Academy of Management Journal 2006, Vol. 49, No. 5, 894–917.

WHEN DOES TRUST MATTER TO ALLIANCE PERFORMANCE?
REKHA KRISHNAN Simon Fraser University XAVIER MARTIN NIELS G. NOORDERHAVEN Tilburg University
We examine how uncertainty moderates the trust-performance relationship in alliances, building on the distinction between behavioral uncertainty, which relates to anticipating and understanding partners’ actions, and externally caused environmental uncertainty. We argue that trust matters more to performance under behavioral uncertainty and less under environmental uncertainty. In data from 126 international alliances, the positive relationship between trust and performance is stronger under high behavioral uncertainty and weaker under high environmental uncertainty. We conclude that partners should concentrate on developing interorganizational trust where potential improvement in alliance performance justifies this effort, which in turn depends on the type of uncertainty faced.

Strategic alliances blur firm boundaries and create mutual dependence between previously independent firms (McEvily, Perrone, & Zaheer, 2003). A distinctive characteristic of strategic alliances is that partners have to deal not only with the uncertainty in their environment but also with the uncertainty arising from each other’s behavior (Harrigan, 1985). Because of partners’ dependence on each other, previous research has emphasized the importance of relational factors for the smooth functioning of strategic alliances (Powell, 1990). Although various relational mechanisms and norms have been studied, including for instance norms of solidarity and flexibility (Poppo & Zenger, 2002: 712), none has received more attention than trust (Gambetta, 1988; Mayer, Davis, & Schoorman, 1995; McEvily et al., 2003; Sako, 1991; Zaheer, McEvily, & Perrone, 1998; Zand, 1972). Accordingly, a great deal of research in this tradition has identified interorganizational trust as a key factor contributing to alliance success, the general view being that trust has a positive effect on alliance performance (e.g., Dyer & Chu, 2003; Mohr & Spekman, 1994; Zaheer et al., 1998). The existence of trust between alliance partners cannot be taken for granted, however; partners may not only have to cultivate trust intentionally (Parkhe, 1998; Sako, 1991), but may also incur substantial real and opportunity costs in its pursuit (McEvily et al.,
The paper benefited greatly from the insights of our associate editor and anonymous reviewers. Any remaining errors are ours. This research was supported by the CentER for Economic Research at Tilburg University.
894

2003; Poppo & Zenger, 2002). Furthermore, interorganizational trust need not always improve alliance performance (McEvily et al., 2003). Indeed, researchers are beginning to recognize that the relationship between trust and alliance performance may be complicated and contingent on other factors. Thus, Carson, Madhok, Varman, and John (2003) argued that the effect of trust on task performance in vertical R&D collaborations strengthens with the clients’ ability to understand the tasks involved. Langfred (2004) argued that the effect of trust on the performance of self-managing teams reverses when individual autonomy is high. These studies suggest that the benefits derived from trust may magnify under certain conditions and diminish under other conditions. However, contingent reasoning has yet to be applied to the effect of uncertainty, two forms of which—behavioral and environmental—are potentially the most fundamental strategic factors in alliances (Harrigan, 1988; Kogut, 1989; Williamson, 1985, 1991). There is extensive support in prior research for the overall beneficial effect of trust. Empirical studies have shown that trust, by bringing about good faith in the intent, reliability, and fairness of partner behavior (Sako, 1991; Zaheer et al., 1998), allows for constructive interpretation of partner motives (Uzzi, 1997), reduces the potential for conflict (Zaheer et al., 1998), and encourages smooth information flow between partners (Sako, 1991; Zand, 1972). Trust thus mitigates uncertainty about partner behavior. Yet the same qualities of trust that mitigate uncertainty about partner behavior and engender its beneficial effects may also limit the cognitive efforts of partners when they consider

2006

Krishnan, Martin, and Noorderhaven

895

their broader environment. Trust researchers have argued theoretically (Webb, 1996: 292) and found empirically (Langfred, 2004) that the perception of reliability of information from a partner and the cognitive comfort that trust brings about also limit variety of thought and action and attentiveness to detail. Therefore, trust may reduce the alertness needed when alliance partners have to respond to environmental uncertainty. The result may be that partners respond to external challenges inadequately or not at all. Thus, as we discuss in more detail below, trust seems to entail a trade-off between the capacity to deal with behavioral uncertainty and environmental uncertainty. In this article, we examine this trade-off by theorizing that trust has different effects on alliance performance, depending on the levels of behavioral and environmental uncertainty present. We define a strategic alliance as any extended cooperative agreement intended to jointly develop, manufacture, and/or distribute products (Gulati, 1998; Zollo, Reuer, & Singh, 2002: 701). Although trust helps alliance partners to cope with uncertainties pertaining to each other’s behavior, trust also tends to constrain partners’ responses to environmental demands, thus hindering them from responding appropriately to environmental uncertainty. Below, we develop these arguments in greater detail and report tests of the resulting hypotheses in a sample of 126 international strategic alliances in India.

situations most prominently include alliances (Gulati, 1995; Mohr & Spekman, 1994). One important concern with alliances is that conflict between partners can occasion high costs or a premature breakdown of relationships (Zaheer et al., 1998). Trust helps defuse such conflict, because trusting partners are more likely to interpret each other’s equivocal actions in a manner conducive to the stability of the relationship. As Noorderhaven (2004) argued on the basis of case studies by Doz (1996), if a firm encounters unexpected actions by its partner that could be ascribed to both good and bad intentions, the presence of trust reduces the likelihood of a negative interpretation. For instance, when confronted with disappointing sales of a product line, a partner might either explain the inadequate performance on the basis of an ineffective promotional campaign, or view the failure as signaling a lack of commitment of the other party’s distributors. In such equivocal situations, trust facilitates mutual understanding and allows for the benefit of the doubt. It thus reduces the costs of interpartner conflict as well as other transaction costs (Dyer & Chu, 2003; Zaheer et al., 1998). Research also shows that such costs are negatively related to alliance performance (Zaheer et al., 1998). Therefore, all else being equal, trust improves alliance performance. Hence: Hypothesis 1. Ceteris paribus, trust is positively related to alliance performance. The above hypothesis does not mean that trust improves the performance of all alliances equally. Next, we argue that the type of uncertainty that alliance partners encounter moderates the relationship between trust and alliance performance. In this respect, we distinguish between two types of uncertainty that are relevant to interfirm relations (Sutcliffe & Zaheer, 1998; Williamson, 1985): environmental uncertainty, which results from exogenous sources outside the scope of the alliance, and behavioral uncertainty, which results from difficulty in anticipating and understanding the actions of an exchange partner. Below, we develop predictions that the trust-performance relationship is likely to be stronger under behavioral uncertainty but weaker under environmental uncertainty. Trust, Behavioral Uncertainty, and Alliance Performance Concern about partner behavior is a predominant source of internal tension in strategic alliances (Park & Ungson, 2001; Parkhe, 1993; Sutcliffe & Zaheer, 1998). Such concerns stem in large measure from behavioral uncertainty—that is, the po-

TRUST AND ALLIANCE PERFORMANCE The concept of trust has received ample attention from various disciplines, and although prior research has put forth diverse interpretations of trust, a common core emerges. Building on this prior research, we define interorganizational trust as the expectation held by one firm that another will not exploit its vulnerabilities when faced with the opportunity to do so (Barney & Hansen, 1994; Mayer et al., 1995; Sako, 1991). This expectation is confirmed when parties (1) demonstrate reliability by carrying out their promises, (2) act fairly when dealing with each other, and (3) exhibit goodwill when unforeseen contingencies arise. Our definition thus bases interorganizational trust on three related components: reliability, fairness, and goodwill (Dyer & Chu, 2003). The goodwill component in the definition of trust extends beyond contractual obligations in that partners commit themselves and make contributions to their relationship that go beyond what was explicitly guaranteed (Sako, 1991: 453). Hence, trust stands to be relevant in situations where firms make substantial and open commitments to a partnership. These

896

Academy of Management Journal

October

tential inherent in a situation for difficulty anticipating and understanding actions. Although this feature of the alliance context is never zero, its magnitude varies among alliances. Specifically, potential inability to anticipate and understand actions is higher in alliances of the following two types (Das & Teng, 2000; Park & Ungson, 2001): (1) alliances involving high interdependence, in which the partners’ contributions are highly intertwined (Nooteboom, 2002; Park & Russo, 1996; Stinchcombe, 1985) and (2) alliances in which each partner is able to further private interests at the expense of collaborative interests (Khanna, Gulati, & Nohria, 1998; Park & Ungson, 2001), especially alliances between potential competitors (Bleeke & Ernst, 1993; Hamel, 1991; Kogut, 1988; Oxley & Sampson, 2004; Park & Russo, 1996). Interdependence. The degree of interdependence in an alliance increases with the importance and extent of the resources shared between partners and with the resulting overlap in division of labor between them (i.e., the resulting shared responsibility for a number of tasks) (Gulati & Singh, 1998; Kumar & Seth, 1998; Thompson, 1967). Alliances that are set up to share production facilities typically create only weak interdependencies (Gulati & Singh, 1998). Resource allocations and role assignments in these partnerships tend to be straightforward and stable, and the division of labor is thus likely to be simple. In contrast, alliances formed to jointly develop new technology or to speed up innovation (an alliance for designing a leading-edge microprocessor, for example) lead to high interdependence (Nickerson & Zenger, 2004: 620; Park & Russo, 1996). These alliances are characterized by substantial overlap between the partners’ responsibilities and involve ongoing mutual adjustment between partners (Gulati & Singh, 1998). In highly interdependent activities, coordination is difficult because the complexity associated with interdependence discourages coordination by standardization. Instead, the overlapping division of labor calls for coordination by mutual adjustment, which precludes the use of standard rules to streamline decision making and regularize interactions between interdependent alliance partners (Thompson, 1967). The variability associated with highly interdependent alliances renders coordination by mutual adjustment highly demanding and difficult in terms of communication and decision efforts (Gulati & Singh, 1998; Thompson, 1967). As a consequence, the higher the interdependence, the more likely that any change one partner makes will affect the other partner in unplanned ways, and the more immediate and severe the adverse impact of any mistake (inten-

tional or not) by a partner (Nooteboom, 2002; Thompson, 1967). In addition, high interdependence in alliances requires partners to share valuable knowledge-intensive resources, exposing these to each other (Kumar & Seth, 1998; Nooteboom, 2002; Park & Russo, 1996; Park & Ungson, 2001). Being harder to observe, value, and protect, shared knowledge-intensive resources increase the potential for misunderstandings concerning each partner’s intents and contributions to the alliance (Oxley, 1999). The difficulty in discerning the closely intertwined partner contributions further threatens the open sharing of resources and information among partners and consequently magnifies the coordination difficulties in high-interdependence alliances (Park & Russo, 1996). Interorganizational trust stands to be especially beneficial in the presence of such behavioral uncertainty. By asserting good faith in the intent and reliability of partner behavior, trust allows partners to engage in constructive interpretation of each other’s actions (Zaheer et al., 1998). It also encourages partners to be aware of the processes and procedures that each partner follows (Gulati & Singh, 1998). Thus, trust encourages partners to remain flexible when managing their interface in the face of interdependence. It also alleviates apprehensions regarding the sharing of valuable information; the resulting information exchange and socialization assist in maintaining effective integration and coordination. Under high interdependence, interorganizational trust is therefore essential for alliance performance, as it facilitates mutual adjustment and allows the smoother synchronization of critical tasks. Conversely, we expect trust to have a weaker effect on alliance performance under conditions of low interdependence, as the overlap in the division of labor between partners is lower and hence the scope for misinterpretations and tensions is likely to be lower as well (Gulati & Singh, 1998). Hence, Hypothesis 2. The positive relation between trust and alliance performance is stronger in alliances with a high degree of interdependence than in alliances with low interdependence. Interpartner competition. Interpartner competition exists when a partner tries to maximize its private interests at the expense of the alliance or the other partner (Baum, Calabrese, & Silverman, 2000; Park & Russo, 1996; Park & Ungson, 2001). In alliances formed between potential competitors, concerns about opportunistic exploitation loom especially large, because the partners may have strong incentives to appropriate each other’s resources (Khanna et al., 1998; Oxley & Sampson, 2004). Prior research has shown that alliances be-

