Concerning trust and information

Industrial Marketing Management 36 (2007) 968 – 982

Concerning trust and information Sara Denize a,⁎, Louise Young b,1 a

School of Marketing, University of Western Sydney, Locked Bag 1797, Penrith South DC, NSW 1797, Australia b School of Marketing, University of Technology Sydney, PO Box 123, Broadway, NSW 2007, Australia Received 15 December 2006; received in revised form 10 June 2007; accepted 16 June 2007 Available online 24 July 2007

Abstract Communication and the associated information exchanges are key drivers of the development of relationships and of the trust embedded within them. This paper considers the development of business relationships in terms of the continuing co-evolution of trust and information exchange and the issues associated with researching these processes. The interconnections of trust and information exchange are examined in a survey of business relationships involving information exchange (n = 355). Analysis of variance shows few, if any, of the aspects of the standard conceptualizations of information exchange are associated with increased levels of trust. It is information exchange norms that have the greatest (positive) association with the level of trust. These norms emerge as part of the long term co-production of the relationship itself. This has important managerial implications. We conclude there are few managerial actions involving managing communication that can “manufacture” trust and improve or develop relationships in the short term. The paper concludes with a discussion of alternative ways of envisaging communication and relationship management and the nature and future of research into the evolution of business relationships. © 2007 Elsevier Inc. All rights reserved. Keywords: Trust; Information exchange; Communication; Interfirm relationships; Relationship evolution

1. Introduction The importance of ongoing relationships, in such diverse areas as industrial marketing (Axelsson & Easton, 1992; Håkansson, 1982; Håkansson & Snehota, 1995), marketing channels (Dwyer, Schurr, & Oh, 1987), services (Gummesson, 1993), and relationship marketing (Bhattacharya & Bolton, 2000; Kavali, Tzokas, & Saren, 1999; Weitz & Jap, 1995), is widely recognized. Committed and trusting relationships — often grounded in a network of personalized associations — are seen as a context in which continuing, effectively operating business is likely to occur (Young & Wilkinson, 1997). Mechanisms that facilitate relationship development are critical for managers' success. Among these is effective communication. The ability to exchange information and use these exchanges to create knowledge are critical parts of relationships' performance (Brown & Duguid, 2000). Via communication, information is acquired about the relationship's ⁎ Corresponding author. Tel.: +61 2 9852 4137. E-mail addresses: [email protected] (S. Denize), [email protected] (L. Young). 1 Tel.: +61 2 9514 3538. 0019-8501/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.indmarman.2007.06.004

functioning, the salient factors impacting on it and the competence and motivation of the principals within it. Decisions are thereby made as to who to work with (or not), how much resource to invest and in what ways to commit. Trust is deeply embedded in the “communication–decisionmaking” process. Trust is that thing that opens our minds to others and in doing so, opens the possibilities of leveraging business relationships and opening networks to achieve competitive advantage (Young, 2006). The considerable body of published work that conceptualizes, operationalizes and tests the importance of trust in business marketing concludes that it is central to business functions. A myriad of causal relationships with other relational constructs have been proposed and their covariance tested; in particular commitment, cooperation, conflict and performance have been foci. In that research trust is often conceptualized as a driver of relational performance (Morgan & Hunt, 1994). Less attention has been paid to the ways in which relationships and performance evolve, and the role that trust and communication play in this (Wilkinson, Young, & Ladley, 2007). Recognition of the evolutionary properties of relationships presents challenges for the relationship researcher; the commonly-used, variance-based structural equation modeling

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

(SEM) and similar types of analysis of survey-based data do not allow multi-directional relationships to be tested. These deep processes can be explored using case methods but this is not always possible. Access, time and resource constraints as well as a desire to move beyond the modest extrapolations that cases allow, all feature as reasons for using other methods. This paper addresses these challenges. To do this we use a dynamic conceptualization of relationship development (which, as is discussed, creates substantial challenges for researchers) throughout the paper. The paper addresses this neglected “relationship-as-process” perspective focusing on the processes of communication and in particular information exchange and how they are both created by and give rise to trust. More specifically, the paper focuses on the ways in which information exchange and trust are linked. This is explored in a framework of relationship evolution. That is, the framework articulates that through time information exchange (and other factors) both create the conditions to facilitate or impede the development of trust; and that trust creates an environment which impacts upon the information exchange that occurs (Young & Wilkinson, 1997). This builds on work that addresses the life cycle of relationships, e.g. Dwyer et al. (1987) and Ford (1980) where the state of the future relationship is conceptualized as a function of the current relational state. The paper begins with a conceptualization which highlights trust as a dynamic experience. Next we consider the nature of communicated information, exchange of information and the connections between trust and the informational properties and processes of relationships. Here too, we present information exchange as emerging from the interplay between its component parts and as embedded in and emerging from an atmosphere of trust. The associations between trust and information exchange are then empirically assessed in an exploratory study. Given our study uses data from a cross-sectional survey, this requires we consider relationship development less directly. Measures are constructed and the validity of their content — relative to our dynamic perspective — is considered. Using transparent methods of analysis, specifically ANOVA rather than SEM, we illustrate ways information exchange is, and is not, associated with trust. In discussing these outcomes we reflect on the ways and extent to which these findings can be interpreted within an evolutionary framework. The paper concludes with a discussion of the implications of these findings and the implications of different research approaches in studying business relationships as well as suggestions for further research.

exchange relationship existing there is trust (Williamson, 1993). This orientation towards rational assessment is reflected in operationalizations that articulate trust in terms of assessment of others' reliability and integrity (Morgan & Hunt, 1994), reliance (Rotter, 1967) and willingness to act in your interests (Anderson & Narus, 1990). These assessments are important but insufficient to explain the central role of trust in linking the processes and functions of relationships. This requires consideration of how trust is generated, specifically what activities open our minds to others, enable empathy and deepen relational capabilities (Young, 2006; Young & Albaum, 2003; Young & Daniel, 2003). Fig. 1 depicts trust as an amalgamation of responses to, and circumstances that occur as a result of, assessments of the relationship (and relationship partner) and communication occurring. As depicted in Fig. 1, emotions and cognitive elements combine to form trust. The latter include the calculation of costs, benefits and risks associated with particular situations (Bagozzi, Gopinath, & Nyer, 1999), the way that others who are recipients of trust are perceived, including their credibility (Young & Wilkinson, 1989) and assessment of the state of the focal relationship (Wilkinson & Young, 1994). These combine to form an overall perception of a relationship. At the same time participants are emotionally responding to these assessments (and assessments are being modified as a result). This process is the evolving context in which affectual trust — a sustainable combination of cognition and emotion — is being built. As the context changes, so too do the “rules of the game” for building trust and what constitutes the sustainable trust that is sought. The process depicted in Fig. 1 is that of relational evolution. This figure shows trust and communication as co-evolving as do communication and relational assessment and relational assessment and trust. This depiction is at variance with considerable literature in marketing that has considered the relationship entirely one way — where communication creates the climate/conditions for trust (e.g. Mohr & Nevin, 1990;

2. Trust in relationship evolution and communication Trust has been the focus of considerable research in industrial marketing. Yet there are few theories of trust itself. In marketing the focus has moved toward a functional conceptualization where trust is presented as, that which emerges from assessment of risks/costs set against the benefits of interacting. This establishes the universality of trust in transactions; as risk is inevitable (Barber, 1983). But it conceptualizes trust primarily as the criteria for exchange and as a result does not discriminate well between different kinds relationships, e.g. by virtue of an

969

Fig. 1. Development of trust and relationship evolution.

