Integrating and extending insights from transaction cost analysis and relational exchange theory

Determinants of commitment and opportunism: Integrating and extending insights from transaction cost analysis and relational exchange theory

Ashwin W Joshi


Effectively governed relationships are characterized by high levels of commitment and low levels of opportunism. In this research, we develop a conceptual model that integrates insights from transaction cost analysis and relational exchange theory to identify three key antecedents of commitment and opportunism: transaction specific assets, environmental uncertainty, and relational norms. Our model extends these prior frameworks by identifing dependence and long-term orientation as the mediating processes through which the antecedents affect the outcomes. We test the model using a scenariobased experimental method with 168 purchasing managers as subjects. Results provide substantial support for the model. The managerial, theoretical, and future research implications of these results are discussed.


Des relations menees efficacement se caracterisent par leur haut niveau d’engagement et leur faible niveau d’opportunisme. Dans la presente recherche, nous etablissons un modele conceptuel qui integre les notions d’Analyse du coat des operations et de Theorie des echanges relationnels afin d’identifier les trois antecedents clis de l’engagement et de l’opportunisme : les actifs specifiques de la transaction, l’insicurite de l’environnement et les normes relationnelles. Notre modele elargit ce premier cadre en identifiant la dependance et l’orientation a long terme comme procede de mediation par lequel les antecedents influent sur les resultats. Nous examinons le modele grace a une methode experimentale que se fonde sur le scenario en prenant comme sujets 168 responsables des achats. Les resultats apportent un rM appi au modele. Nous discutons aussi, a partir des resultats, des consequences sur la gestion, la theorie et les futures recherches.

Conventional governance theory predicts that exchange transactions will be governed using one of two governance mechanisms: markets or hierarchies (Williamson, 1975). Contrary to this prediction, the contemporary reality of governance is that a number of governance mechanisms that are “neither markets nor hierarchies” are flourishing (Powell, 1990). This discrepancy between conventional governance theory and contemporary reality has regenerated research interest in the issue of how exchange relationships are managed (Heide, 1994; Robicheaux & Coleman, 1994; Williamson, 1985). Although originally depicted as the sole alternatives to governing exchanges, spot market transactions (markets) and full vertical integration (hierarchies) are now generally regarded as being the polar ends of a continuum of governance options (Williamson, 1985).

The many new governance mechanisms found in contemporary practice are regarded as being intermediate to markets and hierarchies (Powell, 1990; Williamson, 1985) and have been described by a plethora of terms, including strategic alliances, joint ventures, joint action relationships, value-added partnerships, working partnerships, franchise arrangements, relational exchanges, research consortia, and networks (Achrol, 1991; Anderson & Narus, 1990; Dwyer, Schurr, & Oh, 1987; Heide & John, 1990; Johnston & Lawrence, 1988; Powell, 1990; Ring & Van de Ven, 1992; Webster, 1992). As these new types of governance mechanisms have emerged and spread throughout the industrial landscape, they have motivated theorists to adapt prior frameworks (Joshi, 1995) and develop new ones (Ring & Van de Ven, 1992, 1994) to account for these phenomena.

While there are significant differences amoung these new types of governance mechanisms (Joshi, 1995; Webster, 1992), there are also noteworthy underlying commonalities. In particular, these new governance mechanisms require exchange partners to be committed to each other and refrain from behaving opportunistically against each other (Anderson & Weitz, 1992; Dwyer et at., 1987; Gundlach, Achrol, & Mentzer, 1995; Morgan & Hunt, 1994). Given the centrality of the commitment and opportunism constructs to these new governance mechanisms, it is not surprising that an extensive theoretical and empirical literature documenting a broad range of antecedents of these behaviours has developed over the past decade. For example, idiosyncratic investments, reputation (Anderson & Weitz, 1992), relationship termination costs, relationship benefits, and shared values (Morgan & Hunt, 1994) have been identified as key antecedents of commitment. Similarly, relational norms, relationship termination costs, dependence, relationship benefits, and shared values, have been shown to attenuate opportunism (Gundlach et al., 1995; John, 1984; Joshi & Stump, 1996; Kelley, Skinner, & Ferrell, 1989; Provan & Skinner, 1989).

In identifying the antecedents of commitment and opportunism, this research has drawn from a number of different theoretical perspectives, including transaction cost analysis [TCA] (Anderson & Narus, 1990; Morgan & Hunt, 1994), relational exchange theory [RET] (Heide & John, 1992; Joshi & Arnold, 1997, 1998), power– dependence theory (Stern & Reve, 1980), communication theory (Mohr, Fisher, & Nevin, 1996; Mohr & Nevin, 1990), network theory (Anderson, Hakansson, & Johanson, 1994), social adaption theory (Frazier & Rody, 1991; Hallen, Johanson, & Seyed-Mohamed, 1991), and organizational contingency theory (John, 1984).

In contrast to the theoretical eclecticism of this prior research, we restrict the theoretical focus of the present study by developing explanations from the two predominant theoretical frameworks in the governance literature: transaction cost analysis (Williamson, 1985, 1991 a, 1991b) and relational exchange theory (Dwyer et al., 1987; Macneil, 1980). These frameworks have come to be regarded as critical bases for understanding behavioural dynamics in exchange relationships (Heide, 1994; Robicheaux & Coleman, 1994) and have drawn attention to three antecedent factors: asset specificity, environmental uncertainty, and relational norms.

Although the potential for theoretical underspecification exists, i.e., such restriction means that the full range of antecedents of commitment and opportunism are not considered in the analysis, there is a compelling rationale for deliberately restricting focus. This enables us to better explain the underlying process by exploring how and why these antecedents affect commitment and opportunism, While there is an extensive literature on the antecedents of commitment and opportunism, a significant limitation of this research is that process explanations of how and why antecedent variables affect these outcomes have not been fully developed and tested. For example, research has established that manufacturer investment of idiosyncratic assets (asset specificity) increases manufacturer commitment toward that partner (Anderson & Weitz, 1992). Likewise, Brill’s (1994) reanalysis of John’s (1984) data showed that relational restrictiveness (low relational norms) reduced compliance (increased opportunism). In both instances, however, the process by which this effect occurs is speculated upon but not empirically tested.

In contrast, the present study develops and tests a conceptual model that depicts the linkages-of asset specificity, uncertainty, and relational norms-the antecedent variables-to the outcome variables commitment and opportunism, being fully mediated by two key variables, dependence and long-term orientation. Developing and testing process explanations that connect antecedent variables to outcomes is important because it enhances the validity of the theory test. This expanded approach allows researchers to test both the validity of the prediction of a theory (e.g., and increase in A will lead to an increase in C) as well as the validity of the theory-specified process explanation (e.g., A will increase C through its impact upon B). Since it is conceivable that predictions can be supported for reasons not outlined by theory, testing for theory-driven process explanations allows researchers to rule out such spurious support (Pilling, Crosby, & Jackson, 1994).

From a managerial perspective, process models are also important because they provide intermediate markers for final outcomes. For example, in the context of this research, increased manufacturer commitment and reduced manufacturer opportunism are important outcomes that suppliers seek. Suppliers can foster these behaviours by encouraging manufacturers to invest in specific assets and by developing relational norms with them. However, since there is a lag between the time these antecedent variables are established and the time those desired outcomes occur, suppliers can assess the interim effectiveness of these strategies by monitoring whether or not the manufacturer’s temporal orientation to the relationship has changed from a short-term one into a long-term one.

