Departures in Negotiation: Extensions and New Directions

In my earlier work, I developed a framework for analyzing turning points or critical moments in negotiation that consists of three parts: precipitants, departures, and consequences. Precipitants are the factors inside or outside a negotiation that lead to or trigger change. The three types of precipitants are procedural, decisions made to change the structure or format of the talks; substantive, new ideas or concepts introduced by one or more of the parties; and external, events that occur outside of the talks, including both policy and leadership changes or third-party interventions.

Departures follow the precipitants. Whereas precipitants are suggestions, ideas, or events that occur, departures are the decisions then made by a party or parties to accept or reject the proposed changes as well as the transitions that occur from one stage to another. Departures can be more or less abrupt in terms of their suddenness.

Consequences refer to either progress toward agreements (deescalatory) or movement away from agreements toward impasses (escalatory). They follow immediately from departures. The departure is the reaction to a precipitant, whereas the consequence is the direction of the talks toward or away from positive outcomes as a result of the departure (Druckman 2001).

Departures are indicated by events or decisions that occur during the course of a negotiation. The events or decisions are part of the substance of the talks. Focusing on them enables us to gain observable or measurable indicators of the turning-points concept, facilitating comparative analysis. However, if we expand that analysis to probe the psychological and social dynamics that may occur in conjunction with departures, we could increase our understanding of departures by addressing such fundamental questions as: How do negotiators recognize a departure when it occurs and what actions do negotiators take to bring them about?

First we identify the common features of departures and then inquire about the processes that occur at the individual (psychological) and interactive (social) levels to invoke these features of departures. We pose the question: What are the psychological and social factors that precede and accompany departures? This sets the stage for a review of relevant research that sheds light on how these processes may operate during the course of a negotiation. We conclude with a discussion that considers negotiation as a collectivity or system of actors and identifies factors that influence the system.

To get the discussion started, let us look at three examples that illustrate how the framework analysis applies to a variety of negotiation scenarios: an extended international trade negotiation, the United States–Russia relationship, and a baseball player–management conflict that captured headlines in the sports pages. All three cases share some common features of departures that will be highlighted later in the essay.

Interactions between the U.S. and China over intellectual property rights (IPR) during the Clinton administration showed how a policy maker’s active strategies produced consequences for an important bilateral relationship (for details, see Hulse and Sebenious 2003). Several prior IPR agreements negotiated between the two countries had failed to stop the piracy of U.S. intellectual property. The option of threatening China with trade sanctions was risky due to the lack of unified U.S. domestic support. U.S. Trade Representative Charlene Barshefsky was faced with the challenge of balancing the conflicting interests of a business community and an administration that benefited from smooth relations with China against those of the intellectual property industries that were being harmed by piracy. Knowing this, Barshefsky moved forward toward achieving the goal of an agreement that would work. Her actions can be captured by turning points that occurred in each of three phases: prenegotiation, negotiation, and implementation.

Barshefsky’s early actions were intended to set the stage for negotiation. She worked with the administration to improve the U.S. relationship with China and, then, cleared a legal hurdle that provided the basis for negotiation. She used a “rule of law” principle to address the Chinese trade issue. This principle placed the issue in the narrower context of economics rather than politics.

“Framing both sanctions and the IPR agreement as part of the same issue-area with an easily comprehensible and powerful proportionality argument added to its appeal,” according to Rebecca Hulse and James Sebenius (2003: 321). The tactic helped Barshefsky gain the support of the U.S. business community.

Negotiation developed from three turning points judged to have occurred in this early stage: an executive order to place conditions on China’s most-favored-nation (MFN) status renewal in May 1993 (an abrupt departure); dropping of the conditional MFN policy in June 1994 (an abrupt departure); and designating China as a country subject to a special provision that assured it of fair trade in the summer of 1994 (a nonabrupt departure following from prior decisions).

Barshefsky’s prenegotiation strategy was effective in getting the Chinese team to the table. China agreed to enter negotiations in mid-1994. The negotiation itself is depicted as a turning point (a nonabrupt departure). After eight months of on-again/off-again talks, Barshefsky had an IPR agreement.

Implementation problems surfaced quickly. However, Barshefsky’s active “no-nonsense” approach combined with her effective tactics to persuade the Chinese policy makers to adhere to the terms of the agreement. (Particularly notable in this regard was her sensitivity to the importance of face saving for the Chinese negotiator.) Two turning points during this period led to positive consequences for the IPR issues and for the relationship: a threat was made to key factories in the Chinese textile industry that they would pay a significant price if they did not stop their illegal IPR operations (abrupt departure); and a refusal to accept the Chinese leader’s invitation to consult with him, stating that to do so would put him in an embarrassing position (abrupt departure).

