Conflict Management Toolkit | Approaches | Peacemaking

Theories

There are several different theories analyzing the process and outcome of peacemaking. The study of peacemaking involves an analysis of the actors, the stakes and the factors involved. All theories feed into each other, and combining these theories allows for a more complete understanding of the issues involved in peacemaking.

1. Integrative Approach

The integrative approach divides the negotiating process into three phases:

• Diagnostic
• Formulation
• Details

In the practice of negotiation, these phases are not necessarily linear. Rather, they can be repeated and addressed many times throughout the negotiation process.  The toughness dilemma in the integrative approach calls for tough diagnosis in order to clearly elucidate the true interests of each party. In the formulation phase, negotiators should be more soft, in order to find a common, mutually acceptable solution. In the detail phase, the negotiators should again take a tough stance to assure their party receives as much benefit from the peacemaking process as possible.

Diagnostic Phase

The Integrative Approach focuses heavily on the diagnosis of the underlying issues which have created the conflict, including

• Understanding the nature of the conflict, meaning the context, root causes, relationships between the conflicting parties, and influences of external powers.
• Discerning what the real issues are at stake, and not necessarily those claimed by the parties.
• Highlighting the similarities in each party's positions, which show that the parties do have common objectives that cooperation is possible.

The great value gained in the diagnosis phase of the Integrative Approach was seen in the Israeli-Egyptian dispute over the Sinai desert. Both sides were warring over the Sinai desert, but both sides wanted the land for different reasons: Israel wanted the security afforded by such a buffer zone, while Egypt wanted to maintain it's territorial claims. Through negotiations led by former U.S. President Jimmy Carter, the parties eventually realized that sovereignty and security were not incompatible goals. Thus, an agreement was reached in which Israeli troops were withdrawn in exchange for Egyptian commitment not to militarize the Sinai desert. Analysis of the conflict from the Integrative Approach allowed the negotiations to reach a win-win consensus in which the interests and concerns of both parties were addressed and a viable solution was found.

Formulation Phase

In the formulation phase:

• A structure for the negotiations and a common definition of the problem are agreed upon.
• Terms of trade are decided.
• Notions of justice are determined.

Implementation Phase

This final phase requires the agreement be put into practice, often proving to be a great challenge. Here lower level negotiations play an important role in swaying potential spoilers to abide by the new rules. In this phase peacekeeping, humanitarian assistance and peace building efforts help prevent a re-escalation of violence and support peace.

2. Game Theory

Strategic approaches to conflict management are most often expressed in terms of game theory. Such structure-based approaches form their analysis based on determined end points. The prisoner's dilemma and the chicken dilemma are both variations of game theory. These strategic approaches are outcome-oriented and see the choices one party makes in negotiation as the result of a strategy based on the values of the available outcomes. These approaches were especially popular during the Cold War when American and Soviet nuclear weapons build-up was characterized as a "game of chicken." Game theory assumes that:

• all actors are rational,
• the pay-offs are known,
• the "game" can be played numerous times.

Prisoner's Dilemma

This approach is based on the set-up of two partners in a crime who are separately being questioned by the police. Each is given the option of confessing to the crime or remaining silent. The pay-off of each criminal's action is dependent upon what the other does: If criminal A confesses and B does not, A doesn't serve any time and B has to serve 4 years. If both and A and B confess, they each get 2 years. If neither A nor B confesses, they each serve 1 year. The difficulty lies in that the two cannot communicate with each other so they have to choose their actions on what each thinks the other will do.

From the matrix above, it is evident that the prisoners would both be better off if they both remained silent. However, both sides are rational, and both will try to unilaterally improve their own situation. As both parties move to their Nash solution (in which a move unilaterally improves their situation) by confessing (in the hope of not having to serve any time), both parties worsen their own situation. Both parties confess, and so both parties are forced to serve two years in prison. This end point is not ideal for either party, but it is the most rational solution: if the other happens to confess, then the one who remains silent faces four years in prison.

In the game theoretic analysis, it is the role of the peacemaker to move the parties from the Nash solution, where both parties confess, to the Nash point, where both parties remain silent and each serve only one year. It is here where the free flow of information, trust and relationships are especially useful in changing the endpoints of the game.

• Toughness Dilemma: In the Prisoners Dilemma game, parties should be soft to open. In repeated games, this will increase likelihood that the other party will open softly in the next iteration (tit for tat). If this soft opening is not reciprocated, the negotiator can be punished with a tough stance.

Chicken Dilemma

This dilemma is based on the game of chicken in which the chicken represents a lack of courage or a 'loss of face.' The 'chicken game' is known to be played between two rivals driving two cars towards each other on one road. The rivals speed towards each other in their cars. The first rival who veers the car in order to prevent a collision 'loses face' and is 'chicken.' The rival who did not veer is considered the courageous winner. If both rivals veer, there is no damage, but both sides lose face. If neither side veers, the worst possible outcome- both sides destroyed- occurs.

The role of the peacemaker in the chicken dilemma is to elucidate the consequences of actions. The parties are often compelled to the worst-case scenario in response to political pressure or their own unwillingness to be perceived as cowardly. A third party peacemaker can change the game by introducing incentives for both parties to "veer," or by threatening incentives, for example through the loss of international legitimacy, who parties who fail to "veer."