The conceptual model, which is introduced in the next section, captures the process explanations that link the antecedent variables to the consequences. Following the conceptual model, the scenario-based experimental method that was used to test the model is discussed. Following this, the data analysis procedures and results are described. Finally, theoretical and managerial implications arising from this study, as well as future research directions, are discussed.

Conceptual Framework Background

Transaction cost analysis (TCA). TCA is principally concerned with managing exchange transactions such that the sum of production and transaction costs are minimized (Williamson, 1985). Production costs are those entailed in internalizing the functions of the exchange partner and transaction costs are those entailed in establishing agreements with, monitoring the performance of, and enforcing contractual clauses against the exchange partner. In the context of manufacturer-supplier relationships, for example, a manufacturer can manage the supplier relationship by either routinely submitting the incumbent supplier to market discipline through periodic competitive bidding (the market governance mechanism) or internalizing the functions of the supplier (the hierarchy governance mechanism). Both the market and hierarchy governance mechanisms represent trade-offs between transaction and production costs. Market governance imposes transaction costs on the manufacturer, however it does not create production costs. In contrast, hierarchical governance imposes production costs on the manufacturer and minimizes transaction costs.

TCA identifies two factors that determine whether transaction costs are greater than production costs, or vice versa, in situations of asset specificity and uncertainty. Asset specificity is the nonredeployable investment made by one party in their partner. For example, when a manufacturer invests in training supplier personnel, that manufacturer makes an investment that cannot be redeployed should the relationship with the focal supplier be terminated. The second factor that determines the extent of transaction/production costs is uncertainty, or the degree to which future environmental conditions cannot be readily predicted.

Per TCA, the sum of production and transaction costs is minimized (a) when transactions of the high asset-specificity/uncertainty form are governed by the hierarchy mechanisms, and (b) when transactions of the low asset-specific ity/uncertainty form are governed by using the market mechanism. The transaction-cost problem owes its existence to the assumption of opportunism (John, 1984). A manufacturer’s investment of specific assets in a supplier gives the supplier control over the manufacturer. This power advantage, coupled with the assumption that the supplier is potentially opportunistic (or may pursue its self-interest with guile), makes the manufacturer vulnerable to the supplier (Klein, Crawford, & Alchian, 1978). For example, the supplier can use its power advantage to impose an increase in price for its component after signing a contract. Given its loss of investment in the supplier should the relationship be terminated, the manufacturer may have no choice but to comply. Accordingly, TCA contends that despite the production cost increase associated with internalizing the supplier, the offsetting reduction in transaction costs makes hierarchy more efficient than markets when specific assets are invested. In contrast, when no specific assets are involved, the market mechanism (or the threat of replacement) is sufficient to keep the supplier compliant (Heide, 1994; Rindfleisch & Heide, 1997; Williamson, 1985, 1991a, 1991b).

Desire to avoid partner opportunism (or anticipated transaction costs) is the process explanation that connects the TCA antecedent variables, asset specificity and uncertainty, to the outcome, type of governance. One party’s investment of specific assets in the partner increases the risk arising from potential partner opportunism, thereby making hierarchical governance more economical. This premise has received support in prior research (Pilling, Crosby, & Jackson, 1994). TCA, however, also makes a less well-recognized assumption about the impact of party opportunism on that party’s commitment toward, and opportunism against, the partner. It suggests (Williamson, 1985) that since specific asset investments create delayed payoffs, the investing party has to ensure relationship continuity in order to secure these payoffs. Accordingly, the invested party will display commitment to the partner and refrain from behaving opportunistically against the partner in order to ensure the necessary continuity of the exchange relationship. We subscribe to this premise and elaborate on it as we develop our model after overviewing relational exchange theory (RET).

Relational exchange theory (RET). RET develops the argument that relational norms are a unique class of governance mechanism that serves to prescribe (commitment) and proscribe (opportunism) certain behaviours in exchange relationships (Macneil, 1980; Morgan & Hunt, 1994). The distinguishing characteristic of relational norms as a governance mechanism is that, unlike markets and hierarchies, they are an endogenous form of control. Behaviour in relational norm-based relationships is controlled not through incentives (as in market governance) or fiat (as in hierarchical governance), but internalization (Kelman, 1958) and moral control (Larson, 1992). Behaviour is regulated through a system of mutual and self-regulation (Gundlach et al., 1995).

A key limitation of RET is that the process explanation though which relational norms regulate behaviour is not fully developed in the theory. Indeed, some scholars in this tradition (e.g., Macneil, 1980) define relational norms so broadly that regulated behaviours become part of the definition of relational norms rather than a consequence of them. As Heide (1994) observed, RET is more of a descriptive than a predictive theory. By delimiting the definition of norms to expectations and by specifying processes through which these expectations influence behaviours, we seek to develop the predictive dimension of this theory. (Our conceptual model is depicted in Figure 1 above.)

TCA Processes

Asset specificity, dependence, and long-term orientation. When one party makes a specialized investment in its partner, transaction-specific assets are said to exist (Williamson, 1985). To illustrate, manufacturers may (a) invest resources in training a particular supplier, (b) modify their production processes to effectively incorporate the supplier’s component, or (c) relocate to be physically close to their supplier in order to economize on transportation costs. In each of these cases, should the relationship with the supplier be terminated prematurely, the investment would be foregone (Klein et al., 1978).

Dependence is the expectation that considerable costs would be involved to replace an incumbent supplier (Heide & John, 1988; Morgan & Hunt, 1994). Given the switching costs entailed in specific asset investments, it can be expected that a manufacturer that has invested specific assets in a supplier will experience dependence on that supplier. Consistent with this argument, prior research has found a positive relationship between asset specificity and dependence (Ganesan 1994; John & Weitz, 1989).

Long-term orientation refers to the time horizon over which the manufacturer assesses the performance of that supplier (Ganesan, 1993, 1994; Kelly & Thibaut, 1978). A manufacturer that assesses the relationship with an incumbent supplier on a transaction- by-transaction basis is said to have adopted a short-term orientation toward that supplier. In contrast, a manufacturer that assesses the effectiveness and profitability of the supplier relationship over a series of transactions is described as having a long-term orientation. As a consequence, long-term oriented manufacturers are less concerned with the outcomes of individual transaction and more concerned with monitoring the achievement of longterm strategic objectives, such as continuous cost reduction and quality improvement (Kalwani & Naryandas, 1995; Wilson, Dant, & Han, 1990).

Drawing from TCA, we argue that a manufacturer that has invested specific assets in a supplier should develop a long-term orientation with that supplier in order to secure the rents from these investments (Williamson, 1985).

H1: A manufacturer’s investment of specific assets in a supplier is positively related to the manufacturer’s dependence on the supplier.

H2: A manufacturer’s investment of specific assets in a supplier is positively related to the manufacturer’s long-term orientation toward the supplier.

Technological unpredictability, dependence, and long-term orientation. The TCA framework recognizes that uncertainty may arise from exogenous sources (due to the unpredictability of extra-dyadic events) and endogenous sources (e.g., adverse selection, moral hazard, or performance ambiguity [Williamson, 1985]). Rather than seeking to examine the impact of all sources of uncertainty, we focus on exogenous uncertainty, i.e., that arising from the dynamic external environment.