970

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

Zand, 1972). There is however another body of literature that asserts that trust is an antecedent to communication (based around Morgan & Hunt, 1994). Rather than choosing a one approach Fig. 1 recognizes the importance of both. The figure depicts communication as the way — deliberately or not — we receive information and insights as to the costs, benefits and risks associated with particular situations (Bagozzi et al., 1999), and the state of the focal relationship (Young & Wilkinson, 1997). Information exchange is a particular type of communication which excludes certain social or personal interactions such as phatic communion or discourse relating to social or personal matters (Denize, 2004). Thus, an email revising a delivery schedule is an information exchange but an email reflecting on a social occasion would not be so classified. Both information exchange (that we see as at the heart of business relational development and hence our focus) and other communication can play a role in relationship evolution. Information exchange drives the assessments central to the generation of trust (Young, 2006) but it is embedded in other communications (and hence the two are likely to be difficult to separate). Via communication with others (and information about them), we also make assessments as to their competence and motivation, referred to as “credibility” (Young & Wilkinson, 1989). As Fig. 1 indicates, communication and trust result in certain benefits (and costs) for a relationship and those participating in it. These, in turn, change the nature and context of the relationship by changing the processes that are possible within it (including the communication possibilities), the value placed on the relationship, the expectations of those involved and the conditions surrounding the relationship. Changed relational conditions in turn create changed opportunities for trust (Young & Wilkinson, 1997). In other words, the figure depicts emergence, that is, there is no causal order but rather there is a combination of factors coalescing in a bottom-up way with not always predictable relational evolution the result (Holland, 1995). 3. Credibility: linking communication, information exchange and trust Trust and communication interact but consideration of the role that these interactions play in relational development has been neglected for some time. The early work on trust by communication and social psychology researchers put considerable focus on credibility as a key ingredient in the process. An individual is believed to be credible if they do as they say they will and/or convey information accurately (Rotter, 1971; Schlenker, Helm, & Tedeschi, 1973). Therefore, it was argued, perceived credibility makes it possible to anticipate another's future behavior, reduce perceived risk and thus to trust them (Deutsch, 1958). This work showed strong associations between trust and credibility, in some instances credibility was conceptualized as a component or kind of trust (reflected in work of authors such as Giffin, 1967). One stream of research focused on the informational attributes of credibility. But these researchers found that for the emergence of trust, informational credibility was less important than personal credibility. While

the information itself might be evaluated for its credibility, of greater importance is the perception of the communicator of that information. If information comes from a trusted other it is often not evaluated but is immediately trusted and used (i.e. it is seen as an “agent” of the trusted other). This applies in personal settings and also sometimes when the source is not known to the receiver of the information (Giffin & Patton, 1976; Hovland, Janis, & Kelley, 1953). A related stream of research looked at credibility more as a general social competence. People are judged as credible based on their general communication abilities, specifically their ability to communicate their (trustworthy) intentions (Loomis, 1959). Also considered in this work was the interaction of the communicator and recipient. The capability of a “truster” to receive, process, and act upon another's intentions is seen as a significant part of communication and trust (Giffin, 1967). The gradual disappearance of (explicit) consideration of credibility as a driver of trust has had substantial implications (work by Moorman, Zaltman, and Deshpandé (1992), Moorman, Deshpandé, and Zaltman (1993) is an exception). First, consideration of the direct link between trust and communication largely disappeared, with the result that a major group of behavioral drivers of trust were excluded from consideration as mechanisms generating trust. Instead, more generalized traits of others (e.g. reliability, fairness) and relational states (presence/ absence of contracts, goodwill) became the focus of research concerned with the antecedents of trust (Morgan & Hunt, 1994). These traits and states were more easily conceptualized as (only) trust's antecedents rather than also as outcomes and this no doubt contributed to the current focus on the studying of relationships in a comparative static way (Wilkinson et al., 2007). We conclude that credibility is an important inclusion in conceptualizing and operationalizing trust. The removal of credibility from trust research was due, we believe, not to its lack of importance, but to its ambiguity. We consider this ambiguity as strength. Credibility is ambiguous because can be aligned with either positive or negative behavior, e.g. early researchers point out that one could “trust” in the delivery of punishment if it was believed that their threatener was credible (Schlenker et al., 1973; Tedeschi, Schlenker, & Bonoma, 1973). However the common usage of the term “trust” (in marketing) is positive and it is likely that in part to avoid confusion credibility was gradually excluded from the considered set of things that led to trust (Young, 1992). But this ambiguity reflects the reality of relationship evolution. Relations can evolve in positive or negative ways and close and trusting relationships can have both positive and negative impacts (Wilkinson & Young, 2002, 2005). Second, credibility is empirically ambiguous. When constructing trust scales, communication elements, including credibility, often cross-load on factors (i.e. have high factor loadings on two or more factors) and are dropped as they do not give the “clean” dimensions that will maximize discriminant validity of scales (as originally prescribed by Churchill (1979)). And, as in the past two decades previously-utilized items increasingly have been used to construct trust scales, credibilityrelated items increasingly have become less likely to be included in the set of potential items. However we argue that

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

the factor cross-loading of credibility-related items are a strength rather than weakness. This is particularly the case where a single trust scale is the goal; then items that indicate that the scale has a rich mixture of trust attributes within it are desirable. Fortunately this is being increasingly recognized and is addressed in alternative methods of scale construction, e.g. Rositter's (2002) C-OAR-SE method and use of Rasch modeling (e.g. Andrich, 2002). 4. Towards a conceptualization of information exchange in relationships Information exchange and trust are embedded in and emerge from their interaction with each other. They create and are created by each other. The following discussion illustrates how the various aspects of information exchange interrelate and coevolve with trust. This is the same process that is depicted in Fig. 1, however in Fig. 2 the focus is on the communication and in particular information exchange aspects of relationship development. Mutual exchange of information is at the heart of this evolutionary process (Ford, 1980). Fig. 2 depicts the interaction of the type of information, the way information is transformed, the medium by which it is exchanged, and the transfer of that information. These influence and are influenced by relational factors including the motivations of the entity participating in the exchange, the norms of exchange that have developed and the more general nature of the relationships. The figure indicates a continuing cycle where, if relational performance warrants it and trust builds, possibilities for information exchange will be extended. This facilitates the further development of trust and opens further relational opportunities (Hallén, Johanson, & Seyed-Mahamed, 1990).

971

Fig. 2 extends the conceptualization of information exchange by widening the set of attributes included within it and by considering the rich connections between these attributes. This is in contrast to the commonly-used, simple models of information exchange which focus solely on activity, i.e. the way information is exchanged, or on resource, i.e. type of information that is exchanged. The time frame of information exchange is also extended in our conceptualization. Central is the emergence of information exchange norms. This occurs in the medium to long term, as the relationship develops. In the early stage of the development, firms may engage in superficial information exchange. As the relationship matures, an information exchange “meta-structure” emerges which is based on shared norms and expectations. Information exchange is likely to be characterized by greater scope and depth of topics (Weitz & Jap, 1995). Cunningham and Homse's (1990) findings support this view. As a precursor for operationalizing the attributes of information exchange, the following discussion considers in more detail the rich interconnected nature of the aspects of information exchange and considers how these influence relationship development and are influenced by it. To contextualize, these aspects of information exchange are considered within the framework of the Actor–Activity–Resource (AAR) model (Axelsson & Easton, 1992). The AAR model is a conceptual framework with which to explore the interactions within and beyond relationships. Relationships and networks are considered in terms of the links between actors (bonds), the activities they perform (links) and the resources they control (ties) as well as the connections between the links, bonds and ties. Of particular relevance for this work is a proposed a fourth “dimension” of the AAR model, “schema coupling,” i.e. the idea combinations that result from interaction and communication,

Fig. 2. Information exchange; actor–activity–resource dimensions.

972

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

(Welch & Wilkinson, 2002). This is entirely compatible with the focus of Fig. 2 which sees communication and in particular information exchange as at the core of relationship functioning. The following discussion of information exchange considers it both in terms of each of four aspects of the AAR model and in terms of the way information exchange attributes (sometimes) represent more than one aspect, i.e. that the aspects overlap and interact. At the same time the ways in which various information exchange attributes are developed within relationships and in turn assist in developing relationships are considered. To describe Fig. 2 information is first considered in the traditional way, as (primarily) a resource. Types of information are often differentiated by content (Weill, 1990a,b). Transactional information is that which is relevant to the day-to-day activity of firms, e.g. information about the sale of goods and services. Operational information is that which pertains to the smooth functioning of the firm, e.g. details of inventory levels. Strategic information provides competitive advantage. It is argued that exchanging transactional information has little or no effect on trust but its efficient exchange is likely to reduce costs. The exchange of strategic information is only likely when there is highly developed relationship with trust. From a relationship development perspective there is an implied hierarchy of these three information types. We argue that the exchange of all three types of information also creates the conditions for relational development, including a greater degree of trust but that there is more opportunity for relational development in the exchange of operational than transactional information and in strategic than operational exchange. Relational development in turn creates further possibilities for development of information resources. Information exchange is also conceptualized as activitybased, considered here in particular in terms of transferring activities. Transfer activities are characterized by their direction and reciprocity, complexity, intensity, variability and scope (Arndt, 1985). Transfer direction and reciprocity are concerned with the orientation of the information flow to and/or from one actor to another. As we would expect, congruence between the nature of the flows, the extent to which they are unidirectional or reciprocal has been shown to positively impact on relational factors such as commitment, satisfaction and coordination (Mohr, Fisher, & Nevin, 1996). This is represented in Figs. 1 and 2 in terms of the nature of the relationship (and its maturity). The character of the relationship is determined by the pattern of information exchange events including the reciprocity of exchange and the expectations for future benefits, and in turn the character of exchange events is determined by the state of the relationship. The intensity of transfer activities (i.e. their frequency) and scope of these activities (i.e. extent) are positively associated with trust (Maltz & Kohli, 1996). The complexity of transfer activities (i.e. presence of multiple links, involving multiple messages transferring different types of information) has not been considered in studies involving interfirm relationships. However, work describing the evolution of interfirm relationships suggests that the information exchange process will become more “complex” as the relationship matures and