Taken together, the six turning points identified in this extended case result in the following path:

Inside precipitant (three procedural, two substantive)→abrupt departure (in five of the six departures)→escalatory or de-escalatory consequence at t (three immediate consequences were escalatory, three were de-escalatory)→de-escalatory consequence at t+1 (a later consequence).

This case is similar to the ten trade cases analyzed in my previous work (Druckman 2001). Frequent inside precipitants and abrupt departures occurred in each of those talks, which were mostly organized conferences, and in this less-bounded case. However, this case had more frequent immediate escalatory consequences, reflecting perhaps the difference between bounded conferences and the extended multilevel domestic and international interactions typical of this case.

The ease with which the turning-points framework can be applied is also demonstrated by an examination of events that do not involve formal or official negotiations. Although a number of examples can be given, two articles in the October 4, 2001 issue of the Washington Post caught my attention. One appeared in the news section, the other in the sports section. Both referred to the events as turning points.

The article in the news section was titled “U.S. — Russia Recast Their Relationship” and noted that “a turning point may have occurred when Deputy Secretary of State Richard L. Armitage met Sept. 19 with Russian Foreign Minister Vyacheslav Trubnikov in Moscow.” Indeed, relationship changes have occurred that are maintained to this day.

The article in the Sports section titled “Schottenheimer Reacts, With Changes,” noted that the players “indicated that they were hopeful that a clear-the-air meeting between coaches and players on Monday will be a turning point in the season.”

The impetus for change in these two examples is different. In explaining the notion of a tipping point, Malcolm Gladwell (2000) uses the illustration of Paul Revere’s ride to sound the alarm that “The British are coming,” prior to the U.S. Revolutionary War to indicate how a piece of news can spread quickly through a population. Similar to Gladwell’s (2000) idea, the U.S. and Russia were shocked into action by September 11. The terrorist attack was the precipitant to a departure in process, in this case the way the two nations conducted their relationship.

The Washington Redskins’ coaches and players meeting was a response to a need for change dramatized by a series of losses. The string of lopsided outcomes was the precipitant to the meeting which was a departure from usual procedures in the sense of providing both the coaches and players with a kind of therapeutic encounter that was unique to their experience. After all, something had to be done to change the team’s performance and record.

While the deputy foreign ministers’ meeting was construed by journalists as a turning point, in fact, this meeting was probably only one of several indicators that the relationship between the two nations was changing. The meeting did secure Russian support for stationing the U.S. troops in the former Central Asian republics in order to deal with the American war against terrorism. Other decisions were made in conversations between Presidents Bush and Putin. These included an agreement to share intelligence, to provide an air corridor for humanitarian missions, and a re-framing of the Chechen issue by linking it to Bush’s antiterrorism campaign.

Add to these decisions a noticeable change in the tone of the communications and it may be possible to conceive of these changes as signaling a “real end to the cold war,” as noted by Carnegie Endowment scholar and Russian expert Michael A. McFaul. Developments since these initial meetings suggest that the changes may hold over the long term. However, whether they lead to a re-shaping of the discussion over other divisive issues between these countries, such as antimissile defenses and Russia’s debt to the West, remains to be seen.

The changes in this case occurred as a result of an external precipitant — the events of September 11, 2001. This is similar to the security cases analyzed in Druckman (2001). The path for this example is:

External precipitant (terrorist attacks)→abrupt departure (change in policies toward cooperation)→de-escalatory consequences at t (relationship changes)→de-escalatory consequences at t+1 and beyond (relationship changes have been sustained)

Similar to the international example, the football team meeting also resulted in an improved relationship. It did not, however, lead to a dramatic improvement in performance. Interestingly, while the players “were hopeful that the . . . meeting . . . [would] be a turning point in the season,” neither the players nor the journalists described the meeting as a turning point in the final analysis. Judging the meeting in terms of its short-term consequences (immediate improved play on the field), it would appear that a turning point had occurred. However, they also judged the meeting in terms of its longer-term consequences (no playoffs, jobs not saved), which called into question whether a turning point had indeed occurred.