However, exogenous uncertainty is a diffuse concept whose influence on transaction governance decision-making varies depending on the origin of that uncertainty (Stump, 1995). Such uncertainty may originate in upstream or downstream markets, from the regulatory and competition sectors of the task environment, as well as from elements of the macroenvironment such as social, economic, or technological trends (Achrol, Reve, & Stern, 1983). Uncertainty can arise within a multitude of domains, and it can also be attributable to a number of different causes. For example, a manufacturer may experience uncertainty regarding their end customer (domain) because of rapid changes in downstream market preferences (pace of change) or because the downstream market does not have a standardized technology preference (cause [Achrol, 1991; Weiss & Heide, 1993]). Prior research has shown that uncertainty can have varying impacts on exchange relationship dynamics, depending on its source and type (see Klein, Frazier, & Roth, 1990; Stump & Joshi, 1998).

In this research we focus on technological uncertainty (domain) created by rapidly changing downstream market preferences (cause). Within a manufacturer-supplier context, we define technological unpredictability as the inability of the manufacturer to stipulate technical specifications for input components for anything other than a short period of time (Heide & John, 1990). In order to successfully operate in a technologically unpredictable environment, the manufacturer needs to be able to rapidly upgrade its product technology or risk technological obsolescence in downstream markets (Leenders & Fearon, 1993; Moriarty & Kosnick, 1989). A manufacturer’s dependence on and long-term orientation toward a supplier can limit its adaptability to changing technological standards. Since switching suppliers entails costs, the manufacturer may be constrained from turning to another supplier with the requisite technology (Balakrishnan & Wernerfelt, 1986; Williamson, 1991a).

Consequently, we expect that manufacturers whose downstream markets are technologically unpredictable will avoid becoming dependent on (Jackson, 1985; Stump, 1995; Weiss & Heide, 1993) or developing a long-term orientation toward one particular supplier (Heide & John, 1990). Such manufacturers are more likely to undertake short-term contractual relationships with a cluster of suppliers to seek out those that possess the most advanced technology (Williamson, 1991a). Therefore,

H3 Technological unpredictability is negatively related to a manufacturer’s dependence on a supplier.

H4: Technological unpredictability is negatively related to a manufacturer’s long-term orientation toward a supplier.

RET Processes

Relational norms, dependence, and long-term orientation. Relational norms (Macneil, 1980), an aspect of the relationship context, are the shared values of exchange partners about what constitutes appropriate behaviour within their relationship (Heide & John, 1992; Morgan & Hunt, 1994).

We adopt the tripartite conceptualization of relational norms that features the flexibility, information exchange, and solidarity dimensions proposed by Heide and John (1992). Flexibility refers to the shared expectations that parties will be willing to modify the original terms of the contract to take into account changes in the contracting environment. Information exchange refers to shared expectations that parties will be willing to provide any information if it may be helpful to the partner, whether or not the party is contractually obliged to do so. Finally, solidarity refers to the bilateral expectations that parties will act in such manner as to benefit each other (Heide & John, 1992).

Relational norms evolve from being shared descriptive expectations (e.g., “this is the way we do things in this relationship”) to shared normative expectations (e.g., “this is the way things should be done in this relationship”) [Berger & Luckmann, 1967; Larson, 1992]. As relational norms evolve, they become internalized by the exchange partners (Kelman, 1958) and consequently come to serve as moral controls that promote prorelationship behaviours, such as commitment, and proscribe unilaterally beneficial behaviours, such as opportunism.

While relational norms can contribute to relationship efficiency (Barney & Hansen, 1994), they are difficult to establish. They require substantial up-front investments of time, money, and personnel by both exchange partners (Frazier, Spekman, & O’Neal, 1988). Furthermore, even once they are established, relational norms are fragile and require continuous maintenance and development (Dwyer et al., 1987; Larson, 1992). Having made ongoing commitments of resources to relational norm development and recognizing that transaction cost savings will arise from the presence of these norms, manufacturers can be expected to perceive switching costs and dependence. In order to realize the ongoing stream of rents from relational norms, manufacturers are apt to have adopted a long-term orientation toward their supplier.

H5: Manufacturer-supplier relational-norm development is positively related to the manufacturer’s dependence on the supplier.

H6: Manufacturer-supplier relational-norm development is positively related to the manufacturer’s long-term orientation toward the supplier.

Outcomes of TCA and RET Processes

Dependence, long-term orientation, and commitment. Commitment refers to the willingness of exchange partners to work at maintaining the relationship (Morgan & Hunt, 1994) and to make short-term sacrifices for the long-term stability of the relationship (Anderson & Weitz, 1992).

Since a dependent manufacturer is one whose condition has resulted from self-imposed switching costs, the manufacturer has a greater motivation to sustain the relationship even when it involves sacrificing short-term gains (Anderson & Weitz, 1992; Morgan & Hunt, 1994). However, manufacturers will be unwilling to make the short-term sacrifices necessary to preserve a supplier relationship if they do not expect the relationship to be profitable over the long run (Kronman, 1985). Therefore, having a long-term orientation toward the supplier can be viewed as necessary for the demonstration of commitment by the manufacturer (Kelly & Thibaut, 1978).

H7: A manufacturer’s dependence on a supplier is positively related to that manufacturer’s commitment to the supplier.

H8: A manufacturer’s long-term orientation toward a supplier is positively related to that manufacturer’s commitment to the supplier.

Dependence, long-term orientation, and opportunism. Opportunism is defined as the pursuit of selfinterest with guile (John, 1984; Williamson, 1985). Such behaviour may be exhibited by either party or both parties in the exchange relationship. In this study, we focus our attention on the manufacturer’s opportunistic tendencies.

Dependence implies that recurrent exchange will take place within the relationship, akin to a game. Drawing from game theory (Axelrod, 1984; Heide & Miner, 1992), we argue that opportunism in such ongoing games will be avoided because of the potential it creates for the emergence of tit-for-tat strategies. While retaliation is never easy to cope with, a dependent manufacturer is particularly vulnerable to retaliation by the incumbent supplier since it is constrained from being able to defect to another supplier (Lawler, Ford, & Blegen, 1988). Fearing a retaliatory response by the current supplier, the dependent manufacturer may not behave opportunistically against their suppliers in the first place. Consistent with these arguments, Provan & Skinner (1989) found dependence and opportunism to be inversely related.

While opportunism does have the potential to generate short-term windfalls for a manufacturer who initiates such behaviour, when the manufacturer holds an expectation that their relationship with a supplier will endure for an extensive period, this “shadow of the future” (Axelrod, 1984, p. 124) can curb manufacturer actions in the present. In effect, the immediate gains from opportunism are likely to be limited, since the supplier partner has an extended time frame within which to retaliate. Consistent with this argument, Heide & John (1990) noted that “future interaction between exchange partners provides an opportunity to reward good behaviour and punish opportunism” (p. 26). Accordingly, we expect that manufacturers will desist from opportunistic behaviour against a supplier if they expect to be dealing with this supplier for the foreseeable future.

H9: A manufacturer’s dependence on a supplier is negatively related to that manufacturer’s opportunism against the supplier.