exchange partners make adaptations. By this we mean that the development process is adaptive and emergent. This can be envisaged as a framework where trust and information exchange are constantly “recreating” each other, i.e. trust is necessary for complex information exchange processes to develop (as per Cunningham & Homse, 1990; Ford, 1980, 1984; Hallén et al., 1990; Weitz & Jap, 1995) and information exchange is a means by which further trust is created. The variability of transfer activities describes the extent to which the information exchange process differs from transaction to transaction in a given relationship. There is empirical support for the notion that this is progressive. Mohr with various others (Mohr & Sohi, 1995; Mohr et al., 1996) establish that communication which is structured, planned, and explicit (and which does not vary) is positively associated with cooperation and commitment and negatively associated with opportunism (conceptualized as distorting and withholding information). From this one might infer that exchange variability evolves within an evolving relationship at least up to a point. This has been addressed in life cycle orientated relationship research, which found that mature relationships tend to be characterized by high trust and less variability in information exchange process — as through time adaptations are made between the organizations (Cunningham & Homse, 1990). Presumably this would be institutionalized in very long term relationships (along the lines described by Ford, 1980), i.e. neither the relationship as a whole nor the exchange of information within it would vary much beyond a certain point of development. The AAR model facilitates the consideration of the interaction of activities and resources. Here, the properties of the activity of the transfer of information are inextricably linked to the nature of the information that is exchanged. Thus, information is considered as a resource embedded in activity, with properties similar to those of transfer activities including; complexity, intensity, scope and variability. Information complexity refers to the density/simplicity of information resources. Simple resources will usually be straightforwardly factual, whereas more complex information will be dense, conditional and interrelated. Complex information requires more processing before it can be integrated into the knowledge structures of the actor receiving it. Information intensity describes the range of different types of information in a transaction. As the number of types of information increases, the information intensity rises. Information scope refers to the amount of information. As the volume of information transferred increases, information scope increases. Together these properties describe the richness of the information. As relationships mature and exchange partners increasingly make adaptations to one another, it becomes possible to transmit information of increasingly broad scope, i.e. with more intensity and complexity. This information in turn creates opportunities for further relationship development and specifically for a greater degree of trust (in line with Cunningham & Homse, 1990). Information variability describes the newness (or novelty) of the information and the changes in the type of information that is exchanged. Firms engaged in ongoing trusting relationships have institutionalized their information exchange behaviors.

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

What is known to one actor is largely (already) also known to the other. Conversely, less mature relationships offer the opportunity for variable (novel) information to be exchanged. Theories of relationship evolution, knowledge acquisition, and social networks all suggest that as relationships develop and mature there will be a reduction in the variability of information (Constant, Sproull, & Kiesler, 1996; Granovetter, 1973; Hansen, 1999). The form and character of the way in which information is exchanged are important parts of the information resource — often characterized as media “status” (Kaye, 1995). Here we continue to focus on the activity–resource interaction but now include actor components. Of particular relevance are properties of the information exchange media, its complexity and personalization. The media/resource complexity is concerned with the amount of processing required to make data exchangeable. For example oral media formats are the least complex. Textbased and multimedia documents are more complex. Electronic media formats are the most complex. Previous conceptualization suggests that as relationships mature and adaptations are made, increasingly complex media formats are supported (Dwyer et al., 1987; Ford, 1980). We would add to this that as increasing complex media formats are used, the information that can be communicated means that there are increased possibilities for relational development and specifically for the development of trust. Media/resource personalization relates to the degree of interpersonal contact that occurs when using a particular medium. For example, a telephone is more personalized than computerbased information exchange protocols. There is considerable evidence that using increasingly personalized media facilitates relational development (Cunningham & Homse, 1990). We would add to that relational development constrains and facilitates the choice of exchange media. Next we consider motivations for information exchange, and the actor–resource interplay in which they emerge. Actors exchange information for a purpose; to use it in different ways, i.e. to recombine the ideas into schema such that these have (greater) meaning and relevance for their organization (Welch & Wilkinson, 2002). A standard typology of information (or knowledge) use differentiates between instrumental, conceptual and symbolic utilization. Instrumental utilization refers to the application of information in decision-making. This implies specific action on the part of the receiver — for example, a price change, a call to a customer, calling a sales meeting, adjusting goals and so forth (Menon & Varadarajan, 1992). Conceptual use occurs when information is used indirectly to enhance knowledge (Menon & Varadarajan, 1992; Weiss, 1980). Symbolic information use describes the use of information that is inconsistent with the intended purpose or function of the information (Weiss, 1980). These three forms of utilization are related in somewhat different ways to relational development. We assume that motivations for use of information are grounded in motives for exchange. In other words people exchange information so that received information can be used in one or more ways. Consciously or unconsciously this includes using information to manage relationships. Instrumental use is primarily a risk-

973

reducing device and this, it is argued, increases trust (Easton, 1992). Conceptual use relates to knowledge enhancement which also reduces risk. Furthermore, the combining of new information with pre-existing knowledge to “grow” new knowledge implies that this will often take place within a continuing relationship. While knowledge will in part be grown from combining information from different sources, it also involves updating information previously received from the same source or using knowledge built from the relationship as a context in which to evaluate and use new information. Symbolic use of information is somewhat analogous to “affective use” as described by Menon and Varadarajan (1992, p. 62). “Affective use …is the use of research with the intent of ‘feeling good.’” The use of relationships to enhance our well-being has been noted by a number of social psychologists (Argyle, 1991; Asch, 1952) and by market researchers (Dixon & Wilkinson, 1986). The desire for better quality and more developed and more trusting relationships follow from this (Young, 2006), as does the use of information to facilitate trust. We speculate that the precise nature of relationship-enhancing motivations may evolve as the relationship evolves. In new relationships information is used to build information-sharing norms and the development of trust while in more mature relationships the primary motive may be to sustain the relationship and institutionalize existing practice. Also part of the relational aspects of information exchange depicted in Fig. 2 are information exchange norms. These are joint expectations concerning information exchange behaviors within the relationship (Heide & John, 1992) and incorporate actors acting to create and exchange resources through time. The impact of information exchange norms on interfirm relationships is well documented. Mohr and Nevin (1990) predicted a strong positive impact on relational trust. This has been substantiated empirically (Campbell, 1997; Cannon & Perreault, 1999; Doney & Cannon, 1997; Doucette, 1997; Mohr & Spekman, 1994; Mohr & Sohi, 1995; Mohr et al., 1996). In line with the discussion thus far, we argue that it is entirely likely that relational trust also has a significant impact on the growth of norms. This is consistent with the foundation work in psychology which found significant impacts upon the norm of a “generalized expectancy of trust” that guided behavior (Rotter, 1967, 1971, 1980). The final aspect of information exchange we consider is the actor–activity interplay associated with use of influence. Activities directed at influencing the behavior of others can be regarded as transformational. Such activities have not normally been considered part of information exchange but they have been conceptualized as a type of communication behavior (Mohr & Nevin, 1990; Mohr et al., 1996). Mohr and Nevin (1990) describe two broad types of communication behavior — coercive and cooperative communication. Coercive communication is a transformational activity for gaining compliance that refers to negative consequences for noncompliance. These activities include threats, demands and negative normative communications. Conversely, cooperative communication is a transformational activity that refers to positive consequences for compliance. These activities include

974

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

promises, information persuasion, requests, and positive normative communications (Gundlach & Cadotte, 1994; Hunt & Nevin, 1974; Wilkinson & Kipnis, 1978). Empirical work suggests that influence attempts have negative consequence for relationships, even when cooperative strategies are used. These effects are likely to be far more damaging when the transformational activities are coercive in nature (Frazier & Rody, 1991; Håkansson, Johanson, & Wootz, 1990). Influence use is evolutionary as the application of influence, the response to it and the response to the response are part of a continuing process (Kelman, 1961). Also, the use of power and its influence on trust and vice versa have been considered in the marketing literature (e.g. Morgan & Hunt, 1994; Young & Wilkinson, 1997). 5. A study of trust and information exchange Information exchange relationships were examined using a comprehensive structured questionnaire that closely examined the connections between information exchange partners. The use of this method followed the approach used in the Interfirm Relations Research Program (IRRP) and the IMP2 (Industrial Marketing and Purchasing) project (Blankenburg-Holm, Eriksson, & Johanson, 1996; Wiley, Wilkinson, & Young, 2006; Young & Wilkinson, 1997). These researchers have argued that this approach gives effective and comparable “road maps” of interfirm relationships across different contexts. Interviews were conducted with key informants about the information exchange relationships in which their firm was engaged. Key informants were selected using a convenience sampling strategy but were excluded through screening questions at the beginning of the interview if they were not knowledgeable about at least one information exchange relationship(s). Although not necessarily optimal, the use of key informants to report on relationships is a well established approach in business research (Kumar, Stern, & Anderson, 1993). The focal relationship discussed in the interview was systematically selected. This was done by asking informants to identify information exchange relationships between their organization and four others (a customer, a supplier, a competitor and another non-trading relationship). The interviewer then selected a single, focal relationship for the interview according to a predetermined random framework (in line with methods proposed by Wilson & Mummalaneni, 1990; Young & Wilkinson, 1997). As was the case in these previous studies of IRRP and IMP2 this combination of convenience and systematic selection resulted an intentionally broad crosssection of organizations, industries and informants. A total of 369 interviews were conducted, 14 surveys were removed because they were incomplete. The final sample included reports from 355 informants. A further description of the sample is provided in the results section. Following is an overview of the operationalization of trust and information exchange. Item descriptions are provided in Appendix A (more detailed operationalization information and item performance information is available from the corresponding author).