2006

Krishnan, Martin, and Noorderhaven

897

tween potential competitors engender greater tendencies of partners to engage in such “de facto internalization” (Baum et al., 2000; Hamel, 1991: 84). Moreover, because potential competitors are familiar with the areas that their partners operate in, the potential competitors have superior capacity to absorb and reuse proprietary knowledge (Cohen & Levinthal, 1990; Park & Russo, 1996). In situations that lend themselves to interpartner competition, the potential to misunderstand the motives of a partner is significant, and this exacerbates a firm’s tendencies to protect its own resources, especially knowledge, at the risk of hampering the alliance relationship (Hamel, 1991; Kale, Singh, & Perlmutter, 2000). By detracting partners from contributing fully to the performance of the alliance, these concerns interfere with the realization of the synergistic benefits of the alliance (Grindley, Mowery, & Silverman, 1994; Madhok & Tallman, 1998). Trust can counteract such problems by increasing each partner’s confidence that the other will not abuse its vulnerabilities (Barney & Hansen, 1994; Bradach & Eccles, 1989; Mayer et al., 1995). Faith in the intentions and fairness of the other makes each partner more likely to respect the boundaries of the other’s resources and proprietary knowledge. This mutual respect encourages partners to provide the substantive resources and accurate and timely information that enhance collaborative benefits (Sako, 1991; Zand, 1972). Trust is all the more advantageous when the potential for interpartner competition is high, because it facilitates mutual understanding and counteracts the attendant failure to cooperate. In alliances with low potential for interpartner competition, conversely, the appropriation of resources is less likely to be of strategic concern. Hence, partners’ suspicions regarding each other’s intents within and outside the alliance are less crippling. Thus, the benefits of interorganizational trust are lower. Hypothesis 3. The positive relation between trust and alliance performance is stronger in alliances in which the potential for interpartner competition is high than in alliances in which the potential for interpartner competition is low. Trust, Environmental Uncertainty, and Alliance Performance Environmental uncertainty results from changes in the economic conditions faced by an organization that are outside its control and hard to anticipate (Dess & Beard, 1984; Koopmans, 1957), such as instability or unpredictability in markets (Aldrich,

1979; Dess & Beard, 1984; Wholey & Brittain, 1989). Environmental uncertainty demands speedy and responsive decisions (Huber, Miller, & Glick, 1990: 13; Mintzberg, 1978). This pressure requires organizations to engage in significant scanning of their environment in search of accurate and reliable information that enables them to interpret and act upon the threats and opportunities facing them (Aguilar, 1967; Anderson & Paine, 1975; Hambrick, 1982). Unpredictable changes in the environment stand to affect the performance of an alliance (Harrigan, 1985; Kogut, 1989). To sustain performance in an uncertain environment, alliance partners need to monitor changes and adjust the alliance’s strategy accordingly (Harrigan, 1985). Anderson and Paine argued that in adjusting strategy, “The critical area is not uncertainty per se but the processing of accurate information to deal with uncertainty” (1975: 814). This information processing may be a bottleneck because of problems of information overload (Mintzberg, 1978; Robertson, 1980), which are exacerbated by information unfamiliarity (Park & Sheath, 1975). In highly uncertain environments, cognitive limitations may introduce considerable limitations and biases in the decision-making process, by prompting the application of inappropriate rules-of-thumb (Barnes, 1984; Cyert & March, 1963; Schwenk, 1984). The risk that biases will enter decision making in uncertain environments is greater in the presence of trust, especially in alliances. Scholars are beginning to recognize the heuristic quality of trust (McEvily et al., 2003; Nooteboom, 2002; Uzzi, 1997). Like cognitive heuristics (Bazerman, 1998), trust enables decision making under conditions of uncertainty, but may also produce systematic biases that can result in significant errors (Ferrin & Dirks, 2003). Specifically, when partners trust each other, their tendency to screen the information provided by the other for accuracy decreases, and their inclination to accept the information at face value increases (McEvily et al., 2003; Szulanski, Cappetta, & Jensen, 2004; Uzzi, 1997). Because trust accustoms partners to rely on each other without doubt, each partner tends to rely more extensively on the other’s knowledge of the environment when scanning the opportunities and threats faced by the alliance, while paying less attention to the completeness and veracity of the information thus obtained. Trust encourages partners to minimize redundancies in the search process by exploiting each other’s purported expertise to engage in specialized search. For instance, in international alliances it is common for the local partner to scan the environment for regulatory changes or changes in consumer preferences in the local market, while the foreign partner monitors technological changes, global demand, or new competition from foreign

898

Academy of Management Journal

October

firms (Beamish & Banks, 1987). Such tendencies tend to be more pronounced in the presence of trust. Prior research has also shown that confronting alternative views and diverse information stimulates creativity and constructive criticism (Jehn, Northcraft, & Neale, 1999; Simons, Pelled, & Smith, 1999). Specialized search as described above, on the contrary, reduces variety in information and restricts the cross-fertilization of viewpoints required for crafting well-informed responses to the environment (Webb, 1996). In alliances, because trust functions as a simplifying heuristic that constrains cognitive effort (McEvily et al., 2003), it also stands to bias partners’ efforts to scan and make sense of their environment, and therefore it may result in suboptimal responses. Trusting alliance partners may even experience “strategic blindness”— outright insensitivity to environmental changes (McEvily et al., 2003: 97). Because partners commit resources, effort, and time in the process of cultivating trust, they tend to be wary of actions that may damage the relationship (Nooteboom, 2002). For instance, if responding to environmental change would require major changes, such as bringing in a new partner or ending an alliance (Harrigan, 1985), partners may be apprehensive about the eventuality of having to cultivate trust and adjust to a new partner all over again, or having to go it alone. In the presence of such apprehensions, partners tend to weigh losses resulting from responding to the environment more than the gains that might come about (Bazerman, 1998; Nooteboom, 2002). Partners may prefer “inaction over action and status quo over any alternatives” (Kahneman & Lovallo, 1993: 18), culminating in their alliance failing to respond to demands of its environment. Overall, interorganizational trust stands to result in inadequate response to the challenges posed by an uncertain environment because it limits cognitive efforts or even causes strategic blindness. Notwithstanding good intentions, trust may thus lead partners into making slow and suboptimal decisions for their alliance, or even no decision at all. This places the alliance at variance with environmental demands. We expect that as a result the effect of trust on alliance performance will not be as positive in the presence of high environmental uncertainty. Under low environmental uncertainty, on the contrary, complete and accurate environmental scanning is less critical. Furthermore, less effort is required to adjust to the environment because of its stability and predictability. Therefore, the limiting effects of trust discussed above tend to be less relevant. Scholars have consistently argued that hard-to-predict changes in market environment create critical

uncertainty for organizations (Cameron, Kim, & Whetten, 1987; Delacroix & Swaminathan, 1991; Dess & Beard, 1984: 56). Many researchers have used the concept of instability, which Child (1972) defined as the degree of difference involved in an environmental change, to capture environmental uncertainty (e.g., Bergh & Lawless, 1998; Keats & Hitt, 1988; Snyder & Glueck, 1982). Others have argued that instability is but one dimension of uncertainty. Unpredictability— that is, the degree of irregularity in an overall pattern of environmental change (Child, 1972)—also magnifies the consequences of a changing environment (e.g., Lawrence & Lorsch, 1973; Wholey & Brittain, 1989). Because environmental uncertainty is a function of both instability and unpredictability (Buchko, 1994), these two distinct dimensions of market variation are both relevant to our investigation of the joint consequences of trust and environmental uncertainty. That is, we expect instability and unpredictability to each reduce the trust-performance relationship. Hypothesis 4. The positive effect of interorganizational trust on alliance performance is weaker when market instability is high than when it is low. Hypothesis 5. The positive effect of interorganizational trust on alliance performance is weaker when market unpredictability is high than when it is low. METHODS Data Data were collected through a survey of international strategic alliances operating in India. As stated above, we identified strategic alliances as extended cooperative agreements intended to jointly develop, manufacture, and/or distribute products (Gulati, 1998; Zollo et al., 2002). In the last 15 years, India has become one of the most attractive investment locations in the world (A.T. Kearney, 2004). Strategic alliances have been a prevalent mode of entry into the Indian market, as in the rest of the world, and the number of strategic alliances in India has also risen by more than 50 percent over the past decade (Bhaumik, Beena, Bhandari, & Gokarn, 2003). Furthermore, different industries in India exhibit sharply different patterns in the variation of demand over time. These differences made strategic alliances in India a relevant empirical setting in which to study the role of environmental uncertainty. Thus, India provided a rich and suitable context in which to study the conditions for successful strategic alliances.

2006

Krishnan, Martin, and Noorderhaven

899

To obtain a target population of international strategic alliances, we examined Capitaline, a secondary database, and member lists of various international chambers of commerce in India. We thus identified a sample of 700 dyadic international strategic alliances operating in India. Following Parkhe (1993) and Simonin (1999), we designed the research to aim at respondents highly knowledgeable about their firms’ alliances. The sensitivity of the questions, and the fact that their top executives deal directly with most international alliances in India, demanded that managing directors or chief executive officers fill in the questionnaire. These target respondents’ names were identified using Capitaline and chamber of commerce data. Data Collection The questionnaires were designed and the survey implemented according to Dillman’s (2000) tailored design method, which suggests several ways to encourage response. The measurement items were generated through a review of prior alliance literature. We used university faculty and doctoral students to assess whether the content of the items tapped the conceptual domain of the focal construct (DeVellis, 1991). This assessment yielded a set of fine-tuned questionnaire items that we used in personal interviews and early pretests with managing directors of Indian firms involved in international strategic alliances to verify the clarity of the items. We modified the wording of a few items slightly as a result and enriched one measure, as described below. This process further strengthened content validity. Appendix A reports the survey items used in this study. The first wave of questionnaires was sent to managing directors and senior executives of 700 Indian firms with international alliances. This wave of survey mailings was followed, four weeks later, by a second wave. Of the 700 managing directors and senior executives who received questionnaires, 126 responded, yielding an 18 percent response rate. This rate is comparable to those of recent surveys of alliance managers in other emerging economies, such as 14.4 percent for China (Isobe, Makino, & Montgomery, 2000) and 19 percent for Mexico (Robins, Tallman, & Lindquist, 2002). All the responses to our survey came from individuals directly responsible for the alliances: 80 came from chairpersons and managing directors of the alliances, 30 from presidents, vice presidents, and general managers, and 16 from full-time directors. Nearly 75 percent of the respondents had been with their firms for more than 5 years, and of these almost 25 percent had more than 20 years’ tenure.

