV shows. With regard to Hypothesis 4a, the regression analysis confirms that the transfer of tacit knowledge partially mediates the positive effect of trust on alliance success, for both the overall measure of trustworthiness and each of its different dimensions. Following the Baron and Kenny (1986) procedure to test for mediation, overall trustworthiness had a significant coefficient of 0.50 on alliance success in Model 12. Trustworthiness also had a significant positive effect (0.55 coefficient) on the regression on tacit knowledge in Model 5 reported earlier. When both trustworthiness and tacit knowledge are introduced in Model 13, the highly significant coefficient for trustworthiness decreases from 0.50 to an insignificant 0.28 coefficient, while the tacit knowledge variable has a highly significant coefficient of 0.39. Thus, in support of Hypothesis 4a, the tacit knowledge transfer variable captures a significant part of the explanatory power that was previously attributable to overall trustworthiness, whose coefficient is significantly reduced when tacit knowledge transfer is included in the analysis. Trust, however, does more for the success of the alliance than just facilitating tacit knowledge transfers. Models 17–19 also provide support for Hypothesis 4a, now based on the different dimensions of trustworthiness. Each of the three dimensions is significant when introduced alone in Model 17, though they are not when introduced at the same time because of the high correlation among them. When tacit knowledge is introduced in the next step for each dimension separately, the coefficients for all of them are substantially reduced, becoming insignificant for integrity and ability and reducing the coefficient for benevolence from 0.43 to 0.26, both significant (the results from these regressions are not shown, but their significance is reported in Table IV). Again, tacit knowledge transfer partially mediates the effect of the dimensions of trustworthiness on alliance. In contrast, Hypothesis 4b is not supported. Neither willingness to take risk nor explicit knowledge are significant in Models 14 and 15. The limited effect of these two variables on alliance success is also shown in Models 16 and 19, where all the variables are included. In comparison, overall trustworthiness and tacit knowledge transfer are positively associated with alliance success in Model 16, though only benevolence remains significant in Model 19. © Blackwell Publishing Ltd 2008 706 M. Becerra et al. Table IV. Regression analysis of alliance success Dependent variable: Alliance success Independent variables Model 12 Model 13 Model 14 Model 15 Model 16 Firm size Alliance years Overall trustworthiness Tacit knowledge transfer Willingness to take risks Explicit knowledge transfer -0.04 -0.04 0.50** -0.06 -0.05 0.28 0.39** -0.10 0.15 -0.09 0.10 -0.08 -0.24 0.29 -0.07 -0.06 0.39* 0.31* -0.30* 0.14 R2 Change in R2 R2 change F test 0.238 0.341 0.103 7.83** 5.30** 0.028 0.51 0.084 0.055 3.09+ 0.395 5.23** Dependent variable: Alliance success Independent variable Model 17 Model 18 Model 19 Firm size Alliance years Integrity Benevolence Ability Tacit knowledge transfer Willingness to take risks Explicit knowledge transfer -0.07 -0.02 -0.07s 0.28s 0.36s -0.08 -0.02 -0.05ns 0.23s 0.15ns 0.37* -0.08 -0.03 0.05ns 0.27s 0.13ns 0.30 -0.30 0.18 0.267 0.355 0.088 6.56* 0.404 0.049 1.90 R2 Change in R2 R2 change F test 3.56** Notes: Standardized coefficients reported. ** Significant at the 0.01 level (2-tailed). * Significant at the 0.05 level (2-tailed). s This coefficient was positive and significant when the other two dimensions of trustworthiness were dropped from the analysis. ns This coefficient was not significant when the other two dimensions of trustworthiness were dropped from the analysis. It is interesting to note that the coefficient for willingness to take risks is actually negative and significant (or marginally so) in the models with all the variables. It seems that once the benefits of trustworthiness and the transfer of explicit and tacit knowledge have been taking into account, more willingness to take risks is actually associated with less successful alliances. Our interpretation of this result is that the willingness to take risks is not inherently good or bad; in fact, it has an insignificant zero-order correlation with alliance success. However, it may provide some benefits for the partners, which in these alliances is primarily channelled through the transfer of explicit knowledge. Once the possible benefits of explicit knowledge are controlled for, accepting greater risk may turn into gullibility and have negative consequences. © Blackwell Publishing Ltd 2008 Transfer of Knowledge 707 DISCUSSION The results provide evidence that tacit and explicit knowledge are indeed very different