5.1. Measurement of trust Measures of trust were drawn from earlier work (Morgan & Hunt, 1994; Young, 1992; Young & Wilkinson, 1989, 1997) and included were items that captured beliefs about partners' confidence in each other, reliability, truthfulness and trustworthiness. In line with Young's (1996) recommendation that attention should be paid to the mixture and variability of items included in a trust scale, we determined that parsimony dictated a single scale. To conform to our conceptualization this would need to be broad-based and include credibility as a key component. Therefore the initial item set included both direct (use the word “trust” in items) and indirect measures of trust and both behavioral (e.g. “provide a truthful picture”) and affectual (e.g. “have confidence”) aspects. The items included direct measures of both trust and trustworthiness to allow us to consider relational activity (i.e. an item addressing how the other party behaved and another concerned with the informant's response). The indirect measures of trust are similarly relational and focus on credibility, including both the motivational aspects of credibility, i.e. the honest, truthful aspects of credibility and the competence aspects that are to do with the quality of the information emanating from the source. Interestingly, these items dealing with competence aspects of credibility such as accuracy of information provided had somewhat lower factor loadings and were excluded because they slightly lowered the alpha (0.89). The resulting set items reflected “trust in each other”, “confidence in each others fairness and honesty”, “trustworthy behavior” and “truthfulness in negotiations”. Confirmatory factor analysis indicated that a four-item solution had good fit and reliability (χ2 (2) = 15.14; Δχ2 (12) = 65.45, p = 0.0000; GFI = 0.979; TLI = 0.952; CFI = 0.984; IFI = 0.984, explaining 68% of the variance with a composite reliability of 0.89). Future work will consider the implications of excluding/ including competence aspects of credibility. 5.2. Measurement of information exchange Measures of all previous-discussed aspects of information exchange were developed to enable exploration of their association to trust. Property of information — As we have discussed, transfer activities can be characterized by their direction/reciprocity, complexity, intensity, variability and scope (Arndt, 1985). There were no pre-existing scales to measure these dimensions; accordingly, single-item measures were developed. Direction/ reciprocity was measured on a seven-point rating scale. Informants were asked to consider “who provides most information”, (Respondent Firm = 1, Other Firm X = 7). A mid-point was provided so balanced information exchange could be indicated. The complexity, intensity, scope and variability of the transfer activity were each measured using six-point semantic differential rating scales. Similar single-item measures were used to describe the properties of information (see Appendix A for items used). Type of information — Information content was classified as transactional, operational or strategic in line with Weill (1990a,b).

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

A pool of twelve items measured the exchange of various types of information content. The final, formative, three-dimensional measure revealed the extent/scope to which transactional, operational and strategic information resources were exchanged. A similar approach was used to develop identify the properties of the media as information resources draw many of their attributes from the media used in the exchange. Transfer media (e.g. telephone, in-person, EDI and so forth) were characterized by their complexity and the extent to which they were personalized. These characteristics of information exchange media were measured using a formative multi-item scale capturing the use of various types of information exchange media. See Appendix A for the set of items used. Motivations for information exchange — No items for information exchange motivations have been previously developed but information/knowledge use is a well-developed concept within both the marketing and information utilization literature. This literature was used to develop a multi-item reflective measure of instrumental, conceptual and symbolic information exchange motivations. The final scale comprised twelve items reflecting the three motivational dimensions. Items were measured on a six-point rating scale indicating agreement. A four-factor solution was obtained via exploratory factor analysis. This solution explained 61.8% of the variance. All items loaded on the theorized dimension, but items reflecting symbolic motivations split into two groups (relational and forced motivations). Confirmatory factor analysis provided evidence of fit with the theoretical structure of the motivation construct (χ2 (49) = 137.55, p = 0.000, GFI = 0.94, TLI = 0.86, CFI = 0.90, IFI = 0.90) composite reliability statistics range from 0.61 to 0.73 and variance extracted ranged from 0.35 to 0.58. Future work will focus on improving this new measure. See Appendix A for items used. Transformational activities — The transformational properties of information exchange are reflected in the communication behaviors or influence tactics employed by exchange partners. Influence tactics have generally been operationalized using reflective multi-item scales (for each type of influence tactic, for example, threats, demands, promises etc.). However, given the overall length of the research instrument, an alternative approach was adopted. Two multi-item formative measures were developed (along the lines of Young (1992)) to capture the coercive and non-coercive dimensions of the transformational activity. Six-point rating scales were used to measure the likelihood of using each of these transformational behaviors within the relationship. A pool of ten items was used to form the dimensions of coercive and cooperative communication behaviors discussed in the literature (Gaski, 1984; Gaski & Nevin, 1985; Mohr & Nevin, 1990; Mohr & Spekman, 1994; Mohr & Sohi, 1995; Mohr et al., 1996; Stern, 1972). Three of the items measured particularly strong coercive or confrontational communication behaviors, while the remaining items measured a range of more cooperative, less confrontational communication behaviors. See Appendix A for items used. Information exchange norms are joint expectations about information exchange behaviors within the relationship (Heide & John, 1992). There have been a number of studies using

975

measures of information exchange norms, mostly based on the work of Heide and John (1992) and Heide and Miner (1992). Measures of information exchange norms have typically included items reflecting expectations about information exchange as well as items reflecting information exchange behaviors. As a result, there is some correspondence between the potential item pools used to operationalize information exchange norms and items that reflect aspects of informational credibility in measuring trust (refer to Appendix A for item descriptions). This is scarcely surprising; as discussed trust and information exchange norms are part of the same emergent process of relationship development. Recognizing emergence is at the core of the relationship development process requires recognizing that parts of the process are overlapping and thus are not conceptually distinct (Wilkinson & Young, 2002). The implications of this are discussed in the concluding sections of the paper. While they are similar, face validity suggests that these “trustlike” items are more appropriately identified as reflecting information exchange norms and expert panel evaluations during pre-testing consistently identified them as part of an information exchange norms construct. Furthermore, as discussed, informational credibility (i.e. competence-related) items are not central to the trust scale. The final norms measure comprised five items reflecting a single dimension with good reliability (alpha = 0.82). Confirmatory factor analysis provided further verification of the theoretical structure of the information exchange norm construct (χ2 (5) = 135, p = 0.93; GFI = 0.99 TLI = 1.00 CFI = 1.00, composite reliability = 0.834 and the variance extracted = 0.506). See Appendix A for the set of items used. Nature of the relationship — A number of items describing the nature of the ties between information exchange partners were included in the research instrument to test differences in information exchange and trust in different contexts. The findings of Young and Wilkinson (1997) highlight some possible differences of buyers and sellers in relationship development. Easton and Araujo (1992) suggest that there should be no compelling differences in information exchange and relationship atmosphere in trading versus non-trading relationships however there has been no research done addressing this. Single-item measures were used to distinguish between trading and non-trading relationships, and the nature of the ties between firms. 6. Results The achieved sample (n = 355) comprised both trading, i.e. buyer–seller relationships (75%) and non-trading relationships (25%). Rather more trading relationships were included in the sample than planned (75% compared to planned 50%) as informants were less able to identify a non-trading relationship in which information exchange occurred. Many of the nontrading relations included those between competitors (n = 32), they also included relationships with industry organizations, unions, government departments and so forth (n = 37). Some of these had contractual or ownership ties but did not trade (i.e.

976

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

Table 1 ANOVA trust and information exchange dimensions, excluding norms (dependent variable = trust, main effects only) Source

F

Sig.

Parameter estimate (beta)

Corrected model 13.316 0.000 Intercept 293.707 0.000 4.003 Relational motivations 11.775 0.001 0.139 Coercive communication 22.413 0.000 −0.185 behavior Information exchange 4.094 0.044 −0.075 complexity Information exchange 21.887 0.000 0.151 intensity Transactional information 6.122 0.014 0.134

T (Sig.)

