6

Intergroup Relations in Organizations

Jeanne M. Brett

Jorn Kjell Rognes

In investigating intergroup relations in organizations, we assume that groups are the basic structural building blocks of organizations and that relations between groups in organizations must be managed. We want to know why some intergroup relations are more difficult to manage than others, how to distinguish well-managed intergroup interfaces from those that are poorly managed, and what factors lead to effectiveness.

We have drawn on a diverse literature in investigating these questions, including microtheory that focuses on the enactment of organizational roles, macrotheory that focuses on the internal structures of organizations, and organization-environmental theory that focuses on transactions between organizations. The conceptual framework for research on organizations developed in this chapter recognizes the interplay between these three types of theory.

Our framework assumes that the study of intergroup relations within organizations is important and that the appropriate conceptual approach is an integrative one. These are not idle assumptions. The importance of lateral relations within and between organizations has increased in recent years (Kochan and Bazerman, 1986). Organizations are becoming larger and associated with that growth is increased specialization of groups within organizations (Mintzberg, 1979). Another factor contributing to the specialization of organizational groups is increased complexity in the environment and in technology. A result of this increased specialization is that tasks requiring intensive coordination cannot be grouped into a single functional unit (McCann and Galbraith, 1981).

Lateral relations between organizations are also increasing in importance. There has been substantial growth in the long-term contractual relations between different organizations, for example, in partly integrated distribution channels (Stern and El-Ansary, 1977), in contractual relations in the construction industry (Stinchcombe, 1983), and in the international expansion of the service sector (Kochan and Bazerman, 1986). The recent surge of mergers and acquisitions has, in many cases, transformed lateral relations between organizations into lateral relations within organizations.

The increasing importance of lateral relations within organizations has spawned a plethora of structural forms—matrices, task forces, teams, and so on—that fall between the ideal types that have preoccupied management scholars (that is, bureaucracy) and economists (competitive markets). In hierarchical, bureaucratic systems, conflict is resolved by the dictates of the superior. In the competitive market, conflict can be resolved (dissolved) by changing transaction partners. But when groups, whether inside the same organization or in different organizations, have to engage in long-term, lateral relations, conflict management becomes problematic. Authority, responsibility, and appropriate behavior are unclear (Brown, 1983). Unlike hierarchical relations, there are often no norms about group interaction when conflict breaks out. Furthermore, managers leading such groups and representing them in dealings with each other are seldom selected for their negotiation skills and may not have had an opportunity to learn such skills. Probably the most important factor, however, is that they are not rewarded for the development and maintenance of cooperative lateral relations. Many managers holding key lateral positions talk about lateral relations in terms of problems, not conflict, and fail to recognize that what they are doing to manage those relations is negotiating.

While this chapter focuses on lateral relations within, not between, organizations, there is much to be gained conceptually from recognizing the interplay between levels of theory. The issue of the effective management of transactions between units is central to both intra- and interorganizational relations. The problems of enactment are the same.

This chapter is organized as follows: First intergroup relations are defined and the characteristics of intergroup relations are identified. Second, criteria for effective intergroup relations are presented. Third, propositions about the effectiveness of various techniques of managing intergroup relations are developed.

Intergroup Relations in Organizations

In order to understand what an intergroup relationship within an organization is, we have to be able to know one when we see one. This entails being able to identify a group and a relationship and to determine whether the relationship is between groups in the same organization or between groups in different organizations. The first and last issues having to do with the definition of a group and an organization are relatively easy compared to the issue of defining a relationship.

An organizational group is a set of individuals who perceive themselves and whom nonmembers perceive as constituting an identifiable social aggregate within the organization. The group may be a small, primary work group, a task force whose members are formally assigned to a variety of different functional areas, or a huge functional department like marketing, production or accounting. The only definitional requirement is that members and nonmembers recognize the group as a distinguishable social aggregate that exists within the boundaries of the organization.

In most cases, organizational groups can be further distinguished in terms of specific work tasks and the technologies, goals and values associated with them. If each group were producing its own product or service, there might be little need for significant intergroup coordination. In most cases, however, identifiable groups in organizations are producing only a specialized segment of the organization's product or service. Coordination between such groups is a necessity. Even organizations whose groups have traditionally provided distinct services—for example, banks with their separate areas controlling loans, transfer of funds, and investment—are requiring significant intergroup coordination in order to provide full service to customers in an increasingly competitive environment.

An intergroup relationship within an organization is a coordinated link or bridge between two distinct organizational groups. The link or bridge is necessary because each group needs resources from the other group in order to complete its own task (Aiken and Hage, 1968). On an automobile production line, the trim department (group) needs to receive painted chassis from the paint department. If something goes wrong in paint, trim waits; it cannot put chrome on unpainted chassis even if they did come down the line. Likewise, if something goes wrong in trim, paint waits, for it has no other outlet for its product.

Power-Dependency Relationships. Theoretically, an intergroup relationship is a power-dependency relationship (Emerson, 1962). Each group is dependent on the other for resources. The degree of dependency is a function of the availability from other groups of the resources being exchanged and the value of the resources to each group. In our example, paint and trim are highly interdependent and the dependency is symmetric. Paint has no other place but trim to send chassis, and moving chassis through the paint shop is a major element of the paint group's performance evaluation. Trim has no other source of painted chassis and, like the paint group, is evaluated on how many cars are trimmed per shift.

Some intergroup relations are based on asymmetric dependencies. For example, marketing and manufacturing departments depend on each other to produce the market for the product and to produce the product for the market. But a marketing department that has top management approval to contract out production when a particular campaign generates demand in excess of manufacturing's capacity is less dependent on manufacturing than manufacturing is on marketing. Such asymmetries in dependence imply differences in power between the two departments.

Power, according to Emerson's theory (1962), is a function of dependency. A group is powerful when the resource it provides is needed and scarce. A group is weak when the resource it provides is abundant and/or not very important to the other party. In an asymmetric power relationship, the more powerful party is able to get more from the exchange because the weaker party is willing to give more to maintain the relationship. In our example about marketing and manufacturing, manufacturing should be more willing to acquiesce to marketing's requests than vice versa because its alternatives are not as good as marketing's alternatives.

The example points out several other characteristics of power. First, strategic resources—that is, resources from which a party can derive power—are not limited to materials. Marketing's power in regard to manufacturing is embedded in its expertise and information. Marketing, through its prediction capacity, can smooth market demand for manufacturing that cannot respond quickly to changes in the market. Other resources that may be a source of power include financial and human resources, expertise and technology. For example, in the classic matrix design integrating functional departments and projects, the power of the functional department heads rests on their control of human and technological resources. The power of the project team directors rests in their financial resources.