The news article described the participants in the meeting as confronting sensitive issues and the coaches as being willing to consider making changes in the ways they communicated to the players. The head coach admitted that he made the mistake of not explaining to the players the reasons for his decisions. He vowed that he would try harder to communicate better in order to overcome the mistrust that had developed between him and the players, saying that “in some situations they would rather have a father figure than a general.”

Although, the next two games following the meeting produced less lopsided loses — the team played better — they were not victories. However, performance did move in a positive direction and, for some precipitating events, the impacts on process take time to develop. During the second half of the season, the team enjoyed a winning streak, but their earlier losses prevented them from reaching the playoffs. In the end, the coaches jobs were not saved, perhaps because the expectations of the owners and fans for their team’s performance were — and had been — quite high (long-term consequence).

These events can be depicted in terms of the following path:

Procedural precipitant (a therapeutic meeting)→nonabrupt departure (team plays better)→immediate consequence at t (less lopsided losses)→mid-term consequences at t+1 (victories)→longer-term consequences at t+2 (jobs not saved) (Note that t+1 designates the next time period; t+2 designates two time periods after t.)

The three examples we have just described suggest that negotiation departures have some common features.

Focusing attention on these three examples, what do the departures have in common? The departures in the IPR case consisted of changes in policy in the prenegotiation stage and a series of abrupt threats and reversals following the negotiated agreement. In the second case, changes in policy toward increased cooperation with Russia occurred following the terrorist attack of September 11. And, performance changes occurred following the coaches–players’ meeting.

These events have several features in common. The first is a change from earlier events or patterns, in varying degrees of suddenness or abruptness. The second is that changes occur in the relationship between the parties. A third is that the changes are clear or self evident: Observers are likely to agree that a change occurred. And, a fourth feature is an action taken usually by one of the parties that leads to consequences for both.

The key element is that change, often initiated by one of the parties, is manifest in the interaction process. This kind of process is captured quite nicely by Erving Goffman’s (1969) idea of an expression game where exchanges have implications for a relationship that continues beyond the negotiation. All of the examples discussed deal with long-term relationships.

The examples also highlight the functions of monitoring or observing and acting or influencing as noted by Goffman:

This conception emphasizes a relationship between managing impressions (encoding) and processing information (decoding). The factors affecting the encoder’s actions and the decoder’s attentiveness to those actions have been a subject of considerable research.

Because decoding and encoding are functions performed by individual negotiators, much of the research has been concentrated on this level. Studies include both cognitive (awareness, sorting out) and emotional (expressing, feeling) processes. When they are considered together in the context of an expression game, the focus is on interactions: The negotiator is in easy reach of both the decoder and encoder roles. His or her sense of being more the encoder or the decoder depends on whether he or she is persuading or analyzing during a particular episode.

The negotiation outcome results from the combination of moves made (concessions or offers made and countered) and postures taken (e.g., tough, soft, demanding, accommodating). The research on synchronizing moves, including reciprocal exchanges, posturing, influence tactics, and unilateral moves pertains to this discussion.

The research reviewed in this section is divided into three levels of analysis: changes in individuals, changes in the interaction process, and changes in the collective process. Implications of the findings at each level for the turning-point concept are developed in the form of an agenda for further research in the next section.

One type of individual change that occurs is in emotions, indicated by self reports or by observing nonverbal behaviors. There is a large literature on the role of emotion in interpersonal and intergroup relations which indicates the impact that feelings can have on how events are interpreted (Granberg and Brown 1989; Zajonc 1984), how emotional change can be a precursor to cognitive and behavioral change (Greenberg, Rice, and Elliott 1993), and the conditions under which certain moods can be induced (Druckman and Bjork 1994).

Future research should draw out the implications of these phenomena for the turning-points framework. For example, when the negotiations are highly salient to all parties, are the participants more or less sensitive to key events that can turn the interactions in a different direction? How do the feelings of the parties toward one another determine whether inside or outside precipitants will be more effective in leading to agreements? These issues call attention to such relevant aspects of the concept as awareness, connections between feelings and meaning, and precipitating factors within the interaction process.

Ongoing laboratory research on negotiation provides a glimpse into the way that emotions may influence decisions. From experiments that we have been conducting on the configuration of furniture in the negotiating room, we obtained different results in simulation and field settings. On the one hand, simulation role players obtained significantly more agreements when they negotiated in chairs without tables than when they talked behind tables. On the other hand, small claims court disputants in mediated sessions were not responsive to this difference or to other situational cues manipulated in the experiment.