with regard to their relationship to trustworthiness, risk, and alliance success. The transfer of explicit knowledge is closely associated with the willingness to take risks and both of them have little impact on the evaluations of the success of the alliance. In contrast, the transfer of tacit knowledge is highly related to the perceptions of the partner’s trustworthiness and both have a significant effect on alliance success. These findings have important implications for the management of knowledge in alliances and also for the literature on trust that we would like to discuss below. We will also look at the limitations of our study. The Transfer and Protection of Knowledge in Alliances In contrast to some of the existing empirical research (Dhanaraj et al., 2004), we have found that the transfer of tacit knowledge is more closely associated with alliance performance than explicit knowledge transfers, whose positive effect on the success of the alliance is negligible. This is fully in line with the resource-based view of the firm and the presumed critical importance of tacit knowledge for competitive advantage (Barney, 1991), but it has not received clear empirical support in the alliances literature. Though we have not found performance consequences for the transfer of explicit knowledge in alliances, we believe that this may not be generalizable to all types of alliances. Quite possibly, the firms in our sample may have been primarily concerned with establishing strong social links to the partners rather than just facilitating the transfer of information data among firms. We believe that probably for this reason only the exchange of tacit knowledge had an effect on alliance success, but not the exchange of explicit knowledge. Our results also show that the transfer of explicit knowledge requires greater willingness to take risk than the transfer of tacit knowledge; therefore, it seems that explicit knowledge is more carefully guarded by managers. This finding represents an interesting paradox when it is connected with the results about the greater impact of tacit knowledge transfer on alliance success. Tacit knowledge constitutes a source of competitive advantage for many firms and they frequently participate in alliances to exchange this type of knowledge, but they seem to be more concerned with protecting their explicit knowledge instead. Given the limited effect of explicit knowledge transfer on performance, we should probe further into why managers protect it so much. It is possible that managers may have a somewhat distorted perception about the relative risks of tacit versus explicit knowledge transfers driven by their differences in observability and replicability that we highlighted earlier. In this sense, because explicit knowledge is immediately identifiable but tacit knowledge is harder to detect and to absorb, they may protect themselves against those knowledge transfers whose risks are readily observable and more easily avoided. In other words, transferring explicit knowledge involves a ‘clear and present danger’ that needs to be carefully managed, even though it may be relatively less important for the ultimate competitive advantage of the firm. In any case, the high willingness to accept risk that explicit knowledge transfers require, coupled with its lower effect on alliance success, deserves further research. © Blackwell Publishing Ltd 2008 708 M. Becerra et al. Implications for Research on Trust Mayer et al. (1995) claim that we need to distinguish among the antecedents of trust (i.e. evaluations of the trustworthiness of the other party with regard to its integrity, ability, and benevolence towards the trustor), trust itself (i.e. the trustor’s willingness to take risks when dealing with the other party), and actual risk-taking actions (like the exchange of knowledge between alliance partners). They argue that the need for trust only arises in a risky situation, in which the perceived risk is compared to the level of trust. Based on their model, the trustor will actually engage in risk-taking actions only ‘if the level of trust surpasses the threshold of perceived risk’ (Mayer et al., 1995, p. 726). Though this model seems theoretically reasonable and inspired our research, the role of trust in facilitating risk-taking activities (such as the transfer of knowledge in alliances) needs to be qualified, as the results show with respect to the different mediating roles of perceived trustworthiness and the willingness to take risk. We did find that greater perceptions of a partner’s trustworthiness are positively associated with the willingness to take risks, as Mayer et al. (1995) and Rousseau et al. (1998) suggest. However, the relationship between trustworthiness and actual risk taking (i.e. tacit and explicit knowledge transfers) is not mediated