A second important characteristic of power is that although it is situationally based, it is a psychological construct. The utility of marketing's power depends on manufacturing's perception of its dependence on marketing. Manufacturing is going to act on the basis of its own perceptions. Furthermore, manufacturing's own perceptions of its dependence may be at variance with marketing's perceptions of manufacturing's dependence. We propose that in a dyadic relationship perceived power has four elements:

• Party A's perception of own power
• Party A's perception of B's power
• Party B's perception of own power
• Party B's perception of A's power

The empirical research on situational power (Rubin and Brown, 1975; Bacharach and Lawler, 1981; Greenhalgh, Neslin, and Gilkey, 1985) conceptualizes power as a psychological construct but, with the exception of the Greenhalgh, Neslin, and Gilkey study, treats power as a situational variable for analytic purposes. Results show that in the laboratory, parties with high situational power use it and get higher outcomes (Rubin and Brown, 1975). Greenhalgh, Neslin, and Gilkey (1985) take this research one step further by showing that own perceived power is indirectly related to outcome. We propose that perceptions of others' power may also have a significant impact on a negotiated outcome. So, going back to our marketing and manufacturing example, manufacturing's perception of its own power may affect the negotiation, but so may marketing's perception of manufacturing's power.

We offer a third characteristic of power as a proposition. We propose that power relationships between groups in organizations are stable at least in the short term. There are two reasons why. First, in an organization each group usually has a monopoly on its task. Structural change to break that monopoly either from within the organization or through contracting out is likely to require top management approval. Such change will not occur swiftly. Second, while we concede that in a dyadic relationship within an organization control over resources and the values attached to them varies between parties across issues, we note that power is a perception and thus it carries all the biased baggage of any other perception. In particular, we suspect a large halo effect across issues such that power distinctions between issues tend to blur. We propose that perceptions of power based on accumulated experience across a range of interface issues define stable social influence relationships between groups.

Over time, of course, changes will occur in power-dependency relationships between groups. These changes will be functions of changes in each group's environment that make its re-sources less valued or less scarce to the other group. For example, the power of groups within organizations that offer centralized computing facilities to other organizational groups has diminished significantly with the widespread availability of powerful personal computers.

Intergroup Transactions. An intergroup transaction occurs “whenever individuals belonging to one group interact collectively or individually with another group or its members in terms of their group identification” (Sherif, 1966, p. 9). Most intergroup transactions occur between representatives of groups acting in the interests of their groups, not of themselves. Thus, an intergroup transaction is not the same as an interpersonal transaction. What the representatives “get personally from the encounter is not strictly germane to the question of what they have secured for those persons whom they represent” (Stephenson, 1981, p. 177).

In theory, any member of a group is an equally viable representative of the group's interests. In practice, group members frequently do not have homogeneous values and interests and are therefore not fully interchangeable in the role of group representative. Two factors serve to reduce variability in representative behavior. First, the role of group representative carries with it strong expectations about appropriate role enactment. These expectations are shared by the role incumbent, group members, and members and representatives of the other group. So even when group representatives are representing their own interests and not the group's, the other group's representatives believe they are representing the group. The reward contingencies are such that they are better off representing their group, not themselves, for they are ultimately accountable to their group. Groups that elect or select their own leaders remove leaders whom group members perceive to have negotiated an intergroup relationship that violates the group's interests (Bok and Dunlop, 1970). In organizations, intergroup representatives often also hold roles of hierarchical authority within their own groups (for example, unit manager). They are not typically accountable to their groups for their jobs, but they are highly dependent on group members to implement the intergroup relationship. Group representatives with hierarchical authority are ultimately evaluated based on the effectiveness of their groups. A group that fails to implement an intergroup transaction negotiated by its representatives makes the representatives look bad to higher management.

The justification for theory at the intergroup level rather than the interpersonal level is the fact that group representatives act on behalf of the “group,” not themselves. The reason for the representatives' interaction is an interface issue that exists at the group, not the individual, level.

Interface Issues. There are three types of interface issues: conflict, constraints, and interdependent action (Dale, 1981). In this chapter we are primarily interested in intergroup conflict and its management. Constraints and interdependent action are outcomes of the management of intergroup conflict. Constraints refer to limits placed on a group. For example, marketing could agree to suspend advertising for a product with which manufacturing was having quality control problems. Interdependent action refers to coordinated action between two groups. In our assembly line example, the speed of the line is a coordinated action; if paint moves chassis at the rate of sixty per hour but trim can move them only at the rate of fifty per hour, coordinated action will break down and conflict erupt. One solution, a constraint on intragroup action, might be for paint to slow down.

Intergroup conflict occurs when two groups linked in a power-dependency relationship disagree about one or more aspects of the terms of that relationship. Conflict is endemic to groups in organizations. The products or services of all but the smallest organizations are the result of coordinated action of functionally differentiated groups. Since each group performs only a monopolistic set of tasks, transfer of resources among groups is a critical part of organizational functioning. Coordination between groups is exacerbated by the fact that differentiated groups vary not just with respect to task but also in terms of technology, goal and time orientation, reward systems, and member characteristics (Lawrence and Lorsch, 1967; March and Simon, 1958). The very elements that support differentiation between groups inhibit integration because they cause interdependent groups to take positions on interface issues that maximize their own parochial interests.

While the fundamental causes of intergroup conflict are interdependence and differentiation, the intensity of conflict between groups is partly a function of scarce resources. If one group has excess resources (slack), it can develop excess capacity and thereby buffer its relationships with other groups and reduce intergroup conflict. When slack is not an option, lateral relationships between interdependent groups are likely to be conflictive. U.S. firms, reducing slack under pressures of deregulation and foreign competition, are learning that they have to improve coordination between groups within organizations as well as with suppliers. Interestingly, firms are beginning to recognize that one way to improve intergroup relations is to improve managers' negotiating skills.

While the causes of conflict are structural, the actual management of conflict takes place between individuals representing the groups. The process by which group representatives determine new terms and conditions of their exchange relationship is negotiation or bargaining (these terms are used interchangeably). Negotiation is an intergroup (or interpersonal) decision-making process under conditions of dependence and conflicting interests. Dependence needs no further definition at this point; conflicting interests do. An interest is the reason why a group (party) wants an exchange structured in a particular fashion. For example, manufacturing has two major related interests: labor costs and long production runs. Marketing, in contrast, is interested in a quick response to changes in the market environment. Both groups want to sell the product. So while marketing and manufacturing share an important interest, manufacturing's interest in long production runs and marketing's interest in quick responses are in conflict. This conflict will surface from time to time (depending on the dynamism of the market environment). The process by which the conflict is managed is negotiation. Note that the conflict is not resolved by the negotiation. Conflicting interests are structurally based and do not disappear when an issue has been successfully negotiated. Conflicting interests are simply subordinated to the joint gain of coordinated action (or intragroup constraints), at least until the environment changes and the coordinated action (or intragroup constraints) no longer is more valuable than the parochial interests of at least one of the parties.