An explanation for these findings — which is the basis for further experiments — is that the higher emotion of the real-world disputants made them less responsive than the role-players to visual cues in the setting. Referred to as “emotional flooding,” it may reduce attention to cues in a manner similar to cognitive overload. A question of interest is whether emotional negotiators also miss events that may signal departures.

Although emotions may reduce awareness of cues during negotiation, they seem to have a strong impact on perceptions and judgments prior to negotiating. In her recently-completed dissertation study, Berenike Carstarphen (2003) compared the impacts of personal stories (high affect condition) with factual information (low affect) about a conflict resembling the dispute in Cyprus. After receiving a description of the conflict and its history, simulation role players were given an emotionally charged story of a victim in the conflict or were provided with information about the unfolding events. In two other conditions, role players received both types of information in opposite order; either the personal story first or the factual information first. The four conditions were compared to a pretest condition in which neither type of information was presented.

The impacts of these conditions were evaluated using a variety of questions asked only during a prenegotiation session. These included perceptions about the opponent, the anticipated negotiation climate, preferred strategy, and expected outcomes. The results were generally consistent in showing that the largest changes on each of these types of questions occurred for the personal story-only condition followed, in order, by personal story first, information first, and the information-only condition. These changes were regarded as shifts and they reinforce findings also obtained by Carstarphen in field analyses of nonscripted dialogues.

The impacts of these sorts of personal stories on the negotiation process remain to be discovered. Conceivably, emotional arousal has a stronger impact on perceptions and anticipations prior to interaction than on responsiveness to cues in the actual negotiation.

Another type of change is discussed in the literature on cognitive social psychology and is part of the research on information-processing and framing. These include work on phased team learning and development (Garsick 1988); the role of feedback on team performance (Klein et al. 1992); transactive memory (Wegner 1986); illusions of comprehension due to different subjective experiences (Griffin and Ross 1991); and the development of shared mental models (Orasano and Salas 1993).

More directly related to negotiation is research on framing, including frame-breaking changes (Druckman, Husbands, and Johnston 1991; London 1985), and on the monitoring function (Druckman 1978; Winham 1977). Keeping track of developments is central in bargaining models (e.g., Coddington 1968) and essential for detecting deviations from established patterns of offers or staying alert to concessions that require adjustments (Druckman and Harris 1990).

Detecting the departures or discontinuities in the course of negotiated interactions depends on monitoring what is happening. By combining the monitoring or decoding function with the more strategic function, the negotiator can time his or her actions in order to move the talks to another stage. Together, these roles combine the skills of the analyst-decoder with those of the strategist-encoder. The interaction between the two functions is central to Goffman’s (1969) idea of the expression game discussed previously. The way this occurs in negotiation remains an important topic for further inquiry.

Another angle for thinking about information-processing in negotiation is at the level of psychophysiological processes. Relevant research in this area focuses on a wave-form in the human electroencephalogram (EEG) known as the P300 (the amplitude of the wave-form occurring roughly between 327 and 753 milliseconds in the EEG record). This wave-form has been shown to occur regularly when expectations are disappointed (Donchin 1981). It is evoked in response to low-probability events and hypothesized to precede facial movements associated with surprise (Ekman and Friesen 1975).

The connection between these processes and negotiation was made in an experiment that I conducted with Demetrios Karis and Emanuel Donchin (Druckman, Karis, and Donchin 1983). We showed that the P300 occurred in response to the other bargainer’s unexpected strategy shifts (from stingy to generous or vice versa); but it did not occur in response to the other’s expected moves. Most bargainers adjusted their own concession strategy in response to the other’s shift, but not until sometime later in the session. This suggests that subjects respond immediately at the physiological level but delay their response at the behavioral level. Awareness of the other’s strategy shift preceded adjustment to the shift.

Many precipitants of turning points have an element of surprise. An implication of the P300 research is that they may be noticed (or recorded) but not elicit a response until later. The delayed response may be the critical departure in process needed to turn a negotiation around, from impasse to progress toward an agreement. A period of reflection or “time out” may facilitate the adjustment needed for a process departure to occur.

The research discussed in this section has implications for coordination and change in negotiation. Shifts that occur within individual negotiators may not translate into change in the negotiation process. The emotional appeals that result in (or prevent) individual changes may reduce responsiveness to the other’s reactions or moves. The period of reflection needed to process new information may delay response to communication about changes that the other negotiator has experienced.