by the willingness to take risks, as the results in Tables III and IV show. In other words, trust facilitates the transfer of tacit knowledge regardless of the willingness to take risks of the managers involved; in contrast, the transfer of explicit knowledge requires greater willingness to take risks, but it is not significantly associated with trustworthiness. Though we should be cautious about drawing causality inferences from cross-sectional research, trustworthiness and the willingness to take risks, though positively correlated, do not seem to form a causal chain that leads to the transfer of knowledge. Contrary to Mayer et al. (1995), our findings suggest that risk does not necessarily need to be present for trust to exist, to be meaningful, and to exert its lubricant effect within relationships identified by the literature on trust and alliances. Thus, we may question the definition of trust as ‘the willingness of a party to be vulnerable to the actions of another party’ (Mayer et al., 1995, p. 712). In our opinion, saying ‘I believe that you are trustworthy’ is equivalent to ‘I trust you’, regardless of whether there is more or less risk involved in the relationship. In fact, this conceptualization of trust as the perception of another party’s trustworthiness about specific dimensions has actually been used by abundant research on trust (Butler, 1991; Cummings and Bromiley, 1996; JohnsonGeorge and Swap, 1982; Krishnan et al., 2006). The conceptualization in Mayer et al. (1995) and Rousseau et al. (1998) of trust as willingness to take risks may place too much emphasis on what can be regarded as just one consequence of trust (i.e. willingness to take risk), which sometimes may not be relevant or characteristic of the relationship in which trust is studied. However, willingness to take risk may be very important in other circumstances. Trust, defined as perceptions of high trustworthiness, is driven by variables like the generalized propensity to trust of individuals (Rotter, 1971) and communication frequency (Becerra and Gupta, 2003), and it has many consequences, such as increasing the willingness to take risks that Mayer et al. (1995) stressed. However, it has many other consequences, like facilitating open communication (O’Reilly et al., 1987), reducing © Blackwell Publishing Ltd 2008 Transfer of Knowledge 709 conflict (Zaheer et al., 1998), and enhancing relational opportunities (Huemer, 1998), which are not necessarily related to willingness to take risks. This is also the case for the transfer of tacit knowledge in our study, which is not associated with greater willingness to take risks. In addition to overall trustworthiness, we also explored its three main dimensions. As we focused primarily on the comparison between explicit and tacit knowledge transfers, we leave for trust researchers to probe further into the differences among the three dimensions of trustworthiness. However, some of the results deserve to be highlighted. First, it seems that the willingness to take risks in alliances is only associated with the perceptions about the partner’s general integrity, but not with his/her ability or benevolence towards the trustor. In other words, risks (like those involved in the transfer of explicit knowledge) are avoided with those partners regarded as having low integrity, regardless of their expertise or their attitude towards the firm. Second, though the three dimensions of trustworthiness are associated with greater transfer of tacit knowledge, more tacit knowledge is exchanged with partners perceived as having greater ability. This seems very reasonable because learning should preferably be done from knowledgeable experts, rather than well-intended partners. Finally, we have also detected that benevolence, probably because it is usually associated with interest congruence, is significantly higher in those alliances with greater success. The results confirm that trustworthiness is indeed a multidimensional construct and each of them has unique effects on relationships. Future research may study the interaction among these dimensions of trustworthiness and the ultimate reason why they have these different effects. Limitations There are some limitations that should be identified and their potential impact assessed. Most of them deal with the data used in our study, including a relatively small sample of Norwegian alliances and the perceptual data coming from one side of the dyad. As we explained in the Methodology section, the perceptual measurement of most of the variables does not seem problematic to us due to their inherent perceptual nature. More specifically, we measured the perceptions about the partner’s trustworthiness as well as the trustor’s willingness to take risks from the perspective of the firm making such evaluations through the