Summary. This section provides the conceptual definitions and theoretical basis for the propositions about the effectiveness of various techniques of managing intergroup relations in organizations. The main points are as follows:

• A group is an identifiable social aggregate within an organization.
• An intergroup relationship is a coordinated action between two groups that exchange resources.
• Power is a function of dependency.
• Power is a psychological construct.
• Power relationships are stable, at least in the short term.
• An intergroup theory is necessary because while groups interact through individual representatives, the representatives present their group's interests, not their own, on interface issues that exist at the group level, not the individual level.
• Intergroup conflict occurs because interdependent groups have differentiated parochial interests that are stable only so long as each group's environment is stable.
• Scarce resources exacerbate intergroup conflict.
• Negotiation (or bargaining) is the process by which groups determine the terms and conditions of their exchange relationships.

Effectiveness of Intergroup Relations

In order to develop propositions about the effectiveness of various techniques of managing intergroup relations in organizations, we must develop criteria of effectiveness that are meaningful at the intergroup level of conceptualization. The intergroup effectiveness criteria must be conceptually the product of an intergroup exchange; they must be capable of characterizing the intergroup relationship over time, and they must provide an index of the quality of the exchange.

We found the literature evaluating alternative modes of dispute resolution to be a better source for ideas about intergroup effectiveness criteria than either the group or organizational effectiveness literature. This is because dispute resolution studies focus on the results of a dyadic exchange, whereas the group and organizational effectiveness studies focus on a single entity acting in its environment. A good example of the latter type of evaluation study, which would have benefited from knowledge of the former type of study, is Gladstein (1984). Gladstein was studying group effectiveness. Her dependent variable, however, was the result of an intergroup transaction rather than an intragroup process. It is not surprising, then, that intra-group variables did not predict sales, but market factors did.

In the alternative dispute resolution evaluation studies—for example, Brett and Goldberg's (1983) study of grievance mediation and Pearson's (1982) study of the mediation of child custody disputes—researchers have focused on the cost of the process, the endurance of the settlements, and participants' evaluations of the process (procedural justice) and the outcomes (distributive justice). We have focused on the first two of these criteria. We believe that in organizations, agreements that are perceived by the implementors to be unjust on either procedural or distributive grounds will not endure. This is a testable proposition. If it is true, then justice issues need not be formally included in our conceptualization of intergroup effectiveness. By not including justice in our definitions, we do not mean to denigrate their importance in research on intergroup relations and negotiations. Indeed, we draw heavily on the procedural justice research in developing propositions about the effectiveness of various techniques for managing intergroup relations.

We propose that the effectiveness of an intergroup relation can be evaluated in terms of efficiency and quality. Efficiency refers to the management costs associated with transforming an intergroup conflict into actions agreed upon by the groups. Those actions include constraints on intragroup activities and/or coordinated intergroup activities. Manpower costs associated with developing the transaction terms are part of the efficiency criterion. Management costs associated with implementing an exchange either in terms of constraints or coordinated action are not part of the efficiency criterion. Quality refers to the degree to which the intergroup negotiation results in a well-defined exchange agreement that endures until changes in relevant environmental factors make the exchange agreement obsolete.

Our concept of effectiveness is related to, but not synonymous with, Williamson's concept of transaction costs. Transaction costs are incurred, according to Williamson (1975), when at least one of the parties in an interdependent exchange relationship pursues its own interests rather than maximizing the two parties' joint benefits. Transaction costs can take two forms: Costs are incurred when parties engage in needless haggling over the terms of the transaction and/or when the terms of the transaction are advantageous to one party but suboptimal for the relationship as a whole (Williamson, 1979).

We propose that at the intergroup level of conceptualization, transaction costs have two elements: efficiency and quality. Our efficiency criterion refers to costs associated with the establishment of transaction terms; that is, the manpower and resources expended during the intergroup negotiations. A major challenge associated with operationalizing our efficiency criterion is to be able to distinguish between the costs of negotiating the transaction terms and the costs of implementing them in terms of coordinated action or intragroup constraints. Our quality criterion refers to the nature of the negotiated transaction terms. We propose that transaction costs are incurred when transaction terms are renegotiated even though no relevant environmental factor has changed.

Why do negotiators reopen agreements? We doubt they do so because they have been to a negotiations training workshop and now believe they can better the other party. We doubt they do so because they think they have more power. (Recall we proposed that the power distribution between two parties or groups is stable in the short term.) We expect negotiators reopen negotiations after a settlement because the settlement was poor, either because it was incomplete or ill defined. A complete, well-defined settlement is one that covers all aspects of the interface issue. Procedures, responsibilities, tolerances, and timing are all specified so that new issues, which require new negotiation, do not arise during implementation. Despite the arguments by some theorists of the utility of ambiguity in getting agreement among groups with conflicting interests (Weick, 1969, 1979), we believe that exchanges defined in ambiguous terms will quickly break down because of unmet expectations. The result is increased transaction costs as the details of the intragroup constraints or the coordinated action are worked out.

We do not expect parties to achieve complete, well-defined settlements in all situations. Bounded rationality will often inhibit completeness and it may not be possible to establish terms for all possible contingencies. There may also be occasions when parties wish to come to agreement in principle and work out the details of the exchange as the relationship develops. If a complete, well-defined settlement is not worked out, effectiveness is a question of the parties' abilities to readjust settlements, when needed, without incurring excessive management costs.

We recognize that there will be occasions when negotiations need to be reopened or extended. We are primarily concerned, however, with negotiations that result in ill-conceived settlements due to the parties' failure to make an explicit agreement on one or more of the known intergroup issues.

Efficiency and quality are separate constructs. Intragroup constraints or coordinated actions that endure justify significant costs in the negotiation phase, though enduring settlements are not necessarily the result of significant negotiation expenses. High-quality solutions justify high development costs; low-quality solutions do not. Thus, we propose that, controlling for the rate of environmental change, intergroup relations that incur low negotiation costs relative to their endurance are effective intergroup relations.

The rate of environmental change is an important factor in this criterion. Groups in organizations that face dynamic environments will have to negotiate more frequently and their solutions will be less enduring than those of groups in organizations that face stable environments. Thus, it is essential that the rate of environmental change be considered when evaluating the effectiveness of an intergroup relationship. Groups that negotiate more frequently than the rate of change in their environment are ineffective by our conceptualization and incur transaction costs, according to Williamson (1979).

The rate of environmental change, however, may be much more rapid for one of the groups in the relationship than the other. When there is substantial disparity between groups in the rate of environmental change, we expect the normal rate of negotiation to fall somewhere between the rate expected if both groups were in dynamic environments and the rate expected if both groups were in stable environments.

We recognize that Williamson's transaction costs concept has been difficult to operationalize for interorganizational relations. We know that operationalization of our conceptualization of transaction costs at the intergroup level will be difficult, if only because there are sometimes no price mechanisms for intraorganizational exchanges. But, as Jennergren (1979) points out, transfer prices can be estimated for exchanges between organizational subunits. Furthermore, it is not uncommon that profit centers are charged for use of centralized staff functions (for example, computer services). In principle, prices can be put on the exchange of resources within organizations.