H10: A manufacturer’s long-term orientation toward a supplier is negatively related to that manufacturer’s opportunism against the supplier.


Research Design

A noteworthy feature of our study is our the method of testing the research hypotheses. Unlike the bulk of previous research in this area, which has relied on the survey method (Rindfleisch & Heide, 1997), we use the experimental method, following the precedents of a steadily increasing body of empirical work in businessto-business marketing and channels research (Gundlach & Cadotte, 1994; Gundlach et al., 1995; Jackson, Keith, & Burdick, 1984; Keith, Jackson, & Crosby, 1990; Pilling et al., 1994).

This method affords two advantages over surveys. First, it allows us to have greater control over the research context. By controlling for background factors, we are able to rule out competing explanations to the ones that we propose, thereby enhancing the internal validity of our results. Second, the experimental method allows us to temporally separate the collection of data at each of the independent, mediating, and outcome variable stages. Creating temporal separations between measurement of the independent, mediating, and outcome variables results in a stronger test of the causal sequence implied in the model than would be possible through the traditional survey procedure.

Each of the three antecedent variables-asset specificity, technological unpredictability, and relational norms-were manipulated at low and high levels. The experimental manipulations were presented within a written scenario. All eight versions of the scenario had a common introduction, where subjects were instructed to play the role of a purchasing manager in a manufacturing company in a case that simulated their relationship with Supplier X.

Research Context and Data Collection

The Purchasing Management Association of Canada (PMAC) provided a national mailing list of 1,000 purchasing managers from a cross-section of manufacturing industries. We mailed the research instrument to 500 randomly selected purchasing managers, along with a personalized letter requesting their participation. Respondents were promised a copy of the study results if they provided us with their business card. A letter from the vice-president of research at PMAC endorsing the aims of the research and encouraging their participation and a postage-paid return envelope were enclosed in the mailing.

Three weeks after the mailing, we received 168 usable responses (response rate = 33.6%). The respondents were experienced (mean level of purchasing experience = 9 years) and largely male (70%). They worked in manufacturing industries ranging from chocolates to aviation equipment.

Instrument Development and Pretest

Scale items from established scales of asset specificity (Anderson, 1988; Heide, 1987; John & Weitz, 1989), technological unpredictability (Achrol & Stern, 1988; Heide & John, 1990; Weiss & Heide, 1993), and relational norms (Gundlach et al., 1995; Heide & John, 1992; Kaufmann & Dant, 1992) were appropriately modified to construct the scenario manipulations of the antecedent variables. Established scales were used to measure the mediating variables: dependence (Heide, 1987; John & Weitz, 1989) and long-term orientation (Ganesan, 1994), as well as the consequence variables: commitment (Anderson & Weitz, 1992) and opportunism (John, 1984; Provan & Skinner, 1989) [see the appendix for manipulations and measures].

We discussed the instrument with five purchasing managers during on-site interviews. Based on these conversations, subtle changes were made to the wording of the scenario manipulations and the ordering of information. As part of a pretest, the modified questionnaire was mailed to 96 randomly selected purchasing agents nationwide. Three weeks after the mailing, we had received 35 responses. Despite the small sample size of the pilot study, we were able to establish that the low and high levels of each of the antecedent variables were statistically significantly different from each other. Further, we examined the correlation of the scale items and the scales had reasonable measurement properties.

Antecedent Variable Manipulations

Asset specificity. The scenario manipulation of low asset specificity indicated that the manufacturer had not made any investments with low salvage value outside the supplier relationship in which the investments were presently employed. Conversely, the high asset-specificity manipulation indicated that the manufacturer had dedicated significant investments to the supplier. The manipulation check revealed that the difference in the manipulations on the two-item asset specificity manipulation check was statistically significant (M, low asset specificity = 7.23; M, high asset specificity = 11.62, F(^sub 1,166^) = 121.718, p

Technological unpredictability. In the low technological-unpredictability condition, subjects were informed that customer demand was stable and that both the manufacturer and the suppliers used standardized technology. By contrast, in the high technological-unpredictability condition, subjects read that customer demands changed rapidly and, as a result, the manufacturer constantly had to update its production technology and the production technology of its suppliers in order to stay competitive. The manipulation check revealed that the difference in the manipulations on the one-item technological-unpredictability manipulation check was statistically significant (M, low technological unpredictability = 4.53; M, high technological unpredictability = 5.71, F(^sub 1,1661^ = 19.693, p

Relational norms. We operationalized flexibility, information exchange, and solidarity norms in our manipulations. In the low relational-norms condition, subjects were told that in their supplier relationship, when faced with an unexpected situation, they would hold each other to the original terms of their contract instead of renegotiating their contract to reflect the new situation. By contrast, in the high relation al- norms manipulation, subjects were told that when faced with an unexpected situation, both they and their supplier expected to be able to make adjustments to the original contract to reflect the new realities of their contracting environment. The norms of information exchange and solidarity were also operationalized in a similar fashion. The manipulation check revealed that the difference in the manipulations on the three-item (one item for each norm) relational norms manipulation check was statistically significant (M, low relational norms = 10.70; M, high relational norms 15.75, F(^sub 1.166^) = 95.928, p

Furthermore, to assess the overall quality of the manipulations, we asked the respondents to rate the realism of the scenario (M = 5.2/7, where 7 = very realistic), how interesting they found the situation to be (M = 4.9, where 7 = very interesting), and their confidence in the consistency of their responses with how they would actually behave (M = 6.3, where 7 = very confident). These results suggest that subjects were sufficiently involved in the scenario.

Measures of Remaining Variables

The mediating and consequence variables were all measured using multi-item scales.

Manufacturer dependence on the supplier. The dependence scale measured manufacturers’ expectation of the costs of switching and terminating their relationship with Supplier X. Scale items were drawn from Ganesan (1994) and Morgan and Hunt (1994). The four– item scale had a reliability of .86.

Manufacturer long-term orientation. This scale measured manufacturers’ expectation of continuity relative to the long-run benefits of the supplier relationship. Scale items were drawn from Ganesan (1994). The three-item scale had a reliability of .90.

Both mediating variables were measured following exposure to the antecedent variable manipulations. Before we measured the consequence variables, subjects were exposed to a “critical incident” in the supplier relationship (Bitner, Booms, & Tetreault, 1990). Specifically, subjects read that the supplier would not be able to meet delivery deadlines in the short-term future due to an employee strike at the supplier’s facility. They were also informed that this could have potentially negative implications for their own organizations, i.e., that it could hold up the manufacturer’s production schedule, thus causing the manufacturer to miss delivery deadlines to their own customers that could consequently result in customer ill will.

The critical-incident research method typically involves having subjects recall a particular incident (e.g., a dissatisfying experience) prior to measurement of the antecedents and consequences of this incident. Key limitations of this recall design are that it is susceptible to consistency bias (i.e., the psychological tendency of subjects to achieve consonance between their affective experience of an event and their behavioural responses to the event) and memory lapses (Landon, 1980; cf. Singh & Wilkes, 1996). In order to overcome these limitations, we modified the critical incident method by creating a hypothetical critical incident for subjects and measured their behavioural intentions with reference to this hypothetical incident. From our pretest, subjects rated this hypothetical incident highly on both realism of description and relevance, thus leading us to believe that this approach has validity (see Singh & Wilkes).

