An Experimental Test of Negotiation Strategy Effects on Knowledge Sharing Intentions in Buyer–Supplier Relationships

Authors


Abstract

Negotiations are an essential element of buyer–supplier relationships that form the foundation of modern supply chains. Research has identified two common types of negotiation strategies that are used in buyer–supplier negotiations. A win–win negotiation strategy attempts to maximize mutual gain while a win–lose strategy focuses on obtaining a disproportionate share of benefits. This study investigates the relational costs of adopting a negotiation strategy in terms of adverse effects on knowledge sharing intentions in interdependent buyer–supplier relationships. A between-subjects scenario-based experiment is used to test hypotheses developed from applicable literature and social exchange theory. Results of the experiment indicate that employing a win–lose negotiation strategy may decrease future intentions toward information exchange, communication quality and operational knowledge transfer between buyers and suppliers. The findings inform managerial aspects of supply chain relationship management, extend the negotiation literature to consider longer-term effects of common negotiation strategies and provide insights into social exchange theory.

Introduction

At the most fundamental level, supply chains consist of interdependent buyer–supplier relationships (Cooper, Lambert & Pagh, 1997; Mentzer et al., 2001; Xu & Beamon, 2006). Supply chain members need each other to obtain goods and services to fulfill end-consumer needs (Atkins & Rinehart, 2006; Ramsay, 2004). As part of their efforts to coordinate key supply chain activities, interdependent buyers and suppliers often negotiate essential elements of common costs and services. Agreement on items such as pricing, delivery terms, shipment schedules, carrier selection and quality standards is negotiated between supply chain members. During such negotiations, buyers and suppliers often employ different negotiation strategies.

Two of the most common negotiation strategies include a collaborative win–win approach and a competitive win–lose tactic (Krause, Terpend & Petersen, 2006). To foster the development of longer-term interdependent buyer–supplier relationships, win–win negotiation strategies are often advocated (Zachariassen, 2008). However, evidence suggests that win–lose negotiation strategies often outperform win–win approaches by obtaining a larger share of economic benefits (Graham, Mintu & Rodgers, 1994). As these examples demonstrate, both types of negotiation strategies are often utilized by buyers and suppliers. Regardless of the strategic approach, negotiation is an essential aspect of supply chain management. Specifically, win–win and win–lose negotiation strategies are common elements of interdependent buyer–supplier relationships in modern supply chains. Unfortunately, little is known about the effects of negotiation strategies on future collaborative behaviors in exchange relationships (Atkins & Rinehart, 2006).

Research in negotiation often focuses on the economic outcome of profit or the psychological outcome of satisfaction within a specific buyer–supplier interaction (Mintu-Wimsatt & Graham, 2004). Although such contributions are meaningful and important to negotiation research, these focused outcomes limit the scope of inquiry in exchange relationships by treating negotiation interactions as discrete events rather than part of an ongoing relationship (Dwyer, Schurr & Oh, 1987). Discrete events like individual negotiations are the foundation of critical buyer–supplier relationships that form modern supply chains (Daugherty, 2011). Unfortunately, most negotiation research views buyer–supplier negotiations as isolated incidents instead of a continuing contribution to the overall supply chain relationship. However, insights from social exchange theory (SET) suggest that negotiations are in fact a critical part of ongoing exchange relationships that will influence future intentions and interactions between buyers and suppliers (Emerson, 1976; Thibaut & Kelley, 1959). Therefore, to increase understanding of buyer–supplier relationships, it is important to learn how a discrete negotiation event can impact longer-term collaborative behaviors in exchange relationships (Atkins & Rinehart, 2006).

In spite of the ubiquitous nature of negotiations between supply chain members, recent research has largely ignored the importance of understanding negotiation effects in modern supply chains. Therefore, the purpose of this research is to address this gap in the literature by empirically testing the impact of negotiation strategies on future intentions to engage in collaborative knowledge sharing behaviors in interdependent buyer–supplier relationships. Independent variables in the study include the type of negotiation strategy and the level of relational interdependence. Independent variable effects will be assessed on the dependent variables of information exchange, communication quality and operational knowledge transfer.

This study makes an academic contribution by utilizing SET to extend the scope of negotiation research beyond limited outcomes of discrete events to include intended collaborative knowledge sharing behaviors associated with ongoing interdependent buyer–supplier relationships. Given this new perspective on negotiation research, this study seeks to maximize internal validity and clearly identify causal relationships. Therefore, a scenario-based experimental design was deemed appropriate for the research goals and also answers the calls for additional behavioral experiments in supply chain research (Boyer & Swink, 2008; Eckerd & Bendoly, 2011; Knemeyer & Naylor, 2011; Thomas, 2011; Tokar, 2010; Waller & Fawcett, 2011). From a managerial perspective, supply chain members may gain insights into additional relational costs and benefits associated with common negotiation strategies. Such information will help managers more effectively assess the cost/benefit trade-offs in supply chain relationship management.

Literature Review

Negotiation Strategy

Negotiation is often viewed as a single interaction between two parties where exchange conditions are determined (Mintu-Wimsatt & Calantone, 1996). In contemporary supply chains, negotiations commonly determine the details of product and service exchanges between buyers and suppliers (Atkins & Rinehart, 2006; Fang, 2006). Although the practice has always been an important aspect of business, the current economic environment has placed increased emphasis on successful negotiations as performance pressures continue to increase (Herbst, Voeth & Meister, 2011). Therefore, many firms develop negotiation strategies designed to obtain desirable results. Research on negotiation strategies looks at the interaction patterns parties use to reach an acceptable agreement (Ganesan, 1993). The cumulative evidence in the literature suggests that two common types of negotiation strategy are win–win and win–lose approaches.

A win–win negotiation strategy has been labeled in the literature by a number of terms such as integrative, cooperative, problem-solving approach and collaborative (Campbell, Graham, Jolibert & Meissner, 1988; Krause et al., 2006). The goal of a win–win strategy is to reconcile divergent interests between a buyer and supplier to provide both parties with joint benefits as an outcome of the specific negotiation (Zachariassen, 2008). Win–win negotiators seek information from their counterparts to consider the needs and preferences of all parties involved in the negotiation rather than attempting to exploit differences (Mintu-Wimsatt & Graham, 2004). For long-term relationships, use of a win–win strategy has been advocated to improve overall supply chain performance, but such claims often lack empirical evidence (Graham et al., 1994; Zachariassen, 2008).