17.138 (0.000) 3.431 (0.001) − 4.734 (0.000) − 2.023 (0.044) 4.678 (0.000) 2.474 (0.014)

R-squared = 0.177 (adjusted R-squared = 0.164).

same parent company, subsidiary–parent, franchisee–franchisor) (n = 19). Informant organizations were involved a range of industries including distribution (retail/wholesale trade n = 107, transport and storage n = 23), and manufacturing and production (n = 78). A large group were engaged in services (finance, communication and business services n = 71, hospitality and recreation services n = 26, government and other community services n = 43). The average length of the information exchange relationship was 7½ years. The longest relationship had continued for over 80 years; however 75% of the sample focused on relationships lasting 10 years or less. Throughout the paper we have argued that trust and information exchange are dependant on each other. Yet in considering the relationship between trust and information exchange most variance-based analytical techniques requires that one variable be identified as dependant. Here trust has been selected as the dependent measure. In doing so, it is possible to consider the relatively under-researched role of information exchange in greater depth. Furthermore, many commentators highlight the difficulty in engineering trust (Wilkinson & Young, 2005); however it may be possible to shape the information exchange that occurs in relationships. So by using information exchange to “explain” trust the normative implications of this work are clearer. Analysis of variance was used to investigate the connections between the various dimensions of information exchange and trust. As noted trust has been set as the dependant measure and all information exchange variables except information exchange norms were included in the initial model. Only the main effects were included (interaction terms were omitted). Information exchange dimensions which did not significantly impact on trust were deleted. The final model showed that different dimensions of information exchange had significant, varied and particular impacts on trust. These are set out in the following table. As shown in Table 1, relational motivations, information exchange intensity and the exchange of transactional information have positive effects on trust, while the use of coercive communication behaviors (transformational activities) and the complexity of the information exchange process impact negatively on trust. Together these variables explain approxi-

mately 17% of the variation in trust. (As this is correlation data, the reverse relationship holds, that is trust having only a small effect upon information exchange components, though in such a model we would expect the Betas to be different although in the same direction and of similar magnitude.) Scrutiny of the parameter estimates shows that with the exception of information exchange complexity the magnitude of the effects are similar. These findings are in line with existing theory (discussed earlier) and plausible explanations exist for the reverse relationships. For example, it is likely that high levels of trust in relationships inhibit coercive behaviors. What is interesting is that many of the other “parts” of information exchange are not significantly related to trust. Nor were there any apparent differences between firms engaged in trading and non-trading relationships. No differences were identified when considering the nature of the relational ties (i.e. buyer/seller, competitor etc.). We speculate that the presence of information exchange alone is sufficient to create a trusting relationship and that the particular nature of the relationship has little to no bearing on the process once established (we note that most relationships in our sample had been established for some time). Such a view finds support in literature dealing with non-trading relationships (Easton & Araujo, 1992). We speculate further regarding the impact the type of relationship has on trust (and information exchange) later in this section. Next the role of information exchange norms was considered. The initial analysis was repeated including all information exchange variables as well as information exchange norms. Only variables with significant main effects on trust were retained in the final model. As shown in Table 2, a very different picture emerges when information exchange norms are included. Apart from the use of coercive communication behaviors none of the other information exchange dimensions were retained in the final model. Together, information exchange norms and the use of coercive communication behaviors (transformational activities) explain over 50% of the variation in trust. This means that a change in information exchange norms will provide a relatively large change in trust (and as argued previously the reverse relationship, that is, a change in norms providing a relatively large in trust is to be expected). The use of coercive information strategies has a negative and much smaller impact on trust. No differences were observed between firms engaged in trading and non-trading relationships, nor were there any effects Table 2 ANOVA: trust and information exchange dimensions, including norms (dependent variable = trust, main effects only) Source

F

Sig.

Parameter estimate (beta)

Corrected model 216.412 0.000 Intercept 64.976 0.000 1.585 Information exchange 390.395 0.000 0.779 norms Coercive communication 17.347 0.000 − 0.112 behavior R-squared = 0.554 (adjusted R-squared = 0.552).

T (Sig.)

8.061 (0.000) 19.758 (0.000) −4.165 (0.000)

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982 Table 3 ANOVA: norms and information exchange dimensions, (dependent variable = information exchange norms, main effects only) Source

F

Sig.

Parameter estimate (beta)

Corrected model 17.714 0.000 Intercept 237.825 0.000 3.240 Relational motivations 11.333 0.001 0.122 Coercive communication 7.150 0.008 −0.090 behavior Information exchange 37.541 0.000 0.171 intensity Conceptual motivations 10.714 0.001 0.122 Information exchange 5.895 0.016 −0.064 variability

T (Sig.)

15.422 (0.000) 3.367 (0.001) − 2.674 (0.008) 6.127 (0.000) 3.273 (0.001) − 2.428 (0.016)

R-squared = 0.207 (adjusted R-squared = 0.195).

associated with the nature of the relational ties. This analysis suggests that very few of the dimensions of information exchange appear to have direct effects on trust. Given the pivotal role of trust in interfirm relationships (Morgan & Hunt, 1994; Young & Wilkinson, 1988), the apparently limited direct impact of information exchange on its development raises questions about the extent to which relationships and the trust within them can be actively managed, although this has been proposed (Mohr et al., 1996; Sheng et al., 2006). The work of Heide and John (1992) suggests that information exchange norms might mediate the impact of other dimensions of information exchange on trust, thus explaining the limited direct impacts found in respect to other information exchange dimensions. This was investigated by examining the main effects of information exchange on information exchange norms, and is shown in the following table. As shown in Table 3, the analysis of variance was significant, with aspects of information exchange explaining approximately 20% of the variation in information exchange norms. (As already considered the norms explain aspects of information exchange in a similar way.) Exchange intensity and coercive communication behavior are important predictors of information exchange norms and other information exchange variables also have an effect including the positive impact of motivations for information exchange (in particular, conceptual and relational motives), the negative impact of information exchange activity and the use of coercive communication behavior. When considering the impact the nature of the interfirm relationship has on the formation of relational norms an interesting effect emerges. Information intensity has a positive impact on norm formation (β = 0.273, t = 5.348, p = 0.000)2, and this effect is strongest in non-trading relationships. In trading relationships we see the impact of intensity on exchange norms dampened and

2 This was investigated by considering the interaction effects between each of the information exchange dimensions with the nature of the relationship (trading and non-trading; buyers and sellers and so forth). This parameter estimate varies from Table 3 because here the main and two-way interaction effects are considered.

977

more so for sellers (β = − 0.176, t = − 2.664, p = 0.008) than buyers (β = − 0.126, t = − 1.850, p = 0.065). These findings indicate that the nature of the relationship (i.e. trading or non-trading) and position of the informant, (i.e. buyer and seller) may play an unanticipated role. To investigate more subtle impacts of role we repeated the preceding analysis (Table 1) — this time including interaction effects between the nature of the relationship and each of the other parts of the information exchange process. The overall model was significant (F = 7.97, df = 10, p b 0.001) and there was some evidence that interactions between the nature of the relationship and other aspects of the information exchange process had an impact on the development of trust in the relationship. It appears that the information exchange processes for trading and non-trading relationships may be different — with these differences being captured in variations in other aspects of information exchange (in for example the intensity of information exchange (transfer activity) and the type of information that is exchanged). This would explain the absence of direct effects associated with the nature of the relationship in the analysis undertaken. Comparisons of means (ANOVA) confirms that firms engaged in trading relationships have substantially different information exchange structures than organizations involved in non-trading relationships. Not surprisingly trading relationships exchange more transactional information (X¯ = 1.938 c.f. X¯ = 1.481, p b 0.001) exchange is more intense (X¯ = 4.28 c.f. X¯ = 3.71, p = 0.001), and information exchange norms are stronger (X¯ = 4.530 c.f. X¯ = 4.328, p = 0.031), they also tend to make greater use of coercive communication behaviors (X¯ = 2.46 c.f. X¯ = 1.95, p b 0.001). The nature of non-trading relationships is not well documented in the literature; however recent work establishes that information exchange resources and activities have significant effects on trust in these relationships (Denize & Young, 2006). Here we find that there may be differences in how this emerges for trading and non-trading relationships and we contend that this should be considered in future research. 7. Implications and conclusions 7.1. Research implications This study is intentionally exploratory in that it uses the ANOVA tests to consider the broad pattern of connections (including multi-directionality) between aspects of information exchange, norms and trust, and is speculative in that it considers in terms of complex adaptive systems the relationships in which trust and information exchange evolve. The substance of these findings leads us to the view that exchange relationships are more than the simple sum of their transactions, they emerge in a bottom-up way and are complex (i.e. their components are connected, they display properties that are more than the sum of their parts and small incremental changes can invoke sudden, unexpected consequences) (Holland, 1995). We see this in particular with respect to the importance, indeed dominance of long term-oriented information exchange norms' association with trust. This complexity makes it difficult to predict the way