Assuming that intergroup transaction costs can be estimated, we can then compare transaction costs across different interfaces. Then, controlling for interdependence and rate of environmental change, we can begin to test causal models of intergroup effectiveness.

Our conceptualization of transaction costs provides significantly more precision than Williamson's ecologically oriented framework. From the ecological perspective, intergroup relations that persist over time must be effective. Otherwise, the groups would decouple or be integrated into one unit. While the ecological perspective may be quite useful in showing how groups couple and decouple in response to environmental changes over the long term, it provides no insight into the causes of variation in effectiveness across intergroup dyads in the relatively short term. Since many intergroup relations are stable for relatively long periods of time, it becomes important to study the utility of different ways to manage the interfaces. Our operationalization of the transaction costs concept makes such research possible.

To summarize, we have defined in this section the effectiveness of intergroup relations in terms of transaction costs. We proposed the following:

• A transaction is of high quality when the negotiated exchange agreement endures until changes in relevant environmental factors make it obsolete.
• Groups in dynamic environments will have to negotiate more frequently than groups in stable environments, and their exchanges will be less enduring.
• Assuming that such transaction costs can be measured, then, controlling for environmental change and dependency, it is possible to test causal models of intergroup effectiveness.

Techniques for Managing Intergroup Relations

The major technique for managing conflict between groups in organizations is to dissolve the relationship either by changing the organizational structure, hence redefining intergroup relations (McCann and Galbraith, 1981), or by transferring key people (Brett, 1984). Kolb (1986) presents a particularly rich description of how a division manager, familiarly known as “Reorganize it Rogers,” manages a classic conflict between planning and operational units. Rogers attributes conflict to personalities and individual styles of the managers involved. He says to Kolb: “Did you ever meet such a cast of characters? I can't get them to do anything I want. I've tried.” Kolb continues:

Rogers's responses reflect the two different approaches to conflict management in organizational theory and organizational psychology. The former views the inability to manage conflict as a structural problem; the latter, as a personality problem. If, according to Rogers, people were less rigid and more flexible, problems between managers would not occur. Note that Rogers fails to recognize that his managers' “problems” are conflict associated with their matrix roles.

The view of intergroup relations developed in this chapter is consistent with the organizational theory view that conflict has structural causes. However, we do not believe that structural change is a sufficient condition for managing intergroup conflict. Organizational structures merely determine group interdependencies. The interaction occurs between people. We propose that people's skills in managing intergroup relations, though not their personalities (Greenhalgh, Neslin, and Gilkey, 1985), make a significant contribution to the effectiveness of any particular structural configuration. Furthermore, we propose that negotiation skills can, to some extent, compensate for structure. In this section we discuss various structural techniques for managing intergroup conflict and propose the type of skills necessary to manage intergroup relations effectively in each.

Hierarchy and Negotiations. In organizational theory, intergroup relationships have traditionally been controlled by hierarchy and legitimate authority (McCann and Galbraith, 1981). Conflict between groups is managed by appealing to a common superior. Hierarchy, however, has important shortcomings as a conflict management device. Hierarchy becomes overloaded in dynamic environments when spans of control are too broad or when decision making is centralized with top management (McCann and Galbraith, 1981). The effect of an overloaded hierarchy is slow response to situations that require quick decisions. The other problem with hierarchy is poor-quality decisions about the terms of an intergroup exchange setting up intragroup constraints or coordinated action. There are several reasons why hierarchical decisions may be poor. First, hierarchical leaders may be cognizant of the organization's superordinate goals but lack detailed knowledge and understanding of each group's interests on a particular interface issue. Second, hierarchical managers, either because they are already overloaded or because they perceive that their role in conflict management is to make decisions, fail to investigate conflict issues thoroughly. Sheppard (1983) found that managers making decisions to control conflict act like inquisitors. They investigate the issue by asking questions of their own formulation and then impose their own power-based solution. Kolb writes about Rogers in similar terms: “Rogers' role definition is shaped by his position as manager of the division. Indeed, he reaffirms his authority when he decisively solves the problems. His subordinates contribute to this definition. They complain to him about their colleagues. They ask him to order people to perform particular tasks. They ask him to clarify roles, responsibilities, and objectives. As one manager said, ‘If he would tell us which objective—the product or the methodology—was more important, there wouldn't be any problems.’ They look to him to resolve the situation, which he does. However, in doing so, he seems to create a dependence on him that limits the ability of his managers to deal with conflict and ambiguity in the future” (Kolb, 1986, p. 26).

Rogers, when he is not reorganizing, makes unilateral decisions that resolve the question of which objective is more important on an issue-by-issue basis. The tension between product orientation and method orientation continues to generate “problems” that get sent up the hierarchy to Rogers.

We propose that three other conflict resolution models associated with a hierarchical structure will be superior to this style found by Sheppard (1983) and Kolb (1986). The first alternative is for the hierarchical manager to act as a judge and require that each group make a presentation outlining positions, interests, and potential settlements. In this way the hierarchical manager should be able to make a better-informed decision, for it behooves each group to explain why the other group's position is untenable.

The second alternative is mediation. The manager acting as a mediator will try to facilitate an agreement between conflicting groups. Techniques may include getting the group representatives to recognize their own group's and the other group's interests with respect to the interface issue; getting the groups to generate proposals that package issues so that some of each group's interests are met; and suggesting that a particular proposal be tried on an experimental or nonprecedent-setting basis, putting pressure on powerful groups to look to the long term and “leave a little money on the table” and putting pressure on nonpowerful groups to withdraw. (See Carnevale, 1986, for a theoretical and Shapiro, Drieghe, and Brett, 1985, for an empirical discussion of what effective mediators do.)

Recent research (Brett and Goldberg, 1983; Pearson, 1982) shows that mediation is effective in high-conflict situations in which the disputing parties are also highly interdependent and thus have substantial common interests as well as interests in conflict. Brett and Goldberg (1983) report that 80 percent of grievances between labor and management were settled in one form or another after mediation. The “one form or another” settlements turn out to be an important aspect of mediation. Unlike hierarchical decisions that frequently split the difference between disputing parties, mediation can result in acquiescence by one or the other party, a purely integrative compromise, or a split-the-difference settlement. Pearson's research points out another important aspect of mediation. It appears to be more successful in developing enduring settlements than the judicial style of dispute resolution. Pearson found recidivism rates to be significantly lower when child custody agreements were reached through mediation than when such agreements were imposed by the court. Low recidivism translates into a high quality of solution according to our criteria of effectiveness. An additional potential benefit of mediation is that subordinates might learn some negotiation skills. Mediators use many of the same techniques used by negotiators intent on solving the problems. Advocates of mediation make the social learning theory argument that disputants can learn negotiation skills by modeling the mediator-manager. While there is plenty of research to support social learning theory, we know of no research that has studied the impact of mediation on disputants' negotiation skills.