• Toughness Dilemma: In the chicken dilemma, parties should be tough to open, in order to the other party to make a concession- to veer. Once the other side makes concessions, the parties should be soft in their rewards.

Processual Analysis

The processual approach divides negotiations into separate stages. This approach isolates distinctive elements and chronological events in order to better understand the complexity of the negotiating process. While this division is a useful analytical tool, in practice, these phases can overlap and interact. According to the processual approach, negotiation is comprised of a series of concessions which signal intent while attempting to attract the other party towards one's position. These actions represent steps in the negotiation process, at which the parties have the option to agree, disagree or offer a counter argument.

Concepts key to the processual approach include:

• Bargaining Range: Upon entering into negotiations, each party has a bargaining range of possible options and acceptable outcomes. An overlap of the parties' bargaining ranges is needed to successfully initiate negotiations. If there is no overlap in the bargaining ranges then the ranges must be moved, if the range is too large, then alternatives must be limited.
• Critical Risk: The party with the greatest risk is the most likely to make the greatest concession.
• Security Point: This is based on strike costs, or that which the parties would receive if the negotiations broke down. There are several views on the function of security points:
  o The party with the most to lose if the negotiations break down is the one to make the greatest concession.
  o The party with the lower cost of concession is most likely to concede.
  o The party with the higher cost of holding out is most likely to concede.
  o The party with the higher time costs is more likely to concede. This is also known as the rotting rate. If the negotiation has a time limit, the party dependent on that limit is more likely to make larger concessions as the limit is reached.
• Toughness Dilemma: Processual anaylsis of the toughness dilemma shows that softness breeds softness, while toughness breeds toughness. In the negotiating process where both sides choose to be tough, there will be little chance to converge on a mutually accepted point.

3. Behavioral Theory

The behavorial approach analyzes negotiations through the behavior of personalities or groups of parties. The behavioral school focuses on the importance of individuals: how personalities influence cooperation and conflict. Expectations, perceptions and trust are all fundamental to this analysis. Rationality is key, as rational bargainers are able to interpret new information and to change expectations and perceptions in way which facilitate successful negotiations.

• Power: Behavioral analysis accounts for the personal possession of power. If power is the ability of one actor to move another through a desired range of outcomes, then power can bet he ability of a negotiating party to bring another to the negotiating table on terms which are mutually acceptable. Power is also held in the negotiators ability to moderate the extreme peripheries of their constituencies, maintaining their loyalty while also making the necessary concessions to continue the negotiation process.
• Trust: If both parties are to make the compromises necessary to reach a negotiated agreement, trust is a necessary factor. The history of conflict between the parties means that trust is a rare commodity. It is therefore the role of the third party mediator to provide incentives and assurances that agreements will be honored. This mediation will facilitate the development of a relationship between the parties, where trust can be learned and peace sustained.
• Relationships: Repeated interactions between parties in a negotiating process will lead to the development of relationships. Without a third party mediator, this relationship is likely to be negative and lacking in trust. It is thus the responsibility of the mediator to facilitate the development of a positive and constructive relationship. This can be done with incentives and support of legitimacy of the parties. It is hoped that the parties will learn how to interact with each other in the future, developing a relationship where the norms of conflict management to support a sustainable peace have been learned.
• Personal orientation: Negotiators are either competitive or cooperative. In order to get the most out of a negotiation, parties should match the orientation of the other. Two cooperative or two competitive parties will be much more likely to reach a mutually acceptable agreement.
• Toughness dilemma: The toughness dilemma manifests in the behavioral approach taken by the negotiating parties. Hard-liners, often described as "warriors" are especially sensitive to the dynamics of power. Conflicts and negotiation are perceived as having only zero-sum outcomes, thus asserting and achieving one's interests take precedence over reaching a negotiated outcome. Soft-liners, or "shopkeepers," on the other hand, approach conflict management with the desire for mutual accommodation, flexibility and compromise. Soft-liners view negotiations in positive-sum terms and are more willing to transform their positions to reach a negotiated outcome. For both positions, it is evident that attitudes and personality interact with and impact decision-making processes. To deal with the toughness dilemma, soft behavior calls for soft reactions, while tough behavior should be matched with behavior that is also tough.

Social Psychology

Negotiation can be a potent and effective tool for conflict de-escalation and peace-making, especially when undertaken in the social-psychological approach. In negotiations, the importance of perceptions in terms of power distribution or relative positioning are key. Three factors contribute to the commencement of negotiations; with each factor, perception is key.

1. The occurrence of a recent or incipient national crisis, or the perception of a mutually hurting stalemate. This situation is usually caused by a change in the military situation, or with the perception of a real and immediate threat of new or heightened violent conflict.
2. The perception that continued violence is more costly than the negotiations themselves.
3. The asymmetry that previously facilitated ends attained through conflict may have altered, leading to a symmetry which motivates movement to protect interests. If parties perceive that power is distributed more equitably, they are more likely to come to the negotiating table.

