Implications for conflict management

Practitioners and laypersons have neither the time nor the inclination to read the vast body of empirical research literature on negotiator bias and cognition. In this section, we distill decades of research into a few key principles that are best illustrated by real examples.

Exaggeration of Conflict

People involved in social or political conflict tend to overestimate the extremity of the other side’s beliefs, a reflection of the first bias: opposing forces.

Consider the reactions to the real-life conflict commonly referred to as the Howard Beach incident, in which a young black man, Michael Griffith, was struck and killed by a passing car as he attempted to escape a group of white pursuers in the Howard Beach neighborhood of Queens in New York City. In one study, people who characterized themselves as liberals or conservatives were asked to rate the extent to which they believed in the truth of certain statements about the case (“The white pursuers deliberately chased Michael Griffith into the path of oncoming traffic”; “Michael Griffith had consumed cocaine on the night in question”; Robinson, Keltner, Ward, and Ross, 1995). The same people were then asked to predict how “the other side” would rate the truth of the same statements. That is, conservatives were asked to predict liberals’ ratings for each question, as were liberals asked to predict conservatives’ ratings. Both liberals and conservatives overestimated the difference between their side and the other.

Perhaps most surprising, neutrals (people who described themselves as neither liberal nor conservative) also succumbed to this mistake; they too over­estimated the gap between liberals’ and conservatives’ beliefs about the Howard Beach incident. All three groups (liberals, conservatives, and neutrals) exag­gerated the extent of conflict: all three groups overestimated first the extent to which conservatives would interpret the events in ways that blamed the black victim, and second the extent to which liberals would interpret events in ways that favored the black victim.

The pattern of results is not unique to this incident. Exaggeration of perceived conflict exists in many other domains: abortion, the death penalty, the arms race, and even the “Western canon debate” (the dispute among educators about the choice of books in introductory college-level civilization and literature courses).

It is important to consider the implications of the tendency for people to exaggerate conflict. If the partisans in a conflict perceive their differences as greater than they really are, then they might be overly pessimistic about finding common ground. If people hold erroneous assumptions about the gap between their own position and that of the other side, then people might decide that it is not worth even sitting down at the bargaining table on the grounds that any discussion is fruitless (Robinson, Keltner, Ward, and Ross, 1995).

The fact that we exaggerate the extent of conflict means that information exchange among parties is crucial. Unless both sides to a conflict discuss the nature of their beliefs, assumptions, and concerns, each side continues to per­ceive the other as unreasonable and extreme.

In addition to forming an accurate understanding of the conflict, mediators have an important role to play in helping parties overcome their own percep­tion of exaggerated conflict. Exaggeration of conflict comes in two forms: each side tends to see the other side’s position as more extreme than it really is, and one’s own side is also seen as more extreme than it really is. Mediators can help parties see that their own position does not need to be as extreme as they think it needs to be.

Lose-Lose Outcomes

Sometimes all of the people in an interdependent decision-making situation pre­fer one settlement to another but nevertheless fail to achieve it.

In February 2005, the National Hockey League (NHL) Commissioner Gary Bettman officially cancelled the hockey season after a five-month player lockout. Because a new labor agreement could not be reached between the players union and the league, several billion dollars of revenue were lost. Furthermore, almost four hundred of the league’s seven hundred players decided to move to Europe to play hockey (often earning less than half of what they made annually playing in the NHL)(Lapointe, 2005). This is called a lose-lose agreement, because both sides have settled for an outcome that is clearly worse for each of them, than compared with coming to an agreement.

The frequency with which lose-lose agreements occur is both surprising and alarming. One statistical analysis involving more than five thousand participants revealed that lose-lose agreements occurred 20 percent of the time (Thompson and Hrebec, 1996).

Moreover, it is unlikely that the lose-lose agreement is an artifact of the laboratory, with no real-world significance. Balke, Hammond, and Meyer’s examination (1973) of labor-management negotiations at Dow Chemical is a case in point. Analysis of that dispute revealed that labor and management both preferred the same wage increase; yet neither party realized it until after a costly two-month strike.