The third alternative decision-making style that should be superior to hierarchy is direct lateral negotiation. The manager using this approach to conflict management pushes the conflict back down the hierarchy to the groups that are experiencing it. The hierarchical manager refuses to resolve conflict between subordinates but instead makes it clear that group representatives are the best ones to resolve the matter and must do so. The effectiveness of this approach will depend on the group representatives' negotiation skills. Not all group representatives will be skilled negotiators, and even those who are skilled may have difficulty seeing beyond their group's position to a creative solution or putting an interface issue into the context of the group's long-term relationship. Direct lateral negotiation may be particularly difficult when power is unequally distributed between groups. In this situation, the powerful group may adopt a tough, exploitative negotiation strategy and the low-power party may adopt a soft strategy (Pruitt, 1981). The result of such a negotiation is likely to be a distributive settlement that favors the powerful party to an extent greater than one should expect, given its power. The low-power party, however, is likely to be dissatisfied with the settlement and, thus, the settlement may not survive implementation.

Another problem with direct lateral negotiations is that groups may negotiate exchanges that are optimal for them but distinctly suboptimal for the organization as a whole. Direct lateral negotiation is an extremely decentralized mode of conflict resolution (Hage, 1983). Without control systems such as an appraisal system that rewards integrative intergroup exchanges, direct lateral negotiations may be difficult to manage.

Which of these hierarchical conflict resolution models should a manager choose—inquisition, judicial, mediation, or direct lateral negotiation? The answer depends on a number of factors, including efficiency and quality of the solutions generated by each approach, the skills of the parties, the power distribution between groups, and the organization's culture.

The inquisition model is likely to incur the lowest negotiation costs, since it will take less time than the others. It is difficult to compare the judicial, mediation, and direct lateral negotiation alternatives in terms of negotiation costs. In the dispute resolution literature, judicial proceedings and mediation are used only after parties have reached an impasse. This need not be so in the management context. Rogers's behavior, for example, discouraged subordinates from engaging in direct lateral negotiations.

The important issue from our perspective is the relationship between negotiation costs or efficiency and the quality of the solution. Here the dispute resolution literature is more helpful. Disputants may prefer direct lateral negotiations to any third-party mode of dispute resolution. Among third-party modes of dispute resolution, disputants seem to prefer mediation to arbitration (Brett and Goldberg, 1983) and arbitration to inquisition (but see Sheppard, 1985). The reason for this preference ordering has been the subject of a scholarly dispute (Folger, 1986). However, Brett and Shapiro's (1985) recent data show clearly that the reason for the preference ordering appears to be satisfaction with and control over outcome. We note that both of these factors have been identified in the organizational change literature as factors contributing to commitment to organizational change and ease in the implementation of change. Thus, we propose that mediation and judicial models may be superior to the inquisition model of dispute resolution typically used by managers, because they generate higher-quality exchanges.

When parties are not skilled negotiators, third-party modes of dispute resolution may be more effective than direct lateral negotiations. Direct lateral negotiations may result in impasses that increase groups' adherence to positions. They may result in poor-quality exchanges when power is unevenly distributed. Alternatively, direct lateral negotiations may result in exchanges that are optimal for the two groups but distinctly suboptimal for the organization as a whole.

Finally, some of these modes of conflict management within a hierarchy may be more appropriate to the organization's culture than others. Richard Walton (personal communication, 1985) illustrated this point nicely in telling us about his experience doing third-party workshops to aid in managing border disputes in the Horn of Africa. He noted that the Somalis preferred direct dialogue between equals, because this dispute resolution mode was dominant in their tribal society. The Ethiopians preferred appeals to an authority or elder, since this was the dominant mode of dispute resolution in their society. Organizations whose cultures are tied to traditional hierarchical control mechanisms may only be able to move from an inquisition to a judicial model. But even that change in approach may have significant impact on the ease with which a coordinated action or an intragroup constraint is implemented. An organization that is trying to rely on cultural norms, rather than hierarchy, for control might be able to use the mediation or negotiation mode, if managers have these skills. For example, Walton and Hackman in Chapter Five discuss the role of groups in high-commitment, as opposed to high-control organizations. While they do not discuss intergroup relationships and conflict resolution, an inquisition or even a judicial mode of conflict management would be inconsistent with the way hierarchical management in high-commitment organizations deals with groups.

Lateral Relations and Negotiations. The structural response to an overloaded hierarchy is to institute formal lateral control structures, such as liaison roles, task forces, teams, integrating departments, and matrices (Galbraith, 1977). These structures add a set of lateral interface roles to the organization in addition to those already in the organization's hierarchical structure. The major issue concerning lateral structures in organizational theory is whether or not they can be controlled (McCann and Galbraith, 1981; Pfeffer, 1981). As in direct lateral negotiations, there are two aspects to this issue. First, there is the problem that agreements between groups negotiated at low levels in the organization may be optimal for the groups at the interface but suboptimal for the organization. Second, a structural provision for low-level lateral contact may result in impasses, not settlements. In both situations, subsequent hierarchical management costs must be added to the costs of the lateral structure.

We propose that lateral structures will operate more effectively in managing conflict if the structures are staffed with properly skilled people and if the organization's reward system contingently rewards effective (as previously defined) settlements.

The organizational theory literature treats all lateral structures similarly. We distinguish between temporary structures and permanent ones. The distinction is based on how the person's performance in the lateral structure is evaluated. In temporary structures like task forces, members of the task force are representatives of primary groups. Their performance is evaluated by their primary group manager based on how well the task force's settlement (solution) meets the primary group's interests. In permanent structures (for example, a matrix) the manager's performance is evaluated by two direct superiors and ideally also by the third-level boss. The matrix manager should be evaluated based on our quality criterion, his or her success in facilitating lasting settlements of interface issues.

It is a bit difficult to classify various lateral structures as temporary or permanent, because the classification depends on the performance evaluation relationship. In general, liaison roles and task forces are temporary structures. Teams,¹ integrating departments, ombudsmen, and matrices are permanent. Task force leaders, if they are temporarily pulled from their primary group and assigned to a vice-president for the duration of the task force, should be considered as permanent structures though the members may have only temporary assignments.

Temporary lateral structures. Temporary lateral structures are used as communication vehicles to transform intergroup conflict into intragroup constraints or coordinated action. The roles of group representatives—for example, on a task force—are to represent the preferences and provide the skills of their primary group. Control over performance appraisal lies with the primary group manager, not the task force leader. Politically wise task force members know they will have to justify any settlement to their primary group. As a result, temporary lateral structures tend to stimulate a positional approach to negotiations. There are numerous terms used to identify different negotiation strategies. Positional negotiations, as described by Fisher and Ury (1981), are conceptually similar to distributive negotiations, as described by Walton and McKersie (1965). We prefer Fisher and Ury's term because we believe the term distributive connotes a single-issue negotiation. We contrast positional negotiating with a problem-solving approach. Problem solving is called principled negotiations by Fisher and Ury (1981), and integrative by Walton and McKersie (1965). We use the term integrative interchangeably with problem solving. We do not use the term principled because of its ethical connotation.