The alliance partners of the Indian firms were spread over 21 countries. All alliances in our sample were dyadic, and all belonged to industries in the manufacturing sector, where alliances were more prevalent than in other sectors (e.g., Parkhe, 1993; Simonin, 1999). Tests of proportions showed that the distribution of our responses according to their two-digit manufacturing SIC code was not statistically different from those reported in the two landmark studies by Harrigan (1988) and Ghemawat, Porter, and Rawlison (1986) (chi-squares were 8.11 for 16 degrees of freedom and 10.04 for 17 degrees of freedom, respectively). The top four industries in our sample ranked in the same order as in these studies and, as others have also found (e.g., Parkhe, 1993), were relatively high-technology industries (industrial machinery and equipment, chemicals and allied products, electrical and electronic equipment, and transportation equipment). We checked the potential for nonresponse bias by comparing the characteristics of the respondents to those of the targeted population sample. The results of t-tests for the sizes of the firms (p 0.28) and the age of the local firm (p 0.34) revealed no significant differences between respondent and nonrespondent groups. In line with Mohr and Spekman (1994) and Poppo and Zenger (2002), we also tested for nonresponse bias by comparing early and late respondents. Armstrong and Overton (1977) argued that late respondents are representative of nonrespondents. We found no significant difference between early and late respondents on characteristics such as number of employees of the Indian partner (p 0.18), alliance duration (p 0.29), and investment size (p 0.51). Dependent Variable: Alliance Performance Despite the publication of numerous studies on alliance performance (e.g., Aulakh, Kotabe, & Sahay, 1996; Lane, Salk, & Lyles, 2001; Mohr & Spekman, 1994; Parkhe, 1993), no consensus exists on measuring this construct. The hybrid structures and transitory nature of alliances (Buckley & Glaister, 2002; Olk, 2002) present unique challenges for evaluating performance and hinder the use of two common indicators of firm performance, financial profitability and survival. Most alliances do not report financial performance, which would in any case tend to be biased by partners’ accounting preferences. Survival is an imperfect indicator of success because an alliance can be successful and discontinued—for instance, because it has served its purpose— or unsuccessful and not (yet) discontinued, for instance because the partners still hope to improve the relationship (Yan & Zeng,

900

Academy of Management Journal

October

1999). To circumvent such hurdles, much of the alliance performance research has relied on managers’ evaluations of alliance success (e.g., Aulakh et al., 1996; Isobe et al., 2000; Lin & Germain, 1998; Saxton, 1997). Doing so is appropriate when respondents represent top management (see Olk, 2002). Because the respondents in our sample were very well informed about the alliances in which they were involved, we were confident that it was proper to rely on managerial evaluations of alliance success. Moreover, Geringer and Hebert (1991) found strong correlations between subjective and objective measures of alliance performance. We measured alliance performance using a fiveitem Likert scale reflecting (1) the extent to which the local partner is satisfied with the overall performance of its alliance, (2) the extent to which the local partner perceives the foreign partner to be satisfied with the overall performance of the alliance, (3) the partners’ satisfaction with respect to the attainment of goals, (4) the extent to which the local partner is satisfied with the financial performance of the alliance, and (5) the extent to which the local partner perceives its foreign partner to be satisfied with the financial performance of the alliance. With a Cronbach’s alpha of .90, the performance scale demonstrated high reliability (DeVellis, 1991; Nunally, 1978). Independent Variables Trust. Although no standard scale of interorganizational trust exists, prior studies have typically covered two or more of its main dimensions: reliability, fairness, and goodwill (e.g., Aulakh et al., 1996; Dyer & Chu, 2003; Zaheer et al., 1998). In keeping with this approach, we measured trust using a five-item Likert-type scale adapted from Aulakh et al. (1996) and Sako and Helper (1998) that captures fairness, reliability, and goodwill among alliance partners.1 The measure has high statistical reliability based on Cronbach’s alpha (.85). We also validated our trust scale using field interviews based on a three-point categorization of the interview data, which yielded a high converOur first, second, and third items are specific to the fairness, reliability, and goodwill aspects of trust, respectively. The remaining two items covered all three aspects, indicating the overall level of trust, as opposed to distrust (Sako & Helper, 1998), in the information shared within a relationship. Zaheer et al. (1998) theorized that predictability was a dimension of trust, alongside reliability and fairness, but the predictability items dropped out of their interorganizational trust scale upon factor analysis.
1

gent validity score of .70 (p .05).2 Moreover, the correlation between trust and performance in our data, .52, fitted within the range of .34 and .65 reported in prior studies (e.g., Aulakh et al., 1996; Carson et al., 2004; Geyskens, Steenkamp, & Kumar, 1999; Zaheer et al., 1998). Altogether, these statistics provided evidence that our measure of trust was reasonable and consistent. Interdependence. Strategic rationales for forming an alliance, eight of which Gulati and Singh (1998) identified through an extensive review of the literature, captured the range of value creation motives of the partners. On the basis of our pretest with Indian alliance managers, we added one item to their list: access to technology. As did Gulati and Singh (1998), we assigned the nine strategic rationales to one of the three classes of interdependencies identified by Thompson (1967): pooled, sequential, and reciprocal interdependence. We classified as pooled interdependence three strategic rationales that require limited coordination: sharing costs (e.g., joint materials procurement), sharing production facilities, and sharing financial resources. Three other strategic rationales were classified under sequential interdependence: access to financial resources, access to new markets, and access to technology; these require intermediate levels of coordination. Reciprocal interdependence included the remaining three strategic rationales, which require extensive coordination: sharing complementary technology, joint development of technology, and reduction of time needed for innovation (for additional details, see Gulati and Singh [1998: 796]). Interdependence was the average composite score of the strategic rationales present in a alliance (respondents could identify more than one strategic rationale), weighted by the type of interdependence each rationale represented. We assigned ordinal weights of 3, 2, and 1 for reciprocal, sequential, and pooled interdependence, respectively. Degree of in-

For further detail, see the entry, “Triangulation using field interviews,” in Appendix C. For instance, statements such as the following typically received a score of 3: “Our partner acquired a new brand from ‘company A’ and the distribution was initially with ‘company B.’” “Our partner could have easily lifted their hands and said they wanted to leave the new brand with the reputed ‘company B.’ But they decided to give the distribution rights of the new brand to us. Our partner always goes an extra mile for us.” Statements like the following were assigned a score of 1: “In our alliance with a surgical firm, they sent us a surgical device that we later found had bloodstains in it. We were shocked when our partner failed to take responsibility for this. We did not expect this from a partner we have known for a long time now.”

2

2006

Krishnan, Martin, and Noorderhaven

901

terdependence was thus a composite score, with higher values indicating higher interdependence, calculated as follows:
9

CIi

j 1

w i, j *SR i, j 9 ,

where CIi was the weighted composite interdependence score for alliance i; wi, j were the weights reflecting the level of interdependence implied by each strategic rationale j in alliance i, j 1, . . . , 9; and SRi, j was the presence of strategic rationale j in alliance i. To ensure that the weighting scheme did not affect the robustness of our findings, we reran our models while coding the interdependence variable with alternative specifications based on Gulati and Singh (1998). First, we placed each alliance into one of the three categories of interdependence, on the basis of the highest level of interdependence present in it, where reciprocal was higher than sequential, which was higher than pooled (Thompson, 1967: 55). We obtained results consistent with those below when rank-ordering the categories so that alliances with reciprocal (and possibly other) elements were assigned a value of 2, alliances with sequential (and possibly pooled) elements had a value of 1, and alliances limited to pooled interdependence had a value of 0, as Gulati and Singh (1998: 798) suggested. We also replicated our results with a variable with values between 1 and 9 that were based on the rationale(s) present in an alliance. Interpartner competition. Prior empirical research has classified alliances as being between potential competitors when both partners operated in the same four-digit SIC code (e.g., Mowery, Oxley, & Silverman, 1996; Oxley & Sampson, 2004; Park & Russo, 1996). To capture more accurately the extent of interpartner competition present in the alliances, we refined the implied binary measure into three categories indicating different degrees of competitive overlap. We assigned a score of 2 if an alliance operated in the same four-digit SIC code industry as both partners and neither partner was active in any other four-digit SIC code area. Because the partners in such an alliance are horizontally related and the alliance’s activities are central to their businesses, the concerns about breeding a potential competitor are likely to be very high. We gave a score of 1 if both partners operated in the same four-digit SIC code area as that of the alliance, but one or (most often) both partners were active in other four-digit SIC codes as well. This situation indicates a horizontal relationship among partners, but one in which concerns about potential

competition are somewhat lower than in the situation coded 2 because the overlap is less central to the partners’ businesses. Finally, we assigned a score of 0 if the partners did not operate in the same industries. This category includes pure cases of vertical relationship of partners via alliance (for instance, one partner supplies inputs for the alliance, which operates in the same business as the other partner). Though such an alliance might be strategic for the partners, they operate in different industries, so they are less likely to be potential competitors (Harrigan, 1985). We performed several specification checks. First, we assigned a value of 1 if the primary operations of both partners were in the same four-digit SIC code, and 0 otherwise. Second, we assigned a value of 1 for the highest degree of competitive overlap— that is, if an alliance operated in the same four-digit SIC code as both partners and neither partner was active in any other four-digit SIC code industry, and 0 otherwise. Finally, we used the conventional if less precise operationalization whereby partners are considered rivals if they share an SIC code. The results were robust. Environmental instability. Although environmental instability may have various sources, we chose to concentrate on product-market instability in our empirical analysis. The reason was that, with all of our alliances operating within the same country, instability caused by other sources, such as regulatory regime, was likely to be consistent for the sampled alliances. The dynamics of productmarkets, on the contrary, varied between industries, and were therefore a better source of environmental instability for testing our hypotheses. As in previous research that has followed the same logic, our measure captures instability in the sales of each alliance’s industry over the preceding five years (Bergh & Lawless, 1998; Keats & Hitt, 1988). Data were obtained from Capitaline, an authoritative database of financial information about India. We regressed industry sales on year and divided the standard error of the regression slope coefficient by the mean of industry sales (Bergh & Lawless, 1998; Dess & Beard, 1984). Larger values indicated greater environmental instability. Environmental unpredictability. For the reasons explained above, we concentrated on the unpredictability of product-markets to gauge the effect of environmental unpredictability. Our measure used industry sales data spanning five years, obtained from Capitaline, to capture the extent to which alliance partners could predict future trends in product-markets from the recent past (Glick, Ogilvie, & Miller, 1990; Wholey & Brittain, 1989). We measured unpredictability as the coeffi-