978

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

aspects of information exchange will combine and the impact of information exchange on trust and of trust on information exchange. When considered holistically, these findings indicate that few, if any, of the “traditional” ways in which information exchange is conceptualized (i.e. as a transfer activity) are directly associated with trust, but that the properties of the transfer activity are likely to be linked to information exchange norms and it is these that are more directly associated with trust (and vice versa). While the importance of the development of norms is in line with our expectations, we had not anticipated the lack of association we would see between other aspects of information exchange and trust. The importance of development of norms again highlights the limitations of the standard tools for relationship study. While the study presented has attempted to take account of the emergent processes of relationships in the way instrument and scales are designed and more importantly the way the results are interpreted, it is necessarily limited in its capabilities. Authors who consider methods of studying complex adaptive systems (Abbott, 2001; Easton, 2002; Van de Ven & Poole, 2005; Wilkinson & Young, 2005; Wilkinson et al., 2007) argue that cross-sectional survey data is of limited use and that alternatives such as case studies, narrative event sequence analysis and/or simulation should be used. We agree that there are substantive limitations associated with surveys, whether cross-sectional or longitudinal (which are really comparative static), are used. However we would argue that the reality of research in marketing is that survey use, including in the context of the study of business relationships will continue. There is merit in having “snapshots” of the market for comparison and generalization — as long as we recognize that they are not descriptive of the process/evolution of relationships and we do not limit ourselves to these approaches. Also, these survey snapshots can play a role within case studies. A challenge for researchers is to develop methods of survey research design and data analysis that are more sympathetic to a complex adaptive system stance, enable us to retain awareness of relational process and are intended to contribute to wider, multi-method designs. At present, we contend that the opposite is happening. Ever-increasing use of structural equation modeling as the means of understanding what is driving business relationships provides explanations of simple causality but actively obscures exploration of relational process. The fragility of variance–covariance-based methods to violations of assumptions and the impact of multicollinarity on path estimates are further limitations (Wilkinson et al., 2007). Most concerning we believe is that the use of these analytical tools is leading to “the tail wagging the dog”. Researchers are now likely to reject consideration of views of the world that can not be tested with path models and to believe that these models provide comprehensive depictions of business relationship reality where in fact they provide only a very limited and often inaccurate view. The scale of the work and the need for self-validation by researchers leads to self-perpetuation and research agendas and dangerous paradigms that impede our understanding of relationships and their complex, emergent processes.

7.2. Managerial implications No doubt in part researchers in business marketing avoid considering relationships as complex adaptive systems because of the uncomfortable management implications that such a perspective provides (Wilkinson & Young, 2002). The result reported here is no exception. We move beyond the narrow conceptualizations and resulting operationalizations of information exchange and in doing so find that most of the information exchange dimensions that have any substantive effect on trust are not really controllable by managers. The aspect that has the greatest effect on the level of trust is information exchange norms (positive impact) and these norms emerge as part of the long term co-production of the relationship itself rather than from any managerial action that can “manufacture” trust in the short term. Even in the long term neither party “controls” their development. The impact of most information exchange behaviors instead seems to be in its effect on the long term growth of these norms. The next most influential aspect of information exchange is coercive communication behaviors (transformational activities), which have a negative impact on trust. Here, there is also limited managerial value as managers presumably do not want, in the main, to reduce trust. Despite this, strategies for managers can be identified by considering the evolutionary nature of trust and information exchange. For example if trust is increased (and marketing literature suggests options for how to do this) then the propensity for coercive communication is reduced and this should have impact on relational evolution. These findings provide means to assist in shifting managers' thinking about managing relationships within complex selforganizing systems. Wilkinson and Young (2002, 2005) have argued that the marketing discipline must move from seeing management of marketing functions as firm-centric and centering on leadership and control towards a view of management as including participation, response and “followship” within complex adaptive systems of relationships and networks. Results such as these support such a view, showing that managers cannot control relationships but can participate in their development. Understanding the drivers of a need for a longer term relational perspective may help managers to understand and be comfortable with various anxieties attendant with their recognition of limited control (Stacey, 1996). However this does not undermine the use of traditional management analysis and planning techniques because firms are still required to act and respond in intentionally rational ways in the network if the self-organizing process is to work effectively over time (Wilkinson & Young, 2002). Points of action for managers are suggested by these results including decisions on choosing and getting chosen to participate information exchange relationships (or not), sharing information (or not), distorting or withholding information (or not) and the extent to which and the way in which these actions are used to influence of others. More generally, it is by these actions that managers increase or decrease the likelihood of their survival in relationships. As leading complexity theorist Stewart Kauffman (1996) puts it, “the winning games are the games winners play”

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

(p. 79). These finding indicate ways forward and actions to avoid in building network and relationship competence to facilitate survival (Ritter, Wilkinson, & Johnston, 2004).

979

exchange in terms of the rich interconnections of actors, activities and resources and their links, ties and bonds. Acknowledgement

8. Conclusions We conclude this paper by reflecting on our orientation to understanding relationships and its limitations. Surveys in isolation are not ideal for this, and yet we use a survey. We have focused on only information exchange and trust. It could be argued that this gives an incomplete picture of the relationship and its processes. We would agree that comprehensive, holistic pictures of these processes are invaluable, but perhaps are more suited to case study methods. The purpose here is different. We take a close-up view of the part of the process that we argue connects the central aspects of relationship functioning. So, have we selected the correct parts of the relationship to view? We have argued for the importance of communication — of which we see information exchange as being an integral part. We see communication as at the core of the process of continually reproducing the functioning of a business relationship. The choice of trust as our dependent variable in our analysis is more problematic. There are two prongs to this issue. First, should it be trust or information exchange that is the dependent variable? Our discussion and models throughout highlight the need to consider neither/both as dependent. Their causality is inextricably linked and reciprocal. The findings reported here do not posit causality and the relationships reported should be seen as two-way. Higher trust will lead to enhanced norms as well as the reverse and lower trust is likely to result from and lead to increased coercive communication behavior. Second, is there a better representation of the relationship's state/wellbeing than trust? Numerous authors have argued that cooperation, relational commitment, satisfaction and/or performance are at the heart of what we seek to understand and that trust is useful only in so far as it contributes to our understanding of one or more of these (see Young, 2006 for a review). However this is a side issue. Trust has been shown to play an important role in many relationship states. It is for this reason that we focus upon it here. The empirical context of this paper is the interfirm relationship and our discussions of relationship development and evolution, trust and communication have considered this. Future work will extend our considerations beyond the individual dyad to the network. This is a particularly exciting perspective; communication builds connectedness and that is truly the means by which the network is opened (by which we mean is accessible) to the individuals, firms and collections of firms. Future work also will be geared towards expanding our understanding of relationship processes — using the techniques of survey methods but grounding these into the cases that we hope to build that will provide a more detailed view of the complex interplay of information exchange aspects, the more precise ways in which the act to sustain and change relationships and the underlying reasons for the processes we observe. We will do this by considering communication/information

The authors would like to thank Professor Ian Wilkinson for his substantial contributions to the way we think about relationships and their evolution and the way we must therefore research them. Appendix A. Operationalization (full details are available from the corresponding author) Trust (4-item reflective measure, 1 = disagree strongly, 6 = agree strongly) Parties to this relationship… … behave in a trustworthy manner toward each other … provide a completely truthful picture when negotiating with each other … trust each other … have confidence in the fairness and honesty of each other Properties of transfer activities (single-item measures) Directionality/reciprocity (In this relationship 1 = your firm provides most information, 4 = about the same, 7 = Firm X provides most information); Complexity (The information exchange process is 1 = complex, 6 = straightforward); Variability (The information exchange process 1 = changes infrequently, 6 = changes frequently); Intensity (The information exchange process is characterized by 1 = frequent exchanges, 6 = infrequent exchanges, 1 = regular exchanges, 6 = irregular exchange (mean of two items) Properties of information (single-item measures) Complexity (The information exchanged is 1 = complex, 6 = simple); Variability (The information exchanged 1 = changes infrequently, 6 = changes frequently); Scope (We exchange 1 = many different types of information, 6 = few different types of information); Intensity (We exchange 1 = small amount of information, 6 = large amount of information) Type of information (multi-item formative measures) Transactional information (count of transactional information content exchanged, example types include: information about products, prices, promotion or distribution, transaction related information, electronic funds transfer); Operational information (count of operational information content exchanged, example types include: include inventory and stock details, credit information, statutory information); Strategic information (count of strategic information content exchanged, example types include: information about customers, information about suppliers or distributors, information about competitors, marketing strategy, aggregate sales data, market share information) (These dimensions were created by taking the sum of the relevant items. Higher scores on the summary measure indicated more widely based exchange of this type of information. The minimum and maximum values for transactional information are 0 and 3 respectively. Similarly, the values for operational information are 0 and 3, and for strategic information are 0 and 6.)