A win–lose negotiation strategy is defined as the attempt to resolve conflicts through the implicit or explicit use of threats, persuasive arguments and punishments (Ganesan, 1993). The implicit assumption and justification for this strategy is the desired outcome of winning at all costs without consideration for the other party (Calhoun & Smith, 1999). Characteristics of a win–lose negotiation strategy include behaviors that are competitive, individualistic, aggressive and persuasive (Mintu-Wimsatt & Graham, 2004). Threats and warnings may be used by a negotiator to pursue goals that conflict with their negotiation partner's goals (Mintu-Wimsatt & Gassenheimer, 2000). Win–lose negotiators often force an opposing party to make concessions to achieve more profitable outcomes.

Win–win and win–lose negotiation strategies are contrasting approaches to determining exchange conditions. The literature suggests these negotiation strategies have several common characteristics and behavioral elements. However, these aspects of negotiation are usually diametrically opposed. For instance, win–win negotiations emphasize high degrees of information exchange, but win–lose negotiations are characterized by limited information exchange (Campbell et al., 1988; Herbst et al., 2011; Ramsay, 2004). Win–win negotiators communicate clearly, but win–lose negotiators often communicate deceptively (Graham et al., 1994; Krause et al., 2006). Win–win approaches emphasize mutual gain, but win–lose approaches focus on individual goals (Ramsay, 2004; Rubin & Brown, 1975). Win–win bargaining is open to making concessions, but win–lose bargaining is not willing to concede (Mintu-Wimsatt & Graham, 2004; Neu, Graham & Gilly, 1988). Win–win tactics are cooperative and do not rely on aggressive threats, but win–lose tactics often seek to use intimidation to force concessions (Ganesan, 1993; Herbst et al., 2011; Krause et al., 2006). As these examples demonstrate, negotiation styles differ on several important dimensions.

The type of negotiation strategy impacts the outcome of a negotiation (Krause et al., 2006; Rinehart, Eckert, Handfield, Page and Atkins 2004). Negotiation research has largely utilized negotiation outcomes as dependent variables. The most common dependent variable studied in negotiations is the economic outcome of profit that is viewed as an objective result of the buyer–supplier interaction (Mintu-Wimsatt & Graham, 2004). Given the game theory origins that treat a negotiation as an individual encounter, an economic variable limited to a specific negotiation is reasonable. Profit is the most commonly measured monetary outcome in negotiation strategy and has been measured both individually and jointly (Graham et al., 1994; Neu et al., 1988). However, such measures are limited to a specific individual negotiation rather than an ongoing buyer–supplier relationship.

The second most common dependent variable in negotiation strategy research is the social-psychological outcome of satisfaction. Satisfaction represents a subjective assessment of the negotiation encounter, and it is often measured from the buyer's perspective (Graham et al., 1994). When used in negotiation strategy research, satisfaction assessments measure how a negotiator was mentally and emotionally impacted by the strategy of their bargaining counterpart. As an affective measure, satisfaction determines how managers “feel” about the outcome of a negotiation. However, similar to the economic outcome of profit, current satisfaction measures fail to capture the ongoing nature of an interdependent buyer–supplier relationship, the likelihood that the negotiators will meet again in a future encounter, or how past negotiations impact future relational exchanges and intentions in supply chains.

Negotiations take place within the context of an existing buyer–supplier relationship, and impactful research needs to encompass this expanded view instead of treating negotiation exchanges as isolated discrete events (Daugherty, 2011; Dwyer et al., 1987; Gelfand, Major, Raver, Nishii & O'Brien, 2006). Understanding the impact of negotiation strategy choice on collaborative intentions in buyer–supplier relationships is important for managers and academics alike as many critical components in supply chains are negotiated.

Interdependence

Dependence is an important part of understanding supply chains because, to some degree, all members of a supply chain depend on each other (Stern, El-Ansary & Coughlin, 2001). This type of mutual dependence between buyers and suppliers is described as interdependence (Pfeffer & Salancik, 1978). The concept of interdependence suggests that buyers and suppliers must take each other into account to achieve their individual goals (Gundlach & Cadotte, 1994). Although interdependence can be conceptualized with varying degrees of symmetry or asymmetry, in this study the focus is on high and low levels of mutual symmetric interdependence. Relationships with mutual symmetric interdependence have an equal amount of dependence between a buyer and supplier (Kumar, Scheer & Steenkamp, 1995). In these kinds of interdependent buyer–supplier relationships, neither party has more power over the other. Mutually interdependent buyer–supplier relationships are used in this study to control for the known effects of power in negotiations and instead permit the rigorous investigation of negotiation strategy effects in supply chain relationships. Likewise, interdependent buyer–supplier relationships allow for the development of collaborative behaviors that are dependent variables of interest in this study.

Knowledge Sharing

Sharing valuable knowledge and essential information is at the heart of the supply chain concept (Thomas, Esper & Stank, 2010). Information is one of the primary flows in a supply chain, and it is a critical driver of efficiency, effectiveness and overall supply chain performance (Mentzer et al., 2001; Thomas, Fugate & Koukova, 2011). To coordinate key supply chain activities and ultimately meet end-consumer expectations, buyers and suppliers must share key information (Kannan & Tan, 2002; Kwon & Suh, 2004; Terpend, Tyler, Krause & Handfield, 2008). Given the importance of this type of information flow in supply chain relationships, knowledge sharing–oriented constructs are particularly relevant to buyer–supplier relationship research.

The knowledge sharing–dependent variables used in this study are information exchange, communication quality and operational knowledge transfer. Information exchange is defined as the expectation that supply chain members will provide basic information to each other (Lusch & Brown, 1996). Communication quality is defined as the completeness, credibility, accuracy, timeliness and adequacy of communication flows between supply chain members (Mohr & Sohi, 1995). Operational knowledge transfer is defined as the transfer of knowledge or know-how between supply chain members (Fugate, Stank & Mentzer, 2009; Modi & Mabert, 2007). These three constructs were selected as dependent variables because they tap into the domain of key collaborative knowledge sharing behaviors that are essential to high-performing supply chain relationships.