902

Academy of Management Journal

October

cient of alienation (1 – R2) of the regression of industry sales in the survey year on the industry sales of the five preceding years (e.g., Delacroix & Swaminathan, 1991). Control Variables Investment size. Because prior research suggested that the size of partners’ investment in an alliance may affect their commitment to alliance operations (influencing the performance of the alliance), we controlled for investment size (Parkhe, 1993). Following prior research (Luo, 2002), we operationalized investment size as the total amount of investment by both partners measured on a fivepoint interval scale. Cultural distance. Cultural distance could be related to alliance performance (Luo, 2002; Pothukuchi, Damanpour, Choi, Chen, & Park, 2002). Following extant literature, we controlled for national cultural distance by using Kogut and Singh’s (1988) measure, which is based on Hofstede’s (1980) four cultural dimensions. We used Kogut and Singh’s (1988) formula to compute the distance between India’s national culture and that of the country of origin of each foreign partner. Equity alliance. The governance mode within an alliance (Gulati, 1995; Oxley, 1999; Oxley & Sampson, 2004) may be indicative of the motives of the partners and have a large impact on alliance performance (Osborn & Baughn, 1990; Saxton, 1997). We coded alliance governance mode by a binary variable, assigning 1 to alliances that involved the use of equity and 0 for nonequity alliances (e.g., Gulati, 1995; Saxton, 1997). Alliance duration. Partners in long-lasting alliances have had enough time to develop mutual understanding, and thus conflicts that hamper relationship performance may be less likely (Lin & Germain, 1998; Martin, Swaminathan, & Mitchell, 1998). Duration was measured by an item capturing the number of years an alliance had been in existence at the time of measurement (e.g., Kotabe, Martin, & Domoto, 2003; Simonin, 1999). Quality of information exchanged. A higher quality of information exchange may influence alliance performance and trust, irrespective of the level of uncertainty in an alliance (Aulakh et al., 1996). We measured the quality of the information exchanged in an alliance with a five-item Likert scale capturing the frequency, density, and openness of communication, as distinguished by Gupta and Govindarajan (1991). This construct had a Cronbach’s alpha of .81. Position of respondent. We assigned a value of 1 if a respondent held the CEO’s position or its equiv-

alent in a local firm participating in a sampled alliance with a foreign partner, and 0 otherwise. Local partner size. We controlled for the size of the local partner by using the log of the number of employees (e.g., Deeds & Rothaermel, 2003). Industry dummies. Alliances in certain industries may systematically perform better than those in other industries owing to differences in industry structure (Steensma, Tihanyi, Lyles, & Dhanaraj, 2005). To control for industry differences, we used dummy variables for the major industries in our sample, based on two-digit SIC codes. Analysis Measurement analysis was conducted using LISREL’s 8.3 maximum likelihood program (Joreskog & ¨ Sorbom, 1996). We performed confirmatory factor ¨ analysis using LISREL to check for convergent and discriminant validity. We used ordinary least squares regression analysis to examine alliance performance. To control for the possible endogeneity of the choice between equity or nonequity alliance (Sampson, 2004), we implemented Heckman’s (1979) two-stage technique; Appendix B discusses this technique and provides the results of the first-stage probit model. Before calculating the interaction terms used to test Hypotheses 2 through 5, we mean-centered the variables involved (Aiken & West, 1991). RESULTS Reliability and Validity All constructs displayed satisfactory levels of reliability, as indicated by the composite reliabilities ranging from 0.81 to 0.90 (Nunnally, 1978; see Appendix A for the composite reliabilities). Convergent validity, the extent to which different attempts to measure a construct agree (Campbell & Fiske, 1959), can be judged by looking at the item loadings. Each loading ( ) for the multi-item constructs of trust, quality of information exchanged, and alliance performance was significantly related to its underlying factor, and all standardized item loadings were well above the cutoff of .50 (Hildebrandt, 1987), supporting convergent validity. A series of chi-square difference tests on the factor correlations showed that discriminant validity, the extent to which a construct differs from others, was achieved among all the survey-based constructs: trust, quality of information exchanged, and alliance performance (Bagozzi, 1993; Joreskog, ¨ 1971). It was particularly important that discriminant validity be achieved between the constructs of

2006

Krishnan, Martin, and Noorderhaven

903

trust and performance, as these were our focal latent constructs. We constrained the estimated correlation parameter between trust and performance to 1.0 and then performed a chi-square difference test on the values obtained for the constrained and unconstrained models. The significant difference in chi-square ( 2 45.59, df 1, p 0.000) indicated that the two constructs were not perfectly correlated and that discriminant validity was achieved (see Anderson & Gerbing, 1988). We also compared the comparative fit and goodness-of-fit indexes between the constrained and unconstrained models and found that the difference was moderately large ( CFI .11, GFI .09), again suggesting sufficient discriminant validity (e.g., Bagozzi & Yi, 1990). We carried out the same procedures for other construct combinations, too, with similar results. We collected most of our data using a single survey instrument and a single informant per alliance. To address the potential concerns of common method bias and single informant bias, we used several procedural and statistical remedies. Specifically, we undertook the procedural remedies of protecting respondent anonymity, reducing item ambiguity, separating scale items for the trust and performance measures, and obtaining data from different sources for most of the moderator variables (three of four) and several control variables. Our statistical remedies included triangulation of survey data with data obtained from secondary sources and from field interviews, partial correlation adjustment, and Harman’s (1967) one-factor

test. We also examined the results for significant interactions, which are less likely to occur in the presence of single informant bias (Kotabe et al., 2003). Appendix C reports each of these steps in detail. These procedures left us confident that neither common method nor single informant bias was a serious problem in our study. Tests of Hypotheses Table 1 reports means, standard deviations, and correlations for all variables. Table 2 reports the results of the regression model predicting alliance performance. We tested six regression equations for the alliance performance variable, as reported in Table 2. After including only the control variables in model 1, we introduced the main variables in model 2. We introduced the interactions of trust with the behavioral and environmental uncertainty variables separately in models 3 and 4. In model 5, we included all the interaction terms simultaneously. Finally, we reran model 5 with industry dummies to create model 6. Model 1 is significant (p .001), and the control variables explain 31 percent of the variance in alliance performance. Alliance duration (p .05) and the quality of information exchanged (p .001) are positively related to alliance performance. Because of the strong relationship of the quality of information exchanged with alliance performance and the former’s possible overlap with trust, we replicated our models without the quality of information exchanged. The pattern of our findings did not change

TABLE 1 Descriptive Statistics and Correlationsa
Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Alliance performance Position of respondent Investment size Cultural distance Local partner size Equity alliance Alliance duration Quality of information exchanged Interdependence Interpartner competition Environmental instability Environmental unpredictability Trust b Mean 3.75 0.58 2.42 1.86 5.46 0.58 12.10 3.12 0.50 1.28 0.04 0.57

s.d. 0.78 0.49 0.95 0.42 1.63 0.49 9.33 0.74 0.38 0.63 0.03 0.32

1

2

3

4

5

6

7

8

9

10

11

12

13

.04 .17 .09 .15 .14 .18 .49 .11 .24 .03 .11 .52 .13

.06 .03 .20 .03 .19 .07 .03 .15 .01 .03 .09 .03

.18 .39 .17 .15 .30 .23 .06 .07 .04 .15 .00

.23 .02 .02 .04 .03 .18 .01 .10 .01 .05

.06 .25 .19 .02 .22 .06 .05 .05 .00

.09 .34 .14 .06 .00 .15 .27 .62

.06 .12 .03 .01 .00 .17 .11 .32 .14 .03 .03 .43 .27

.03 .03 .04 .14 .00

.01 .11 .12 .00

.19 .14 .00 .07 .00

3.74 0.77 0.00 0.76

.29

a n 126. Correlations with absolute value greater than .17 are significant at the .05 level. Means and standard deviations reported here are for raw scores. b Correction for the endogeneity of the alliance governance choice (equity vs. nonequity). See Appendix B for an explanation.

TABLE 2 Results of Regression Analysis for Alliance Performancea

Variables
0.58 (0.56) 0.07 (0.12) 0.12 (0.11) 0.13 (0.15) 0.04 (0.05) 1.06† (0.61) 0.02* (0.01) 0.58*** (0.08) 0.92 0.15 0.11 0.10 0.03 0.78 0.02* 0.29** (0.72) (0.11) (0.11) (0.14) (0.05) (0.81) (0.01) (0.09) 1.02 0.13 0.12 0.12 0.04 0.91 0.02* 0.29** (0.72) (0.10) (0.11) (0.14) (0.05) (0.80) (0.01) (0.09) 1.06 0.17 0.14 0.07 0.05 0.97 0.02* 0.27** (0.73) (0.11) (0.11) (0.13) (0.05) (0.81) (0.01) (0.09) 1.17 0.15 0.15 0.09 0.05 1.08 0.02* 0.27** (0.75) (0.10) (0.11) (0.13) (0.05) (0.81) (0.01) (0.09)

Hypothesis Number and Prediction Model 1 Model 2 Model 3 Model 4 Model 5

Model 6

Intercept Position of respondent Investment size Cultural distance Local partner size Equity alliance Alliance duration Quality of information exchanged

0.14 0.16 0.10 0.06 1.19 0.02* 0.28**

(0.10) (0.11) (0.14) (0.05) (0.80) (0.01) (0.09)

0.19 0.22* 0.17 0.45† (0.26) 0.47*** (0.09) (0.26) 1, 0.51*** (0.09) 0.48† 0.49†

(0.32) (0.10) (1.13)

0.24 0.22* 0.12

(0.31) (0.09) (0.89)

0.19 0.23* 0.98

(0.31) (0.09) (1.12) (0.26)

0.25 0.23* 1.03 0.52†

(0.32) (0.09) (0.97) (0.27)

0.25 0.23* 0.99 0.53†

(0.33) (0.09) (1.04) (0.27)

Main effects Interdependence Interpartner competition Environmental instability Environmental unpredictability Trust

0.51*** (0.09)

0.46*** (0.08)

0.47*** (0.08)

2, 3, 4, 5,

0.62* 0.30*

(0.27) (0.12) 1.75† 0.43† (0.90) (0.23)

0.64* 0.29* 1.31† 0.49*

(0.31) (0.11) (0.77) (0.21)

0.65* 0.30* 1.33† 0.50*

(0.31) (0.12) (0.78) (0.20)

Interactions Trust interdependence Trust interpartner competition Trust environmental instability Trust environmental unpredictability

Industry dummiesb 0.62 .32 (0.40) (0.50) (0.51) 0.57 (0.50) .49 .17 6.80*** 0.67 (0.51) .53 .04 5.25** 0.71 .51 .03 3.25* 0.80 (0.51) .56 .07 4.25** 0.81 .56 .00 0.16

c

R2 R2 F

a n 126. The changes in R2 in models 3 through 5 are in comparison to the value for R2 in model 2. The coefficients reported are unstandardized estimates, with standard errors in parentheses. b The industry dummies were not significant. c Correction for the endogeneity of the alliance governance choice (equity vs. nonequity). See Appendix B for an explanation. † p .10 * p .05 ** p .01 *** p .001

2006

Krishnan, Martin, and Noorderhaven

905

appreciably. The equity alliance dummy is marginally and negatively related to alliance performance (p .10), but only in the base model. Hypothesis 1 predicts that trust will be positively related to alliance performance. The coefficient for trust in model 2 is positive and significant (b 0.51, p .001), thus supporting Hypothesis 1 and past findings reported in the literature. Given mean centering, the trust coefficient shows the magnitude of the relationship between trust and alliance performance, with other variables held constant at their mean values. Trust, Behavioral Uncertainty, and Alliance Performance The incremental variance accounted for by the interactions between the behavioral uncertainty variables and trust is significant in model 3 ( R2 0.04, p .01). Hypothesis 2 predicts that alliances will benefit more from interorganizational trust when the degree of interdependence is higher. Hypothesis 3 predicts that alliances will benefit more from interorganizational trust when the degree of potential interpartner competition is higher. The coefficient of the interaction of trust with interdependence is significant and positive (b 0.62, p .05 in model 3), supporting Hypothesis 2. We also find a significant interaction between trust and interpartner competition (b 0.30, p .05), thereby supporting Hypoth-

esis 3. These results show that the positive impact of trust on alliance performance increases with interdependence and interpartner competition. To further assess the implications of the regression results, we plotted the relationship of trust and alliance performance over the observed range of trust, with separate regression lines representing different levels of interdependence. We created a similar plot with interpartner competition. The plotted lines represent the performance values expected on the basis of unstandardized regression coefficients from the complete regression (model 5). The low interdependence and low interpartner competition lines indicate values one standard deviation below the mean, and the high interdependence and high interpartner competition lines indicate values one standard deviation above the mean. Figures 1 and 2 graphically support Hypotheses 2 and 3, respectively. The simple slope test (Aiken & West, 1991) reveals that the magnitude of the slope of alliance performance regressed on trust is nearly twice as large for high interdependence (simple slope: b 0.57, t 5.29) as that for low interdependence (simple slope: b 0.32, t 3.55). The slope for high interpartner competition (simple slope: b 0.66, t 6.35) is nearly thrice as large as that for low interpartner competition (simple slope: b 0.26, t 2.41). These results show that the trust-performance relationship