Both these implications can interfere with noticing possible critical moments as well as with the coordination needed to take advantage of them when they occur. Thus, a research challenge is to develop links between individual and interaction processes, which includes identifying the conditions when shifts at one level of analysis facilitate or hinder shifts at the other level. We turn now to changes that occur in the negotiated-interaction process.

A second level of analysis for detecting change concerns interaction or relational processes. Departures are treated here as a feature of the interactions that occur during negotiation. The research on group cohesion (Zander 1979), metacognition (Klein et al. 1992), and emergent group behavior (Sherif and Sherif 1956) are relevant. Emphasizing the importance of reciprocity, social-exchange theorists have recognized a patterning in dyadic and triadic interactions from the interpersonal to international levels. The empirical work in this tradition has documented the factors that influence exchange processes including timing, sense of obligation, evaluation of equivalence, and norms of reciprocity and fairness (Larson 1998). These variables have been shown to influence synchronization, also referred to as responsiveness, in international communication, including negotiation and mediated interactions. The extent to which synchronized (or nonsynchronized) responding precipitates turning points has yet to be explored.

Synchronized responding can propel a negotiation toward agreement if the parties exchange concessions as they do in many game-like bargaining experiments. This kind of bargaining process consists of small accommodations made by each party spiraling toward agreement in a kind of concession making dance (Druckman 1997). Synchronized responding can also propel a negotiation away from agreement if the parties exchange barbs and threats rather than concessions.

Through time, this kind of synchronized exchange may take the form of an escalatory spiral. Actions taken by one, both or all of the parties, or by a third party, are needed to reverse the trend or resolve the impasse. A particular kind of action that has received considerable attention is the unilateral initiative. Because this initiative consists of cooperatively responding to the other’s noncooperative behavior, it is a form of mismatching or nonsynchronized exchange.

Originally proposed by Charles Osgood (1962) as a strategy to break out of the cold war arms race, unilateral initiatives — which I call “starting mechanisms” consist of one party’s willingness to explore the possibility of cooperation. Well-publicized applications of unilateral initiatives or “starting mechanisms” (see Druckman 1990 and 1997) in international politics include Sadat’s 1977 trip to Jerusalem followed by the Camp David talks; Kennedy’s 1963 announcement of a suspension of U.S. nuclear tests and his call for a relaxation of East–West tensions leading to the signing of the nuclear test-ban treaty; and Gorbachev’s stream of unilateral moves leading to the Intermediate-Range Nuclear Forces agreement, a changed relationship between the superpowers and then finally, the fall of the Soviet Union.

Each of these moves was an important departure or turning point in history. These historical events can be placed in our turning-points framework by considering the unilateral move as a precipitant to departures in an ongoing negotiation. However, the key challenge for further research is to understand the conditions under which these dramatic nonsynchronized moves are effective.

Experiments by Sven Lindskold and his collaborators (Lindskold, Bennett, and Wayner 1976; Linskold, Betz, and Walters 1986) identify some ways that a strategist can more effectively implement initiatives. These include adding words to clarify the intention behind the strategy, not requiring reciprocity of cooperative moves too soon, rewarding reciprocity when it occurs, and making it difficult for the noncooperative party (target of the initiative) to explain away the initiatives. These actions are more likely to be effective when the target party perceives that it is dependent on the initiating party, realizes that the initiator cannot be exploited, and judges that the benefits of cooperation outweigh those of continuing competition. These conditions existed for the Sadat and Kennedy initiatives described above. Interestingly, they were reversed for Gorbachev’s initiatives, where the initiator rather than the target had these perceptions and made these judgments.

Unilateral initiatives may be less effective when a negotiating party’s security interests are at stake. William Rose (1988) showed that a nation that perceived itself at risk did not reciprocate the cooperate move made by its adversary. This finding is consistent with the results of our comparative study of turning points: Change was precipitated primarily by external events in security cases but by internal events, including unilateral initiatives, in political and trade cases (Druckman 2001). More effective strategies in this arena are those that prevent the noncooperative negotiator from exploiting (nonsynchronous responding) the other’s cooperative initiative. Back-channel communications and the use of intermediaries may prove to be more effective ways of inducing the desired cooperation than taking a direct approach.