Theoretical Foundation

Social exchange theory serves as the theoretical foundation that informs this study and enables a priori hypothesis development. SET proposes that individuals or corporate groups attempt to obtain desirable results in social interactions by maximizing rewards and minimizing costs (McDonald, 1981; Thibaut & Kelley, 1959). This subjective cost/benefit approach is driven by the basic motivation to gain profitable outcomes in social interactions (Emerson, 1976). Therefore, relational behaviors are determined by rewards of interaction minus the costs of interaction (Griffith, Harvey & Lusch, 2006). This comparison of relational costs and benefits suggests that corporate group behaviors are driven by a quasi-economic mode of analysis (Emerson, 1976). Within SET, the reciprocity principle also suggests that actors will reciprocate the actions of others in exchange relationships (Gouldner, 1960). Therefore, people tend to respond to each other in similar fashions and mirror the actions of others in exchange relationships. Reciprocity evokes obligations to others based on past behavioral interactions. In a positive sense, reciprocity can be described as the mutually contingent or gratifying exchange of goods, services or benefits. In a negative sense, reciprocity can include sentiments of retaliation where emphasis is placed on the return of injuries rather than benefits.

Social exchange theory was selected as an appropriate theoretical lens in this study for several reasons. First, given the social interactions involved in both negotiations and exchange relationships, our research topic is included within the theoretical scope of SET. Second, SET has been utilized in previous negotiation research (Bottom, Holloway, Miller, Mislin & Whitford, 2006; Buchan, Croson & Johnson, 2004; Wolfe & McGinn, 2005). Third, buyer–supplier relationship research is often informed by SET (Cahill, Goldsby, Knemeyer & Wallenburg, 2010; Narasimhan, Nair, Griffith, Arlbju & Bendoly, 2009; Wagner, Coley & Lindemann, 2011). For these reasons, SET was selected as the appropriate theory to study negotiation strategy effects in buyer–supplier relationships.

Hypotheses

In highly interdependent relationships, buyers and suppliers need each other to achieve their individual and collective goals. Interdependent relationships often have a longer-term orientation and lack opportunistic behaviors because of the high costs of destroying the relationships (Kumar et al., 1995). Conversely, with low levels of interdependence, opportunistic behaviors are more likely between supply chain members because switching costs are much lower. The reciprocity principle of SET suggests firms will mirror the actions of others in exchange relationships. In a positive sense, reciprocity predicts that positive actions from one firm will elicit favorable responses from another firm such that buyers and suppliers with high levels of mutual interdependence will likely engage in future collaborative behaviors that increase levels of knowledge sharing due to the symbiotic nature of the relationship. In a negative sense, the reciprocity principle suggests that negative actions will lead to retaliatory sentiments. In buyer–supplier relationships with low levels of mutual interdependence, opportunistic behaviors are more likely (Jambulingam, Kathuria & Nevin, 2011). Such opportunistic behaviors deviate from expectations of mutual benefit and encourage punitive actions that could jeopardize future information flows between buyers and suppliers. Therefore, the following hypotheses are presented based on the reciprocity principle of SET:

H1: An increase in levels of interdependence leads to an increase in intended (a) information exchange, (b) communication quality and (c) operational knowledge transfer activities between buyers and suppliers.

Prior research has advocated the use of win–lose negotiation strategies in transactional relationships characterized by low levels of interdependence (Zachariassen, 2008). In situations where a negotiator is focused purely on cost or profit without regard for a buyer–supplier relationship, a win–lose strategy may seem appropriate. However, in ongoing interdependent buyer–supplier relationships, SET suggests that the use of a win–lose negotiation strategy may have relational costs. Bargaining partners subjected to a win–lose strategy may be less inclined to engage in future collaborative behaviors such as information exchange, communication quality and operational knowledge transfer. The reciprocity principle of SET suggests that buyers and suppliers will mirror the actions of others. Therefore, aggressive “winner take all” negotiation tactics will likely invoke retaliatory sentiments and limit future collaborative knowledge sharing behaviors. Based on applicable literature and theory, the following hypotheses are proposed:

H2: A win–lose negotiation strategy leads to a decrease in intended (a) information exchange, (b) communication quality and (c) operational knowledge transfer activities between buyers and suppliers.

Negotiation strategy and interdependence are independent variables in this study and are hypothesized to affect the dependent variables of information exchange, communication quality and operational knowledge transfer. Beyond these simple main effects, SET suggests an interaction may exist between the level of interdependence in a buyer–supplier relationship and the type of strategy utilized in negotiations. In supply chain relationships with low levels of interdependence, a win–lose negotiation strategy may not be regarded as costly given the likely transactional nature of the relationship. Neither organization has high expectations for the relationship because the buyer and supplier do not truly need each other to achieve their goals. Therefore, the aggressive tactics of a win–lose negotiation strategy have less of an impact on future intentions in buyer–supplier relationships with low levels of interdependence. However, in a highly interdependent buyer–supplier relationship, a win–lose negotiation strategy may be viewed as a violation of the mutually gratifying expectations associated with interdependence. Firms that employ a win–lose negotiation strategy run the risk that their actions will alienate partners that had higher expectations for the relationship. In these situations, a win–lose negotiation strategy violates the quasi-economic profit-seeking motives of SET as relational costs begin to outweigh benefits. Based on the theoretical insight from SET, the following hypotheses are proposed:

H3: A win–lose negotiation strategy leads to a greater decrease in intended (a) information exchange, (b) communication quality and (c) operational knowledge transfer activities in highly interdependent relationships than in lower interdependent relationships.