FIGURE 1 Trust, Interdependence, and Alliance Performance

906

Academy of Management Journal

October

FIGURE 2 Trust, Interpartner Competition, and Alliance Performance

strengthens at high levels of the two variables indicating behavioral uncertainty. Trust, Environmental Uncertainty, and Alliance Performance The additional variance accounted for by the interactions between the environmental uncertainty variables and trust is significant in model 4 ( R2 0.03, p .05). Hypothesis 4 predicts that the benefits that alliances derive from interorganizational trust diminish when environmental instability is higher. Hypothesis 5 predicts that the benefits that alliances derived from interorganizational trust diminish when environmental unpredictability is higher. Marginally significant negative effects are found for the coefficient of the interaction of trust with environmental instability in models 4 (b –1.75, p .10) and 5 (b –1.31, p .10), giving some support to Hypothesis 4. Model 4 also shows a marginally negative interaction effect between trust and environmental unpredictability (b – 0.43, p .10), an effect that becomes stronger in the full model 5 (b – 0.49, p .05). Overall, this pattern of findings supports Hypothesis 5. These results suggest that the positive impact of trust on alliance performance diminishes with high instability and unpredictability in the environment of an alliance. To illustrate these interactions, we created plots for the trust–alliance performance relationship,

with separate regression lines representing different levels of environmental instability and unpredictability one standard deviation above and below the mean. Figures 3 and 4 graphically support Hypotheses 4 and 5, respectively. The simple slope test (Aiken & West, 1991) reveals that the magnitude of the slopes of alliance performance regressed on trust is nearly two times smaller for high environmental instability (simple slope: b 0.33, t 3.44) as for low environmental instability (simple slope: b 0.59, t 5.06). The slope for high environmental unpredictability (simple slope: b 0.30, t 3.1) is more than two times smaller than for low environmental unpredictability (simple slope: b 0.61, t 5.75). Further probing of the interactions (Aiken & West, 1991) revealed that for very high levels of environmental unpredictability, the relationship between trust and alliance performance becomes insignificant. The sign of the simple slope coefficient of trust may reverse for very high environmental unpredict– 0.05, t – 0.28, at the high end of ability (b the observable range). These results indicate that the trust-performance relationship weakens and may disappear altogether at high levels of the environmental uncertainty variables. In model 6, we find that the addition of fixed effects for the main industries in the sample does not substantially add to the explanatory power of the regression ( R2 0.00, n.s.). The industry dum-

2006

Krishnan, Martin, and Noorderhaven

907

FIGURE 3 Trust, Environmental Instability, and Alliance Performance

FIGURE 4 Trust, Environmental Unpredictability, and Alliance Performance

mies are not significant as a set ( F 0.16, n.s.). Furthermore, the results reported above, including all tests of hypotheses, do not change substantially. Thus, our results appear robust across industries.

All in all, the results consistently support our argument that the positive relationship between interorganizational trust and alliance performance strengthens under conditions that foster behavioral

908

Academy of Management Journal

October

uncertainty but weakens under conditions of environmental uncertainty. DISCUSSION By now there appears to be general support for the idea that trust is beneficial to alliances. However, recent studies have suggested that the impact of trust on alliance performance may be contingent on other factors. Yet previous research has not yielded a general theory regarding the conditions under which trust facilitates or fails to facilitate alliance performance. We have presented such a theory, one based on the distinction between behavioral and environmental uncertainty, and shown empirically that, apart from the positive direct relationship between trust and alliance performance, more subtle interaction effects can be distinguished. The relationship between trust and alliance performance is moderated by the type of uncertainty prevailing in a particular alliance, with behavioral uncertainty strengthening, and environmental uncertainty weakening, the relationship between trust and performance. Contributions and Implications Our research makes several contributions. First, we extend the interorganizational trust-performance literature by demonstrating that the type of uncertainty facing alliance partners conditions the relationship between trust and alliance performance. Specifically, behavioral and environmental uncertainties have opposite moderating effects on that relationship. The quite distinct challenges posed by these two types of uncertainty bring about these differing effects. Trust essentially reduces the likelihood of negative interpretations of partner actions by allowing for the benefit of the doubt. This allowance facilitates openness in sharing knowledge and reduces fear of opportunistic behavior by partners. Hence, the benefits from trust are magnified when behavioral uncertainty is high. In contrast, the benefits from trust are reduced when environmental uncertainty is strong, because overconfidence in the information provided by each partner restrains the vigilant environmental scanning and cross-fertilization of views that is of vital importance under this condition. This contrast has implications for research on trust, in that it shows that trust can be a double-edged sword, with a performanceenhancing potential that increases under certain conditions but decreases under other conditions. Second, extant research suggests that potential competition and high interdependence between partners are likely to hamper alliances. Behavioral uncertainty related to partner actions, such as the potential

for misappropriation of proprietary know-how, is considerable in alliances that are potentially competitive and/or where partners are highly interdependent. We contribute to this stream of research by empirically showing that trust brings about benefits by attenuating the effects of behavioral uncertainty in alliances where potential interpartner competition and interdependence are high. This finding underscores the potential benefits of investing in trust when behavioral uncertainty is considerable and also suggests that trust figures among the relational mechanisms and norms that can support alliance performance by allowing partners to realize their potential synergies (see Madhok & Tallman, 1998). Moreover, our findings underline the necessity of taking into account both operational features (e.g., interdependence) and behavioral characteristics (e.g., trust) in studying alliance success. Advising such a dual focus runs counter to the emphasis on the isolated influence of either tangible alliance features or behavioral patterns in much prior research (Contractor, 2005; Doz, 1996; Yan & Zeng, 1999). Likewise, our study shows the benefits of examining industry factors simultaneously with alliance- and partner-level effects. Third, the challenges environmental uncertainty poses for firms are well documented. However, researchers’ knowledge about the role of environmental uncertainty in strategic alliance performance is limited. This gap is all the more relevant as environmental uncertainty is commonly advanced as a leading reason that firms form alliances in the first place (e.g., Harrigan, 1988; Pfeffer & Nowak, 1976). Because alliances involve the interests of more than one firm, relational norms such as trust also shape the manner in which partners respond to various external challenges. Therefore, it is vital for alliance researchers to understand the role of trust in shaping responses to the challenges posed by environmental uncertainty, and the implications of this role for performance. Prior research, for the most part, has stressed the beneficial effects of trust. However, our study shows that this beneficial effect decreases with environmental uncertainty and suggests that, at extreme levels of environmental uncertainty, trust may even have a detrimental effect on alliance performance. The cognitive comfort trust provides may reduce the alertness and cross-fertilization needed in the presence of strong environmental uncertainty. Hence, alliance partners should exercise caution in relying on interorganizational trust and accelerate scanning and search efforts under environmental uncertainty rather than relying unduly on each other to deal with exogenous uncertainty. Thus, our study adds to the research agenda on the limits of trust (e.g., McEvily et al., 2003). As for relevance to practice, we have established

2006

Krishnan, Martin, and Noorderhaven

909

that successful international alliances take into account that trust can have different impacts under different types of uncertainty. Because the intentional cultivation of interorganizational trust involves costs as well as opportunities forgone (Poppo & Zenger, 2002; Sako, 1991), such an effort should be undertaken only when the expected payoff is positive. Our results reveal that, at very high levels of environmental unpredictability, the trustperformance relationship disappears and possibly reverses. The costs of cultivating trust might thus outweigh the expected benefits. Hence, firms ought to expend efforts in developing interorganizational trust specifically when it has the potential to help address behavioral uncertainty (in the presence of high interdependence and/or latent competition among partners), and exercise caution when relying on trust under conditions of strong environmental uncertainty (as caused by product-market instability or unpredictability). Limitations and Suggestions for Future Research This study has some limitations. First, we collected data concerning the perspectives of both partners on alliances through a survey of the Indian partners only. Geringer and Hebert (1991: 252, 256) found a significant, positive correlation between a focal parent’s satisfaction with alliance performance and the perception by the other partner of this focal parent’s satisfaction. However, it would be valuable to gain both partners’ perspectives on each alliance. Yet gathering such information could be very challenging, especially with parent firms originating from many countries, as in our sample (21 partner countries). Second, the operationalization of our moderating constructs, behavioral and environmental uncertainty, does not preclude other sources. We do believe that interdependence and interpartner competition are key aspects of behavioral uncertainty (Das & Teng, 2000; Park & Ungson, 2001), though there may be others. Likewise, product-market instability and unpredictability are critical sources of environmental uncertainty and are perhaps its most commonly used indicators in organizational research (e.g. Buchko, 1994; Delacroix & Swaminathan, 1991; Glick et al., 1990; Wholey & Brittain, 1989). Nevertheless, uncertainty stemming from regulatory and political instability may be expected to also matter in the case of alliances located in transition economies (e.g., Delios & Henisz, 2003). Third, single-informant bias is a potential shortcoming of our research. Yet although single informants provided data on both trust and alliance performance, we have several reasons to believe that

single-informant bias is not a serious concern in our study. As outlined in Appendix C, we undertook multiple procedural and statistical remedies, including protecting respondent anonymity, reducing item ambiguity, separating scale items for trust and performance measures, obtaining data from different sources for most of the moderator variables (three of the four) and for several control variables, making partial correlation adjustments, and applying Harman’s (1967) one-factor test; we also validated our trust and alliance performance measures, using field interviews and archival data, respectively. Furthermore, all our interaction effects are significant, especially those that involve two survey-based measures (trust and interdependence). The significance of such interactions is unlikely to be an artifact of the singleinformant method, as the respondents are unlikely to have consciously theorized the moderated relationships when responding to the survey (Brockner, Siegel, Daly, Tyler, & Martin, 1997; Kotabe et al., 2003). Moreover, the interactions also make reverse causality less plausible. Our study suggests a number of interesting opportunities for future research. First, our approach to the study of trust may be generalized beyond interfirm alliances. For instance, trust has been suggested to benefit knowledge sharing in intraorganizational contexts too (e.g., Makino & Inkpen, 2003; Tsai & Ghoshal, 1998). It would be interesting to explore to what extent the moderating effects of behavioral and environmental uncertainty can also be found in the intraorganizational context. Our reasoning suggests that under conditions of strong environmental uncertainty, the effect of high intraorganizational trust may be weakened (and perhaps reversed), following the same course we found in the interorganizational context. Relatedly, our results on environmental uncertainty and trust may have implications for research on team performance. Langfred (2004) showed that teams that rely on trust were less likely to monitor their team members. Future research could examine the implications of reliance on trust for the performance of teams operating under different levels of environmental uncertainty. Second, we focused our analysis on interorganizational trust, which describes the aggregate relationship between partner firms and is the predominant form of trust to affect performance and satisfaction at an organizational level (Zaheer et al., 1998). Trust may also arise between the top managers of partner firms. These individuals generally shape initial alliance strategies (such as whether to use equity governance, a feature that we controlled for) but not ongoing activities (Krishnan, 2006). As Zaheer and colleagues (1998) found, interpersonal trust is unlikely to matter nearly as much to our outcome of

910

Academy of Management Journal

October

interest as interorganizational trust. Nevertheless, it would be interesting to examine any effects of uncertainty on the relationships between interpersonal trust and alliance decisions (and perhaps performance). In this respect, we would expect that individual managers, too, are prone to be blinded by trust in the presence of environmental uncertainty. Third, uncertainty related to appropriation of proprietary knowledge and complete sharing of information is more salient in knowledge-intensive alliances. It has been argued that trust enables firms to cooperate despite such uncertainty (e.g., Dyer & Chu, 2003). Nevertheless, in knowledge-intensive contexts such as R&D and new-product development, environmental uncertainty may be high—and may occur alongside behavioral uncertainty in the case of allied firms (Harrigan, 1988; Martin & Salomon, 2003a, 2003b). Research examining such contexts could yield further insights into the conditions under which the net benefits of trust can be sustained. Fourth, we examined the impact of trust on alliance performance. Other relational mechanisms and norms may have similar contingent effects on alliance performance. For instance, Parkhe (1991) argued that tension within alliances may also result from factors other than behavioral uncertainty, such as cultural differences. He suggested that routines such as training—rather than trust—may go a long way toward reducing cultural conflict and improving alliance performance. Moreover, relational governance may involve “norms of flexibility, solidarity, bilateralism and continuance” (Poppo & Zenger, 2002: 712), norms that can operate in the presence of uncertainty. Further research exploring such alternatives and complements to trust, then, would be well warranted.3 Partner reputation, especially when amplified through a network of alliances, can have potent effects too (Gulati, 1998; Nooteboom, Berger, & Noorderhaven, 1997). Finally, prior ties between partners are a potent source of shared understanding (Zollo et al., 2002), and their role deserves attention alongside trust.