Further laboratory evidence, reviewed by Martin Patchen (1987), suggests that a strategy combining synchronous responding (as in tit-for-tat) with unilateral (nonsynchronous) initiatives may be most effective in inducing cooperation. The posturing is intended to convey the message of being firm-but-fair. This strategy is similar to other concepts in the negotiation literature. Examples are Morton Deutsch’s (1985) firm but friendly attitude, Dean Pruitt and Steven Lewis’s (1977) flexible rigidity, and Roger Fisher and William Ury’s (1981) invocation to be “soft on the people but hard on the problem.”

Each of these concepts refers to a problem-solving approach characterized by a posture of firm nonbelligerence, self-confidence, and a friendly attitude. This posture is intended to transform an adversary into a collaborator. The transformation is a departure or turning point. The strategies and the way they are communicated are precipitants. At issue is how a strategist can, by his or her posturing, produce the desired changes which have the optimal consequences for the direction taken by a process.

Desired changes are usually regarded as positive turning points. It is the case, however, that turning points can occur in the other direction, moving a process away from agreement. Laboratory research by Mara Olekalns and Philip Smith (2003) showed that facilitating or positive turning points improved the relationship between the two parties, increased their perceptions of trust, and led to better, value-creating outcomes. Inhibiting or negative turning points had the opposite effect: trust was decreased and worse, value-claiming, outcomes occurred. The critical connection suggested by these results is between type of departure, trust, and the course of the talks.

In terms of our framework, trust may be regarded as a consequence of a departure with implications for future departures. The relationship between these processes may be circular: negative (positive) departures reduce (increase) trust which, in turn, leads to further negative (positive) departures, leading to impasses (value-creating agreements) and worsened (improved) relationships. Investigating these sorts of spirals is another new direction for research at the interactional level of analysis.

The research reviewed previously regarding emotional framing of messages and information-processing at the individual level also bears on issues of strategy and trust. But, so too is research on process dynamics where the group is the unit of the analysis. The focus at this level consists of shared understandings by group members. When thought about in the context of negotiation, these emergent concepts would seem to apply more directly to long-term talks such as institutionalized collective bargaining within industries or the multilateral conferences that take place in the United Nations (UN). In these shared professional cultures, negotiator diversity with regard to individual-level variables may yield to uniformity in a manner similar to the way that individual differences, evident early in negotiations, are masked by interaction patterns that develop later (e.g., see Plous 1987). A comparative study of departures in short and long-term negotiations would shed light on issues concerning the impact of group culture on change — including the kinds of initiatives taken, the way they are responded to, and their consequences.

When negotiation cases are studied in terms of group culture, the analyst construes the negotiating parties or delegations as a collectivity. The level of analysis is that of a negotiation system rather than individuals or small interacting groups of negotiators. When thought about in this way, departures may be regarded as changes in the state of the system rather than changes in negotiators themselves or in negotiating interactions. In this section, I provide some conceptual and analytical ideas that may contribute to our understanding of departures as system change.

One of the few frameworks for thinking about international negotiation as a system is offered by Victor Kremenyuk (2002). He conceives negotiation as part of an international network with its own rules and laws. Negotiators are constrained by this network (which may also be regarded as a regime or culture) to behave cooperatively: rules of conduct include “nonviolence, joint problem solving, cooperation, and common values and language” (Kremenyuk 2002: 36).

The institutionalized process described by Kremenyuk probably depicts long-term negotiations among delegations to international organizations. Examples are the conference on disarmament talks in the UN, global environmental conferences, and the trade talks of GATT and NAFTA. These are largely routine activities with few crises, surprises, or abrupt departures.

In other negotiations conducted outside of institutional frameworks, delegates are less constrained by the norms of those systems. These kind of talks are less subject to routine and, thus, more likely to encounter crises, surprises, and abrupt departures. An issue for further research involves whether more frequent and abrupt departures occur in negotiations that are less subject to the constraints of institutional systems.

There is little doubt, however, that negotiations of all types are influenced by events occurring outside the conference room. In this way they are a part of a larger system. The more institutionalized talks — international and domestic — share properties of, and may even be indistinguishable from, that system. The negotiation is one of several activities of the institution. Less institutionalized talks are relatively autonomous or self-contained systems. Although not subject to the regulations and codes that gear the process toward cooperation, these talks are influenced by a variety of external events. They may influence the process indirectly, for example, through effects on the relationship among the parties or be used tactically as when attempts are made to link the prospects of a negotiated agreement to the settlement of related issues outside the negotiation.