980

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

Properties of media (multi-item formative measures) Personalized media (count of use of particular media, example types include: use of phone, face-to-face conversation); Complex media (count of use of particular media, example types include: fax, written documents, report/document, EDI/computer-mediated, intermediary) (These dimensions were created by taking the sum of the relevant items. Higher scores on the summary measure indicated greater personalization/complexity. The minimum and maximum values for personalized media are 0 and 2 respectively. Similarly, the values for complex media are 0 and 5.) Motivations for information exchange Instrumental motivations (3-item reflective measure; 1 = disagree strongly, 6 = agree strongly) (r = reverse coded) My firm would make the same decisions with or without the information we share with FIRM X (r). The information we share with FIRM X materially influences the decision-making of my firm. My firm shares information with FIRM X in order to improve our decision-making. Conceptual motivations (3-item reflective measure; 1 = disagree strongly, 6 = agree strongly) Sharing information with FIRM X enhances my firm's general knowledge rather than facilitating specific decisions. The information my firm shares with FIRM X provides my firm new insights. My firm's general information resources are increased by sharing information with FIRM X. Symbolic relational motivations (4-item reflective measure; 1 = disagree strongly, 6 = agree strongly) My firm shares information with FIRM X to build commitment with FIRM X. Information is shared with FIRM X to enhance our general relationship with FIRM X. My firm shares information with FIRM X as a way of maintaining a pattern of communication with FIRM X. Sometimes the process of sharing information with FIRM X is more important than the information itself. Symbolic forced motivations (2-item reflective measure; 1 = disagree strongly, 6 = agree strongly) Information is shared with FIRM X to satisfy legal or professional requirements. My firm shares information with FIRM X because of the dictates of management or company policy. Transformational activities Cooperative communication behavior (7-item formative measure; 1 = very unlikely to use, 6 = very likely to use) Simply state a request or suggest, e.g. “we would like you to do this” Promise rewards for compliance, e.g., “if you do this we will …” Appeal to common business practice, e.g. “it is normal for you to do this” Discuss the matter, e.g. “let us talk this over” Offer a compromise, e.g. “let us meet half way” Appeal to the relationship, e.g. “do this, we've been doing business a long time …”

Use their experience or information, “if you do this you will benefit because ….” Coercive communication behavior (3-item formative measure; 1 = very unlikely to use, 6 = very likely to use) Demand compliance with requests, e.g. “we insist you do this” Threaten to punish non-compliance, e.g. “if you don't do this we will …” Appeal to the contract, e.g. “the contract requires you do this” Information exchange norms (5-item reflective measure, 1 = disagree strongly, 6 = agree strongly): Parties to this relationship… … rely on each other to be informed … keep the other informed about the things they ought to know … are informed about all useful information by the other … have confidence in the accuracy of information from each other … search for solutions to any joint problems they might have. Nature of the relationship (single-item measures) Trading/non-trading; Type of relationship: (buyer–seller, subsidiary–parent, franchisee–franchisor, joint venture/partnership, competitor, coded as dummy variables for analysis); Relationship duration: years. References Abbott, A. (2001). Time matters: On theory and method. Chicago, IL: University of Chicago Press. Anderson, J. C., & Narus, J. A. (1990). A model of distributor firm and manufacturer firm working partnerships. Journal of Marketing, 54(1), 42−58. Andrich, D. A. (2002). Understanding Rasch measurement: Understanding resistance to the data-model relationship in Rasch's paradigm: A reflection for the next generation. Journal of Applied Measurement, 3(3), 325−359. Argyle, M. (1991). Cooperation: The basis of sociability. London: Routledge. Arndt, J. (1985). The anthropology of interorganisational networks in marketing. Scandinavian Journal of Management Studies(February), 163−180. Asch, S. E. (1952). Social psychology. New York: Prentise-Hall. Axelsson, B., & Easton, G. (Eds.). (1992). Industrial networks: A new view of reality. London: Routledge. Bagozzi, R. P., Gopinath, M., & Nyer, P. U. (1999). The role of emotions in marketing. Journal of the Academy of Marketing Science, 27, 184−206. Barber, B. (1983). The logic and limits of trust. New Brunswick, New Jersey: Rutgers University Press. Bhattacharya, C. B., & Bolton, R. N. (2000). Relationship marketing in mass markets. In J. N. Sheth & A. Parvatiyar (Eds.), Handbook of relationship marketing (pp. 327−354). Thousand Oaks, CA: Sage Publications. Blankenburg-Holm, D., Eriksson, K., & Johanson, J. (1996). Business networks and cooperation in international business relationships. Journal of International Business Studies, 27(5), 1033−1053. Brown, J. S., & Duguid, P. (2000). The social life of information Hard Cover edn. Harvard Business School Press. Campbell, A. J. (1997). Buyer–supplier partnerships: Flip sides of the same coin? Journal of Business and Industrial Marketing, 12(2), 417−434. Cannon, J. P., & Perreault, Jr., W. D. (1999). Buyer–seller relationships in business markets. Journal of Marketing Research, 36(November), 439−460. Churchill, Jr., G. A. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(February), 64−73. Constant, D., Sproull, L., & Kiesler, S. (1996). The kindness of strangers: The usefulness of electronic weak ties for technical advice. Organization Science, 7(2), 119−135. Cunningham, M. T., & Homse, E. (1990). Controlling the marketing-purchasing interface: Resource development and organisational implications. In D. Ford

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982 (Ed.), Understanding business markets (pp. 104−123). London, UK: Harcourt Brace & Company. Denize, S. (2004), Information Exchange in Interfirm Relationships, PhD, University of Technology, Sydney. Denize, S., & Young, L. (2006). Non-trading relationships. Australian and New Zealand Marketing Academy (ANZMAC) Conference 2006: Advancing theory, maintaining relevance. Brisbane, Australia: Queensland University of Technology. Deutsch, M. (1958). Trust and suspicion. Journal of Conflict Resolution, 2(4), 265−279. Dixon, D. F., & Wilkinson, I. F. (1986). Toward a theory of channel structure. In L. P. Bucklin, & J. M. Carmen (Eds.), Distribution channels and institutions, Research in Marketing, Vol. 8 (pp. 27­70): JAI Press. Doney, P. M., & Cannon, J. P. (1997). An examination of the nature of trust in buyer–seller relationships. Journal of Marketing, 61(2), 35−51. Doucette, W. R. (1997). Influences on member commitment to group purchasing organizations. Journal of Business Research, 40, 183−189. Dwyer, F. R., Schurr, P. H., & Oh, S. (1987). Developing buyer–seller relationships. Journal of Marketing, 51(April), 11−27. Easton, G. (1992). Industrial networks: A review. In B. Axelsson & G. Easton (Eds.), Industrial networks: A new view of reality. London: Routledge. Easton, G. (2002). Marketing: A critical realist approach. Journal of Business Research, 55(2), 103−109. Easton, G., & Araujo, L. (1992). Non-economic exchange in industrial networks. In B. Axelsson & G. Easton (Eds.), Industrial networks: A new view of reality (pp. 62−84). New York: Routledge. Ford, D. (1980). The development of buyer–seller relationships in industrial markets. European Journal of Marketing, 14(5/6), 339−353. Frazier, G. L., & Rody, R. C. (1991). The use of influence strategies in interfirm relationships in industrial product channels. Journal of Marketing, 55(1), 52−69. Gaski, J. F. (1984). The theory of power and conflict in channels of distribution. Journal of Marketing, 48(Summer), 9−30. Gaski, J. F., & Nevin, J. R. (1985). The differential effects of exercised power and unexercised power sources in a marketing channel. Journal of Marketing Research, 22(May), 130−142. Giffin, K. (1967). The contribution of studies of source credibility to a theory of interpersonal trust and the communication process. Psychological Bulletin, 68, 104−120. Giffin, K., & Patton, B. R. (1976). Fundamentals of interpersonal communication, 2nd edn. New York: Harper & Row. Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360−1380. Gummesson, E. (1993). Broadening and specifying relationship marketing. Australasian Marketing Journal, 2(1), 31−44 (Formerly Asia-Australia Marketing Journal). Gundlach, G. T., & Cadotte, E. R. (1994). Exchange interdependence and interfirm interaction: Research in a simulated channel setting. Journal of Marketing Research, 31(November), 516−532. Håkansson, H. (Ed.). (1982). International marketing and purchasing of industrial goods. Chichester, United Kingdom: John Wiley & Sons Ltd. Håkansson, H., Johanson, J., & Wootz, B. (1990). Influence tactics in buyer–seller processes. In D. Ford (Ed.), Understanding business markets (pp. 27−41). London, UK: Harcourt Brace & Company. Håkansson, H., & Snehota, I. (1995). Developing business relationships in business networks. London, UK: Routledge. Hallén, L., Johanson, J., & Seyed-Mahamed, N. (1990). Relationship strength and stability in international and domestic industrial marketing. In D. Ford (Ed.), Understanding business markets (pp. 178−190). London, UK: Harcourt Brace & Company. Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organizational subunits. Administrative Science Quarterly, 44(1), 82−111. Heide, J. B., & John, G. (1992). Do norms matter in marketing relationships? Journal of Marketing, 56(April), 32−44. Heide, J. B., & Miner, A. S. (1992). The shadow of the future: Effects of anticipated interaction and frequency of contact on buyer–seller cooperation. Academy of Management Journal, 35(2), 265−291.