Parties naturally fear the compromises involved in attaining negotiated settlements. Fears of concessions threaten the sense of security for both sides, making parties less likely to enter into negotiations in the first place. The social psychological approach can assist in bringing contending parties to the peacemaking process. If each side is able to recognize the fears and perceptions of the other, thus accepting mutual legitimacy, negotiations can create positive change. Third party peacemakers can also support the process by providing reassurance in the form of acknowledgements and confidence building measures. Outside of negotiations, conflict resolution workshops are useful for changing negative perceptions, acknowledging history, and addressing destructive divisions in identity. By addressing the fears and perceptions of threats, these peacemaking initiatives can support the transition from conflict to sustainable peace.

For negotiations to be truly successful, a fundamental transformation needs to occur for both parties. Druckman (1980) maintains that mutual cognitive change- and not just mutual compromise- is necessary if negotiated agreements are to be sustained. Such changes become possible when history is acknowledged and empathy for the other party is allowed.

4. Power Theory

Power can be generally defined as the ability of one party to successfully impose its will on another. The realist school of international relations theory is based heavily on the conceptions and distributions of power. Nations hold power militarily, politically, geographically or economically. While power can be conceived in material terms, such as in the size of a military budget or in national economic indicators, power is also a function of perception. In the realist school, for example, perceptions of relative power positions are key to the relations between states and the outcomes of conflicts.

Violent conflict occurs when there is disharmony of power among either states or groups. This may occur when one group attempts to impose its will upon another and the other party resists; for example, when a government attempts to bring a minority group under unwanted control. In another instance, a weak or collapsed state may lose its power to constrain opposition or rebel movements, thus leading to violence as the opposition asserts their interests and the state attempts to contain them.

Tools for asserting power or to change the progression of the negotiation:

• Side payments or promises of financial rewards
• Military or political security guarantees
• Legitimizing parties through diplomatic recognition
• Diplomacy including the use of incentives and pressures
• Economic sanctions
• Military Force

Key concepts of the Power Theory approach include:

• Use of Power: Conflict management can utilize power and the perceptions of power as a tool for mitigating violence. While asymmetries of power usually facilitate success on the part of the more powerful party, there are tools in conflict management which can redistribute power, thus equalizing power between parties. "Borrowing power," as this redistribution is termed, can be especially useful in negotiations, where the use of timing by either stalling or pushing a deadline, politics in responding to one's constituency, or notions of justice by lauding equity and fairness as opposed to inequality or oppression, can all be skillfully manipulated to advance one's own best interests.
• BATNA: The Best Alternative To a Negotiated Agreement or BATNA, is an important tool when entering into negotiations. If a party knows their BATNA, it has the power to not compromise more than it desires. If the other party knows that there is a BATNA, it may be more willing to make concessions than when it believes the opposing party has no other alternatives.
• Toughness Dilemma: Power also plays into the toughness dilemma. While the stronger party may be more prone to enter into negotiations fully asserting its relative power, there may be more value in soft positioning, thus increasing the probability of reaching mutually satisfactory negotiated outcome.

5. Optimal Agent Independence

The simplest models of negotiation treat it as the contest between two monolithic and unmediated parties.  While these idealized conditions facilitate analysis and yield important insights, they also ignore factors critical to negotiation in the international arena. International negotiation is generally brokered through agents, and these agents typically represent multiple principals. Certain agencies or parties will have principal ownership over the process, but others may also be involved, supporting or opposing the strategies and objectives laid out for the negotiators. A negotiation team will often contain agents who speak on behalf of these disparate principals, and these agents will serve not only to aid in the process, but also to represent the views of their sponsors and to monitor the proceedings. Subordinates on the team may attempt to steer the negotiation by opening up back channels of communication, exerting influence over the agenda, and even attempting to wrest control of the process away from the lead negotiator.

To capture the complexity of these internal dynamics, a more nuanced negotiation model is required, one that takes into consideration the effect of a lead agent’s independence of action with respect to the principals, the other agents, and the opposition.  The purpose of this paper is to identify the parameters that characterize this margin of maneuverability and to present a negotiation model that explores how shifts to these parameters will alter the outcome of the process. Existing models of negotiation that focus on agent independence fail to address adequately the evolution of the process over time. Making use of empirical examples to guide inductive reasoning, this paper will arrive at a set of generalizable hypotheses about agent independence as it pertains to phases in the negotiation in order to refine these models.
 
An agent’s margin of maneuverability, or independence, can be defined as the aggregate of three individual components: flexibility in the authorizing stage, autonomy in the representation stage, and authority in the ratification stage (Nicolaïdis 1999, pp. 94–98).  In the authorizing stage, principals provide the agent with an initial mandate governing their objectives, priorities, and principles.  They may also present information about the aspiration and reservation points as well as what side payments may potentially be available.  Flexibility entails strategic control over these conditions. The more flexible the mandate, the more freedom an agent has to define these points independently, either in advance or over the course of the negotiation (Nicolaïdis 1999, pp. 95–96).