Methodology

Overview

To test the theoretically derived a priori hypotheses concerning negotiation strategy and interdependent buyer–supplier relationships, a scenario-based experimental design (Rungtusanatham, Wallin & Eckerd, 2011) was used in this study. Experimental methodology was selected for several reasons. First, experimentation is a commonly accepted method in both negotiation research (Bottom et al., 2006; Buchan et al., 2004; Krause et al., 2006; Wolfe & McGinn, 2005) and buyer–supplier relationship research (Huang, Gattiker & Schwarz, 2008; Nair, Narasimhan & Bendoly, 2011; Thomas et al., 2010). Second, experimentation is ideally suited to isolate cause-and-effect relationships and systematically test scientific theory (Siemsen, 2011; Thye, 2007). Third, experimental methods provide a great deal of control, maximize internal validity and provide an “unequivocal assessment of causality” (Beatty & Ferrell, 1998, p. 186; McGrath, 1982). Fourth, experimental designs enable the investigation of buyer–supplier relationship phenomena that is difficult to duplicate because managers are often unwilling to share specific details of actual business relationships (Atkins & Rinehart, 2006; Day & Klein, 1987; Pilling, Crosby & Jackson, 1994). Finally, scenario-based experimentation “reduces biases from memory lapses, rationalization tendencies and consistency factors” (Grewel, Roggeveen & Tsiros, 2008, p. 428). For these reasons, a scenario-based experimental design was deemed appropriate for this research.

In this study, a 2 (interdependence: high vs. low) × 3 (negotiation strategy: win–lose vs. none/control vs. win–win) factorial design resulted in six treatment conditions. The dependent variables used to assess intended collaborative knowledge sharing behaviors in buyer–supplier relationships were information exchange, communication quality and operational knowledge transfer. Due to the categorical nature of the experimental treatment conditions, the hypothesized simple main effects and interaction effects were analyzed through MANOVA and ANOVA (Creswell, 2003; Hair, Black, Babin & Anderson, 2010; Kerlinger & Lee, 2000). The psychometric properties of the measures used in this experiment were validated through confirmatory factor analysis with structural equations modeling (Garver & Mentzer, 1999).

Sample

Senior undergraduate logistics majors enrolled in the capstone supply chain management course at a large southeastern university comprised the sample of research participants. As part of the standardized curriculum, participants had completed formal experiential training in business negotiations, relationship management and supply chain management. The total sample size was 78 with 13 participants per cell and exceeded minimum sample size requirements (Hair et al., 2010). The sample was 80 percent male, and the average age was 24. Over 85 percent of the participants self-reported at least 1 year of work experience with the average work experience being 4.6 years.

Although many objections have been raised regarding the use of student samples (Stevens, 2011), the practice is often justified for several reasons. First, student samples are widely accepted and frequently used in behavioral experiments in areas such as negotiation (Bowles & Flynn, 2010; Krause et al., 2006; Miller & Karakowsky, 2005) and buyer–supplier relationship research (Fugate, Thomas & Golicic, 2012; Srivastava & Chakravarti, 2009; Thomas et al., 2011; Tokar, Aloysius, Waller & Williams, 2011). Second, many studies also show that there are no significant differences in experimental results between undergraduate student samples and professional managerial samples (Croson & Donohue, 2006; Ganesan, Brown, Mariadoss & Ho, 2010; Machuca & Barajas, 2004). Third, undergraduate participants serve as a desirable control mechanism in experiments due to the consistent classroom delivery setting and the relative homogeneity of the sample that limits confounding effects of unknown demographic variables (Thomas et al., 2010). As these examples show, there are compelling reasons to consider student samples in experimentation.

In addition to the broad justification for student sample usage in experimental designs, we think student samples are justified in this specific study for several additional reasons. First, given the initial stage of this research stream, we are most interested in maximizing internal validity and establishing causal relationships. Homogeneous student samples support these goals (Thomas, 2011). Second, students are included within the boundary scope conditions of SET that informs this work and are therefore subject to theoretically derived hypotheses. Perhaps more importantly, SET has universalistic theoretical scope and therefore holds regardless of the sample characteristics (Stevens, 2011). Third, student samples are the de facto standard in negotiation research given that over 96 percent of existing negotiation research utilizes student samples (Buelens, van de Woestyne, Mestdagh & Bouckenooghe, 2008). This de facto standard is due to the universal nature of negotiations in basic exchanges that form the foundation of modern societies. Students inherently understand negotiations due to the universal nature of negotiation phenomena. Finally, students that participated in this study were able to understand and respond to the experimental treatment conditions given their prior life experiences with negotiation both within and beyond their work place experience as well as their formal educational training on business negotiations, relationship management and supply chain management. For these reasons, the sample utilized in this experiment is appropriate and meets guidelines for student sample use in supply chain research (Stevens, 2011; Thomas, 2011).

Procedure

Participants were randomly assigned to one of the six treatment conditions in the factorial design. Random assignment was utilized to decrease the likelihood of systematic between-group differences and maximize internal validity of the experiment (Huang et al., 2008). A scenario describing a buyer–supplier negotiation in an interdependent relationship was read by each of the participants. The independent variables (interdependence and negotiation strategy) were manipulated via the scenario. The scenarios were initially developed based on construct descriptions in the literature. Insights from academic and managerial subject matter experts were then utilized to further refine the scenarios and ensure realistic treatment conditions. Written scenarios are commonly used to operationalize independent variables and facilitate role playing in experimental designs (Rungtusanatham et al., 2011). After reading the scenario, participants were asked to complete a questionnaire that assessed likely responses to the business situation described in the scenario. The scenario manipulations are provided inAppendix A, and the individual scale items from the questionnaire are provided in Appendix B.

Demand Characteristics

Demand characteristics refer to “all aspects of the experiment which cause a subject to perceive, interpret and act upon what he believes is expected or desired of him by the experimenter” (Sawyer, 1975, p. 20). Although researchers can never completely rule out demand characteristics, they can limit the degree of contamination of this commonly accepted methodological limitation (Agyris, 1968; Allen, 2004). In this study, several aspects of the research design and experimental procedure were developed to reduce demand characteristics and their potential confounding effects. First, the study utilized a between-subjects design that is less prone to demand characteristics (Grice, 1966; Sawyer, 1975). Second, as part of the standard debriefing process at the conclusion of the experiment, this study utilized a postexperimental inquiry and determined demand characteristics were not likely present (Allen, 2004). Finally, several aspects of the experimental procedure were designed to reduce participant apprehension and limit demand characteristics (Sawyer, 1975). Responses were anonymous, participants were instructed that there were no “right” or “wrong” answers and a less threatening projective technique was utilized in the experimental treatment conditions. Based on these procedures and assessments, it is unlikely that participant behaviors were impacted by demand characteristics.