Another example is illustrated in Walton and McKersie’s analysis (1965) of the Cuban missile crisis, which stemmed from the Soviet Union’s buildup of missile bases in Cuba during the Cold War. The crisis had reached dangerous proportions when the United States threatened to retaliate against the Soviet Union when Cuba fired on American airplanes. Meanwhile, the Soviet Union, unbeknown to the United States, also preferred that Cuba refrain from provoking the United States, because there was a danger that Cuba’s behavior would incite a war over issues not important to Soviet interests. The parties that had come to the brink of nuclear war shared compatible interests without realizing it.

Why does this happen? As discussed earlier, people sometimes adopt a fixed-pie perception in which they believe that the other person’s interests are completely opposed to their own. This belief is established at the outset, before people even have the opportunity to meet or talk with each other. In addition, the fixed-pie per­ception is remarkably durable; it remains even when people have high incentives and ample feedback is available to challenge the perception.

But sometimes people do realize their preferences are compatible with the other party’s, and yet still fail to capitalize on shared interests. Political pressures, situational norms, and organizational constraints can prevent people from opti­mizing their compatible interests.

Invalid Perceptions of Control

We all like to think that we have control over our environment. To some degree, this is certainly true; when we pull on a door it usually opens, and when we say hello to a colleague we usually receive a hello in response. Theorists from several domains conclude that this sense of personal control is necessary to a healthy self­concept; without some sense of control over outcomes in our environment, we would feel helpless and worthless. But we easily become so accustomed to feel­ing in control that we automatically, egocentrically assume we have causal influ­ence over certain events in our environment when, in fact, we have no such influence. As an intriguing example, craps players act as though they control the dice by throwing them softly to produce low numbers or throwing hard for high numbers (Henslin, 1967). As another, a baseball player on a hitting streak might wear the same pair of socks in every game for fear of jinxing the streak.

Inflated perceptions of control generally stem from the same sort of egocen­tric bias that we described earlier. In short, people are acutely aware of their own actions, thoughts, and feelings and less aware of those of others. Moreover, we have a poor memory for contingency information, and our judgments are biased by self-serving motivations. Although people who are given a lottery ticket with a preassigned number gladly accept the opportunity to trade the ticket for one from a different lottery with a better chance of winning (Langer, 1975), those who choose their own number prefer to hold on to the ticket and forgo the opportunity to enter a lottery with a better chance of win­ning.

Conflict situations often present circumstances that encourage people to adopt invalid perceptions of control. In one study, negotiators faced an oppo­nent who was (unbeknown to the negotiator participating in the study) actu­ally a confederate of the experimenter. During the negotiation, the confederate strictly follows a preset schedule of concessions. When the preprogrammed schedule called for the confederate to make an extreme initial demand followed by retreats to a moderate demand, the participants felt they had control over the opponent and the final settlement. But when the schedule called for the con­federate to make a moderate initial demand and then refuse to retreat, negotia­tors reacted unfavorably because they felt they had no causal influence on the outcome of the negotiation (Benton, Kelley, and Liebling, 1972).

In another study, participants played a game in which they had a choice of cooperating or competing with the other party. The other party faced a similar choice. The worst outcome for a player is to cooperate while the opponent com­petes. The best outcome for both players together is if both cooperate. As a player in this game, you are told either that your opponent has already made a choice or that your opponent will make a choice at the same time you do. In either case, you do not know your opponent’s choice before making your own, and your opponent does not know your choice before making his own. Faced with this situation, people are more likely to cooperate if the opponent’s deci­sion is made at the same time as their own (Morris, Sim, and Girrotto, 1998). In this situation, players have the causal illusion that they can influence oppo­nents who have not yet decided by “showing” their opponent that they are cooperating. On the other hand, if the opponent has already made a decision, the player knows there is no way of influencing it, because causation cannot work backward in time. Of course, in either case, the opponent has no way of knowing the player’s choice until after making his, so the player is unable to show the opponent anything about his choice.

