Georgetown University's Health Systems Administration
Decision Analysis in Healthcare
        

 

Conflict Analysis


 

Attitudes toward conflict have shifted. Conflict, once considered a problem, is now seen as a potentially positive and creative force. This understanding has led us to prefer to manage, rather than eliminate, conflict. Effective management of conflict depends on an analysis that points the way to constructive outcomes—to win-win solutions. 

This chapter describes a conflict analysis methodology (CAM) that is designed to resolve conflicts and build consensus among competing constituencies. The steps in CAM are identifying the constituencies, their assumptions, their goals, and the appropriate spokes-people to represent them; identifying the issues underlying the conflict and their possible levels of resolution (possible solutions); developing and analyzing “treaties” (resolutions with a separate agreement on every contested issue); and following a structured process of negotiation and consensus building to agree on a final treaty.

The idea of a decision analysis framework for examining and resolving conflicts is not new. Theoretical models with specific prescriptions have long been discussed in the literature (Raiffa 1982; Von Neumann and Morgenstern 1944). This chapter, however, presents not a specific model but an entire methodology to increase all parties’ understanding of the conflict.  It includes not only a mathematical formula but also behavioral advice on what to do when.  It is a methodology to support the process of conflict resolution by addressing the basic sources of conflict: lack of understanding, lack of information, and distorted communication. After these roadblocks are disposed of, we attempt to break down the conflict into component issues that can be traded off. If we succeed, each party achieves a victory on the issues it deems important. 

Conflict Analysis may be applied in different situations:

  • To assemble conflicting parties to find general areas of agreement and a solution that meets the concerns of all parties
  • To help a neutral observer or mediator understand the issues and priorities of the parties to a conflict
  • To help one party clarify its position and perhaps role play the opposing positions (This role play clarifies the opposition’s values and perceptions, enabling one side to understand the opposition’s viewpoint and to develop a negotiating strategy.)

In any conflict analysis, a central question is who is the decision maker?  CAM can be used in several situations.  First, if several constituencies recognize that a conflict must be resolved and are willing to take the necessary steps.  An example is when department heads meet to resolve their conflict about budget.  Second, CAM can also be used if one party wants a deeper understanding of the conflict.  For example, a clinician negotiating an agreement with a Health Maintenance Organization might want to explore various contract provisions.  

If a constituency cannot or will not participate in the analysis, a group of “objective outsiders” should be asked to role-play it. While such refusal or inability to participate may reduce the chances that a chosen treaty will be implemented, having proxies is better than omitting pertinent viewpoints from the analysis. Note that it is helpful if the objective outsiders are highly regarded by the missing constituency.

Who should be invited to participate in the analysis depends on the purpose of the analysis. If a manager wants to explore how the other side might react to various proposals without committing to an action, the analysis must be performed with a group of objective outsiders. But if the purpose is to raise awareness about the problem or reach an agreement on it, the analysis should involve as many actual parties to the conflict as possible.

A related point of discussion is our preference to meet alone with each constituency. Conflict analysis is a process to increase the understanding of each constituency’s position, increase the consistency of its choices in subsequent negotiations, and identify different ways of solving the conflict by finding areas of possible agreement with the opponents. Many of these goals can be realized better in private meetings of constituencies.  Group meetings can follow but should not precede individual model building sessions.  When people in conflict meet before conflict is understood, they may stress their disputes instead of their agreements and escalate the conflict to personal issues.  They may force each other into a corner reducing the possibility of later compromise.  It is better to wait and meet together when various points of views are better understood and the possibility of miscommunications is reduced.

The premises underlying our methodology are:

  1. People have cognitive biases that become more acute under conflict or crisis.   When negotiations are complex (meaning they are based on many issues), these cognitive limitations prevent full consideration of possible solutions.
  2. Preconceptions and false assumptions impair the ability to make the trade-offs that can lead to a solution.  Each party may erroneously assume they know the priorities of the other.   Asking for the priorities can reduce these misconceptions.  When two parties negotiate some conflict is due to their differences but other conflicts are due to miscommunications.  CAM reduces conflict due to poor communication.
  3. A conflict is easier to grasp if it is broken into components (some of which may have little or no conflict).  Many parties that see each other in conflict are surprised to find out that they have large areas of agreements.
  4. Individuals can specify their values and prioritize them by using a structured process.   In the chapter on modeling preferences we have shown that this can be done.  In CAM, the premise is that preferences can be modeled even when people are in conflict.
  5. The analyst can learn about every viewpoint by modeling each party separately.  When one party is not available, the analyst can interview observers who can role-play the values, preferences, and priorities of absent parties.  
  6. The analyst can (a) reduce conflict due to miscommunication, (b) provide insight into the thinking of each party and (c) identify solutions overlooked by both parties.  A conflict analyzed is a conflict understood. The analyst can emphasize that there are wide areas of agreements, so the conflict does not escalate.  The analysis injects rationality in highly charged situations and helps each party gain new insights.  The analyst can use models of parties to identify better solutions.  Solutions that are often overlooked by both parties.