Projective Technique

A projective technique was utilized in this study to limit the potential effects of social desirability biases. A projective technique is a form of indirect questioning that asks participants to respond to structured questions from the perspective of another person or group (Fisher, 1993). When responding to these types of indirect questions, participants “project” their own values and behaviors (Mick, Demoss & Faber, 1992). Scenario-based experimental designs that utilize projective techniques provide insights into managerial behaviors and corporate strategies (Antia, Bergen, Dutta & Fisher, 2006; Thomas et al., 2010). Projective techniques are often utilized in research contexts where participants may be tempted to distort self-reported information to manage impressions and present themselves in the best possible light (Knemeyer & Naylor, 2011). Given the reciprocity theoretical foundations of this research, retaliatory behaviors were likely responses to some of the experimental treatment conditions. Therefore, the researchers wanted to allow participants to respond to the treatment conditions in a realistic manner without feeling threatened or forced to respond in a socially desirable manner.

Measures

Measures for the independent and dependent variables were adapted from existing multi-item scales. Minor modifications were made to maintain subject/verb agreement that was consistent with the scenario manipulation. A seven-point Likert scale was used for all items with a range from “strongly disagree” to “strongly agree.” The independent variable manipulation check items for negotiation strategy were adapted from Graham et al. (1994), and the interdependence scale was adapted from Golicic and Mentzer (2006). Dependent variable measures for information exchange were adapted from Lusch and Brown (1996), operational knowledge transfer measures were adapted from Modi and Mabert (2007) and communication quality measures were adapted from Mohr and Sohi (1995). In prior research, all scale items were found to be valid and reliable measures.

Analysis

Scale Purification

To ensure the adapted measures were valid and reliable in the context of this study, scale purification procedures were utilized. Convergent validity and unidimensionality were tested using confirmatory factor analysis with structural equations modeling. The fit of the confirmatory factor analysis measurement model was adequate with a RMSEA of 0.079 and CFI of 0.967 (Hair et al., 2010). The estimated parameter loadings were all significant, and there were no cross-loadings. The factor loadings are presented in Table 1. Reliability was assessed with Cronbach's Alpha (Appendix B). All alpha values exceeded the recommended value of 0.70 (Nunnally & Bernstein, 1994). To demonstrate discriminant validity, the average variance extracted (AVE) was assessed and all values were >0.5 threshold (Hair et al., 2010). Furthermore, the squared correlation between all pairs of constructs was <AVE (Appendix C), thus indicating discriminant validity (Fornell & Larcker, 1981). Based on this analysis, the measures used in this study are acceptable.

Table 1. Confirmatory Factor Analysis Loadings (CFI = 0.967; RMSEA = 0.079)
InterdependenceWin–Lose StrategyCommunication QualityOperational Knowledge TransferInformation Exchange
0.969    
0.932    
0.916    
 0.869   
 0.823   
  0.820  
  0.798  
  0.794  
  0.758  
   0.621 
   0.622 
   0.601 
    0.458
    0.555
    0.456

Manipulation Checks

To determine if the participants responded as planned to the independent variable manipulations in the experimental treatment conditions, manipulation checks were performed (Bachrach & Bendoly, 2011; Rungtusanatham et al., 2011). Tests show that the manipulations in this experiment worked as intended. There was a significant effect of the interdependence manipulation (F = 127.45; Mhigh interdependence = 5.40 > Mlow interdependence = 1.99; p < 0.001) as well as a significant effect of the negotiation strategy manipulation (F = 94.68; Mwin–lose strategy = 6.06 > Mnone (control) = 2.94 > Mwin–win strategy = 2.12; Bonferroni test: p < 0.05 for all pairwise comparisons). Based on these results, participants did perceive significant differences between each treatment condition.

Confounding Checks

In addition to manipulation checks, confounding checks were performed on the interdependence and negotiation strategy experimental manipulations. Confounding checks were used to assess the discriminant validity of manipulations and insure that one experimental manipulation was not influenced by another manipulation (Purdue & Summer, 1986). Based on the analysis, the experimental manipulations were clean and results of this research can be interpreted in a straightforward manner.

Realism Checks

Realism of the experimental treatment conditions in this study was assessed with both qualitative and quantitative approaches. For the qualitative approach, scenario manipulations were originally developed based on insights from the academic literature. Once the initial scenarios were developed, they were reviewed by four academic experts with experience in behavioral experimentation, buyer–supplier relationships and negotiations. The academic experts provided feedback regarding readability, length, credibility and realism. After making appropriate adjustments, the scenarios were deemed acceptable by the academic panel. The scenarios were then shared with three experienced purchasing professionals. Feedback was again solicited regarding each treatment condition. The managers provided their assessments and recommended some minor changes. After the changes were implemented, the managers agreed that the scenarios captured and manipulated the constructs of interest in a realistic and understandable manner.

The reliability of an experimental design depends on participants being able to recognize and respond to treatment conditions (Louviere, Hensher & Swait, 2000; Rungtusanatham et al., 2011). Therefore, a quantitative realism check was also performed in this study. Items developed by Dabholkar (1994) were used to evaluate the realism of the scenarios in the experimental design (Appendix B). Participants were asked if the scenario was realistic and if they could imagine themselves in the situation. The realism check indicated that participants considered the scenarios to be engaging and realistic with an average score of 5.56 on a seven-point scale. This finding suggests that participant perceptions of the scenario manipulations were realistic enough to evoke authentic behavioral responses. Therefore, due to the participants' perceived realism, results of this experiment do provide insights into basic behavioral responses to negotiation strategies.

Main Analysis

A MANOVA was conducted on the dependent variables in this study with interdependence and negotiation strategy as factors. As anticipated, a main effect of interdependence was observed (Wilks' λ = 0.678; F = 10.91; p < 0.001). Additional univariate tests show that an increase in interdependence leads to an increase in intended information exchange (F = 24.56; p < 0.001), communication quality (F = 7.64; p < 0.01) and operational knowledge transfer (F = 17.65; p < 0.001). These results offer support for H1a–H1c. The overall ANOVA results are provided in Table 2, and the dependent variable cell means are provided in Table 3.