These results suggest a tactic for concession making during conflict resolution: inform the opponent of an imminent concession simultaneously with, or just after, the opponent makes a concession. This fosters the illusion of control in your opponent; he reasons that by showing you he is making a generous concession, you will reciprocate and do the same. In the meantime, regardless of what you decide to do, you can benefit from the opponent’s generous concession. Similarly, mediators can suggest to each party that the other party is ripe for influence and suggest that a concession will greatly influence the other party.

Biased Judgments of Fairness

We have made the point that fairness is not an absolute construct; instead, it is socially defined. What is fair to one person may not be fair in the eyes of another. Consider what happened when two vice presidents of a major For­tune 100 company were promoted to senior vice president at about the same time in the late 1990s (Klein, 2003). Both of them moved into new offices, but one of them suspected an inequity. He pulled out blueprints and mea­sured the square footage of each office. His suspicions were confirmed when it turned out the other’s office was bigger than his by a few feet. A former employee said, “He blew a gasket.” Walls were removed, and his office was reconfigured to make it as large as his counterpart’s. This example illustrates that in any situation, there are as many interpretations of fairness as there are parties involved. Here, equality, equity, and need are all plausible prin­ciples on which a decision can be made. Hence, in conflict resolution, two people may both truly want a fair settlement, but they may have very differ­ent and equally justifiable ideas about what is fair.

Although people generally want what is fair, their assessments of fairness are often self-serving (Messick and Sentis, 1979). Moreover, the fact that we have little or no self-awareness of this influence on our otherwise sound judgment heightens the intransigence of our views. Suppose you have worked for seven hours and have been paid $25. Another person has worked for ten hours doing the same work. How much do you think the other person should get paid? If you’re like most people, you believe the other person should get paid more for doing more work—about $30, on average (Messick and Sentis, 1979). This is hardly a self-serving response. Now, consider the reverse situation: the other person has worked for seven hours and been paid $25. You have worked for ten hours. What is a fair wage for you to be paid? Messick and Sentis found the average response to be about $35. The difference is about $10, and it illustrates the phenomenon of egocentric bias: people pay themselves substantially more than they are willing to pay others for doing the same task.

Consider another example. You are told about an accident in which a motor­cyclist was injured after being hit by a car. After learning all the facts, you are asked to make a judgment of how much money you think is a fair settlement to compensate the motorcyclist for his injuries. Then, you are asked to play the role of either the injured motorcyclist or the driver of the car and to negotiate a settlement. Most of the time, people in this situation have no trouble coming to an agreement (Babcock, Loewenstein, Issacharoff, and Camerer, 1995).

Now imagine doing the same thing, except that your role assignment comes first. That is, first you are asked to play the role of the motorcyclist or the dri­ver, and then you learn all the facts, decide on a fair settlement, and finally negotiate. In this situation, the only thing that changes is that you learn the facts and make a fair settlement judgment through the eyes of one of the parties, instead of from the standpoint of a neutral observer. As it turns out, this differ­ence is crucial. Instead of having no trouble coming to an agreement (as do the people who do not know their roles until just prior to the negotiation), people who know their roles from the beginning have a very difficult time coming to an agreement (Babcock, Loewenstein, Issacharoff, and Camerer, 1995). The high impasse rate among people who know their roles from the beginning is linked to self-serving judgments of fairness. The more biased the prenegotiation fair­settlement judgment, the more likely the later negotiation will result in impasse. Thus, a person who knows she is playing the role of motorcyclist before mak­ing a fair-settlement judgment is likely to assess a large damage award (in her own favor). A person who knows he is playing the role of the car driver before making a fair-settlement judgment is more likely to assess a small damage award (in his own favor). The result is that these two people have quite a hard time negotiating an agreement, because their assessments of what is fair are so far apart.