For an example, consider conflict resolution at end of life.  Family members will have different perspectives on what needs to be done.  The health care provider bears the cost while it waits for a consensus to develop.  CAM can help by breaking down the decision into its components and reducing the conflict due to miscommunications.  CAM can also help by producing solutions overlooked by the parties. 

When thinking about managing conflict, we must always consider escalation (the movement of conflict to a higher level of tension where potential for losses are increased), a debilitating syndrome with many deleterious effects. The sources of escalation are a lack of understanding among the parties, a lack of information about the opponent’s position, and an emphasis on bargaining behavior. CAM counteracts these problems by increasing communication, emphasizing a problem-solving attitude, offering a joint definition of the problem, reducing the influence of ideology, and pointing the parties toward win-win solutions.

Conflicts can become so heated and the constituencies so stubborn that rational approaches cannot manage them. Often, however, all sides realize a decision must be made and that a better understanding of one another’s position is essential. In that case, it becomes helpful for the parties to know the goals, attitudes, values, motivations, and levels of aspiration of the parties; the nature of the vital issues in the conflict and the options available to resolve them; and the consequences of taking each possible action. Unfortunately, research suggests that people are inconsistent in their judgment, unaware of their values, and unable to explain accurately not only their opponents’ positions but their own as well (Balke et al. 1973; Janis and Mann 1977). 

Conflict analysis must represent each constituency’s value structure without distortion, Difficult though it may be, a successful outcome depends on all constituencies’ regarding the analyst as neutral and objective. By objectivity, we do not mean a cold and aloof tone but an ability to feel the concerns and values of all constituencies. Distorting or failing to integrate into the model every viewpoint might persuade the excluded constituency to refuse to participate or to sabotage the efforts at resolution. Distorting a constituency’s values will also undermine the model’s value as a description of the conflict and a tool to generate treaties. Thus, practitioners must develop not only model-building skills but also interpersonal abilities to interview, assimilate, and role-play other people.

The Methodology

So what is this methodology that promises to do so much? CAM consists of the three major phases (shown in Table 1), which can be broken down into 10 steps.

The first phase helps the analyst understand the underlying issues and gain a perspective on the background of the conflict. In the second phase we explore the problem by refining and decomposing the preliminary goals into specific components, which we call issues. In the third phase, we explore solutions to the entire conflict by analyzing treaties. The third phase of analysis begins by asking the constituencies to weight the relative importance of the various issues and to state preferences for possible levels of resolution. We then package one level of resolution for each issue into a “treaty” and score it to see how well it meets each constituency’s needs. In the next step we search for the optimum treaty for all parties (or for a specific party if that is our purpose). Then we convene all parties to search for a consensus resolution. Because the environment can be explosive at this stage, we determine the sequence in which to debate the issues on the basis of criteria such as the sources of disagreement and the levels of resolution.

Step Phase Action Purpose
1 Understand the problem Hold informal interviews with people informed about the conflict. To get an overview of the conflict and write a preliminary list of goals and their importance.
2 Identify constituencies and their spokespeople. To define the coalitions and identify individuals who will later develop the conflict model and choose treaties.
3 Analyze assumptions. To obtain a general understanding of the problem, identify ideological and technical sources of conflict.
4 Structure the problem Perform an in-depth interview with one or more objective outsiders. To refine the goals identified in step 1, check understanding of issues and different set of values involved, and begin exploring grounds for resolution. 
5 In separate sessions with each constituency, identify issues, and levels of resolution. To develop the conflict model and identify key levels of resolution that lead to an overall compromise.
6 Assess importance of issues & value of resolution levels To quantify the importance and preference of the component issues and the levels of resolution,
7 Explore the solution Form and score treaties To generate a set of feasible solutions and explore Pareto optimal solutions.
8 Analyze the treaties To generate treaties that are likely to resolve the conflict.
9 Develop a strategy of negotiation To increase the likelihood of a positive outcome.
10 Present results to all constituents. To generate an acceptable treaty and develop guidelines to implement the treaty or agree on further actions in the resolution process.