Table 2. ANOVA Results for Main and Interaction Effects
EffectsInformation Exchange F-StatisticCommunication Quality F-StatisticOperational Knowledge Transfer F-Statistic
Interdependence24.55 (< 0.001)7.64 (< 0.01)17.65 (< 0.001)
Win–lose strategy15.13 (< 0.001)10.52 (< 0.001)1.53 (= 0.223)

Interdependence × 

Win–lose strategy

5.75 (< 0.01)0.70 (= 0.499)2.42 (= 0.096)
Table 3. Dependent Variable Cell Means
Dependent VariableInterdependenceNegotiation StrategyMeanStandard Error
Information ExchangeLowWin–Lose3.5380.304
None3.9520.293
Win–Win4.4440.316
HighWin–Lose3.7380.293
None6.1940.316
Win–Win5.7220.316
Communication QualityLowWin–Lose4.1730.340
None5.8930.328
Win–Win5.8330.354
HighWin–Lose5.4110.328
None6.4380.354
Win–Win6.3750.354
Operational Knowledge TransferLowWin–Lose3.1790.386
None2.8570.372
Win–Win3.1670.402
HighWin–Win3.5480.372
None4.7780.402
Win–Lose4.8890.402

A main effect of negotiation strategy was also observed (Wilks' λ = 0.613; F = 6.38; p < 0.001). Additional univariate tests indicate that a win–lose negotiation strategy leads to a decrease in intended information exchange (F = 15.13; p < 0.001) and communication quality (F = 10.52; p < 0.001). The results provide support for H2a and H2b.

As predicted, the main effects were qualified by the hypothesized interaction between interdependence and negotiation strategy (Wilks' λ = 0.812; F = 2.52; p = 0.024). Additional univariate tests indicate that a win–lose negotiation strategy leads to a greater decrease in intended information exchange (F = 5.75; p < 0.01) and operational knowledge transfer (F = 2.42; p = 0.096) in highly interdependent relationships. The results provide support for H3a and partial support for H3c. Figures 1–3 show the effects of interdependence and win–lose negotiation strategy on information exchange, communication quality and operational knowledge transfer. Table 4 provides a summary of the tests of hypotheses from the study.

Table 4. Summary Tests of Hypotheses
HypothesisPredictionFinding
H1An increase in levels of interdependence leads to an increase in
(a) information exchange (< 0.001)Supported
(b) communication quality (< 0.01)Supported
(c) operational knowledge transfer activities (< 0.001)…between buyers and suppliersSupported
H2A win–lose negotiation strategy leads to a decrease in
(a) information exchange (p < 0.001)Supported
(b) communication quality (p < 0.001)Supported
(c) operational knowledge transfer activities (p = 0.223)…between buyers and suppliersNot Supported
H3A win–lose negotiation strategy leads to a greater decrease in
(a) information exchange (p < 0.01)Supported
(b) communication quality (p = 0.499)Not Supported
(c) operational knowledge transfer activities (p = 0.096)…in highly interdependent relationships than in lower interdependent relationshipsPartially Supported
Figure 1.

Information Exchange

Figure 2.

Communication Quality

Figure 3.

Operational Knowledge Transfer

Discussion

The purpose of this research was to test theoretically derived hypotheses related to negotiation strategy effects on future knowledge sharing behaviors in interdependent buyer–supplier relationships. The effects of win–win and win–lose negotiation strategies were assessed in two types of relationships and shown to impact intended information exchange, communication quality and operational knowledge transfer. Insights from SET suggest that buyer–supplier interactions in discrete negotiation events will have effects on future collaborative efforts in ongoing relationships. Overall, results indicate that extending the scope of negotiation research beyond isolated outcomes of a specific bargaining interaction is an important issue for consideration in supply chain relationship research.

The experimental data in this study suggest that negotiation strategies and relational interdependence are directly related to several essential knowledge sharing behaviors. Specifically, as interdependence increases, intended information exchange, communication quality and operational knowledge transfer also increase. Results of this experiment also show that the use of win–lose negotiation strategies decreases intended information exchange and communication quality in ongoing buyer–supplier relationships. These findings provide empirical support for the notion that negotiation strategy outcomes are not limited to the typical negotiation research outcomes of profit achievement or satisfaction. Rather, as part of an ongoing exchange relationship, isolated negotiation incidents have a broader impact. Thus, the study provides additional empirical support for the future use of the reciprocity principle of SET in negotiation and relational contexts.

Beyond the simple main effects of negotiation strategy and relational interdependence, results of this study also show a significant interaction that is consistent with SET predictions. Specifically, in highly interdependent buyer–supplier relationships, a win–lose negotiation strategy results in less intended information exchange and operational knowledge transfer than when win–win approaches are utilized. This interaction suggests several interesting insights. First, win–lose negotiation strategies do not have noticeable adverse effects on some knowledge sharing intentions in relationships characterized by low levels of interdependence. Such evidence suggests that negotiation strategy selection has no effect on intended information exchange or operation knowledge transfer in situations where buyers and suppliers do not depend on each other. However, in highly interdependent relationships, bargaining partners view the use of win–lose negotiation strategies as a substantial relational cost that jeopardizes the future exchange of information and knowledge transfer. Therefore, results of this experiment show that win–lose strategies may be accepted without penalty in less interdependent relationships, but when buyers and suppliers depend on each other, winner take all negotiation tactics reduce knowledge sharing intentions between supply chain members.

Another interesting finding in this research is the impact of win–lose negotiation strategies on the dependent variable of communication quality. A win–lose negotiation strategy has a simple main effect on communication quality, but no interaction effects. Unlike the dependent variables of information exchange or operational knowledge transfer, communication quality intention is reduced by a win–lose negotiation strategy regardless of the level of interdependence in a buyer–supplier relationship. This finding suggests that information exchange and operational knowledge transfer may be somewhat robust to adverse effects of win–lose negotiation strategies in less interdependent relationships, but communication quality intentions may be jeopardized when aggressive bargaining tactics are implemented.