Conclusion Our research provides significant insights into the advantages and limitations of interorganizational trust for strategic alliances. Specifically, the study underscores the need to move beyond a focus on the direct link between trust and alliance performance in seeking to understand the conditions under which trust promotes or inhibits alliance performance. Researchers (and managers) ought to take into account the type of uncertainty facing alliance partners—that is, whether the source of uncertainty is internal or external to their alliance. In our study, the type of uncertainty moderated the relationship between trust and alliance performance in such a way that the trust-alliance performance relationship strengthened under behavioral uncertainty and weakened under environmental uncertainty. We hope that our study triggers future studies that will look in more detail at the complicated and contingent role of trust in inter- and intraorganizational relationships.

REFERENCES
Aguilar, F. J. 1967. Scanning the business environment. New York: Macmillan. Aiken, L. S., & West, S. G. 1991. Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage. Aldrich, H. E. 1979. Organizations and environments. Englewood Cliffs, NJ: Prentice-Hall. Anderson, C. R., & Paine, F. T. 1975. Managerial perceptions and strategic behavior. Academy of Management Journal, 18: 811– 823. Anderson, J. C., & Gerbing, D. W. 1988. Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103: 411– 423. Armstrong, S. J., & Overton, T. S. 1977. Estimating nonresponse bias in mail surveys. Journal of Marketing Research, 14: 396 – 402. Aulakh, P., Kotabe, M., & Sahay, A. 1996. Trust and performance in cross-border marketing partnerships: A behavioral approach. Journal of International Business Studies, 27: 1005–1032. Bagozzi, R. P. 1993. Assessing construct validity in personality research: Applications to measures of self-esteem. Journal of Research in Personality, 27: 49 – 87. Bagozzi, R. P., & Yi, Y. 1990. Assessing method variance in multitrait-multimethod matrices: The case of selfreported affect and perceptions at work. Journal of Applied Psychology, 75: 547–560. Balasubramanyam, V. N. 2003. India: Trade policy review. World Economy, 26: 1357–1368.

It should be noted that distinguishing trust from alternative norms is not straightforward. In Poppo and Zenger (2002: 715), three items yielded one “composite factor with the following underlying norms and dimensions: open communication and sharing of information, trust, dependence, and cooperation” ( .78). Gibson and Birkinshaw (2004), examining intraorganizational functioning, found that nine items meant to measure “support” and “trust” (Ghoshal & Bartlett, 1994) loaded onto a single factor ( .93). Still, we believe that further scale development could yield finer differentiations to support the type of research suggested here.

3

2006

Krishnan, Martin, and Noorderhaven

911

Barnes, J. H. 1984. Cognitive biases and their impact on strategic planning. Strategic Management Journal, 5: 129 –137. Barney, J. B., & Hansen, M. H. 1994. Trustworthiness as a source of competitive advantage. Strategic Management Journal, 15: 175–190. Baum, J. A. C., Calabrese, T., & Silverman, B. S. 2000. Don’t go it alone: Alliance networks and startups’ performance in Canadian biotechnology. Strategic Management Journal, 21: 267–294. Bazerman, M. 1998. Judgement in managerial decision making. New York: Wiley. Beamish, P. W., & Banks, J. C. 1987. Equity joint ventures and the theory of the multinational enterprise. Journal of International Business Studies, 18(2): 1–16. Bergh, D. D., & Lawless, M. W. 1998. Portfolio restructuring and limits to hierarchical governance: The effects of environmental uncertainty and diversification strategy. Organization Science, 9: 87–102. Bhaumik, S. K., Beena, P. L., Bhandari, L. P., & Gokarn, S. 2003. Survey of FDI in India. Working paper no. 6, London Business School, Centre for New and Emerging Markets. Bleeke, J., & Ernst, D. 1993. Collaborating to compete. New York: Wiley. Bowonder, B., & Richardson, P. K. 2000. Liberalization and the growth of business-led R&D: The case of India. R&D Management, 30: 279 –288. Bradach, J. L., & Eccles, R. G. 1989. Price, authority and trust: From ideal types to plural forms. In W. R. Scott & J. Blake (Eds.), Annual review of sociology, vol. 15: 97–118. Palo Alto, CA: Annual Reviews. Brockner, J., Siegel, P. A., Daly, J. P., Tyler, T., & Martin, C. 1997. When trust matters: The moderating effect of outcome favorability. Administrative Science Quarterly, 42: 558 –583. Buchko, A. A. 1994. Conceptualization and measurement of environmental uncertainty: An assessment of the Miles and Snow perceived environmental uncertainty scale. Academy of Management Journal, 37: 410 – 425. Buckley, P. J., & Glaister, K. W. 2002. What do we know about international joint ventures? In F. Contractor & P. Lorange (Eds.), Cooperative strategies and alliances: 49 – 69. Amsterdam: Pergamon. Cameron, K. S., & Kim, M. U., & Whetten, D. A. 1987. Organizational effects of decline and turbulence. Administrative Science Quarterly, 32: 222–240. Campbell, D. T., & Fiske, D. W. 1959. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56: 81–105. Carson, S. J., Madhok, A., Varman, R., & John, G. 2003. Information processing moderators of the effectiveness of trust-based governance in inter-firm R&D collaboration. Organization Science, 14: 45–56.

Child, J. 1972. Organizational structure, environment and performance: The role of strategic choice. Sociology, 6: 2–21. Cohen, W. M., & Levinthal, D. A. 1990. Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35: 128 –152. Contractor, F. J. 2005. Alliance structure and process: Will the two research streams ever meet in alliance research? European Management Review, 2(2): 123–129. Cyert, R. M., & March, J. G. 1963. A behavioral theory of the firm. Englewood Cliffs, NJ: Prentice-Hall. Das, T. K., & Teng, B. S. 2000. Instabilities of strategic alliances: An internal tensions perspective. Organization Science, 11: 77–101. Deeds, D. L., & Rothaermel, F. T. 2003. Honeymoons and liabilities: The relationship between age and performance in research and development alliances. Journal of Product Innovation Management, 20: 468 – 484. Delacroix, J., & Swaminathan, A. 1991. Cosmetic, speculative and adaptive organizational change in the wine industry: A longitudinal study. Administrative Science Quarterly, 36: 631– 661. Delios, A., & Henisz, W. J. 2003. Political hazards, experience, and sequential entry strategies: The international expansion of Japanese firms, 1980 –1998. Strategic Management Journal, 24: 1153–1164. Dess, G. G., & Beard, D. W. 1984. Dimensions of organizational task environments. Administrative Science Quarterly, 29: 52–73. DeVellis, R. F. 1991. Scale development. Newbury Park, CA: Sage. Dhanaraj, C., Lyles, M.A., Steensma, H. K., & Tihanyi, L., 2004. Managing tacit and explicit knowledge transfer in IJVs: the role of relational embeddedness and the impact on performance. Journal of International Business Studies, 35: 428 – 442 Dillman, D. A. 2000. Mail and Internet surveys: The tailored design method. New York: Wiley. Doz, Y. L. 1996. The evolution of cooperation in strategic alliances: Initial conditions or learning processes. Strategic Management Journal, 17: 55– 83. Dyer, J. H., & Chu, W. 2003. The role of trustworthiness in reducing transaction costs and improving performance: Empirical evidence from the United States, Japan, and Korea. Organization Science, 14: 57– 68. Ferrin, D. L., & Dirks, K. T. 2003. The use of rewards to increase and decrease trust: Mediating processes and differential effects. Organization Science, 14: 18 – 31. Gambetta, D. 1988. Trust: The making and breaking of cooperative relationships. Oxford, U.K.: Basil Blackwell. Geringer, J. M., & Hebert, L. 1991. Measuring perfor-

912

Academy of Management Journal

October

mance of international joint ventures. Journal of International Business Studies, 22: 249 –264. Geyskens, I., Steenkamp, J. B. E. M., & Kumar, N. 1999. A meta-analysis of satisfaction in marketing channel relationships. Journal of Marketing Research, 36(2): 223–239. Ghemawat, P., Porter, M. E., & Rawlinson, R. A. 1986. Patterns of international coalition activity. In M. E. Porter (Ed.), Competition in global industries: 345– 365. Boston: Harvard Business School Press. Ghoshal, S., & Bartlett, C. A. 1994. Linking organizational context and managerial action: The dimensions of quality of management. Strategic Management Journal, 15: 91–112. Gibson, C. B., & Birkinshaw, J. 2004. The antecedents, consequences and mediating role of organizational ambidexterity. Academy of Management Journal, 47: 209 –226. Glick, W. H., Ogilvie, dt, & Miller, C. C. 1990. Assessing dimensions of task environments: Intra-industry and aggregate industry measures. Paper presented at the annual meeting of the Academy of Management, San Francisco. Greene, W. H. 1997. Econometric analysis. Upper Saddle River, NJ: Prentice-Hall. Griffith, D. A., & Myers, M. B. 2005. The performance implications of strategic fit of relational norm governance strategies in global supply chain relationships. Journal of International Business Studies, 36: 254 –269. Grindley, P., Mowery, D. C., & Silverman, B. S. 1994. SEMATECH and collaborative research: Lessons in the design of high-technology consortia. Journal of Policy Analysis and Management, 13: 723–758. Gulati, R. 1995. Does familiarity breed trust? The implications of repeated ties for contractual choice in alliances. Academy of Management Journal, 38: 85–112. Gulati, R. 1998. Alliances and networks. Strategic Management Journal, 19: 293–317. Gulati, R., & Singh, H. 1998. The architecture of cooperation: managing coordination costs and appropriation concerns in strategic alliances. Administrative Science Quarterly, 43: 781– 814. Gupta, A. K., & Govindarajan, V. 1991. Knowledge flows and the structure of control within multinational corporations. Academy of Management Review, 16: 768 –792. Hambrick, D. C. 1982. Environmental scanning and organizational strategy. Strategic Management Journal, 3: 159 –174. Hamel, G. 1991. Competition for competence and interpartner learning within international strategic alliances. Strategic Management Journal, 12: 83–103. Hamilton, B. H., & Nickerson, J. A. 2003. Correcting for