Table 1: Summary of the Conflict Analysis Methodology

The third phase concludes when we assemble the spokespeople to discuss the conflict model and suggested treaties and decide on further action. Note that even if the representatives agree on a treaty, they still must “sell” it to their constituencies, so agreement among spokespeople does not guarantee the conflict will end.  

We will illustrate the application of CAM by discussing the issue of whether parental consent should be required before family planning services are provided to adolescents. We assume that a hospital administrator is developing a family planning service that could meet the community needs —one that does does not alienate any major interest groups and yet meets a growing market demand.

Phase 1. Understand the Problem

 

Step 1:  Informal interviews

During phase I, the analyst should obtain an overview of the conflict, identify the decision makers, and select who will represent them. In our terminology, we will identify the constituencies, spokespeople, and goals.  To accomplish this, the analyst sets up informal interviews with various constituencies. 

Step 2:  Identify constituencies and spokespersons

Constituencies - people or groups that stand to gain or lose as a result of a conflict - can be identified in several ways, such as by examining lists provided by lobbying organizations, by reviewing testimony on past legislation, by inviting local chapters of national organizations to a discussion and by interviewing key individuals by telephone. Commonly, many groups with various persuasions must be taken into account. However, the value systems of these differing groups toward family planning may be similar enough to be characterized by just two or three models. Such a simplification is a significant help in devising a useful solution. In our example, parents, ministers, physicians, health educators, social workers, and women’s rights advocates might be grouped in ways they are comfortable. The analyst can decide whether groups can be lumped into one constituency by asking several prominent organizations to assign priorities to a set of goals about the conflict. This empirical approach works well if all relevant groups are surveyed.

A common mistake is to canvass only those groups that have lined up against each other. Although the first glance might suggest that only two groups are in opposition, further analysis may identify other important players. Using our example, at first it appears that just two constituencies are involved in the conflict: the “antiabortion forces” and the “family planning advocates.” However, further examination might unearth a third group, which we could call the “concerned parents”—people who believe their teenage children are mature enough to make correct decisions if given balanced, unbiased information, These concerned parents care mostly about their children’s well-being, and if they have an ideology on abortion, they do not want it to influence their children’s actions.

Now that the preliminary analysis is finished, the analyst must identify a spokesperson to represent each constituency participating in the conflict analysis. The analyst needs at least three types of spokespeople: the proponents and opponents of the conflict and objective outsiders. The third group is valued for having a perspective that differs from those of the constituencies. In contrast, if policymakers do not want to negotiate with a constituency but merely wish to increase their understanding of the conflict, the analysis can be performed using only objective outsiders instead of a spokesperson.

Spokespeople should be good at identifying issues and solutions and comfortable with the task of quantifying preferences. While later in the process we may want to include individuals with institutional power to implement a compromise, during the analysis phase the analyst needs people who are willing to break the problem into its components, are also sensitive, insightful, and articulate.

A nomination process to identify spokespeople begins by identifying five or six nominators—people who know the leaders and insightful people in the field and can identify individuals who might adequately represent a constituency.  It is good practice to select only nominees who are suggested by several nominators because this indicates the person is widely respected.  While in general having many people consider somebody is an expert will signify that person’s credibility, even the most widely respected will have opponents and the most knowledgeable his or her cynics. Remember, what we seek is not respect across all constituencies as much as simply the ability to represent how a specific constituency perceives a given situation, and to offer some indication of how it will act. Universal acceptance is not essential.

When talking with the nominees, the analyst can motivate them to participate by mentioning:

  • Who nominated them (make sure to get permission to use the nominator’s name)
  • What the project is about
  • Why their participation is important
  • What will and will not be done with the results
  • How their names will be used
  • What tasks will expected of them
  • How long each task will take and when it will occur
  • What payment (if any) they will receive

In general, three spokespeople from each constituency are sufficient, although there may be good reasons, like attrition, to identify more nominees. During a long process, some spokespeople may drop out because of other commitments or a loss of interest. The analyst must beware of running out of spokespeople before concluding the analysis. A second reason for more than one representative is that in a complex analysis, especially one involving a series of technical and ethical issues, a single spokespeople may be unable to convey the full spectrum of a constituency’s position.