In summary, results of this research highlight several managerial and theoretical issues that are essential to the management and study of supply chain negotiations in buyer–supplier relationships. Given the ubiquitous nature of negotiations and the buyer–supplier relationship building blocks of supply chains, understanding the interface between these two critical areas is indeed relevant. Therefore, the findings of this research have meaningful implications for both the study and practice of supply chain management.

Implications

Although negotiation is widely researched in a number of disciplines, investigated outcomes are generally limited to economic variables like profit or social-psychological variables like satisfaction. Unfortunately, such an approach limits the potential scope of negotiation research in the buyer–supplier relationship domain. Such a gap in the literature is noticeable given the importance of both negotiations and buyer–supplier relationships to supply chain management. Therefore, gaining greater insights into how a discrete negotiation event can impact future collaboration between buyers and suppliers is warranted. This research represents a critical first step toward quantifying the effects of negotiation styles on intended knowledge sharing in buyer–supplier relationships. Results of the experiment suggest that negotiation interactions are not isolated incidents with limited outcomes. Rather, negotiations can have a profound effect on critical information flows between supply chain members. The implication of this finding is that a new stream of research can and should emerge in this area of study. As key supply chain elements continue to be negotiated between interdependent buyers and suppliers, understanding the negotiation/relationship interface is essential.

This research also has managerial implications for relationship management tactics and strategies. Buyers and suppliers who utilize the win–lose negotiation strategy should understand the relational costs of decreased information exchange, communication quality and operational knowledge transfer intentions. This finding is especially salient given that prior research has shown that win–lose negotiators will often beat win–win negotiators by gaining a larger portion of monetary benefits in a negotiation (Graham et al., 1994). When viewed as a discrete event, it would seem that firms should use win–lose strategies in their negotiations. However, given the findings of this research, utilizing a win–lose negotiation strategy seems somewhat shortsighted. This research contributes to our understanding of buyer–supplier relationships and shows that negotiations are not only discrete events, but a continuing part of ongoing exchange relationships. Managers would be well advised to consider both immediate and longer-term effects of negotiation strategies in their supply chain relationships.

Limitations and Directions for Future Research

As the results of this experiment demonstrate, additional research is needed and justified in several areas. Although this study focused on intended knowledge sharing–dependent variables, future research should consider other relational variables. Constructs like trust, commitment, cooperation, joint investment or loyalty may also be affected by negotiation strategies. As the broad theoretical scope of SET informed this work, it is likely other collaborative intentions are also viable dependent variables for future research. Given the importance of collaboration in modern supply chains and given the preliminary evidence in this experiment, additional research is needed that looks at other key relational constructs that are relevant to both the study and practice of supply chain management.

Another area future research could address is different types of relationship structures as well as relationships with international partners. Supply chains are comprised of a wide portfolio of relationship types (Rinehart, Eckert, Handfield, Page & Atkin, 2004) with a broad variety of geographic dispersion. Negotiation strategies may very well take on different meanings in different types of relationships or different international cultures. In this study, the scope of inquiry is limited to interdependent relationships in the United States. However, given the ongoing global sourcing and outsourcing trends in modern supply chains (Kusaba, Moser & Rodrigues, 2011; Trautmann, Turkulainen, Hartmann & Bals, 2009), research is needed that looks beyond traditional American buyer–supplier relationships.

This study represents a critical first step in extending the scope of negotiation research and identifies new potential causal relationships of interest. However, this initial experimental design has several methodological limitations that future research should address. Due to the student sample utilized in this study and the laboratory treatment conditions in the experiment, generalizability and realism are not maximized. Therefore, additional research is needed that capitalizes on the causal relationships established in this study and utilizes other methodologies to offset inherent weaknesses of experimentation. This experiment also measured behavioral intentions rather than actual behaviors. Although the Theory of Planned Behavior (Ajzen, 1991) and meta-analysis suggest that behavioral intentions and subsequent behavior are highly correlated (Sheppard, Hartwick & Warshaw, 1988), other factors can still influence actual behaviors. Therefore, additional research is needed to investigate the proposed behavioral implications of this study.

In this study, common aspects of basic win–win and win–lose negotiation strategies were evaluated. However, negotiation strategies are not always as simply polarized and other approaches exist. Now that potential causal relationships have been identified, researchers should move beyond basic negotiation strategies and examine more nuanced aspects of negotiation such as negotiator skill, personal likeability, perceived competence and satisfaction with negotiated outcomes. These more subtle aspects of negotiation may also play a critical role in influencing behavioral intentions. Future research should address these types of independent variables and determine if they affect ongoing exchange relationships.

Although this study primarily focused on win–lose negotiation strategy effects, future research should also consider effects of win–win tactics. Specifically, additional research is needed to investigate why the win–win negotiation strategy had little impact on knowledge sharing intentions relative to the control group in this experiment. Although manipulation checks for negotiation strategy type showed statistically significant differences between the win–win groups and control groups, these groups had similar levels of collaborative knowledge sharing intentions. There are several potential explanations for these results. Perhaps participants viewed win–win negotiation styles as the standard default approach in modern negotiations. Perhaps win–win negotiations are stressed more in coursework and presented in a positive light when participants receive training. Perhaps SET has unidentified boundary conditions that affect the reciprocity principle and create a bias toward retribution over reward. Although these potential explanations are interesting, they are conjectural and speak beyond the results of the data. Therefore, additional research is needed to understand the longer-term effects of win–win negotiation strategies on behavioral intentions in supply chain relationships.

Appendix A: Directions and Scenarios

Directions

Imagine that The Eagle Company (TEC) is a manufacturer that supplies products to a specific retailer. The business interactions of TEC and the retailer are described below. Assume all scenario descriptions are accurate and trustworthy. After reading the scenario, please answer each question. As you answer each question, predict how TEC would work with the retailer in the future based on the scenario. Please do not base your answers on how you think TEC should work with the retailer, but rather on how they actually would work with the retailer.