person involved in managing the alliance relationship. Similarly, this person evaluated the transfer between alliance partners of explicit and tacit knowledge, as well as the overall satisfaction with the results obtained from the alliance. The construct validity analysis showed that these are distinct concepts. However, they are perceptual variables and we should acknowledge that we are truly investigating how these variables, most of them cognitive-emotional, relate to each other in the minds of the people in charge of these alliances. We should also note the limitations about the explicit knowledge transfer variable, which relies on just two items after moving another one to the tacit knowledge transfer scale. Though this variable achieved an acceptable level of validity (Cronbach alpha of 0.76) and in the exploratory analysis was clearly distinguishable from the tacit knowledge scale, we wanted to explore how much of the paper’s results depended on this particular variable. We did the entire analysis without the explicit knowledge transfer variable and © Blackwell Publishing Ltd 2008 710 M. Becerra et al. obtained very similar results for the transfer of tacit knowledge. Very briefly, Hypothesis 1 is still supported (the greater the trust, the greater the willingness to take risks), the part of Hypothesis 2 that deals with tacit knowledge is still supported (trust and tacit knowledge transfer are positively associated), Hypothesis 3 also holds (willingness to take risks is not associated with the transfer of tacit knowledge), and Hypothesis 4a still receives support (transfer of tacit knowledge significantly mediates the positive effect of trust on alliance). Thus, though the results from the transfer of explicit knowledge only build on two items, the findings are quite robust, even when the variable for explicit knowledge transfer is dropped from the analysis. Overall, though we need to take into consideration these limitations, the results provide empirical evidence that the transfer of tacit vs. explicit knowledge are quite different in terms of their profiles regarding trust and risk. Both types of knowledge are associated with different relational contexts, which have different effects on alliance performance. These results have important implications for both research on trust and the management of knowledge in alliances. ACKNOWLEDGMENTS We would like to thank Mark Kriger for comments received on an earlier version of this article, and the Norwegian Research Council for financial support for this research. APPENDIX: VARIABLES IN THE STUDY AND THEIR RELIABILITY (CRONBACH ALPHA) Explicit knowledge transfer (0.76) a. We do not provide this company with information about our current products and services (reverse coded). b. We do not exchange ideas and information about industry trends with this company (reverse coded). Tacit knowledge transfer (0.77) a. We share activities that provide learning with this company. b. We regularly visit each other’s facilities and observe onsite how operations are conducted. c. This company and ours have learned much from the direct contact between our two organizations. d. We exchange knowledge about customers, suppliers, and competitors with this company (moved from the explicit knowledge transfer scale). Willingness to take risks (0.75) a. We try to minimize risk when cooperating with this firm. b. We try to defend ourselves against the influences of this firm. c. We avoid sharing risky projects with this company. © Blackwell Publishing Ltd 2008 Transfer of Knowledge 711 Alliance success (0.73) a. Last year’s net results from our business with this partner were satisfactory. b. Our cooperation with this partner must be characterized as a success. c. Our cooperation with this partner has contributed to growth in our firm. Trustworthiness (0.91) – scale adapted from Mayer and Davis (1999) a. Integrity (0.83) – This firm has a strong sense of justice. – We never have to worry about whether this firm will keep to its word. – The management team in this firm tries hard to be fair in dealings with others. – Sound principles seem to guide the behaviour of this firm. b. Benevolence (0.83) – This firm really looks out for what is important for us. – This firm is very concerned about our welfare. – Our needs and desires are very important to the management in this firm. – This firm will go out of its way to help us. c. Competence (0.89) – This firm is very capable of performing its job. – We feel very confident about this firm’s skills. – This firm is known to be successful at the things it tries to do. – This firm has much knowledge about the work it does. REFERENCES Almeida, P., Song, J. and Grant, R. M. (2002). ‘Are firms superior to alliances and markets? An empirical test of cross-border knowledge building’. Organization Science, 