Manufacturer commitment. The commitment scale items reflected the manufacturer’s willingness to stand by their supplier through the supplier’s difficult times, despite the potential costs associated with this decision. The four-item scale (Anderson & Weitz, 1992; Morgan & Hunt, 1994) had a reliability of .84.

Manufacturer opportunism. The opportunism scale measured the manufacturer’s willingness to take advantage of the supplier’s difficulty using deceitful means. Scale items were drawn from John (1984) and Provan and Skinner (1989). The three-item scale had a reliability of .73.

Missing Values and Nonresponse Bias

Five respondents did not respond to all of the items. Nonresponse ranged from three to seven items, with no scale missing more than two items. We decided to retain these cases for our analyses by substituting the missing values with the respondent’s average score for the scale (Tabachnick & Fidell, 1989).1

One of our concerns regarding nonresponse bias was that we may not get relatively equal numbers at each of the two levels of the three antecedent variables. Examining the frequencies, we found there were 78 (90) respondents in the low (high) asset-specificity condition, 86 (82) respondents in the low (high) technological– unpredictability condition, and 80 (88) respondents in the low (high) relational-norms condition. Thus, we had sufficient respondents at each level of the antecedent variables, which suggests that there was no systematic nonresponse bias.

Nonresponse bias was also assessed via wave analysis, following procedures recommended by Armstrong and Overton (1977). We compared early respondents (the first 65% of returned questionnaires) with those of the late respondents (the last 35%, based on a natural break in the returns), since research has shown that the characteristics of late respondents are similar to those of nonrespondents. These groups were compared on their ratings of (a) scenario realism, (b) how interesting they found the exercise, (c) their level of confidence that their responses would be consistent with their actual behaviour, and (d) their years of work experience. The results from the MANOVA indicated that early and late respondents were not different on the above characteristics. Accordingly, we concluded that nonresponse bias was not a significant concern.2

Measurement Model

Before estimating the substantive model, we ascertained the validity of our measures of the mediating and consequence variables. Following the procedure recommended by Gerbing and Anderson (1988), we estimated a measurement model in which each of the items was restricted to load on the prespecified factor and the factors themselves were free to correlate. The elliptical reweighted least squares (ERLS) procedure in EQS was used to estimate the measurement model (Sharma, Durvasula, & Dillon, 1989). The results of this model suggested an adequate model fit to the data (Chi^sup 2^ = 149.80, df = 60, p


Structural Model

To test our substantive hypotheses, we estimated the structural model again using the ERLS technique. Overall, our results showed a satisfactory model fit (Chi^sup 2^ = 226.87, df = 95, p

H1: Consistent with our expectation, asset specificity had a significant positive impact on dependence (t = 12.16, p

H2: Our hypothesis of a positive relationship between asset specificity and long-term orientation was not supported. Instead, our results directly contradict this hypothesis (t = – 1.78, p

H3: There was no support for our expectation of a negative relationship between technological unpredictability and dependence (t = -.98, P > .10).

H4: Our hypothesis of an inverse relationship between technological unpredictability and long-term orientation was supported (t 1.85, p

H5: As predicted, relational norms had a significant positive impact on dependence (t = 2.79, P

H6: Relational norms had a significant positive impact on long-term orientation (t = 7.54, p

H7: Our hypothesis of a positive relationship between dependence and commitment was not supported (t = .78, p > . 10).

H8: Long-term orientation was positively associated with commitment (t = 5.99, p

H9: Consistent with our expectation, dependence and manufacturer opportunism were inversely related (t = -2. 10, p

H10: As predicted, long-term orientation was inversely related to manufacturer opportunism (t = -2.47, p

The structural model results are presented in Table 2.

Testing Perfect Mediation

Implicit in our model is the assumption that dependence and long-term orientation are perfect mediators (Baron & Kenny, 1986) of the effect of the three antecedent variables-asset specificity, technological unpredictability, and relational norms-on the two consequence variables: commitment and opportunism. In other words, the six direct paths from the antecedent variables to the consequence variables are restricted to zero in our model. We tested these overidentifying restrictions by estimating each of the six direct paths individually in a series of alternative models. The addition of a structural path results in the loss of one degree of freedom; thus, a significant difference in the chisquare of any one of these six new structural models compared with that of our original model means that the effect of the independent variables is not perfectly mediated (Anderson & Gerbing, 1988; Patterson, Johnson, & Spreng, 1997).

Of the six possible direct paths, only two, the effect of technological unpredictability on opportunism and the effect of relational norms on commitment, reached significance. Consequently, our results show that dependence and long-term orientation perfectly mediate the effects of (a) asset specificity on both commitment and opportunism, (b) technological unpredictability on commitment, and (c) relational norms on opportunism. In the two cases where perfect mediation did not hold, we examined the increase in explained variance in the dependent variable as a result of the addition of the direct path. The increase in explained variance in opportunism is minimal (change in r^sup 2^ = .02) and the increase in explained variance in commitment is marginally higher (change in r^sup 2^ =.03). We conclude, based on these results, that while perfect mediation did not hold, the mediating effects of dependence and long-term orientation are nevertheless substantial.


We drew from TCA and RET to identify three key antecedents of commitment and opportunism in exchange relationships. Further, we extended both theories by explaining the processes by which these antecedents are linked to the outcomes, i.e., dependence and long-term orientation are the process variables that mediate the impact of asset specificity, technological unpredictability, and relational norms on commitment and opportunism. The implications of our results for both TCA and RET are discussed below.

Consider first the relationships between asset specificity and the outcomes. Contrary to the TCA-based expectation, the total effect of asset specificity on commitment is negative (see Table 3). Examination of the individual processes (i.e., [a] asset specificity to dependence, [b] dependence to commitment, [c] asset specificity to long-term orientation, and [d] long-term orientation to commitment) shows that the negative effect of asset specificity on long-term.orientation is what drives this negative total effect.

Consistent with TCA theory, however, the total effect of asset specificity on manufacturer opportunism is negative (see Table 3). However, this aggregate result masks differential underlying processes. While the (a) asset specificity to dependence and (b) dependence to opportunism process is consistent with the expected negative relationship, the (c) asset specificity to long-term orientation, and (d) long-term orientation to opportunism process actually indicates that asset specificity and opportunism are positively related (see Table 3).

This underscores the importance of developing and testing process explanations. Had no process explanations been advanced, we could not have determined how and why asset specificity affects these outcomes. Our explanation for the nonintuitive total effect of asset specificity on commitment would have been speculative in the absence of concrete process measures. We are now, however, in a position to clearly discern how and why this nonintuitive result arose, Similarly, with respect to opportunism, overlooking process variables would have masked the important and interesting differential effects that we uncovered.

The negative effect of asset specificity on long-term orientation is particularly intriguing and holds potentially significant implications for TCA. This begs the question: “Why would invested manufacturers not adopt a long-term orientation with their suppliers?” Reactance theory provides a useful explanation for this unexpected inverse relationship (Brehm, 1966). According to this theory, an invested manufacturer is expected to feel constrained from pursuing its interests unilaterally because of the specific asset investments. This constraint provides the motivation for resistance (or reactance). Reduced long-term orientation vis-a-vis the supplier may then be the first stage in the manufacturer’s attempt to free itself from what may be perceived as an inequitable supplier relationship.