Where flexibility determines how binding the initial instructions will be, autonomy determines the extent to which the negotiator will be able to deviate unilaterally from these instructions over the course of the negotiation. The representation stage begins when agents reach the point of open discussions and continues through the formal negotiations that follow. Autonomy measures the degree of freedom during this stage over such issues as the concession rate or the divulgence of information regarding red lines or other private matters. Autonomy can be restricted through monitoring by other agents at the negotiating table or by requiring reports to the home office during the process (Nicolaïdis 1999, pp. 96–97).

Finally, authority refers to the power an agent has to give approval to the terms of an agreement.  The ratification stage encompasses all specific details of a negotiated settlement, beginning with the framework for resolution and including agreements in principal, even though that may not ultimately dictate the terms of the process. Authority entails the power to make commitments, offer concessions, and approve a final deal (Nicolaïdis 1999, pp. 97–98).

External Determinants of the Margin of Maneuverability

While there are means by which the principals or the agents can manipulate the amount of independence an agent has in the negotiation process, such choices are also subject to the constraints of external factors. To understand where and to what extent the actors are in control of agent independence, one must examine these constraints in more detail.  These factors include the nature of the bureaucratic and political system charged with the negotiation, the specifics of the issue under discussion, the background of the agents chosen to negotiate, and the technology used by the participants.

Though negotiations often occur at a remove from events on the ground, the home environment still strongly affects the principal/agent dynamic.  The context of a negotiation includes broad characteristics such as the political and legal system of the government as well as more narrow ones such as the culture of the agency or department in charge of the process.  For example, during the Cold War, marked differences were observed in the negotiating styles of the Americans and the Soviets. The Americans would typically make the opening offers, while their communist counterparts rarely made proposals of their own, instead chipping away at what was presented it until it was made acceptable. Communist agents had little authority in the process and were rarely willing to exceed their instructions. Such styles may have resulted from ideological differences, from the specifics of national culture, or from differences in the bureaucratic systems in place (Hopmann 1996, pp. 153–160).

Negotiating culture is influenced by institutional history and the legal system. There may be regulations governing the staffing of the negotiating team, laws binding available concessions, and standards over who has the ultimate authority to make agreements. Within a nation, different government agencies, and even the departments within them may take a different approach to the negotiation process (Solomon interview).  With as many parameters as the actions they describe, these constraints cannot predict the outcome of a particular negotiation.  Nevertheless, these structural constraints do provide some guidance as to how agents will behave.  One group may be more likely to over-promise in the beginning stages, another to resort to theatrics, still another to stretch out the process by re-treading old ground at each session (Solomon interview).

The characteristics of the issue under discussion also have a significant effect on the degree of independence that can be granted to an agent. Issues of national security face more scrutiny than those of trade or politics, and the negotiator must pay particular attention to the reentry problem (Gallucci interview). For any one government, some matters will be more sensitive than others and subject to more scrutiny from the public. Examples include coastal fishing rights for Iceland, drilling rights for the Caspian countries, or ownership of historical and cultural artifacts for Egypt.  The more non-negotiable planks comprise a dispute, the less independence the negotiator will have to make decisions about concessions.  Similarly, for an issue that overlaps many jurisdictions or touches on many related topics, there will be more parties jockeying to shape the outcome, which will place restrictions on the independence of the agent. Public opinion is also galvanized more by certain issues than by others, apart from the relative importance of the discussions. Opinion varies not only from issue to issue and country-to-country but also with the timing of external events such as an election cycle, an international summit, or a trade conference. These external events may shift the internal distribution of power and affect the agent’s margin of maneuverability (Putnam 1988, pp. 444–446). The Palestinian/Israeli talks at Camp David and Taba in 1999 illustrate this point, with negotiators scrambling to reach an agreement before Ehud Barak’s mandate expired and he was replaced by the hardliner Ariel Sharon.  Barak’s plummeting polling numbers eroded the legitimacy of the process, and it was ultimately abandoned in the days leading up to the election (Matz 2006, pp. 555–556).

Who the individual negotiator is and what his experience has been will also play a large part in shaping the constraints that he will face as a negotiator.  Seniority and status will determine the institutional resources available for the process, affect the level of respect accorded by the opponent, and help shape the form of the interagency debate that may arise during the process. An agent’s relationship to the principal party making the appointment will have important implications as well.  An independent negotiator will be less inclined to maintain bridges within the agency or to worry about the effect of the negotiation on future promotion, while a negotiator pulled from the ranks will have more lateral connections and informal ties upon which to draw in order to navigate at home. Markers of institutional loyalty such as length of tenure or political affiliation will affect perceptions of trustworthiness. Professional expertise on the part of the negotiator will also determine what his role will be, especially in a highly detailed negotiation that relies on an understanding of the technical issues at hand (Solomon interview).

Finally, technology plays a part in the structural constraints on the negotiator.  As communication has become more facile, the barriers between the principals and the agent have been lowered, placing new restrictions on autonomy and authority. Improved lines of communication reduce the cost for a principal of monitoring the agents and controlling the process.  Better communication frees principals from the information filters of the field officers.  In the past, a foreign service officer involved in a negotiation could make the case for a position in part through the selective emphasis or de-emphasis of bits of information evident in their analysis of the situation. Now this analysis may as easily come from third-party sources. With less ability to manipulate the opinions held by the principals, the agent loses independence over the process (Laipson interview).