We note that unless one or the other group completely capitulates on an interface issue, settlements achieved through either positional or problem-solving negotiations that transform intergroup conflict into intragroup constraints or coordinated action are all compromise settlements. A manager assigned to a task force is likely to stake out a position on the conflict issues that is favorable to his or her group and only move from it when more powerful groups exert other power. The task force member may work to build coalitions with representatives of other groups but is unlikely to enter into problem-solving discussions, because such discussions are time-consuming and a creative, integrative solution would probably be more difficult to sell to the primary group than a solution, albeit a compromise solution, which is a variant of the group's original preferences.

Task forces provide structural opportunities for powerful groups to exploit less powerful groups in ways that may not be consistent with overall organizational strategy. We have three suggestions for controlling temporary lateral structures. First, in order to minimize exploitation, we think it is important for group representatives to task forces to recognize that intergroup negotiations have simply been transformed into intra–task force negotiations. Regardless of the language used in creating the task force, group representatives who do not approach task force deliberations as a negotiation are setting their group up to be exploited. Groups should select representatives to task forces who can expertly present and defend the group's interests. This task requires individuals who have both positional and problem-solving negotiation skills.

Secondly, in order to reduce the likelihood of impasses, task force members need to be motivated to complete the task force's mission. The most direct way of accomplishing this is to make the achievement of a high-quality settlement a central element of each task force member's performance appraisal. Ideally, the hierarchical manager, who would have to manage the conflict if the task force were unsuccessful, should be involved in each task force member's performance appraisal. Contingent rewards and visibility are two factors contributing to a problem-solving mode of task force deliberation. Another factor may be the task itself. Hackman's (1984) model of group decision making emphasizes the fact that decision making should be enhanced when the decision-making task itself has motivational properties. Task forces that provide opportunities for personal skill development or challenges unavailable in normal humdrum jobs are also likely to be motivating and cause group representatives to look beyond the parochial interests of their own groups to the broader interface issues.

If motivation factors are not sufficient to stimulate the task force to move from a positional to a problem-solving approach, it may be necessary for groups to actually instruct their representatives to take a problem-solving approach to task force deliberations. There is support for this proposition in the research on representative-constituency relationships (Benton and Druckman, 1974). In general, negotiators representing constituencies take a tougher positional approach to negotiations than negotiators who do not represent constituencies even when the constituency gives the representative no instructions. Instructions from constituencies to take a problem-solving approach do have an effect on negotiator behavior and settlements. The problem with extrapolating from this research to the organizational setting is that the research was done in the laboratory and the leaders were elected or appointed. Instructions to take a problem-solving approach may have no effect if the group representative also holds a position of hierarchical authority in the group.

Permanent lateral structures. Integrating roles, matrices, and ombudsmen (see Kolb, 1986, for a discussion of the latter) are permanent intergroup interface positions.² They are an expensive structural alternative, since they are full-time conflict management positions, but they may also be cost-effective in terms of long-term settlements. “Reorganize it Rogers,” our example of a hierarchical manager, was managing a matrix. The degree of conflict in Rogers's matrix illustrates that permanent lateral structures, when they do not operate properly, can add significantly to transaction costs. Kolb does not tell us much about the lower-level matrix managers in her case, other than that they were engineers and that their method for managing intergroup conflict was to push it up the hierarchy. This conflict management style violates the whole concept of the matrix. Obviously, design is not enough. What else is necessary to make permanent lateral structures operate as intended? We believe organizational systems that demand and reward high-quality intergroup settlements and role incumbents who have good negotiation skills are essential for making these lateral structures effective.

We have already discussed reward systems in some detail. Incumbents in permanent lateral structures should be evaluated on the quality of the settlements of interface issues that they negotiate. One factor that makes the tasks particularly difficult is that group representatives, with whom a matrix manager or ombudsman interacts, are evaluated on how well their groups' interests are maintained in an interface settlement, not on integrative criteria. Furthermore, incumbents in permanent lateral roles have interests of their own. High-quality interface solutions may be the means for achieving an outcome, but in many organizations the outcome is the most visible criterion and is what is rewarded. As a result, it may be difficult for incumbents in permanent lateral structures to distinguish effective and ineffective means.

We propose that problem solving is a more effective style of negotiation than a positional or distributive approach in permanent lateral structures. These structures seldom have the formal authority (power) that is necessary to successfully negotiate a position. For example, product managers in marketing departments (an integrating structure) frequently have no budgetary control. Their power comes from their ability to absorb uncertainty by predicting or managing markets and to manage conflict with quality interface solutions. Quality interface solutions either on their face or over the long run are most likely to be those arrived at through integrative negotiation.

Do managers have integrative bargaining skills? Some obviously do, but even those who do may lack a structure or framework for planning a negotiation. The current popularity of negotiation training courses in master of business administration (M.B.A.) and executive programs suggests that these are skills managers want to acquire.

The most accessible description of the problem-solving approach to negotiation is in Fisher and Ury (1981). They suggest: (1) focus on interests, not positions; (2) separate the people from the problem; (3) invent options for mutual gain; (4) invent objective criteria; and (5) know your Best Alternative to a Negotiated Agreement (BATNA). The Fisher-Ury principles provide a useful guide in preparing for negotiations. The negotiator intent on approaching the conflict from a problem-solving perspective first anticipates what the groups' positions are likely to be on the interface issues. The next step is to think through the reasons underlying those positions and consider possible hidden agendas. Underlying reasons or hidden agendas are called interests by Fisher and Ury (1981). Then analyze what will happen if no settlement is reached. Will the groups continue the status quo, interacting on the issue using old procedures? Will the conflict get escalated up the hierarchy? This analysis step is called BATNA. An assessment of groups' BATNAs is really an assessment of their power. The group with the better BATNA is the more powerful group. The reasoning goes like this: A's power is a function of B's dependency on A, and B's power is a function of A's dependency on B. If A is relatively less dependent on B—that is, has a better BATNA—than B is on A, then A has relatively more power than B.

These three steps—positions, interests, and BATNAs—provide the information necessary to analyze whether or not there is likely to be an area of agreement for the groups. The analysis of interests is particularly useful in searching for alternative settlements that maximize both groups' interests, regardless of their original positions. All of this analysis is done before the negotiation. This means that chance meetings in the hall, on an elevator, or by a coffeepot are not appropriate for negotiations, unless one is fully prepared. Such meetings might be appropriate for checking out anticipated interests or exploring BATNAs, but a problem-solving negotiation session needs to be a serious joint attempt to develop a mutually beneficial settlement. The chances of negotiating an agreement that a group representative can successfully sell to his or her primary group are better when that representative was fully involved in the deliberations.