TCA operates squarely within the rationalist psychological paradigm (Ghoshal & Moran, 1996). The invested manufacturer is expected to value the transaction-cost minimizing motivation above all else. Yet our results show this not to be the case. Our process results suggest that specific asset investments can have psychological consequences for the invested party that are quite different from those envisioned by TCA. Specifically, these investments can unleash behaviours that run contrary to transaction-cost minimizing precepts. Recall that TCA operates on the assumption that specific asset investments reduce the potential for dysfunctional behaviour by the invested party since relational continuity is a valued outcome for that party. However, it seems plausible that in some situations, e.g., when the negative effect of asset specificity through long-term orientation on outcomes is greater than its positive effect through dependence, relational continuity may not in fact be highly valued by the investing party. The implication for governance in this situation is dramatically different from that proposed in TCA. Contrary to the standard TCA prescription, the invested party may actually seek to establish market governance with their partner, as this presents them with an effective exit option. In sum, TCA rests on a bed of psychological assumptions of the economic rationalist variety. Should these assumptions be violated, as our results suggest they may be, then dramatically different outcomes from those predicted by TCA can ensue. Thus, TCA theory can be usefully developed by systematically reexamining its psychological assumptions.

The second set of TCA processes considered are those that link technological unpredictability to the outcomes. Although TCA itself does not formally posit main effects of uncertainty, we used TCA reasoning to posit that technological unpredictability would lead to reduced commitment and increased opportunism. Both total effects are in the predicted directions (see Table 3), and the process explanations are consistent with TCA reasoning.

Consistent with RET precepts, relational norms increased commitment and decreased manufacturer opportunism (see Table 3). Although RET has been characterized as being more of a descriptive than predictive theory (Heide, 1994), recent research has begun to systematically develop the predictive dimension of this theory (see Gundlach et al., 1995; Heide & John, 1992). We contribute to this endeavour by tracing the how and why linkages between relational norms and the outcomes. Our results provide full support for the proposed processes.

Managerial Implications

A key finding from this study is that manufacturers that are in high relational norm-based relationships with suppliers develop a long-term orientation in these exchange relationships. This, in turn, will increase their commitment toward and reduce their opportunism against their exchange partners. A noteworthy theme that underlies these results is the importance of relationshipbuilding.

A truism in marketing is that reducing customer defection rates has positive implications for the profitability of a selling firm (Jacob, 1994). From the supplier’s perspective, reduced manufacturer (i.e., customer) opportunism and increased manufacturer commitment are valued outcomes since they reduce the probability of manufacturer defection. As our results indicate, a supplier can cultivate these outcomes by striving to develop high relational norms in their manufacturer relationships.

The development of relational norms is essentially a socialization process whereby both parties come to understand and endorse each other’s expectations. Suppliers can enact this socialization process by consistently displaying relational selling behaviours (Crosby, Evans, & Cowles, 1990; Johnston & Lawrence, 1988). Specifically, they should strive to develop a track-record of (a) enabling customers to cope with downstream market turbulence, (b) keeping customers informed of any upstream market developments that may directly or indirectly affect the customer, and (c) consistently acting in a manner that puts the interest of the relationship ahead of their unilateral interests. Since reciprocity is a powerful social force (Gouldner, 1960), customers are apt to respond favourably to relational selling behaviours on the part of their supplier with similar behaviours of their own, thereby setting in motion the process of relational norm development/reinforcement.

The conventional prescription for reducing customer defection has focussed largely on building switching costs for the customer as a means of tying the customer down (Jackson, 1985; Porter 1990). Our results contradict this prescription. Manufacturers that had invested specific assets became dependent on the supplier and this dependence curtailed their opportunism against the supplier. But at the same time, that dependence did not increase their commitment to the supplier. Further, these specific asset investments actually reduced their long-term orientation with the supplier which, in turn, reduced their commitment toward and increased their opportunism against, the supplier. Our results suggest, therefore, that from the supplier’s perspective, the costs of this prescription (in terms of reduced long-term orientation and its associated outcomes) can outweigh its benefits (increased dependence and its associated outcome). Accordingly, we caution suppliers against adopting the conventional prescription in a heavy-handed way, and encourage them to also actively focus on relationship-building.

Limitations and Future Research

While the strength of the experimental method is that it enables us to control extraneous background factors, its weakness is that it can do so at the expense of external validity. The antecedent variables were manipulated using scenario descriptions. While there is precedence for using scenario manipulations in industrial marketing research (Andaleeb, 1995; Jackson et al., 1984; Pilling, Crosby, & Jackson, 1994; Schurr & Ozanne, 1985), there is always the concern that scenario descriptions, in striving for clarity, can sacrifice subtlety and richness. A survey research approach with purchasing managers serving as key informants might be better able to capture the complexity of the real world. Such a study would be a useful complement to the present research.

In addition, we examined the model from only one side of the buyer-supplier dyad. Increasingly however, researchers are calling for a simultaneous examination of relationship models from both sides of the dyad (e.g. Anderson & Weitz, 1992). Testing our model within the context of a simulation such as “The Market Place” could allow the collection of data simultaneously from both sides of the dyad (Cadotte, 1990).

In addition to addressing the above identified limitations, future research should explore the two hypotheses that were not supported in this research. One possible explanation for the absence of a main effect of technological unpredictability on dependence is that this relationship may be moderated by certain personality variables. We explored this avenue by examining, in an ANOVA context, the interaction effect of technological unpredictability, gender, and work experience on dependence but found null results. We acknowledge that other variables, such as risk propensity (Sitkin & Pablo, 1992), could also moderate this relationship, and we leave the task of identifying additional moderators for future research.

We also did not find support for the positive relationship between dependence and commitment. We transformed dependence into a categorical variable using a median split and assessed its interaction with gender and work experience on commitment using ANOVAs, and again found null results. It seems plausible. however, that additional personality variables such as Machiavellianism (Christie & Geis, 1970; Corfman & Lehmann, 1993) and orientation to collectivism (Wagner, 1982, 1995) could have an impact on the dependencecommitment relationship. Future research should systematically analyze the impact of these personality variables on the relationship between dependence and commitment.


Note that the measurement and structural models were run with and without the five cases with missing values. The results in each case for model fit and parameter estimates were the same,

2. The Armstrong and Overton (1977) procedure, although widely used, is a weak test of nonresponse bias. Contacting nonrespondents to collect data on these variables and then comparing their responses with those of the respondents would have provided a more rigorous assessment of this type of bias. We acknowledge this as a limitation of this research and thank a reviewer for raising this observation.


Achrol, R. (1991). Evolution of the marketing organization: New forms for turbulent environments. Journal of Marketing, 55, 77-93.

Achrol, R., Reve T, & Stem, L.W. (1983). The environment of

marketing channel dyads: A framework for comparative analysis. Journal of Marketing, 47, 55-67.

Achrol, R., & Stem, L.W. (1988). Environmental determinants of decision-making uncertainty in marketing channels. Journal of Marketing Research, 25, 36-50.

Andaleeb, S.S. (1995). Dependence relations and the moderating role of trust: Implications for behavioral intentions in marketing channels. International Journal of Research in Marketing, 12, 157-172.