13, 147–61. Anderson, E. (1990). ‘Two firms, one frontier: on assessing joint venture performance’. Sloan Management Review, 31, 19–30. Anderson, J. and Narus, J. (1990). ‘A model of the distributor firm and manufacturing firm working partnerships’. Journal of Marketing, 54, 42–58. Badaracco, J. L. (1991). The Knowledge Link: How Firms Compete Through Strategic Alliances. Boston, MA: Harvard Business School Press. Barney, J. B. (1991). ‘Firm resources and sustained competitive advantage’. Journal of Management, 17, 99–120. Baron, R. M. and Kenny, D. A. (1986). ‘The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations’. Journal of Personality and Social Psychology, 51, 1171–82. Becerra, M. and Gupta, A. K. (2003). ‘Perceived trustworthiness within the organization: the moderating impact of communication frequency on trustor and trustee effects’. Organization Science, 14, 32– 44. Brown, J. C. and Duguid, P. (2001). ‘Knowledge and organization: a social-practice perspective’. Organization Science, 12, 198–213. Butler, J. K. (1991). ‘Towards understanding and measuring conditions of trust: evolution of a condition to trust inventory’. Journal of Management, 17, 643–63. Chakravarthy, B., McEvily, S., Doz, Y. L. and Rau, D. (2003). ‘Knowledge management and competitive advantage’. In Easterby-Smith, M. and Lyles, M. A. (Eds), Handbook of Organizational Learning and Knowledge Management. Oxford: Blackwell, 305–23. © Blackwell Publishing Ltd 2008 712 M. Becerra et al. Churchill, G. A. (1979). ‘A paradigm for developing better measures of marketing constructs’. Journal of Marketing Research, 16, 64–73. Cummings, L. L. and Bromiley, P. (1996). ‘The organizational trust inventory (OTI): development and validation’. In Kramer, R. M. and Tyler, T. R. (Eds), Trust in Organizations: Frontiers of Theory and Research. Thousand Oaks, CA: Sage. Das, T. K. and Teng, B. S. (2001). ‘A risk perception model of alliance structuring’. Journal of International Management, 7, 1–29. Dhanaraj, C., Lyles, M. A., Steensma, H. K. and Tihanyi, L. (2004). ‘Managing tacit and explicit knowledge transfer in IJVs: the role of relational embeddedness and the impact of performance’. Journal of International Business Studies, 35, 428–42. Doz, Y. L. (1996). ‘The evolution of cooperation in strategic alliances: initial conditions or learning processes?’. Strategic Management Journal, 17, Special Issue, 55–84. Dussauge, P., Garrette, B. and Mitchell, W. (2000). ‘Learning from competing partners: outcomes and durations of scale and link alliances in Europe, North America and Asia’. Strategic Management Journal, 21, 99–126. Grant, R. M. (1996). ‘Prospering in dynamically competitive environments: organizational capability as knowledge integration’. Organization Science, 7, 375–87. Gulati, R. (1998). ‘Alliances and networks’. Strategic Management Journal, 19, 293–17. Gulati, R. (1999). ‘Network location and learning: the influence of network resources and firm capabilities on alliance formation’. Strategic Management Journal, 20, 397–420. Hamel, G. (1991). ‘Competition for competence and interpartner learning within international strategic alliances’. Strategic Management Journal, 12, 83–103. Hansen, M. T. (1999). ‘The search-transfer problem: the role of weak ties in sharing knowledge across organization subunits’. Administrative Science Quarterly, 44, 82–111. Heide, J. B. and Miner, A. S. (1992). ‘The shadow of the future: effects of anticipated interaction and frequency of contact on buyer-seller cooperation’. Academy of Management Journal, 35, 265– 91. Hladik, K. J. (1988). ‘R&D and international joint ventures’. In Contractor, F. J. and Lorange, P. (Eds), Cooperative Strategies in International Business. Lexington, MA: Lexington Books, 187–203. Huemer, L. (1998). Trust in Business Relations: Economic Logic or Social Interaction? Boréa: Umeå. Inkpen, A. C. (1998). ‘Learning and knowledge acquisition through international strategic alliances’. Academy of Management Executive, 12, 69–80. Inkpen, A. C. and Beamish, P. W. (1997). ‘Knowledge, bargaining power, and the instability of international joint ventures’. Academy of Management Review, 22, 177–202. Inkpen, A. and Dinur, A. (1998). ‘Knowledge management processes and international joint ventures’. Organization Science, 9, 454–68. Johnson-George, C. and Swap, W. C. (1982). ‘Measurement of specific interpersonal trust: construction and validation of a scale to assess trust in a specific other’. Journal of Personality and Social Psychology, 43, 1306–17. Kale, P., Singh, H. and Perlmutter, H. (2000). ‘Learning and protection of proprietary assets in alliances’. Strategic Management Journal, 21, Special Issue, 217–38. Khanna, T., Gulati, R. and Nohria, N. (1998). ‘The dynamics of learning alliances: competition, cooperation and relative scope’. Strategic Management Journal, 19, 193–210. Kogut, B. and Zander, U. (1992). ‘Knowledge of the firm, combinative capabilities and the replication of technology’. Organization Science, 3, 383–97. Krishnan, R., Martin, X. and Noorderhaven, N. G. (2006). ‘When does trust matter to alliance performance?’