Although biased perceptions of fairness are quite common (Babcock, Loewenstein, Issacharoff, and Camerer, 1995), cultural factors can worsen or lessen these biases. In comparison to negotiators from individualistic cultures (for example, the United States), negotiators from collectivist cultures (for exam­ple, Japan) are less likely to hold such extremely biased perceptions of fairness (Gefland and others, 2002). In one investigation, Gefland and associates (2002) found that, because of their higher levels of biased perceptions of fairness, U.S. negotiators experienced impasse more often compared to Japanese negotiators.

In general, there are often as many proposed solutions as there are parties to the conflict. Each party sincerely believes its own proposed outcome is fair for everyone. At the same time, each party’s conception of fairness is tainted by self-interest, so that each solution is most favorable to the party proposing it.

The Illusion of Transparency

In our daily interactions with others, we tend to assume that when we express particular emotions or preferences, the people with whom we are interacting will accurately pick up on these cues. Yet, this is not necessarily the case. We often overestimate the degree to which others have access to our internal states. The tendency to overestimate the degree to which others understand what is on our minds is referred to as the illusion of transparency (Gilovich, Savitsky, and Medvec, 1998). In other words, the illusion of transparency means that nego­tiators believe that they are revealing more than they actually are. (In other words, they believe that others have access to information about them when in fact they do not.) For example, in one investigation, negotiators judged whether an observer to the negotiation would be able to accurately discern their nego­tiation goals from their behavior (Vorauer and Claude, 1998). Negotiators con­sistently overestimated the transparency of their objectives. Thus, people feel more like an “open book” with respect to their goals and interests in negotia­tions than they actually are.

In certain negotiations, one may stand to receive a better outcome if one does not show too much excitement over closing a deal. For instance, if a prospective home buyer is so enamored with the house that he would be willing to pay the asking price, then communicating this level of enthusiasm to the seller would put the buyer at a disadvantage. If the seller were aware of the buyer’s exuber­ance, she may simply refuse to lower the price at all, knowing that the buyer would likely be willing to settle for the asking price. In this case, it is in the buyer’s best interest to not let his enthusiasm for the home “leak out.” Yet, even when we are consciously trying to hide our true preferences, we still experience the illusion of transparency. In one investigation, Van Boven, Gilovich, and Medvec (2003) found that negotiators who were consciously trying to conceal their preferences imagined that they had “tipped their hand” to the other party more than they actually had. “The illusion of transparency is thus due to the sense that one’s specific actions and reactions that arise in the give-and-take of negotiation—a blush here, an averted gaze there—are more telling than they actually are” (Van Boven, Gilovich, and Medvec, 2003, p. 128).

Sometimes, there are advantages to revealing information. For example, when negotiators have different priorities, negotiators who share information are more likely to reach integrative agreements than those who do not (Thompson, 1991). This is partly because of our inclination to reciprocate: if you share information, chances are that the other party will share information as well (Thompson, 1991). But the illusion of transparency can discourage beneficial information sharing, because it can mislead negotiators into falsely believing that the other party has enough (or too much) information already (Van Boven, Gilovich, and Medvec, 2003). Alternatively, a negotiator who is aware that he might fall prey to the illu­sion of transparency can make a conscious effort to share information with the other party, thereby increasing the likelihood of resolving a conflict in a mutually beneficial way.

Liking, Respect, and Collective Interests

Choices made by organizational decision makers often have implications for the choices made by other organizations within the same industry (for exam­ple, product pricing, employee salaries, and marketing strategy). These decision makers often must decide between pursuing a course of action that benefits their own organizations only and choices that also provide benefit to other orga­nizations. Rather than engaging in mutually competitive actions that ultimately hurt each other (for example, engaging in a negative advertising campaign), some organizations have attempted to find points of cooperation that can align their competitive goals. For example, in 1998, two national dairy companies using separate advertising campaigns agreed to create a single marketing plan to increase milk sales in the United States. Dairy Management, Inc., which used the “Got Milk?” campaign and the National Fluid Milk Processor Promotion Board, which used a popular collection of advertisements in which celebrities wear milk mustaches, coordinated campaigns to increase total fluid milk sales by 4 percent by the year 2000 (Elliot, 1998).