endogeneity in strategic management research. Strategic Organization, 1: 51–78. Harman, H. H. 1967. Modern factor analysis. Chicago: University of Chicago Press. Harrigan, K. R. 1985. Strategies for joint ventures. Lexington, MA: Lexington Books. Harrigan. K. R. 1988. Joint ventures and competitive strategy. Strategic Management Journal, 9: 141–158. Heckman, J. 1979. Sample selection bias as a specification error. Econometrica, 47: 153–161. Hildebrandt, L. 1987. Consumer retail satisfaction in rural areas: A reanalysis of survey data. Journal of Economic Psychology, 8: 19 – 42. Hofstede, G. 1980. Culture’s consequences: International differences in work-related values. London: Sage. Huber, J. P., Miller, C. C., & Glick, W. H. 1990. Developing more encompassing theories about organizations: The centralization-effectiveness relationship as an example. Organization Science, 1: 11– 40. Isobe, T., Makino, S., & Montgomery, D. B. 2000. Resource commitment, entry timing and market performance of foreign direct investments in emerging economies: The case of Japanese international joint ventures in China. Academy of Management Journal, 43: 468 – 484. Jehn, K. A., Northcraft, G. B., & Neale, M. A. 1999. Why differences make difference: A field study of diversity, conflict, and performance in workgroups. Administrative Science Quarterly, 44: 741–763. Joreskog, K. G. 1971. Statistical analysis of congeneric ¨ tests. Psychometrika, 36: 109 –133. Joreskog, K. G., & Sorbom, D. 1996. LISREL 8: User’s ¨ ¨ reference guide. Chicago: Scientific Software International. Kahneman, D., & Lovallo, D. 1993. Timid choices and bold forecasts: A cognitive perspective on risk taking. Management Science, 39(1): 17–31. Kale, P., Singh, H., & Perlmutter, H. 2000. Learning and protection of proprietary assets in strategic alliances: Building relational capital. Strategic Management Journal, 21: 217–237. Kearney, A. T. 2004. FDI confidence index. Global Business Policy Council, 7: 1– 44. Keats, B. W., & Hitt, M. A. 1988. A causal model of linkages among environmental dimensions, macro organizational characteristics and performance. Academy of Management Journal, 31: 570 –598. Khanna, T., Gulati, R., & Nohria, N. 1998. The dynamics of learning alliances: Competition, cooperation, and relative scope. Strategic Management Journal, 19: 193–210. Kogut, B. 1988. Joint ventures: Theoretical and empirical

2006

Krishnan, Martin, and Noorderhaven

913

perspectives. Strategic Management Journal, 9: 319 –332. Kogut, B. 1989. The stability of joint ventures: Reciprocity and competitive rivalry. Journal of Industrial Economics, 38(2): 183–198. Kogut, B., & Singh, H. 1988. The effect of national culture on the choice of entry mode. Journal of International Business Studies, 19: 411– 432. Koopmans, T. C. 1957. Three essays on the state of economic science. New York: McGraw-Hill. Kotabe, M., Martin, X., & Domoto, H. 2003. Gaining from vertical partnerships: Knowledge transfer, relationship duration, and supplier performance improvement in the U.S. and Japanese automotive industries. Strategic Management Journal, 24: 293–316. Krishnan, R. 2006. When is trust produced and when does it matter? Taking a closer look at trust and alliance performance. Tilburg, the Netherlands: CentER Dissertation Series. Kumar, S., & Seth, A. 1998. The design of coordination and control mechanisms for managing joint venture parent relationships. Strategic Management Journal, 19: 579 –599. Lane, P. J., Salk, J. E., & Lyles, M. A. 2001. Absorptive capacity, learning and performance in international joint ventures. Strategic Management Journal, 22: 1139 –1161. Langfred, C. W. 2004. Too much of a good thing? Negative effects of high trust and individual autonomy in self-managing teams. Academy of Management Journal, 47: 385–399. Lau, C. M., & Woodman, R. W. 1995. Understanding organizational change: A schematic perspective. Academy of Management Journal, 38: 537–554. Lawrence, P. R., & Lorsch, J. W. 1973. A reply to Tosi, Aldag & Storey. Administrative Science Quarterly, 18: 397–398. Lee, T. W., Mitchell, T. R., Wise, L., & Fireman, S. An unfolding model of voluntary employee turnover. Academy of Management Journal, 39: 5–36. Lin, X., & Germain, R. 1998. Sustaining satisfactory joint venture relationships: The role of conflict resolution strategy. Journal of International Business Studies, 29: 179 –196. Lindell, M. K., & Whitney, D. J. 2001. Accounting for common method variance in cross-sectional designs. Journal of Applied Psychology, 86: 114 –121. Luo, Y. 2002. Product diversification in international joint ventures: Performance implications in an emerging market. Strategic Management Journal, 23: 1–20. Madhok, A., & Tallman, S. B. 1998. Resources, transactions and rents: Managing value through interfirm collaborative relationships. Organization Science, 9: 326 –339.

Makino, S., & Inkpen, A. C. 2003. Knowledge seeking FDI and learning across borders. In M. Easterby-Smith & M. A. Lyles (Eds.), The Blackwell handbook of organizational learning and knowledge management: 233–252. Oxford, U.K.: Blackwell. Martin, X., & Salomon, R. 2003a. Tacitness, learning, and international expansion: A study of foreign direct investment in a knowledge-intensive industry. Organization Science, 14: 297–311. Martin, X., & Salomon, R. 2003b. Knowledge transfer capacity and its implications for the theory of the multinational corporation. Journal of International Business Studies, 34: 356 –373. Martin, X., Swaminathan, A., & Mitchell, W. 1998. Organizational evolution in the interorganizational environment: Incentives and constraints on international expansion strategy. Administrative Science Quarterly, 43: 566 – 601. Mayer, R. C., Davis, J. H., & Schoorman, F. D. 1995. An integrative model of organizational trust. Academy of Management Review, 20: 709 –734. McEvily, B., Perrone, V., & Zaheer, A. 2003. Trust as an organizing principle. Organization Science, 14: 91– 103. Mintzberg, H. 1978. Patterns in strategy formation. Management Science, 24: 934 –948. Mohr, J., & Spekman, R. 1994. Characteristics of partnership success: Partnership attributes, communication behavior and conflict resolution techniques. Strategic Management Journal, 15: 135–152. Mowery, D. C., Oxley, J. E., & Silverman, B. S. 1996. Strategic alliances and interfirm knowledge transfer. Strategic Management Journal, 17: 77–91. Nickerson, J. A., & Zenger, T. R. 2004. A knowledgebased theory of the firm—The problem-solving perspective. Organization Science, 15: 617– 632. Noorderhaven, N. G. 2004. Hermeneutic methodology and international business research. In R. MarschanPiekkari & C. Welch (Eds.), Handbook of qualitative research methods for international business: 84 – 104. Cheltenham, U.K., & Northampton, MA: Edward Elgar. Nooteboom, B. 2002. Trust: Forms, foundations, functions, failures and figures. Cheltenham, U.K., & Northampton, MA: Edward Elgar. Nooteboom, B, Berger, H., & Noorderhaven, N.G. 1997. Effects of trust and governance on relational risk. Academy of Management Journal, 40: 308 –338. Nunnally, J. 1978. Psychometric theory. New York: McGraw-Hill. Olk, P. 2002. Evaluating strategic alliance performance. In F. Contractor & P. Lorange (Eds.), Cooperative strategies and alliances: 119 –143. Amsterdam: Pergamon. Osborn, R. N., & Baughn, C. C. 1990. Forms of interor-

914

Academy of Management Journal

October

ganizational governance for multinational alliances. Academy of Management Journal, 33: 503–519. Oxley, J. E. 1999. Institutional environment and the mechanisms of governance: The impact of intellectual property protection on the structure of interfirm alliances. Journal of Economic Behavior and Organization, 38: 283–309. Oxley, J. E., & Sampson, R. C. 2004. The scope and governance of international R&D alliances. Strategic Management Journal, 25: 723–750. Park, C. V., & Sheath, J. N. 1975. Impact of prior familiarity and cognitive complexity on information processing rules. In M. L. Ray & S. Ward (Eds.), Communicating with consumers: The information processing approach: 71–78. Beverly Hills, CA: Sage. Park, S. H., & Russo, M. 1996. When competition eclipses cooperation: An event history analysis of alliance failure. Management Science, 42: 875– 890. Park, S. H., & Ungson, G. R. 2001. Interfirm rivalry and managerial complexity: A conceptual framework of alliance failure. Organization Science, 12: 37–53. Parkhe, A. 1991. Interfirm diversity, organizational learning, and longevity in global strategic alliances. Journal of International Business Studies, 22: 579 – 601. Parkhe, A. 1993. Strategic alliance structuring: A game theoretic and transaction cost examination of interfirm cooperation. Academy of Management Journal, 36: 794 – 829. Parkhe, A. 1998. Building trust in international alliances. Journal of World Business, 33: 417– 437. Pfeffer, J., & Nowak, P. 1976. Joint ventures and interorganizational interdependence. Administrative Science Quarterly, 21: 398 – 419. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & 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. Poppo, L., & Zenger, T. 2002. Do formal contracts and relational governance function as substitutes or complements? Strategic Management Journal, 23: 707–726. Pothukuchi, V., Damanpour, F., Choi, J., Chen, C. C., & Park, S. H. 2002. National and organizational culture differences and international joint venture performance. Journal of International Business Studies. 33(2): 243–265. Powell, W. W. 1990. Neither market nor hierarchy: Network forms of organizations. In B. M. Staw & L. L. Cummings (Eds.), Research in organizational behavior, vol. 12: 295–336. Greenwich, CT: JAI Press. Robertson, R. D. 1980. Small group decision making: The uncertain role of information in reducing uncertainty. Political Behavior, 2(2): 163–188. Robins, J. A., Tallman, S., & Lindquist, K. F. 2002. Autonomy and dependence of international coopera-

tive ventures: An exploration of the strategic performance of U.S. ventures in Mexico. Strategic Management Journal, 23: 881–901. Sako, M. 1991. The role of “trust” in Japanese buyersupplier relationships. Ricerche Economiche, 45: 449 – 474. Sako, M., & Helper, S. 1998. Determinants of trust in supplier relations: Evidence from the automotive industry in Japan and the United States. Journal of Economic Behavior and Organization, 34: 387– 417. Sampson, R. C. 2004. The cost of misaligned governance in R&D alliances. Journal of Law, Economics and Organization, 20: 484 –526. Saxton, T. 1997. The effects of partner and relationship characteristics on alliance outcomes. Academy of Management Journal, 40: 443– 460. Schwenk, C. R. 1984. Cognitive simplification processes in strategic decision-making. Strategic Management Journal, 5: 111–128. Shaver, J. M. 1998. Accounting for endogeneity when assessing strategy performance: Does entry mode choice affect FDI survival? Management Science, 44: 571–585. Simonin, B. L. 1999. Ambiguity and the process of knowledge transfer in strategic alliances. Strategic Management Journal, 20: 595– 624. Simons, T., Pelled, L. H., & Smith, K. A. 1999. Making use of difference: Diversity, debate and decision comprehensiveness in top management teams. Academy of Management Journal, 42: 662– 673. Snyder, N. H., & Glueck, W. F. 1982. Can environmental volatility be measured objectively? Academy of Management Journal, 25: 185–192. Steensma, H. K., Tihanyi, L., Lyles, M. A., Dhanaraj, C. 2005. The evolving value of foreign partnerships in transitioning economies. Academy of Management Journal, 48: 213–235. Stinchcombe, A. L. 1985. Contracts as hierarchical documents. In A. Stinchcombe & C. Heimer (Eds.), Organization theory and project management: 121– 171. Bergen: Norwegian University Press. Sutcliffe, K. M., & Zaheer, A. 1998. Uncertainty in the transaction environment: An empirical test. Strategic Management Journal, 19: 1–23. Szulanski, G., Cappetta, R., & Jensen, R. J. 2004. When and how trustworthiness matters: Knowledge transfer and the moderating effect of causal ambiguity. Organization Science, 15: 600 – 613. Thompson, J. D. 1967. Organizations in action: Social science bases of administration. New York: McGraw-Hill. Tourangeau, R., Rips, L. J., & Rasinski, K. 2000. The psychology of survey response. Cambridge, U.K.: Cambridge University Press.