A considerable amount of case-study analyses, performed by P. Terrence Hopmann and his colleagues (e.g., Hopmann and Walcott 1977; Hopmann and Smith 1978; Hopmann and King 1980) documents the impacts of both external events and the international environment on negotiation progress and outcomes. Perhaps the most interesting finding involves the discovery of reciprocal effects of actions taken inside and outside of negotiation. Particularly notable is the effect of tension, suggesting a research question about the relationship between external tension and departures: Are departures sensitive to the level of tension in the system?

The key theme of this research is that negotiation can be viewed as a system which, in varying degrees, is a microcosm of a larger institutional context. Of particular interest to us is how to think about departures in terms of these system properties. How might departures be analyzed in the context of a system? One approach calls for treating change in negotiation as a stochastic process where actions are sequentially dependent on prior actions taken by the same or different individuals in a system.

Within the family of stochastic models, a continuous-time Markov process seems appropriate for analyzing departures in negotiation: A Markov chain assumes that the outcome of one event places the system in a particular state, and the probabilities of new events depend on that state. Central to a Markov chain is the idea of transition rates, which are the rates of movement from one state of the system to another. A relevant question for negotiation asks whether the rates are constant, as Markov’s process assumes.

James Coleman (1973) provided several types of evidence to show that transition rates can change in systematic ways depending on both initiated and terminating processes. With regard to initiated processes, which refer to events that mark entry into a state, research shows that the rate of change out of a state seems to decline as the length of time in that state increases. This is expressed as an exponential decay in the transition rate according to the equation:

where q is the transition rate at the time of the initiating event, a is the rate of decline, and q(t) is the transition rate at time t after the initiating event.

This equation describes a variety of social processes documented by studies reviewed by Coleman (1973) including decline in marriageability from the point of first eligibility; changing jobs as age increases; and residential mobility since the last move. It may also describe a decline in negotiation departures from the time since the last departure (initiating event) occurred. Another way of stating this relationship is that departure rates decrease as a function of time since the last departure took place.

With regard to terminating processes, which refer to processes terminated by an event at a predetermined time, it has been shown that transition rates increase over time in a systematic fashion. This is expressed as an exponential growth function as follows:

where q(0) is the value of q when t= 0 and the other terms are same as that in the first equation.

Coleman uses the example of an auction where bidding could occur until a candle flickered out and the last bid before this terminating event was the bid taken. It may also describe an increase in negotiating departures as a predetermined deadline approaches. A related empirical question is whether the departure rate increases after a deadline has been announced during the negotiation process, rather than before it begins. Does the exponential growth function describe departure rates for both a priori (before the process has begun) and a posteriori (after the process has begun) deadlines?

These functions call attention to the reversed processes of a decline and an increase in transition rates through time. The mechanism may be attention: Initiated events set in motion a process of declining attention to a past (initiated) event; terminating events set in motion a build up of attention to a forthcoming (terminating) event. If this process of augmentation and decay is general, then it should apply also to negotiation. Initiated events are the precipitants in our framework, transition rates are departures, and terminating events are deadlines. These processes can be explored through experimentation. The question to be addressed is whether the functions shown in the above equations describe the negotiation process as it unfolds through a series of precipitants, departures, and consequences. It would be necessary to construe departures as bounded states with entering and exiting as transitions into and out of them. Then it would be possible to assess time spent in a state as well as rates of change. Transition rates can be calculated in a manner similar to the way this is done for concession rates and then, be transformed into probabilities.

A question raised by Coleman is whether these decline and growth functions are useful for all empirical situations: Do they capture the many variations and forms taken by negotiation? These issues can be addressed by experimental research. One approach would involve model fitting. The two functions can be compared in terms of their relative effectiveness in predicting transition rates [q(t)] with and without deadlines or terminating events. Estimates for the parameters of the two models would be provided by experimental data. (For an example of the way experimental data are used to evaluate formal models, see Krause et al. 1975.)

Our distinction between abrupt and nonabrupt departures suggests that, in addition to rate, type of transition may vary for different circumstances. And, other aspects of our turning-points framework may influence transition rates or the frequency (and type) of departures. The most direct — or proximal — influence on departures is from precipitants, which can also be regarded as initiated events: Does the type of precipitant — external, substantive, or procedural — make a difference? Interestingly, Coleman calls attention to “numerous other ways in which transition rates may change due to external and unpredictable events” (Coleman 1973: 28).