This classic problem of choosing between furthering one’s own interests and those of the larger group or collective is at the heart of many conflicts. Our ear­lier discussion of egocentric judgment processes revealed that people tend to serve their own interests not necessarily because they are selfish and greedy, but because their own interests are immediately accessible to them. Thus, even people who truly desire to be evenhanded often behave in what may seem, to a neutral observer, as self-interest. The question, of course, is how to mitigate this undesirable impulse. Merely telling people to monitor their own behavior is surely insufficient—and it may even increase self-interested behavior. Rather, to the extent that people can develop and enhance a sense of social identity— that is, derive meaning and identity from relationships with others—the egocentric effect may be mitigated.

The importance of social relationships in curbing self-interested behavior was illustrated in a study in which people were asked to decide between keep­ing resources for themselves and contributing to the organization as a whole (Thompson, Kray, and Lind, 1998). Not surprisingly, the better the relationship among the decision makers (the more they knew and liked each other), the more likely they were to agree to contribute resources to the organization as a whole. Remarkably, there was another important factor that affected people’s decisions to contribute to the organization: how much respect was bestowed by organizational authorities. People who were granted high respect from relevant organizational authorities (and who also knew and liked their fellow decision makers) contributed the most resources to the organization as a whole. Thus, in the team management example considered earlier, a group of managers who know and like each other, and who also feel that upper man­agement respects and carefully considers their decisions, contributes the most team resources for the good of the organization as a whole. On the other hand, if managers do not know and like each other, or if they are not accorded respect from upper management, they are likely to contribute the least to organization wide concerns.

I’m Happy Only If You’re Sad

When a conflict is resolved, the parties often assess how satisfied they are with the outcome of the resolution. But parties to a conflict do not measure their out­comes on an absolute scale. Instead, success is a socially determined construct, measured by many factors, including comparison with similar others, views of significant others, and the outcomes of one’s opponent. In fact, there is little or no relationship between how good people feel and their actual outcomes. In conflict situations in which parties’ interests are not completely aligned, how good people feel is a converse function of the emotions displayed by the other person: when the other is sad, we feel good; when they are happy, we feel bad (Thompson, Valley, and Kramer, 1995).

Consider what happened to Joseph Bachelder, who was representing a grocery executive in critical negotiations. Bachelder demanded that his client— the grocer—get a 4.9 percent stake in the business. The words barely left his mouth when the company’s controlling shareholder jubilantly agreed. Says Bachelder, “I just died. I knew right away that I had underbid. We could have had more if I had just asked for it” (Anders, 2003).

Our feelings of satisfaction after the fact are not the only way we are affected by our opponent. Our preferences during an ongoing dispute can also be affected by the opponent’s expressed preferences. For example, in a survey of opinions regarding possible arms reductions by the United States and the Soviet Union, respondents were asked to evaluate the terms of a nuclear disarmament proposal, a proposal that was allegedly initiated by either the United States, Soviet Union, or a neutral third party (Ross and Stillinger, 1991). In all cases, the proposal was identical; however, reactions to it depended upon who allegedly initiated it. The terms were seen as unfavorable to the United States when the Soviets were the initiators, even though the same terms appeared moderately favorable when attributed to a neutral third party and quite favor­able when attributed to the United States. (See also Oskamp, 1965.)