Modeling Principal/Agent Dynamics in International Negotiation

The exogenous factors outlined in the previous section define ex ante the independence of the lead negotiator, and these constraints will interact with any deliberate calibration of independence by the principals. But not all parameters are beyond the control of the interested parties. Among their various options, the principals can specify the degree of freedom in the initial mandate, control which members make up the negotiating team, require proposals be pre-screened and authorized, and maintain veto authority over the final resolution. Understanding how these constraints affect the likelihood of reaching a negotiated settlement requires a more sophisticated model of the process than that afforded by the two-person, cooperative game concept. 

We begin with the model of the two-level game developed by Robert Putnam to explore the domestic-international linkages that characterize international negotiation. In Putnam’s model, bargaining occurs at the international level (Level I) and the negotiated package is either approved or rejected at the domestic level (Level II).  Putnam labels the range of Level I agreements that would be acceptable to Level II actors the “win-set,” which he constructs from the intersection of the ZOPA’s of the Level II players.  These win-sets provide the constraints on the bargaining behavior of the Level I negotiators. Putnam’s characterization of win-sets roughly follows the logic of the toughness dilemma—a small win-set will limit the available concessions and be less likely to overlap with that of the adversary. As a win-set shrinks, a negotiated settlement becomes less likely, but an agreement that is reached should come with fewer concessions on the part of the Level I negotiator (Putnam 1988, p. 440).

Putnam’s analysis captures a portion of the dynamic represented by the margin of maneuverability, but it is limited by its static interpretation of the negotiation process. At the Level II negotiation, agreements are accepted or rejected but they cannot be altered. Bottom lines at the domestic level are firm and fully determine the range of options available at the international level.  Putnam’s analysis also does not explore the dynamics of concessions within the Level I negotiations beyond that determined by the size of the win-sets. This stance is reinforced by his observation that, “Level I negotiators are often badly misinformed about Level II politics” (Putnam 1988, p. 452). This myopia has the advantage of curtailing feedback within the model, but it is unrealistic in its constraint over the influence of the chief negotiator, who is powerless to alter the bottom lines of domestic actors except insofar as he is able to directly intervene in the domestic arena. Such a model does not allow for a thorough analysis of the influence of agent independence on the negotiation process.

This uncertainty on the part of the Level I actors about the true bounds of the win-set also prevents understanding of the interplay between the interests of the agents and the principals.  In principle, the win-set simply demarcates the bargaining space, within which the agent is free to negotiate an outcome according to his own interpretation of the problem.  In practice, the agent has only a probabilistic grasp of the win-set bounds and must navigate the boundary regions according to his best interpretation of what will be acceptable to the Level II actors.  This is further complicated by the fact that institutional regulations at Level II are rarely well defined. For example, in the United States, negotiations often fall within the purview of the executive branch and must navigate the foggy interagency process before reaching the desk of the President. Putnam lays out several principles that the agent can follow to address this problem (Putnam 1988, pp. 456–459), but he does not provide a systematic approach to resolving this uncertainty.

By disaggregating the role of independence into flexibility, autonomy, and authority, Kalypso Nicolaïdis begins to explore this dynamic element of the two-level model. Her typology of independence enables a more nuanced exploration of the negotiation process with respect to the role of the lead agent. Flexibility acts as an ex ante filter over the size of the win-set, cutting off a portion that lies within the Pareto optimal frontiers.  Autonomy controls the rate of concession and level of information disclosure, governing movement within the bargaining space.  Authority affects the level of certainty an agent has over the behavior of Level II actors, giving their objections more or less power over the outcome of the process (Nicolaïdis 1999, pp. 95–98). 

From her model, Nicolaïdis deduces three hypotheses about the relative amount of flexibility, autonomy, and authority that should be granted in a particular negotiation situation: 1) If the principals have highly aligned interests, then the flexibility of the agent should be low; 2) If the principals and the agent have loosely aligned interests, then the authority of the agent should be low; 3) restrictions on autonomy achieve the same effect as those on the other two parameters, and can be used to strike the best balance among the three (Nicolaïdis 1999, pp. 116–119).

The first hypothesis is based on the contention that flexibility has both an informational and a strategic effect. With respect to the former, reducing flexibility will clarify the structural constraints on the margin of maneuverability. With respect to the latter, reducing flexibility will alter the size of the bargaining space.  The second hypothesis is based on an intuitive understanding of the interaction between veto power and agent behavior, a position supported by the testimony of negotiators in the field (See for example Neu interview).  While Nicolaïdis does not explore what constitutes a difference of interests or when these differences might affect the negotiated outcome, implicit in both of these first two hypotheses is the assumption that there is an ideal range of interest alignment that best satisfies the toughness dilemma and the reentry problem. Modifications to an agent’s flexibility or authority can artificially provide these same benefits in situations when interests do not fit into that range (Nicolaïdis 1999, pp. 116–118).