. Academy of Management Journal, 49, 894–917. Lane, P. J., Salk, J. E. and Lyles, M. A. (2001). ‘Absorptive capacity, learning, and performance in international joint ventures’. Strategic Management Journal, 22, 1139–61. Lei, D. J., Slocum, W. J. and Pitts, R. A. (2001). ‘Building cooperative advantage: managing strategic alliances to promote organizational learning’. Journal of World Business, 32, 203–23. Liebeskind, J. P., Oliver, A. L., Zucker, L. and Brewer, M. (1996). ‘Social networks, learning and flexibility: sourcing scientific knowledge in new biotechnology firms’. Organization Science, 7, 428– 43. Lubatkin, M., Florin, J. and Lane, P. (2001). ‘Learning together and apart: a model of reciprocal interfirm learning’. Human Relations, 54, 1353–82. Luo, Y. (2002). ‘Building trust in cross-cultural collaborations: towards a contingency perspective’. Journal of Management, 28, 669–94. © Blackwell Publishing Ltd 2008 Transfer of Knowledge 713 Lyles, M. A. and Salk, J. E. (1996). ‘Knowledge acquisition from foreign parents in international joint ventures: an empirical examination of the Hungarian context’. Journal of International Business Studies, 27, 877–904. Madhok, A. (1995). ‘Revisiting multinational firm’s tolerance for joint ventures: a trust-based approach’. Journal of International Business Studies, 26, 117–37. Mayer, R. C. and Davis, J. H. (1999). ‘The effect of the performance appraisal system on trust for management. A field quasi-experiment’. Journal of Applied Psychology, 84, 123–37. Mayer, R. C., Davis, J. H. and Schoorman, D. (1995). ‘An integrative model of organizational trust’. Academy of Management Review, 20, 709–34. Nunnally, J. C. (1978). Psychometric Theory, 2nd edition. New York: McGraw-Hill. O’Reilly, C.A., Chatman, J. A. and Anderson, J. C. (1987). ‘Message flow and decision making’. In Jablin, F. M., Putman, L. L., Roberts, K. H. and Porter, L. W. (Eds), Handbook of Organizational Communication: An Interdisciplinary Perspective. Newbury Park, CA: Sage. Parkhe, A. (1993). ‘Strategic alliance structuring: a game theoretic and transaction cost examination of interfirm cooperation’. Academy of Management Journal, 36, 794–829. Peteraf, M. (1993). ‘The cornerstones of competitive advantage: a resource based view’. Strategic Management Journal, 14, 179–91. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y. and Podsakoff, N. P. (2003). ‘Common method biases in behavioral research: a critical review of the literature and recommended remedies’. Journal of Applied Psychology, 88, 879–903. Polanyi, M. (1967). The Tacit Dimension. Garden City, NY: Anchor. Ring, P. S. and Van de Ven, A. (1994). ‘Development processes of cooperative interorganizational relationships’. Academy of Management Review, 19, 90–118. Rotter, J. B. (1971). ‘Generalized expectancies for interpersonal trust’. American Psychologist, 26, 443– 52. Rousseau, D. M., Sitkin, S., Burt, R. and Camerer, C. (1998). ‘Not so different after all: a cross discipline view of trust’. Academy of Management Review, 23, 393–404. Simonin, B. (1999). ‘Transfer of marketing know-how in international strategic alliances: an empirical investigation of the role and antecedents of knowledge ambiguity’. Journal of International Business Studies, 30, 463–90. Simonin, B. (2004). ‘An empirical investigation of the process of knowledge transfer in international strategic alliances’. Journal of International Business Studies, 35, 407–27. Sitkin, S. B. and Pablo, A. L. (1992). ‘Reconceptualizing the determinants of risk behavior’. Academy of Management Review, 17, 9–38. Szulanski, G. and Capetta, R. (2003). ‘Stickiness: conceptualizing, measuring, and predicting difficulties in the transfer of knowledge within organizations’. In Easterby-Smith, M. and Lyles, M. A. (Eds), Handbook of Organizational Learning and Knowledge Management. Oxford: Blackwell, 513–34. Teece, D. and Pisano, G. (1994). ‘The dynamic capabilities of firms: an introduction’. Industrial and Corporate Change, 3, 537–55. Yan, A. and Gray, B. (1994). ‘Bargaining power, management control, and performance in United States– China joint ventures: a comparative case study’. Academy of Management Journal, 37, 1478–517. Zaheer, A., McEvily, B. and Perrone, V. (1998). ‘Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance’. Organization Science, 9, 123–42. Zander, U. and Kogut, B. (1995). ‘Knowledge and the speed of the transfer and imitation of organizational capabilities: an empirical test’. Organization Science, 6, 76–92. © Blackwell Publishing Ltd 2008