Interdependence Scenario Manipulations

High Interdependence

The Eagle Company and the retailer have been doing business with each other for several years. The retailer is one of TEC's larger customers and represents a meaningful portion of TEC's overall sales volume. Likewise, TEC is one of the retailer's larger suppliers and a meaningful portion of the retailer's overall revenue comes from selling TEC products. Obtaining TEC's products from another supplier would be somewhat difficult for the retailer. Replacing the retailer's sales volume would also be somewhat difficult for TEC.

Low Interdependence

The Eagle Company and the retailer have been doing business with each other for less than a year. The retailer is one of TEC's smaller customers and represents an insignificant portion of TEC's overall sales volume. Likewise, TEC is one of the retailer's smaller suppliers and an insignificant portion of the retailer's overall revenue comes from selling TEC products. Obtaining TEC's products from another supplier would not be difficult for the retailer. Replacing the retailer's sales volume would also not be difficult for TEC.

Negotiation Strategy Scenario Manipulations

Win–Win Negotiation Strategy

The retailer and TEC recently conducted their annual negotiation to determine what TEC products the retailer would carry in their stores over the next year. During these negotiations, the retailer shared information, communicated clearly and focused on achieving mutually acceptable goals. The retailer was not aggressive and did not attempt to threaten or intimidate TEC. The retailer was also open to making concessions to solve problems.

Win–Lose Negotiation Strategy

The retailer and TEC recently conducted their annual negotiation to determine what TEC products the retailer would carry in their stores over the next year. During these negotiations, the retailer did not share information, communicated deceptively and focused on achieving their own goals. The retailer was aggressive and attempted to threaten and intimidate TEC. The retailer was not open to making concessions to solve problems.

Appendix B: Measurement of Dependent and Manipulation Check Variables

Information Exchange (Lusch & Brown, 1996)

Cronbach's α = 0.89

  • TEC would share information with this retailer about changes that may affect them.
  • TEC would share information that might be helpful to this retailer.
  • TEC would share information with this retailer frequently and informally, and not only according to a prespecified agreement.

Communication Quality (Mohr & Sohi, 1995)

Cronbach's α = 0.97

  • TEC would ensure that their communication with this retailer was accurate.
  • TEC would ensure that their communication with this retailer was adequate.
  • TEC would ensure that their communication with this retailer was complete.
  • TEC would ensure that their communication with this retailer was credible.

Operational Knowledge Transfer (Modi & Mabert, 2007)

Cronbach's α = 0.90

  • TEC personnel would visit the retailer's premises to help them improve performance.
  • TEC would invite the retailer's personnel to TEC sites to increase the retailer's awareness of how their product is made.
  • TEC would conduct development and education programs for the retailer's personnel.

Negotiation Strategy (Graham 1985; Graham et al., 1994)

Cronbach's α = 0.83

  • The retailer had a “winner take all” approach to their negotiation with TEC and focused only on their own self-interests.
  • The retailer utilized a “win–win” negotiation style with TEC and focused on joint problem solving.

Interdependence (Golicic & Mentzer, 2006)

Cronbach's α = 0.96

  • TEC and the retailer could not easily replace each other.
  • TEC and the retailer are dependent upon each other.
  • TEC and the retailer believe they are crucial to each other's success.

Realism Checks (Dabholkar, 1994)

Cronbach's α = 0.84

  • The situation described in the scenario was realistic.
  • I can imagine myself in the described situation.

Appendix C

Average Variance Extracted

 IEOPKTCQWLSID

Notes

  1. Diagonal: average variance extracted.

  2. Lower matrix: squared correlations.

Information Exchange 0.73     
Operational Knowledge Transfer0.36 0.76    
Communication Quality0.190.07 0.89   
Win–Lose Strategy0.550.180.26 0.72  
Interdependence0.350.460.120.13 0.88

Biographies

  • Stephanie P. Thomas (M.B.A., University of Tennessee) is a doctoral candidate in the Department of Marketing and Logistics/College of Business Administration at Georgia Southern University in Statesboro, Georgia. Prior to her graduate academic work, she held positions at companies including IBM, Lowe's and Stanley Tools; at the latter company, she worked in warehousing and distribution. Ms. Thomas brings managerial-level experience to the study of customer service, purchasing, negotiation, logistics planning, category management and supply chain strategy. Her research interests currently focus on negotiations, buyer–seller relationships and retail supply chain management. Ms. Thomas' research has been published in the International Journal of Logistics Management, as part of several conference proceedings, and is under review at a variety of marketing and logistics journals.

  • Rodney W. Thomas (Ph.D., University of Tennessee) is an assistant professor in the Department of Marketing and Logistics/College of Business Administration at Georgia Southern University in Statesboro, Georgia. His research focuses on the behavioral aspects of supply chain management, with a particular emphasis on buyer–supplier relationships, time pressure sourcing behavior and retail logistics. He also brings practical experience to this research, having held several positions with the Lowe's Home Improvement Company including Merchandise Buyer and Director of Logistics. Among the publications in which Dr. Thomas' research has appeared are the Journal of Business Logistics, the Journal of Retailing, Industrial Marketing Management, the International Journal of Relationship Marketing and the Transportation Journal.

  • Karl B. Manrodt (Ph.D., University of Tennessee) is a professor in the Department of Marketing and Logistics/College of Business at Georgia Southern University in Statesboro, Georgia. His most recent research projects involve strategic sourcing, relationship management and vested relationships. Dr. Mandrodt has written five books on logistics, metrics, strategic sourcing and procurement. His research articles have been published in a variety of journals, including the Journal of Business Logistics, Omega, the Journal of Supply Chain Management, and Strategic Sourcing: An International Journal.

  • Stephen M. Rutner (Ph.D., University of Tennessee) is a professor of logistics and transportation in the Department of Marketing and Logistics/College of Business Administration at Georgia Southern University in Statesboro, Georgia. He is also the Commander, Army Reserve Element, Defense Logistics Agency, with the rank of Colonel. He has served as Transportation Officer in the US Army and the Army Reserve, and spent two tours in Iraq/Kuwait. Dr. Rutner's primary areas of research have been shipper–carrier alliances, airport and airline management, logistics measurement and information technology. Articles by Dr. Rutner have been published in Transportation Quarterly, the International Journal of Physical Distribution and Logistics Management, among other outlets.

Ancillary