The third hypothesis attempts to characterize the interactions among the three parameters.  Nicolaïdis asserts that autonomy can substitute for flexibility as well as for authority.  Increased oversight during the representation stage reduces the need to reformulate the initial mandate when an agent’s position must change. Limiting autonomy throughout the process will allow the principals to be more certain that a given settlement will lie within the ZOPA. Nicolaïdis argues that it is less costly to increase or decrease autonomy during a negotiation than it is to reformulate the mandate or to add to the uncertainty of commitment and thus recommends using it as a proxy for the other two (Nicolaïdis 1999, pp. 118–119).

Gaps and Limitations of the Model

Despite the insights that Nicolaïdis’ model provides, it does not fully capture the phased dynamics of negotiation, which progresses through a diagnosis, formula, and details stage followed by an implementation process (Zartman and Berman 1982, p. 9).  Within this framework, the toughness dilemma can best be satisfied by toughness in the diagnosis and details phase, when specificity and candor will determine an acceptable outcome, and softness in the formula phase, when there is a need for creativity, openness, and the construction of a common concept of justice to frame the issue. Only by recalibrating the mandate and the degree of oversight will the principals directly affect an agent’s approach to these phases of the negotiation process. Perhaps it is Nicolaïdis’ view that the solution to the toughness dilemma must be an independent assessment made by the agent, but principals do in fact have a great deal of control over this aspect of bargaining behavior (Gallucci interview).

Fisher and Davis provide a more detailed analysis of the role of agent independence during these negotiation phases, although without the benefit of Nicolaïdis’ typology. Rather than employ a more formal model, the authors rely on their intuitive grasp of the negotiation process to reach their conclusions. They argue that a principal should recalibrate an agent’s authority during the process based on the level of learning that both agent and principal are able to achieve. They also recommend that instructions to the agents inform them of the principals’ interests, priorities, and alternatives without specifying settlement points in advance. The authors conclude that commitment authority should begin weak, with the agent primarily sounding out the adversary for common ground, and gradually increase in proportion to the learning that takes place (Fisher and Davis 1999, pp. 68–69).

The prescriptive value of Nicolaïdis’ three hypotheses is also limited by the difficulty of measuring alignment between agent and principals. Absent an internal negotiation process between the agent and the principals, questions of agent/principal alignment cannot be easily answered. Internal beliefs may very well run counter to the official position, but a professional must adopt these opinions as his own. The specific interests of the agent as distinct from the party he represents are of questionable import.  The process of being a negotiator entails role-playing and the ability to keep personal opinions from dictating professional behavior (Gallucci interview). Another way to view this issue is to ask whether the negotiator has legitimacy. For example, were the principals empowered to appoint the agent or is has he been appointed by an external source (Docherty and Campbell 2006, p. 501)? Even if legitimacy is at least high among some parties, the question remains how to determine the measure of alignment.  Is their higher principal/agent alignment when an agent is moderately aligned with all principals or when he is strongly aligned with some and weakly aligned with others? In practice, it may be impossible to measure trust apart from observing the authority vested in an agent, which would make the issue circular.

In general, flexibility, autonomy, and authority as Nicolaïdis describes them are blunt tools for tinkering with the negotiation process. As Fisher and Davis note, instructions that guide are different from instructions that define and can have opposite effects on the process, but both are included in Nicolaïdis’ definition of flexibility. Similarly, autonomy refers to both oversight and control, but the two are also very different concepts. Sharing information is quite distinct from ceding authority. These gaps in the explanatory power of the model make it useful to conduct a closer examination of the effect of these parameters during each of the negotiation phases in order to refine Nicolaïdis’ hypotheses.

Assessing Agent Independence Through Negotiation Phases

Diagnosis Phase
In the diagnosis phase, information sharing and idea formulation are of critical importance. Negotiation post-mortems consistently highlight the importance of preparation to this the process (Gallucci, Laipson, Neu, Solomon, and Zartman interviews, inter alia).  A patient exploration of interests and priorities will maximize the probability of reaching a satisfactory resolution and of leaving little on the table.  A negotiator empowered by information will be better able to move the bottom line of his adversary, to issue credible threats and promises, and to make use of Homan’s maxim to achieve gains at low cost.  It is the negotiation team, with its singular focus on the process, which is best suited for conducting this type of preparation. The principals may themselves not be fully aware of what their true bottom lines are until confronted by a fuller understanding of opposing views. Where multiple principals are involved, it is often the agent who must clarify competing positions in the preparation phase, interpreting the bargaining space based on their own analysis. In the end, bottom lines are elevators, not floors. The true reservation point may only become obvious when faced with the strength of the adversary’s opposition (Zartman interview).

This preparation will be insufficient unless the agent is properly informed about the priorities of the principals.  This requires a clear articulation of their interests through specific instructions as to what may or may not be negotiated. But there is a difference between specificity and intractability. Agents must be made aware of what is on the table and what is not without being pressed to accept fixed positions that tie their hands in the negotiation process.  Given that caveat, the less flexible a mandate is, the better an agent will be able grasp the positions of the principals and the more likely it will be that the negotiation process reaches an acceptable outcome.  The more autonomy an agent has to interact with the adversary, the easier it will be to establish trust and gain information.  Autonomy provides cover for a candid exchange concerning principles and values.  Increased authority also makes trust easier to gain.  An adversary that recognizes the decision making power of the agent will pay closer attention to promises or threats and will be easier to convince of the viability of the process.