Anderson, E. (1988). Transaction costs as determinants of opportunism in integrated and independent sales forces. Journal of Economic Behavior and Organization, 9, 247264.

Anderson, E., & Weitz, B. (1992). The use of pledges to build and sustain commitment in distribution channels. Journal of Marketing Research, 25, 18-34.

Anderson, J.C., & Gerbing, D.W. (1988). Structural equation modeling in practice: A review and recommended twostep approach. Psychological Bulletin, 103 (4), 411-423.

Anderson, J.C., Hakansson, H., & Johanson, J. (1994). Dyadic business relationships within a network context. Journal of Marketing, 58, 1-15.

Anderson, J.C., & Narus, J.A. (1990). A model of distributor firm and manufacturer firm working partnerships. Journal of Marketing, 54, 42-58.

Armstrong, J.S., & Overton, T.S. (1977). Estimating nonresponse bias in mail surveys. Journal of Marketing Research, 14, 396-402.

Axelrod, R. (1984). The evolution of cooperation. New York, NY: Basic Books.

Bagozzi, R.P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16 (1), 74-94.

Balakrishnan, S., & Wernerfelt, B. (1986). Technical change, competition, and vertical integration. Strategic Management Journal, 7, 347-359.

Barney, J.B., & Hansen, M.H. (1994). Trustworthiness as a source of competitive advantage. Strategic Management Journal, 15, 175-190.

Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51 (6), 1173-1182.

Berger, P.L., & Luckmann, 1 (1967). The social construction of reality. New York, NY: Bantam Doubleday Dell Pub[ishing Group.

Bitner, MI, Booms, B.H., & Tetreault, M.S. (1990). The service encounter: Diagnosing favorable and unfavorable incidents. Journal of Marketing, 54, 71-84.

Brehm, J.W. (1966). A theory of psychological reactance. New York, NY: Academic Press.

Brill, J.E. (1994). Beyond managerial opportunism: Supplier power and managerial compliance in a franchised marketing channel. Journal of Business Research, 30, 211-223.

Cadotte, E.R. (1990). The market place. Homewood, IL: Richard D. Irwin.

Christie, R., & Geis, Fl. (1970). Studies in Machiavellianism. New York, NY: Academic Press.

Corfman, K.P., & Lehmann, D.R. (1993). The importance of others’ welfare in evaluating bargaining outcomes. Journal of Consumer Research, 20, 124-137.

Crosby, L.A., Evans, K.R., & Cowles, D. (1990). Relationship quality in services selling: An interpersonal influence perspective. Journal of Marketing, 54, 19-31.

Dwyer, FR., Schurr, P., & Oh, S. (1987). Developing buyerseller relationships. Journal of Marketing, 51, 11-27. Frazier, G.L., & Rody, R.C. (1991). The use of influence strate

gies in interfirm relationships in industrial product channels. Journal of Marketing, 55, 52-69.

Frazier, G.L., Spekman, R.E., & O’Neal, C.R. (1988). Just-intime exchange relationships in industrial markets. Journal of Marketing, 52, 52-67.

Ganesan, S. (1993). Negotiation strategies and the nature of channel relationships. Journal of Marketing Research, 30, 183-203.

Ganesan, S. (1994). Determinants of long-term orientation in buyer-seller relationships. Journal of Marketing, 58, 1-19.

Gerbing, D.W., & Anderson, J.C. (1988). An updated paradigm for scale development incorporating unidimensionality assessment. Journal of Marketing Research, 25, 186-192.

Ghoshal, S., & Moran, P. (1996). Bad for practice: A critique of the transaction cost theory. Academy of Management Review, 21 (1), 13-47.

Gouldner, A.W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25, 161-179. Gundlach, G.T., Achrol, R.S., & Mentzer, J.T. (1995). The

structure of commitment in exchange. Journal of Marketing, 59, 78-92.

Gundlach, G.T., & Cadotte, E.R. (1994). Exchange interdependence and interfirm interaction: Research in a simulated channel setting. Journal of Marketing Research, 31, 516532.

Hallen, L., Johanson, J., & Seyed-Mohamed, N. (1991). Interfirm adaptation in business relationships. Journal of Marketing, 55, 29-37.

Heide, J.B. (1987). Explaining ‘closeness’ in industrial purchasing relationships: The effects of dependence symmetry on inter-organizational coordination patterns. Unpublished doctoral dissertation, University of Wisconsin, Madison, WL

Heide, J.B. (1994). Interorganizational governance in marketing channels. Journal of Marketing, 58, 71-85.

Heide, JR, & John, G. (1988). The role of dependence balancing in safeguarding transaction-specific assets in conventional channels. Journal of Marketing, 52, 20-35.

Heide, J.B., & John, G. (1990). Alliances in industrial purchasing: The determinants of joint action in buyer-supplier relationships. Journal of Marketing Research, 27, 24-36.

Heide, J.B., & John, G. (1992). Do norms matter in marketing relationships? Journal of Marketing, 56, 32-44.

Heide, JR, & Miner, A.S. (1992). The shadow of the future: Effects of anticipated interaction and frequency of contact on buyer-supplier cooperation. Academy of Management Journal, 35 (2), 265-291.

Jackson, B.B. (1985). Winning and keeping industrial customers. Lexington, MA: Lexington Books.

Jackson, D.W., Jr., Keith, J.E., & Burdick, R.K. (1984). Purchasing agents’ perceptions of industrial buying center influence: A situational approach. Journal of Marketing, 48,75-83.

Jacob, R. (1994, September). Why some customers are more equal than others. Fortune, p. 216.

John, G. (1984). An empirical investigation of some antecedents of opportunism in a marketing channel. Journal of Marketing Research, 21, 278-289.

John, G., & Weitz, B. (1989). Salesforce compensation: An empirical investigation of factors related to use of salary versus incentive compensation. Journal of Marketing Research, 26, 1-14.

Johnston, R., & Lawrence, P.R. (1988, July-August). Beyond vertical integration: The rise of the value-adding partnership. Harvard Business Review, pp. 94- 101.

Joshi, A.W. (1995). Long-term relationships, partnerships, and strategic alliances: A contingency theory of relationship marketing. Journal of Marketing Channels, 4 (3), 75-94.

Joshi, AW., & Arnold, S.J. (1997). The impact of buyer dependence on buyer opportunism in buyer-supplier relationships: The moderating role of relational norms. Psychology and Marketing, 14 (8), 823-45.

Joshi, A.W., & Arnold, S.J. (1998). How relational norms affect compliance in industrial buying. Journal of Business Research, 41 (2), 106-14.

Joshi, A.W., & Stump, R.L. (1996). Supplier opportunism: Antecedents and consequences in buyer-supplier relationships. In C. Droge & R. Calantone (Eds.), 1996 summer AMA educators’ proceedings: Vol. 7. Enhancing knowledge development in marketing (pp. 129-35). Chicago, IL: American Marketing Association.

Kalwani, M.U., & Narayandas, N. (1995). Long-term manufacturer-supplier relationships: Do they pay off for supplier firms? Journal of Marketing, 59, 1-16.

Kaufmann. PJ., & Dant, R.P. (1992). The dimensions of commercial exchange. Marketing letters, 3, 175-185.