Incumbents in permanent lateral roles may also frequently find themselves in a mediation role between conflicting groups. Recent research (Kolb, 1986; Shapiro, Drieghe, and Brett, 1985) has demonstrated that successful mediators are not artists who approach each conflict uniquely, but systematic practitioners of a style. Shapiro, Drieghe, and Brett (1985) found, further, that several different styles are effective in getting settlements.

Managers have much to learn from mediators. Mediators use the Fisher-Ury techniques, particularly the one about separating the people from the problem. Many mediators literally separate the opposing groups and shuttle back and forth between them. This works very well when there are emotional tensions between people that are interfering with the negotiation. It is not, however, necessary in all situations.

Other mediator techniques make settlements more palatable. Three of these are as follows: (1) add issues to the conflict so that every group can get something out of the settlements, (2) get the groups to agree that a particular settlement is a onetime agreement and cannot stand as a precedent, and (3) get the groups to try a settlement for an experimental period. Another technique used in mediating multi-issue conflicts is the one-text procedure (see Fisher and Ury, 1981; also Carter, 1982). In this procedure, the mediator provides a settlement proposal written in simple language that focuses on the primary issues. The mediator then shuttles back and forth between the parties, asking them to discuss the proposal. For example, “What might you be willing to do to change the undesirable clause?”, and so on.

We are more confident of our prescriptions for mediation techniques for incumbents of permanent integrating roles than of the problem-solving negotiating style. We know successful mediators use these techniques (Shapiro, Drieghe, and Brett, 1985).³ There is no research contrasting the problem-solving and positional styles of negotiation. A major problem with such research would be to get advocates of both styles to agree on effectiveness criteria. Given the effectiveness criteria developed in this chapter, we propose that the problem-solving approach is more likely to be effective.

Summary and Research Issues. This section integrates the organizational theory and negotiations literature in a series of propositions about the effective management of intergroup conflict. The main points are as follows:

• In hierarchical structures, superiors need mediation skills, subordinates need negotiation skills.
• In temporary lateral structures (for example, liaisons, task forces, or teams), when performance is evaluated in the primary group and not in the task force or team, group representatives may use positional negotiation skills.
• In permanent lateral structures (for example, ombudsmen, task force, or team leaders, integrating departments or matrices), incumbents need problem-solving negotiation skills and mediation skills.
• The balance of power between the groups, the organization's culture, and the reward/incentive system must be considered when designing a management form for an intergroup relationship.

There are numerous research issues identified in this section. A major challenge is to test the effectiveness of various negotiation (positional as opposed to problem-solving) and third-party approaches (inquisition, judicial, mediation) in different intergroup structural arrangements and under different distributions of power and organizational cultures. In this chapter we propose that in general in organizations, (1) problem-solving is more effective than positional negotiation and (2) third-party procedures that place control over the interface exchange with the interfacing groups will be most effective. But we caution that under conditions of unbalanced power, or when the organization's culture is tied to hierarchical control, direct lateral negotiations and mediation may not be effective.

A second challenge is to determine whether problem-solving negotiation skills actually affect the quality of intergroup exchanges. Lawrence and Lorsch (1967) found that a “confrontation style” of decision making was crucial for interdepartmental effectiveness. Can such necessary confrontations be implemented more effectively in a problem-solving rather than a positional style? Lawrence and Lorsch's research does not tell us. The framework developed here suggests researchers should examine skill, structure, power, and the organization's culture and its incentive system in studying the causes of effective intergroup relations.

It is interesting to speculate about the substitutability of skill for structure. Since lateral structures are permanent management devices, they are costly. The structure should be chosen by evaluating the interdependence between the groups, environmental uncertainty, and the negotiation skills of managers. Lateral structures required by the environment, but staffed by unskilled managers, will only increase transaction costs. On the other hand, managers skilled in negotiations may be able to cope with less hierarchy and few permanent lateral structures, holding environmental dynamism and interdependence constant.

One research issue that has not been fully discussed is the long-term viability of an organization that manages conflict via the techniques discussed in this chapter. In periods of dramatic environmental change, organizations must survive revolutions to remain viable. The approaches to managing conflict discussed in this section are not appropriate for revolutionary change during which the organization questions its fundamental operating assumptions. We think the approaches to managing conflict discussed here are appropriate under conditions of moderate change. All the techniques discussed recognize a multiplicity of interests, all recognize that conflict is legitimate, and all are focused on lasting solutions. While it is possible that a solution will outlast its viability in the environment, this seems less likely to occur when the situation requires the concurrence of groups that will be differentially affected by environmental change.

Conclusion

The dominant approach to examining intergroup relations has been structural (for example, Galbraith, 1977). But as Strauss (1978) points out, many structural theories have negotiation at the heart of their framework without examining the process of negotiation itself. To understand the process of intergroup relations in organizations, structural considerations must be linked to the process of negotiation. Different structural arrangements allow for different negotiation models. The weak link is that structural interfaces are enacted by group representatives with different degrees of negotiation skills. In this chapter it is argued that research integrating the structural approaches from organization theory with negotiation theory is the most fruitful avenue to pursue for intergroup research.

In presenting our ideas for this integration of theory, we have drawn no causal models, stated no formal hypotheses, proposed no formal research, and presented no computerized literature review. We have only written a theoretical statement about intergroup relations: why they are conflictive, how they can be evaluated, and how they can be managed. We have focused on intergroup conflict, though we know many managers talk about intergroup problems, not conflict. We have also emphasized the conflictive aspect of intergroup relations, though we recognize that for an intergroup interface to be effective, the majority of the interface time should be consumed in coordinated intergroup action or on constraints on intragroup behavior.

We have also focused on negotiation skills, though there is no research showing that successful managers act as we prescribe. Certainly Kipnis, Schmidt, and Wilkinson's (1980) research that reports that managers use power when they have it and Sheppard's (1983) research showing that managers deal with conflict like inquisitors are discouraging. Yet why should we expect managers to have negotiation skills honed as finely as their financial and market analysis skills or even their skills in managing people? Most people learn negotiating skills in the course of everyday life. They take a problem-solving approach with their families and a positional approach with their car dealers. The problems come in the transfer of skills from the everyday to the organizational setting. It is not legitimate to treat your counterpart in another group the way you do your spouse or a car dealer. So what do you do? Few managers have models, few have had formal training. Most managers are unprepared to manage intergroup conflict effectively. So while the research task of confirming or disconfirming our theory of managing conflict between groups is formidable, the training task, assuming a confirmed theory, is greater.

References

Aiken, M., and Hage, J. “Organizational Independence and Intraorganizational Structure.” American Sociological Review, 1968, 33, 912–930.