This case is an example of reactive devaluation, which is the tendency for a party to undervalue an offer just because it was the other party who offered it (Oskamp, 1965; Ross and Stillinger, 1991). The reasoning behind reactive deval­uation might sound like: “My opponent wouldn’t make this offer unless it’s good for him. But if it’s good for him, then it’s probably bad for me, so I’ll refuse to accept it.” The examples of respondents to the arms reduction pro­posals and Bachelder’s displeasure in the other party’s immediate acceptance of his offer illustrate that our preferences and our evaluation of a dispute can be determined by the other party’s preferences and reactions. When the other party is happy, we are sad; when the other party seems to favor a particular outcome, we devalue it.

Curhan, Neale, and Ross (2004) have demonstrated that in typical negotia­tion formats, in which negotiators begin by simply exchanging initial proposals with one another, reactive devaluation of initial offers is a common response. By engaging in reactive devaluation—assuming the other party’s offer was primarily beneficial for the other party—many of these negotiators failed to reach agreement with one another. Two techniques helped to prevent reactive devaluation. First, asking negotiators to assign a rating to a variety of propos­als reduced reactive devaluation by motivating negotiators to remain consistent with their original assessment even after one of the previously rated proposals is endorsed by the opponent. Second, asking negotiators to have a general dis­cussion about the issues on the table, without making any proposals or offers, also reduced reactive devaluation. During these discussions, negotiators were able to express their needs and priorities, leading their counterparts subsequently to make more charitable attributions about the offers they received. Both prerating proposals and prior discussion are techniques that can be easily implemented by a mediator; if no mediator is present, the parties can try to structure the negotiation so that prediscussion of needs and priorities takes place before any specific offers are exchanged.

The Schmooze Effect

When people negotiate face-to-face, they rely on subtle behavior cues—posture, intonation, facial expression—to understand what the opponent is thinking and feeling. In fact, the credibility of a message is much affected by whether the nonverbal cues are consistent or inconsistent with the verbal message. We understand our opponent’s behavioral synchrony cues automatically; this is not something we are necessarily aware of. Yet this subtle process helps nego­tiators build rapport and trust, which can lead to favorable agreements for all parties.

When behavioral synchrony cues are absent—as when negotiators com­municate via some method other than face-to-face contact—then rapport and trust must be built in some other way. In the absence of any rapport-building mechanism, negotiators endure strained and tense interactions, and they attribute their own feeling of ill ease to the malevolent intentions of the other party.

In recent years, changing organizational structures have created social net­works composed of organization members who are physically separated, as in multinational corporations. Members of these organizations come to rely increasingly on electronically mediated communication methods (e-mail, fax, and phone) to conduct negotiations. But unless negotiators make a conscious effort to develop mutual cooperation and rapport, the absence of social cues in electronically mediated negotiations can lead to uncertainty about social norms and increased risk taking.

Developing rapport before plunging into an electronic negotiation (“e-negotiation”) can help avoid misunderstandings and misattributions. Strangers conducting e-negotiations are more likely to avoid impasse, to feel good about their negotiation counterpart, and to want to negotiate with the same counterpart again in the future if they make efforts to develop rapport prior to negotiating (Moore, Kurtzberg, Thompson, and Morris, 1999; Morris, Nadler, Kurtzberg, and Thompson, 2002; Nadler, 2004). Developing rapport can be accomplished by using surprisingly simple techniques. In one study, e-negotiators exchanged pho­tographs and biographical information and had an initial “getting to know you” chat with their counterpart via e-mail prior to beginning the negotiation (Moore Kurtzberg, Thompson, and Morris, 1999). In other studies, e-negotiators simply picked up the telephone and had a five-minute “getting to know you chat” with their counterpart prior to negotiating (Morris, Nadler, Kurtzberg, and Thompson, 2002; Nadler, 2004). In all three studies, these simple “schmoozing” techniques had dramatic effects: negotiators who did not “schmooze” were much less suc­cessful than those who did. Nonschmoozers were angry and frustrated with their counterpart, trusted their counterpart less, and were less likely to reach an agree­ment than schmoozers were.