The consequences of departures can also influence transition rates. An escalatory consequence may increase the rate as parties move more quickly out of previous states as shown in earlier research on the relationship between crises and turning points (Druckman 1986). A de-escalatory consequence, on the other hand, may decrease the rate, causing parties to remain in a state longer and thus, either delaying or foreclosing the transition to another state. These hypotheses about consequences suggest that what happens after departures can also influence transition rates or departure frequency in a negotiation that is modeled as a continuous stochastic (multivariable) process. All of these considerations from our turning-points framework provide opportunities to adapt and evaluate a general model of stochastic processes to the specific empirical situation of negotiation.

The review of literature regarding these processes suggests a number of ideas for additional research summarized in this section. Two key ideas are presented at each level of analysis.

• (1)

  It would be interesting to learn more about the role emotions play in attending to cues that signal change in negotiation. A research issue concerns discovering the emotional-intensity thresholds or levels that would interfere with recognizing important changes.

• (2)

  With regard to information processing, it would be interesting to learn more about the way decoding or monitoring and encoding or acting combine during the course of negotiation. For example, how do negotiators use their interpretations (or attributions) of each other’s conversational acts to decide on a next move or strategy? How does this connection influence the likelihood of a departure in the process?

• (1)

  It would be interesting to learn more about the role played by synchronized responding in producing departures. Does a departure occur primarily when the parties are coordinating their moves? What is the difference between competitive and cooperative synchronized exchanges in precipitating departures? What is the role of trust in maintaining synchronized, cooperative exchanges? How important are initiatives, taken to increase synchronized responding, in precipitating departures?

• (2)

  Another interesting research issue deals with the effects of group cultures in negotiation. A preliminary (but difficult) question concerns how they develop. A follow-up question concerns the implications of cultures for departures. Comparing departures in samples of short and long-term negotiations may reveal differences in type, frequency, or rate.

• (1)

  An issue that has received limited attention is the role played by institutions in negotiation. Negotiations that are conducted within the context of institutions (e.g., industries, the UN) are more likely to be constrained by norms and rules than those that are independent of institutional frameworks. What are the impacts of institutional structures on types of precipitants and frequencies or rates of departures?

• (2)

  Additional research questions are raised when negotiation is construed as a stochastic process. Focusing on state transition rates, it would be interesting to explore the effects of initiated and terminating events on the rate (speed through states) and form (as a decline or growth function) of change in departures. Does the type of initiated event (as substantive, procedural, or external) or terminating event (amount of deadline pressure) influence the rate of change? What is the role of consequences of departures (as escalatory or de-escalatory) in transition rates?

This paper began with a review of the turning-points framework, illustrated by three cases, two international and one domestic. The extended negotiation between China and the U.S. over IPR was shown to pass through stages where internal precipitants led to abrupt departures that culminated in escalatory consequences early and de-escalatory consequences later in the process. The case of the U.S.–Russia relationship showed that an external precipitant produced an abrupt departure in policies with both short and long-term de-escalatory consequences. The example of the football team meeting provided an additional opportunity to show the relevance of the framework for depicting process departures: In this case, a procedural precipitant led to a nonabrupt departure that resulted in several consequences through time.

These three examples set the stage for a discussion of the psychological and social processes that occur in conjunction with departures. After identifying features of departures common to a variety of cases, we explored the processes that occur at the individual, interactional, and collective levels that may precede and accompany change in negotiation.

The levels of analysis approach taken in this paper suggests a distinction between the occurrence of particular departures and the rate at which departures occur through the course of a negotiation or extended set of interactions. Any departure during negotiation follows from one or another type of precipitating event and results in escalatory or de-escalatory consequences. I suggest that this sequence is influenced by the way negotiators decode (monitor/process information) and encode (manage impressions) around decisions and events as well as the pattern of synchronizing concessions, proposals, and verbal behavior. These are the individual and interactive processes that are assessed over the short term of each departure.

On the other hand, the frequency or rate of departures seems to be a function of time, declining with the amount of time spent in a state (or time since the most recent departure) and increasing with time to a terminating event such as a deadline. Rates are understood in relation to entering, staying in, and exiting from states. They refer to the collectivity of negotiators (or the negotiating system) and are calculated over the long term of an unfolding negotiation. This distinction between short- and long-term processes can be bridged by answers to the question: How do the processes that lead to departures influence the rate at which departures occur? The key is to understand the factors that influence time in state (or time since the last departure). It is the question that forms the basis for further interesting research on turning points.