Bacharach, S. B., and Lawler, E. J. Bargaining: Power, Tactics, and Outcomes. San Francisco: Jossey-Bass, 1981.

Benton, A. A., and Druckman, D. “Constituents' Bargaining Orientation and Intergroup Negotiations.” Journal of Applied Psychology, 1974, 4, 141–150.

Bok, D. C., and Dunlop, J. T. Labor and the American Community. New York: Simon & Schuster, 1970.

Brett, J. M. “Managing Organizational Conflict.” Journal of Professional Psychology, 1984, 15, 664–678.

Brett, J. M., and Goldberg, S. B. “Wildcat Strikes in the Bituminous Coal Mining Industry.” Industrial and Labor Relations Review, 1979, 32, 467–483.

Brett, J. M., and Goldberg, S. B. “Grievance Mediation in the Coal Industry: A Field Experiment.” Industrial and Labor Relations Review, 1983, 37, 49–69.

Brett, J. M., and Shapiro, D. “Procedural Justice: A Test of Two Theories.” Working paper, Kellogg Graduate School of Management, Northwestern University, 1985.

Brown, L. D. Managing Conflict at Organizational Interfaces. Reading, Mass.: Addison-Wesley, 1983.

Carnevale, P. “Strategic Choice by Third Parties: A Theory of Dispute Resolution.” In R. J. Lewicki, B. H. Sheppard, and M. H. Bazerman (eds.), Research on Negotiations in Organizations. Greenwich, Conn.: JAI Press, 1986.

Carter, J. Keeping the Faith: Memoirs of a President. New York: Bantam Books, 1982.

Dale, A. “Interface Issues.” In R. Payne and C. Cooper (eds.), Groups at Work. Chichester, England: Wiley, 1981.

Emerson, R. M. “Power Dependence Relations.” American Sociological Review, 1962, 27, 31–41.

Fisher, R., and Ury, W. Getting to Yes. Boston: Houghton Mifflin, 1981.

Folger, R. “Process and Outcome in Dispute Resolution.” In R. J. Lewicki, B. H. Sheppard, and M. H. Bazerman (eds.), Research on Negotiations in Organizations. Greenwich, Conn.: JAI Press, 1986.

Galbraith, J. R. Organization Design. Reading, Mass.: Addison-Wesley, 1977.

Gladstein, D. “Groups in Context: A Model of Task Group Effectiveness.” Administrative Science Quarterly, 1984, 29, 499–517.

Greenhalgh, L., Neslin, S., and Gilkey, R. “The Effects of Negotiator Preferences, Situational Power and Negotiator Personality on Outcomes of Business Negotiations.” Academy of Management Journal, 1985, 25, 9–33.

Hackman, J. R. “A Normative Model of Work Group Effectiveness.” Technical Report No. 2, Program on Group Effectiveness, Office of Naval Research, 1984.

Hage, J. “Communication and Coordination.” In S. M. Shortell and A. K. Kalunzy (eds.), Health Care Management. New York: Wiley, 1983.

Jennergren, L. P. “Decentralization in Organizations.” In P. C. Nystrom and W. H. Starbuck (eds.), Handbook of Organizational Design. New York: Oxford University Press, 1979.

Kipnis, D., Schmidt, S., and Wilkinson, I. “Intraorganizational Influence Tactics: Explorations in Getting One's Way.” Journal of Applied Psychology, 1980, 65, 440–452.

Kochan, T., and Bazerman, M. H. “Macro Determinants of the Future of the Study of Negotiations in Organizations.” In R. J. Lewicki, B. H. Sheppard, and M. H. Bazerman (eds.), Research on Negotiations in Organizations. Greenwich, Conn.: JAI Press, 1986.

Kolb, D. “Who Are Organizational Third Parties and What Do They Do?” In R. J. Lewicki, B. H. Sheppard, and M. H. Bazerman (eds.), Research on Negotiations in Organizations. Greenwich, Conn.: JAI Press, 1986.

Lawrence, P. R., and Lorsch, J. W. Organization and Environment. Boston: Graduate School of Business Administration, Harvard University, 1967.

McCann, J., and Galbraith, J. R. “Interdepartmental Relations.” In P. C. Nystrom and W. H. Starbuck (eds.), Handbook of Organizational Design. New York: Oxford University Press, 1981.

March, J. G., and Simon, H. A. Organizations. New York: Wiley, 1958.

Mintzberg, H. The Structuring of Organizations. Englewood Cliffs, N. J.: Prentice-Hall, 1979.

Pearson, J. “An Evaluation of Alternatives to Court Adjudication.” Justice System Journal, 1982, 7, 420–444.

Pfeffer, J. Power in Organizations. Marshfield, Mass.: Pitman, 1981.

Pruitt, D. G. Negotiation Behavior. Orlando, Fla.: Academic Press, 1981.

Rubin, S. B., and Brown, B. R. The Social Psychology of Bargaining and Negotiations. New York: Academic Press, 1975.

Shapiro, D., Drieghe, R., and Brett, J. M. “Mediator Behavior and the Outcome of Mediation.” Journal of Social Issues, 1985, 41 (2), 101–114.

Sheppard, B. H. “Managers as Inquisitors: Some Lessons from the Law.” In M. H. Bazerman and R. J. Lewicki (eds.), Negotiating in Organizations. Beverly Hills, Calif.: Sage, 1983.

Sheppard, B. H. “Justice Is No Simple Matter: A Case for Elaborating Our Model of Procedural Fairness.” Journal of Personality and Social Psychology, 1985, 49, 1–10.

Sherif, M. In Common Predicament: Social Psychology of Inter-group Conflict and Cooperation. Boston: Houghton Mifflin, 1966.

Stephenson, G. M. “Intergroup Bargaining and Negotiation.” In J. C. Turner and H. Giles (eds.), Intergroup Behavior. Oxford, England: Basil Blackwell, 1981.

Stern, L. W., and El-Ansary, A. I. Marketing Channels. Englewood Cliffs, N.J.: Prentice-Hall, 1977.

Stinchcombe, A. “Contracts as Hierarchical Documents.” Working paper, Stanford University/Department of Sociology, Northwestern University, 1983.

Strauss, A. Negotiations: Varieties, Context, Processes, and Social Order. San Francisco: Jossey-Bass, 1978.

Walton, R., and McKersie, R. A. A Behavioral Theory of Labor Negotiations. New York: McGraw-Hill, 1965.

Weick, K. E. The Social Psychology of Organizing. Reading, Mass.: Addison-Wesley, 1969.

Weick, K. E. The Social Psychology of Organizing. (2nd ed.) Reading, Mass.: Addison-Wesley, 1979.

Williamson, O. E. Markets and Hierarchies. New York: Free Press, 1975.

Williamson, O. E. “Transaction Cost Economics: The Governance of Contractual Relations.” Journal of Law and Economics, 1979, 22, 233–261.