981

Holland, J. H. (1995). Hidden order: How adaptations build complexity. Reading, MA: Addision-Wesley. Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion: Psychological studies of opinion change. New Haven: Yale University Press. Hunt, S. D., & Nevin, J. R. (1974). Power in a channel of distribution: Sources and consequences. Journal of Marketing Research, 11(May), 186−193. Kauffman, S. (1996). Investigations: The nature of autonomous agents and the worlds they mutually create. Working paper 96-08-072. Santa Fe, NM: Santa Fe Institute. Kavali, S. G., Tzokas, N. X., & Saren, M. J. (1999). Relationship marketing as an ethical approach: Philosophical and managerial considerations. Management Decision, 37(7), 573−581. Kaye, D. (1995). Sources of information, formal and informal. Management Decision, 33(5), 13−15. Kelman, H. C. (1961). Processes of opinion change. Public Opinion Quarterly, 25(1, Spring), 57−78. Kumar, N., Stern, L. W., & Anderson, J. C. (1993). Conducting interorganizational research using key informants. Academy of Management Journal, 36(6), 1633−1651. Loomis, J. L. (1959). Communication, the development of trust and cooperative behaviour. Human Relations, 12, 305−315. Maltz, E., & Kohli, A. K. (1996). Market intelligence dissemination across functional boundaries. Journal of Marketing Research, 33(February), 47−61. Menon, A., & Varadarajan, P. R. (1992). A model of marketing knowledge use within firms. Journal of Marketing, 56(October), 53−71. Mohr, J. J., Fisher, R. J., & Nevin, J. R. (1996). Collaborative communication in interfirm relationships. Journal of Marketing, 60(3), 103−115. Mohr, J., & Nevin, J. R. (1990). Communication strategies in marketing channels: A theoretical perspective. Journal of Marketing, 36−51. Mohr, J. J., & Sohi, R. S. (1995). Communication flows in distribution channels: Impact on assessments of communication quality and satisfaction. Journal of Retailing, 71(4), 393−416. Mohr, J., & Spekman, R. (1994). Characteristics of partnership success: Partnership attributes, communication behaviour, and conflict resolution techniques. Strategic Management Journal, 15, 135−152. Moorman, C., Deshpandé, R., & Zaltman, G. (1993). Factors affecting trust in marketing research relationships. Journal of Marketing, 57(January), 81−101. Moorman, C., Zaltman, G., & Deshpandé, R. (1992). Relationship between providers and users of marketing research: The dynamics of trust within and between organizations. Journal of Marketing Research, 29(August), 314−329. Morgan, R. M., & Hunt, S. D. (1994). The commitment–trust theory of relationship marketing. Journal of Marketing, 58(July), 20−38. Ritter, T., Wilkinson, I. F., & Johnston, W. J. (2004). Managing in complex business networks. Industrial Marketing Management, 33(3), 175−183. Rositter, J. R. (2002). The C-OAR-SE procedure for scale development in marketing. International Journal of Research in Marketing, 19, 305−335. Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35(4), 651−665. Rotter, J. B. (1971). Generalised expectancies for interpersonal trust. American Psychologist, 26, 443−451. Rotter, J. B. (1980). Interpersonal trust, trustworthiness, and gullibility. American Psychologist, 35(1), 1−7. Schlenker, B. R., Helm, B., & Tedeschi, J. T. (1973). The effects of personality and situational variables on behavioural trust. Journal of Personality and Social Psychology, 25, 419−427. Sheng, S., Brown, J. R., Nicholson, C. Y., & Poppo, L. (2006). Do exchange hazards always foster relational governance? An empirical test of the role of communication. International Journal of Research in Marketing, 23(1), 63−77. Stacey, R. D. (1996). Complexity and creativity in organizations. San Fransisco, CA: Berrett-Koehler Publishers. Stern, L. W. (1972). Communication with resellers. Division of research reprint series. Ohio: College of Administrative Science, Ohio State University. Tedeschi, J. T., Schlenker, B. R., & Bonoma, T. V. (1973). Conflict, power and games. Chicago, IL: Aldine Publishing Company. Van de Ven, A. H., & Poole, M. (2005). Alternative approaches to studying organizational change. Organization Studies, 26(9), 1377−1404.

982

S. Denize, L. Young / Industrial Marketing Management 36 (2007) 968–982

Weill, P. (1990a). The relationship between investment in information technology and firm performance in the manufacturing sector. University of New South Wales. Weill, P. (1990b). Strategic investment in information technology: An empirical study. Information Technology, 13(3), 141−147. Weiss, C. H. (1980). Knowledge creep and decision accretion. Knowledge: Creation, Diffusion, Utilization, 1(3), 381−404. Weitz, B. A., & Jap, S. D. (1995). Relationship marketing and distribution channels. Journal of the Academy of Marketing Science, 23(4), 305−320. Welch, C., & Wilkinson, I. F. (2002). Idea logics and network theory in business marketing. Journal of Business to Business Marketing, 8(3), 27−48. Wiley, J., Wilkinson, I. F., & Young, L. (2006). The impact of connected relations on relationship performance: A comparison of European and Chinese international business relations. Journal of Business and Industrial Marketing, 21(1), 3−13. Wilkinson, I. F., & Kipnis, D. (1978). Interfirm use of power. Journal of Applied Psychology, 63(3), 315−320. Wilkinson, I. F., & Young, L. (1994). The space between: The nature and role of interfirm relations in business. Research conference on relationship marketing; Atlanta, Georgia. Wilkinson, I. F., & Young, L. (2002). On cooperating: Firms, relations and networks. Journal of Business Research, 55(2), 123−132. Wilkinson, I. F., & Young, L. (2005). Toward a normative theory of normative marketing theory. Journal of Marketing Theory, 5(4), 363−396. Wilkinson, I. F., Young, L., & Ladley, D. (2007). Group selection versus individual selection and the evolution of cooperation in business networks. Industrial Marketing and Purchasing Conference. Manchester UK. Williamson, O. E. (1993). Calculativeness, trust, and economic organization. Journal of Law and Economics, 36, 453−486. Wilson, D. T., & Mummalaneni, V. (1990). Bonding and commitment in buyer– seller relationships: A preliminary conceptualisation. In D. Ford (Ed.), Understanding business markets (pp. 408−420). London, UK: Harcourt Brace & Company. Young, L. (1992). The role of trust in interorganisational relationships in marketing channels. Unpublished doctoral thesis, University of New South Wales. Young, L. (1996). Are our measures of trust trustworthy? Australia and New Zealand marketing educators conference; Auckland, New Zealand.

Young, L. (2006). Trust: Looking forward and back. Journal of Business and Industrial Marketing, 21(7), 439−445. Young, L., & Albaum, G. (2003). Developing a measure of trust in salesperson/ customer relationships in direct selling. Journal of Personal Selling and Management, 23(3), 253−269. Young, L., & Daniel, K. (2003). Affectual trust in the workplace. International Journal of Human Resource Management, 14(1), 139−155. Young, L., & Wilkinson, I. F. (1988). The role of trust and cooperation in marketing channels: A preliminary study. European Journal of Marketing, 23, 109−122. Young, L., & Wilkinson, I. F. (1989). The role of trust and cooperation in marketing channels: A preliminary study. European Journal of Marketing, 23(2), 109−122. Young, L., & Wilkinson, I. F. (1997). The space between: Towards a typology of interfirm relations business. Journal of Business to Business Marketing, 4 (2), 53−97. Zand, D. E. (1972). Trust and managerial problem solving. Administrative Science Quarterly, 17(2), 229−239.

Sara Denize is Associate Head of School in the School of Marketing at the University of Western Sydney. Dr Denize's research interests centre around information exchange in interfirm relationships and networks and most recently has focused on the use of agent-based modeling to explore innovation networks. Her work has been published in Journal of Business and Industrial Marketing.

Louise Young is Professor of Marketing at the University of Technology Sydney. Her current research focuses on the evolution and management of business relationships and networks and in particular on the psychology of the individuals participating using methods of case study, deep qualitative interview, lexicographic analysis and agent-based modeling. She has published over 60 research papers in a wide range of business and social science journals, including European Journal of Marketing, Industrial Marketing Management, International Journal of Research in Marketing, Journal of Business Research, Journal of Marketing Theory, Journal of Personal Selling and Management among others.