Formula Phase
The need for autonomy is especially apparent in the formula phase, where the first attempts are made to form a joint framework for a settlement.  As the diagnosis achieves sufficient information to move forward, it is incumbent upon the two parties to find creative solutions to bridge the gap in demands and to expand the pie to be divided. Parties must be able to cast out a variety of options to see which will appeal to the interests of the opponent. Often this will entail using tentative offers to feel each other out, making use of backchannels or private discussions, or in other ways creating the latitude necessary to reach compromise. The lead negotiator must have the independence to reach out in this stage of the process (Zartman and Berman 1982, pp. 70–74). 

One illustrative account of the power of such autonomy to create breakthroughs is Paul Nitze’s private discussions with his counterpart Yuli Kvitsinsky during the SALT II talks.  As a result of more than a dozen such exchanges, the two reached an agreement in principal for a joint reduction in intermediate range ballistic missiles that appeared to satisfy all of the remaining sticking points in what had been a long and fruitless process (Talbot 1985, pp. 116–151).  There are many reasons why agents are more likely to achieve resolving formulas through creative construction than are the principals they represent. Formality and protocol serve as natural barriers to departure from prepared positions to compromise solutions. Face-to-face conversation helps to break down these barriers, as do some of the other techniques for trust-building that negotiators employ such as caucusing and informal meetings in bars and restaurants.  An individual agent can also establish a track record of reliability and honesty apart from that of his government whose past policies may be the source of the present conflict (Zartman and Berman 1982, p. 29). By building trust, the agent will lower barriers to information exchange which creates opportunities to parse an issue, probe for implied utilities, and make use of Homan’s maxim to move toward an agreement.

It is less clear whether authority will contribute or detract from the likelihood of reaching a resolving formula. An agent with weak authority can use the threat of a veto from his principals in order to bargain tough while appearing soft.  The veto prospect helps turn agents into constructive negotiators, figuratively seated on the same side of the table and fighting against implacable resistance from domestic forces immune to persuasion (Nicolaïdis 1999, p. 108).  However, a party that makes concessions will expect something in return, and the less authority an agent has, the more difficult it will be to convince an adversary that an offer is legitimate. To reach an agreement there must be room for mutual concessions, and the agent must have the power to authorize them (Zartman and Berman 1982, pp. 221–224). Without authority, it is difficult to establish trust, and without trust, it is difficult to view the adversary as a collaborator instead of a competitor in the process. Such a perception makes relative gains appear more important than absolute gains and will preclude a coherent effort to expand the pie.  This reduces the chances that any agreement will be reached and makes the achievement of a resolving formula unlikely. The less trust a party has in a negotiation, the less utility they will perceive in compromising and the more tolerable the status quo will become. 

Apart from the conclusions suggested by the principal/agent dynamic, because an agent will most often have greater experience as a negotiator than the principals he represents, he should be given the benefit of increased flexibility, autonomy, and authority in the formula stage.  An agent will typically be selected based on prior experience with the issue and a technical understanding of the bargaining process.  This background will aid in determining points of agreement and in finding a potential way forward in the negotiation.  The same experience provides agents with the professional detachment to make difficult compromises on behalf of the principals. As such, they should be empowered to contribute to the framing of a settlement.

Details Phase
When it comes to actually specifying compromise positions, the negotiator must be able to fit the information gleaned in the diagnosis phase within the framework of the formula.  To do so, he must have a clear understanding of the relative utilities that both sides attach to the issues under discussion. In this phase, cognitive and judgment errors, though a potential hazard throughout the process, have the greatest influence.  With increased independence, particularly increased autonomy, comes a heightened risk that these errors will occur, as more power is vested in a single person who for issues of pride, desire for professional advancement, or inability to ignore sunk costs, has excessive interest in reaching a negotiated outcome. Individuals acting alone are more prone to lapses in attention, retention, and recall—failings that create misunderstanding and lead to missed opportunities.  Individuals are more likely to cling to information that supports their beliefs and to discard that which conflicts with them. These deficiencies tend to be an artifact of individual thinking, and they are best overcome through group engagement and behavior.  Thus, limitations on an individual negotiator’s autonomy, particularly through the presence of dissenting voices at the table able to draw on the same first-hand information, can help to overcome cognitive biases and enable a more positive assessment of the process (Hopmann 1996, pp. 121–129).

Empowered agents are also more prone to errors of judgment.  Individual negotiators tend to place undue confidence in their acumen and persuasiveness, which can lead to an excessive opening bid, an aversion to making concessions, and an unwillingness to refine their theory of justice.  Such tendencies contribute to, and are exacerbated by, a propensity to hoard information and strategic plans from teammates and principals, something that undermines processes that are ongoing and precludes institutional learning.  An empowered negotiator is more likely to escalate commitment to previous positions in order to avoid admitting mistakes. Problems of framing and loss aversion identified by prospect theory are likely to have a stronger influence on the outcome of a negotiation when an individual lacks the oversight of outside observers (Hopmann 1996, pp. 129–132).