Keith, J.E., Jackson, D.W., Jr., & Crosby, L.A. (1990). Effects of alternative types of influence strategies under different channel dependence structures. Journal of Marketing, 54, 30-41.

Kelley, S.W., Skinner, ST, & Ferrell, O.C. (1989). Opportunistic behavior in marketing research organizations. Journal of Business Research, 18, 327-340.

Kelly, H.H., & Thibaut, J.W. (1978). Interpersonal relations: A theon, of interdependence. New York, NY: John Wiley & Sons.

Kelman, H.C. (1958). Compliance, identification, and internalization: Three processes of attitude change. Journal of Conflict Resolution, 2 (1), 51-60.

Klein, B., Crawford, R.A., & Alchian, A.A. (1978). Vertical integration, appropriable rents, and the contracting process. Journal of Law and Economics, 21, 297-326.

Klein, S., Frazier, G.L., & Roth, V.J. (1990). A transaction cost analysis model of channel integration in international markets. Journal of Marketing Research, 27, 196-208.

Kronman, A.T. (1985). Contract law and the state of nature. Journal of Law, Economics, and Organizations, 1, 5-32. Landon, E.L., Jr. (1980). The direction of consumer complaint

research. In Jerry Olson (Ed.), Advances in Consumer Research (Vol. 7, pp. 335-338). Ann Arbor, MI: Association of Consumer Research.

Larson, A. (1992). Network dyads in entrepreneurial settings: A study of the governance of exchange relationships. Administrative Science Quarterly, 37, 76-104.

Lawler, E.J., Ford, R.S., & Blegen, M.A. (1988). Coercive capability in conflict: A test of bilateral deterrence versus conflict spiral theory. Social Psychology Quarterly, 51 (2), 93-107.

Leenders, M.R., & Fearon, H.E. (1993). Purchasing and material management (10th ed.). Homewood, IL: Richard D. Irwin.

Lusch, R.R, & Brown, J.R. (1996). Interdependency, contracting, and relational behavior in marketing channels. Journal of Marketing, 60, 19-38.

Macneil, I.R. (1980). The new social contract: An inquiry into modern contractual relations. New Haven, CT: Yale University Press.

Mohr, J.J., Fisher, R.J., & Nevin, J.R. (1996). Collaborative communication in interfirm relationships: Moderating effects of integration and control. Journal of Marketing, 60, 103-115.

Mohr, J.J., & Nevin, J.R. (1990). Communication strategies in marketing channels: A contingency perspective. Journal of Marketing, .50, 36-51.

Morgan, R.M., & Hunt, S.D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58,20-38.

Moriarty, R.T, & Kosnik, TJ. (1989). High-tech marketing: Concepts, continuity, and change. Sloan Management Review, 30, 7-17.

Patterson, PG., Johnson, L.W., & Spreng, R.A. (1997). Modeling the determinants of customer satisfaction for businessto-business professional services. Journal of the Academy of Marketing Science, 25 (1), 4-17.

Pilling, B.K., Crosby, L.A., & Jackson, D.W., Jr. (1994). Relational bonds in industrial exchange: An experimental test of the transaction cost economic framework. Journal of Business Research, 30, 237-251.

Porter, M. (1980). Competitive strategy. New York, NY: Free Press.

Powell, W.W. (1990). Neither market nor hierarchy: Network forms of organization. In Barry M. Staw & Larry L. Cummings (Eds.), Research in Organizational Behavior (Vol. 12, pp. 295- 336). Greenwich, CT. JAI Press.

Provan, K.G., & Skinner, SJ. (1989). Interorganizational dependence and control as predictors of opportunism in dealer-supplier relations. Academy of Management Journal, 32 (1), 202-212.

Rindfleisch, A., & Heide, J.B. (1997). Transaction cost analysis: Past, present, and future applications. Journal of Marketing, 61, 30-54.

Ring, P.S., & Van De Ven,A.H. (1992). Structuring cooperative relationships between organizations. Strategic Management Journal, 13, 483-498.

Ring, P.S., &Van de Ven,A.H. (1994). Developmental processes of cooperative interorganizational relationships. Academy of Management Review, 19 (1), 90-118.

Robicheaux, R.A., & Coleman, J.E. (1994). The structure of marketing channel relationships. Journal of the Academy of Marketing Science, 22 (1), 38-51.

Schurr, P.H., & Ozanne, J.L. (1985). Influences on exchange processes: Buyers’ preconceptions of a seller’s trustworthiness and bargaining toughness. Journal of Consumer Research, 11, 939- 953.

Sharma, S., Durvasula, S., & Dillon, W.R. (1989). Some results on the behavior of alternate covariance structure estimation procedures in the presence of non-normal data. Journal of Marketing Research, 26, 214-221.

Singh, J., & Wilkes, R.E. (1996). When consumers complain: A path analysis of the key antecedents of consumer complaint response estimates. Journal of the Academy of Marketing Science, 24 (4), 350-365.

Sitkin, S.B., & Pablo, A.L. (1992). Reconceptualizing the determinants of risk behavior. Academy of Management Review, 17 (1), 9-38.

Stem, L.W., & Reve, 1 (1980). Distribution channels as political economies. Journal of Marketing, 44, 52-64.

Stump, R.L. (1995). Antecedents of purchasing concentration: A transaction cost explanation. Journal of Business Research, 34, 145-157.

Stump, R.L., & Joshi, A.W. (1998). To be or not to be [locked in], that is the question: An investigation of buyers’ commitments of dedicated investments to support new transactions. Journal of Business- to-Business Marketing, 5 (3), 33-64.

Tabachnick, B.G., & Fidell, L.S. (1989). Using multivariate statistics Ord ed.). New York, NY: HarperCollins.

Wagner J.A., III (1982). The organizing dilemma: Individualism, collectivism, and the control of organizations. Unpublished doctoral dissertation, University of Illinois, Urbana,IL.

Wagner, J.A., III (1995). Studies of individualism-collectivism: Effects on cooperation in groups. Academy of Management Journal, 38 (1), 152-172.

Webster, FE., Jr. (1992). The changing role of marketing in the corporation. Journal of Marketing, 56, 1-17.

Weiss, A.M., & Heide, J.B. (1993). The nature of organizational search in high technology markets. Journal of Marketing Research, 30, 220-233.

Williamson, O.E. (1975). Markets and hierarchies: Analysis and antitrust implications. New York, NY: Free Press.

Williamson, O.E. (1985). The economic institutions of capitalism. New York, NY: Free Press.

Williamson, O.E. (1991a). Comparative economic organization: The analysis of discrete structural alternatives. Administrative Science Quarterly, 36, 269-296.

Williamson, O.E. (1991b). Strategizing, economizing, and economic organization. Strategic Management Journal, 13, 483-498.

Wilson, D.T., Dant, S.P., & Ran, S.-L. (1990). State-of-practice in industrial buyer-supplier relationships (Report 6

1990). University Park, PA: Institute for the Study of Business Markets.

Ashwin W. Joshi

York University

Rodney L. Stump

Morgan State University, Baltimore, MD

Address all correspondence to Ashwin Joshi, Schulich School of Business, York University, 4700 Keele Street, North York, ON, Canada, MM IP3.

Copyright Administrative Sciences Association of Canada Dec 1999

Provided by ProQuest Information and Learning Company. All rights Reserved