2006

Krishnan, Martin, and Noorderhaven

915

Tsai, W., & Ghoshal, S. 1998. Social capital and the value creation: The role of intrafirm networks. Academy of Management Journal, 41: 464 – 476. Uzzi, B. 1997. Social structure and competition in interfirm networks: The paradox of embeddedness. Administrative Science Quarterly, 42: 35– 67. Webb, G. 1996. Trust and crises. In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research: 288–301. Thousand Oaks, CA: Sage. Wholey, D. R., & Brittain, J. 1989. Characterizing environmental variation. Academy of Management Journal, 32: 867– 882. Williamson, O. E. 1985. The economic institutions of capitalism. New York: Free Press. Williamson, O. E. 1991. Comparative economic organization: The analysis of discrete structural alternatives. Administrative Science Quarterly, 36: 269 –296. Yan, A., & Zeng, Z. 1999. International joint venture instability: A critique of previous research, a reconceptualization, and directions for future research. Journal of International Business Studies, 30: 397– 414. Zaheer, A., McEvily, B., & Perrone, V. 1998. Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization Science, 9: 141–159. Zand, D. E. 1972. Trust and managerial problem solving. Administrative Science Quarterly, 17: 229 –239. Zollo, M., Reuer, J. J., & Singh, H. 2002. Inter-organizational routines and performance in strategic alliances. Organization Science, 13: 701–713.

4. Our firm is generally doubtful of the information provided to us by our foreign partner. (reverse-coded) 5. Our foreign partner firm is generally doubtful of the information we provide them. (reverse-coded) Quality of Information Exchanged ( .80; 1 “strongly disagree”; 5 “strongly agree” unless otherwise indicated) 1. Our foreign partner firm has provided relevant information whenever we asked them for it. 2. We are promptly notified by our foreign partner whenever any major change occurs at their firm. 3. We get clear information about the plans of our foreign partner concerning the collaboration well in advance. 4. How often do senior managers from your firm communicate with their counterparts in the foreign partner firm? (1 “daily”; 5 “once a month or less”) 5. How often do senior and middle managers in your company make business trips to your foreign partner firm? (1 “twice a month or more”; 5 “once a year or less”). Alliance Duration We asked the year in which an alliance was formed and used it to calculate duration. Interdependence Indicate which of the following describe the value creation rationales of the alliance. (The questionnaire indicated that the respondents could check more than one rationale of the nine listed in the text.)

APPENDIX A
Survey Items Alliance Performance ( 5 “strongly agree”) .90; 1 “strongly disagree”;

APPENDIX B
Controlling for the Endogeneity of the Equity/Nonequity Governance Choice Motivation Our sample included both equity and nonequity alliances. Prior research states that managers consciously select from these two alliance governance modes the one that is likely to enhance the performance of their particular alliance (e.g., Gulati, 1995; Sampson, 2004). Therefore, it is likely that alliance performance depends on unobservable characteristics that determine alliance governance choices. To account for this potential endogeneity, we used Heckman’s (1979) two-stage technique (see Hamilton & Nickerson, 2003; Shaver, 1998). Implementation and Results Using Heckman’s two-stage procedure, we first estimated a probit model of alliance governance choice (nonequity alliance vs. equity alliance) and generated the inverse Mills ratio. We then estimated the alliance performance model using the inverse Mills ratio from the first stage as a control variable. Incorporating this correction term into the second-stage model yields unbiased estimates of the predictors of alliance performance (Greene, 1997). In

1. The objectives for which the collaboration was established are being met. 2. Our firm is satisfied with the financial performance of the collaboration. 3. Our foreign partner firm seems to be satisfied with the financial performance of the collaboration. 4. Our firm is satisfied with the overall performance of the collaboration. 5. Our foreign partner firm seems to be satisfied with the overall performance of the collaboration. Trust ( agree”) .85; 1 “strongly disagree”; 5 “strongly

1. Sometimes our foreign partner changes facts slightly in order to get what they want. (reverse-coded) 2. Our foreign partner has promised to do things without actually doing them later. (reverse-coded) 3. Our foreign partner has given us truthful and valuable information even when it did not form part of the contract.

916

Academy of Management Journal

October

our first-stage model, we used two variables suggested in prior research: the tenure of respondent (e.g., Poppo & Zenger, 2002) and alliance formation before liberalization (e.g., Hamilton & Nickerson, 2003). Firstly, a respondent’s experience within her or his firm was a proxy for corporate knowledge about interorganizational relationships (Poppo & Zenger, 2002). Respondent tenure was likely to influence the choice of alliance governance mode, all the more so as each of our respondents held a position of high responsibility. Secondly, before a major liberalization of foreign trade relationships through legislation in 1991, the Indian government exerted pressure on foreign firms to contribute some form of equity to the Indian economy (Balasubramanyam, 2003; Bowonder & Richardson, 2000). Hence, alliances formed before liberalization were more likely to be equity alliances, though not all were. There is no theoretical basis to link either of these variables directly with alliance performance (Hamilton & Nickerson, 2003; Poppo & Zenger, 2002). Other variables from our work were also included in the first-stage model. Table B1 reports the firststage results. Preference for an equity alliance is negatively related to tenure of respondent (p .01) and marginally related to environmental unpredictability (p .10), and positively related to investment size (p .05) and alliance formed before liberalization (p .01). Because correcting for self-selection of governance mode is important in theory, we included the inverse Mills ratio from this model into the second-stage models. The ratio’s nonsignificance in the second stage indicated that the potential endogeneity of governance mode was not adversely affecting our estimated results about alliance performance. The hypothesized re-

sults are similar with and without the inclusion of the inverse Mills ratio.

Rekha Krishnan (rekhak@sfu.ca) is an assistant professor of international business in the Faculty of Business Administration at Simon Fraser University. She received her Ph.D. in business from the CentER Graduate School at Tilburg University. Her research interests include strategic alliances, trust, and emerging markets. Xavier Martin (x.martin@uvt.nl) is a professor of strategy and international business in the Faculty of Economics and Business Administration at Tilburg University and a fellow of the CentER for Economic Research. He received his Ph.D. in business administration from the Graduate School of Business Administration at the University of Michigan. His research interests include corporate strategies for international expansion, the dynamics of interorganizational relationships and alliances, knowledge transfer and accumulation strategies, innovation and new-product introduction strategies, and the performance implications of these phenomena. Niels G. Noorderhaven (n.g.noorderhaven@uvt.nl) is a professor of international management in the Faculty of Economics and Business Administration and the head of the Department of Organization and Strategy at Tilburg University. He received his Ph.D. in business administration from Groningen University. His research focuses on interorganizational cooperation, with special attention to issues related to culture and trust.

TABLE B1 Results of Probit Analysis for the First-Stage Governance Modela
Variable Intercept Tenure of respondent Position of respondent Investment size Cultural distance Local partner size Interdependence Interpartner competition Environmental instability Environmental unpredictability Alliance formed before liberalization n
2

Model 1 1.86* (0.75) 0.05** (0.02) 0.08 (0.25) 0.39* (0.19) 0.13 (0.30) 0.14 (0.09) 1.06 (0.68) 0.02 (0.21) 2.06 (2.42) 0.68† (0.38) 0.05** (0.02) 126 160.66***

Please note: This article continues with Appendix C, on the following page.

a The dependent variable was set to 0 for nonequity alliances and 1 for equity alliances. Standard errors are in parentheses. † p .10 * p .05 ** p .01 *** p .001

2006

Krishnan, Martin, and Noorderhaven

917

APPENDIX C Remedies Undertaken against Common Method Bias and Single Respondent Bias
Remedy and Rationale
Procedural Protecting respondent anonymity. This technique decreases respondents’ tendency to make socially desirable responses and/or be acquiescent or lenient when crafting their responses (Podsakoff et al., 2003: 888). Reducing item ambiguity. Careful attention to the wording of items helps reduce item ambiguity (Tourangeau, Rips, & Rasinski, 2000). Our cover letter assured respondents complete anonymity.

Implementation

We were careful to avoid vague concepts and doublebarreled questions and to keep questions simple, all of which reduce item ambiguity (Tourangeau, Rips, & Rasinski, 2000). We pretested the survey with Indian managers, which helped us identify and replace a few ambiguous words. In our questionnaire, the trust and alliance performance items were placed far apart from each other—about 50 percent of the relevant questionnaire pages apart. Items were not grouped by variable, and the variables were not labeled on the basis of the reported constructs (trust, etc.). We obtained data on most (three-quarters) of the moderator variables and on several control variables from archival sources.

Separating scale items. Reduces the likelihood of respondents guessing the relationship between predictor and criterion variables and consciously matching their responses to the two measures (Parkhe, 1993).

Data from different sources. Measures based on different sources help control common method or single-informant bias (Podsakoff et al., 2003). Statistical Partial correlation adjustment. If a variable can be identified that is theoretically unrelated to at least one other variable in a study, preferably the dependent variable, then it can be used as a marker variable in controlling for common method variance (Lindell & Whitney, 2001).

We used tenure of the respondent as the marker variable, as it was theoretically unrelated to many other variables and especially to alliance performance. The use of a variable such as this is consistent with existing research (e.g., Griffith & Myers, 2005). All our significant zero-order correlations remained significant after the partial correlation adjustment, suggesting that common method bias was not a serious problem in our study (Lindell & Whitney, 2001). The correlation between the subjective and secondary alliance performance measures (i.e., return on capital employed) available for 35 equity alliances was highly significant (r .38, p 0.02). Note that our measure of alliance performance had a general focus, but return on capital was a focused indicator, preventing a stronger correlation. We used interview data available for ten alliances to validate the trust measure. Two independent coders categorized the interview responses using three-point scales to indicate the extent to which interorganizational trust existed in a relationship. We did not exceed threepoint categorization for ease of interpretation of the interview data (e.g., Lau & Woodman, 1995; Lee, Mitchell, Wise, & Fireman, 1996). The correlation between the trust scale obtained from the survey and the interview notes coded by the independent raters was .70 (p .05). No discrepancy was noted regarding variable content. An unrotated principal components factor analysis on all the variables measured using the survey instrument revealed four factors with eigenvalues greater than 1.0, which together accounted for 59 percent of the total variance; also, the first (largest) factor did not account for a majority of the variance (19.71%). All our interaction effects were significant, including the interaction that involved two survey-based items. Support for interaction hypotheses is unlikely to be an artifact of single-respondent bias, as it is implausible that respondents will consciously theorize moderated relationships when responding to a survey (Kotabe et al., 2003).

Triangulation using archival sources. Triangulating survey data with data from secondary sources is often used to check the convergent validity of a construct (Dhanaraj, Lyles, Steensma, & Tihanyi, 2004; Keats & Hitt, 1988; Parkhe, 1993).

Triangulation using field interviews. Interview-based data can be coded to establish the reliability and validity of variables. We used the interview notes to validate the trust variable since our interview focused mainly on the quality of the relationship between partners, thus providing usable information about trust. Because the interview notes did not provide enough useful data on how well an alliance performed, we used archival data instead to validate the alliance performance construct (see “Triangulation using archival sources”).

Harman’s one-factor test. If a substantial amount of common method bias exists in data, a single or general factor that accounts for most of the variance will emerge when all the variables are entered together (Podsakoff et al., 2003).

Significance of the interaction terms. A pattern of significant interaction terms suggests that results are unlikely to have resulted from single-informant bias (Kotabe, Martin, & Domoto, 2003).