As in the formula phase, authority in the details phase is of ambiguous benefit in the process.  More authority weakens the negotiator’s power to effect concessions but strengthens his ability to close an agreement.

Reentry Process
Agent latitude tends to yield advantages at Level I of the negotiation, but it creates an added likelihood of an agreement being rejected by the Level II negotiators. Agents who engage in trust building and information exchange with the adversary are vulnerable to accusations of disloyalty. Exposure to the views of the opponent increases risk of role conflicts, and agents may even be labeled as traitors if they attempt to lobby for unpopular positions put forward by the other side (Babbitt 1999, p. 145). Such was the case with Nitze during the SALT II talks, whose secret agreement raised accusations of having “gone off the reservation.”  Ultimately both the Soviets and the Americans rejected the proposal for being too generous to their opponents (Talbot 1985, pp. 116–151). 

Even if a negotiated settlement is endorsed by the government, it may be unacceptable to a population that has not been prepared for it in advance. The 1993 Oslo accords, presented as a fait accompli, were rejected by influential political groups in both camps and ultimately rendered ineffective as a result (Cutcher-Gershenfeld and Watkins 1999, pp. 44–46).  Similarly, agent latitude allows the principals to distance themselves from the negotiation process, maintaining the political cover to later reject an agreement. Such negotiations are not just less likely to succeed in the long run, they are also less likely to generate meaningful compromises, as one side will be wary about making firm commitments if the other lacks enforcement capability.

In general, then, higher levels of either autonomy or flexibility will tend to increase uncertainty about whether a negotiated agreement will be palatable to the principals when it is finally brought home. A broadly defined mandate may give an agent more room to construct a suitable agreement, but it makes him more likely to exceed the bounds of what a principal is willing to tolerate.

Ownership over the process also has a significant role to play in determining the support of a Level II player.  A public or government agency not primed or engaged in the process will be less supportive of the outcome that has been reached.  It is far simpler to sell an outcome to a constituency that feels it has been adequately represented during the proceedings, especially in matters of civil or international conflict (Zartman and Berman 1982, pp. 215–219).  To achieve this level of inclusiveness, it may be beneficial to restrain an agent’s autonomy in the negotiation process even when he is trusted to act competently on his own.  A steady give and take of information will prepare the parties for reentry and remove the excuse that their views were not adequately represented. 

Conclusions

These observations lead to several conclusions that refine our understanding of agent independence in international negotiation.  A negotiator is most likely to succeed when given a mandate that clearly articulates the priorities, interests, and alternatives of the principals without placing undo constraints on the agent by defining acceptable and unacceptable solutions.  This is particularly true in the diagnosis phase, when such information will facilitate signaling and probing, and in the details phase, when an agreement must be brokered such that it will survive reentry.  Autonomy over the process is crucial in all stages, but especially in the formula stage, when the way forward must be constructed between the two parties.  However, autonomy in the formula and the details phases threatens the likelihood of successful reentry by reducing ownership on the part of the principals and obscuring the continuity between initial positions and final outcomes.  Finally, authority at each phase increases the likelihood of a settlement.  It helps build trust in the diagnosis phase, encourages innovation in the formula stage, resolves sticking points in the details phase, and silences competing voices in the reentry phase. It expands the bargaining space and encourages concessions.  This same authority will reduce an agent’s ability to extract favorable concessions.  Such authority also presupposes that the agent has a full understanding of the positions of the principals throughout the negotiation process and that there is a high level of trust established between the principals and the agent. 

Returning to Nicolaïdis’ model, these observations give insight as to the viability of her three hypotheses.  With respect to the first, flexibility must be more sharply defined in terms of interests, priorities and alternatives on the one hand and settlement points on the other.  In any situation, an agent is best off having a great deal of flexibility over the latter, subject to the details of the former. But the agent’s mandate not only defines the dimensions of the bargaining space, it also determines how clearly this information will be conveyed to the agent.  As such, it may be necessary to reduce flexibility below the level recommended by the size of the bargaining space alone to ensure that the boundaries are clear enough to discern.

In the case of the second hypothesis, this discussion has demonstrated that in the in international negotiations, it is difficult to determine the alignment of principals and agents or to distinguish this from the alignment of principals to one another.  In either case, the propensity of certain principals to reject a negotiated agreement decreases with their ownership over the process. Thus, as Nicolaïdis suggests, with loosely aligned interests, autonomy and/or authority should be lowered. 

Finally, regarding the third hypothesis, while Nicolaïdis suggests that changes to autonomy be used as a catchall to substitute for changes in flexibility or authority, the merits of such a recommendation become less obvious after examining it during the different negotiation phases.  Of the three parameters, autonomy has the most wide-reaching effect on the negotiation process, aiding trust building in the diagnosis stage and creativity in the formula stage while contributing to judgment errors in the details stage and reentry problems after a resolution has been reached.  As such, it is preferable to adjust agent autonomy to contend with direct implications on the process rather than to use it as a proxy for changes to other actions.