Step 3:  Analyze assumptions

After the preliminary analysis is finished and the spokespeople have agreed to participate, the analyst must do some homework. This is because conflict about the definition or solution of a problem often arises from sharp disagreement over a fundamental factor that is not obvious to anybody, even the stakeholders themselves. We are referring to conscious or unconscious assumptions.

Now the analyst must bring these assumptions to the surface for examination. This step has three purposes: to compare and systematically evaluate the assumptions, to examine the relationship between assumptions and each constituency’s policies, and to formulate proposals that are acceptable in light of the assumptions.

To elicit the assumptions, the analyst must gain as much information as possible about each constituency’s views, Leaflets, brochures, advertisements, position papers, legislative hearings, and data used by a constituency to buttress its position are all valuable clues. Interviews may also be useful, but more often than not assumptions are so ingrained that people accept them unconsciously and have difficulty articulating them. Commonly, individuals are surprised by their assumptions once they are made explicit. This can lead to reconsideration.

Throughout this process, look for recurring catchwords or slogans in a constituency’s statements. Slogans are chosen for their emotional content; they contain a wealth of information about the constituency’s values. In the example of parental notification about family planning services for adolescents, constituency I said, “I want my children to have the courage to say no.” Constituency II said, “Let’s stop children from having children.” While such statements simplify the conflict, they certainly give the flavor of the competing positions. This tone allows the analyst to devise whether the debate is emotional or technical, and whether it concerns ethics or money.  Again, remember that at this stage the analyst is trying to understand where the different constituencies are coming from and to summarize their positions.

Table 2 shows some assumptions behind the world views of constituencies I and II. A close reading shows that assumptions can be classified as contradictory and non-contradictory. Constituency I’s assumption that access to contraceptives lures teenagers into sex directly contradicts constituency II’s assumption that access to contraceptives does not increase sexual activities. Clearly, the two assumptions cannot both be true at the same time. A non-contradictory pair of assumptions is also shown. While constituency I assumes that administration and red tape will soak up as much as 90 percent of the funds, constituency II assumes that the cost-benefit ratio of family planning programs is excellent. These assumptions seem to be in opposition to each other, but both can be true at once.

Catchwords

Constituency I Constituency II
Too much government Let's stop children form having children
I want my children to have the courage to say no Contraception is better than unwanted pregnancies
   

Assumptions

Constituency I Constituency II
Administration and red tape will eat as much as 90 percent of the funds The cost benefit ratio in family planning programs is excellent
Contraception is dangerous and people are misinformed about its effects Lack of pregnancy allows minors to take advantage of other possibilities (education, employment, etc.)
Parents provide a better job providing sexual education Parents do not provide adequate sexual education
Morality is the best contraceptive Counselors in family planning agencies provide the most persuasive influence against premarital sex
Access to contraceptives lures teenagers into sexual activities Access to contraceptives does not increase sexual activities
The decision to have sex is a good opportunity for establishing communication between parents and children Confidentiality is crucial in obtaining planning services

Table 2:  Catchwords and Assumptions of Two Constituencies

It is crucial that the constituencies perceive the analyst as a fair individual who is sensitive to their values. To accomplish this, the analyst should use the same terminology as the constituency and restrain from evaluating the validity of ascertains.  The analyst can ask clarifying questions but not make any judgment regarding the response.

Phase 2. Structure the Problem

After the analyst has identified the constituencies and their general goals and assumptions, and the spokespeople have agreed to participate, it is time to model the conflict. In this phase, we break down the conflict into its components and then quantify each one. These components are: the goals of each constituency, the issues that form the heart of the conflict and that must be addressed to resolve it, and the possible levels of resolution for each issue.

Step 4:  In-depth interviews

During this phase, the analyst will have two sets of meetings, one with objective outsiders and one with the spokespeople for each constituency. Performing in-depth interviews with a few objective outsiders is an excellent way of preparing and rehearsing before the actual sessions with the spokespeople. Objective outsiders need not be experts in the subject matter—they can be trusted associates who are acquainted with the problem and not intimately involved with its solution. These people are chosen for their analytical skills, candor, and willingness to participate. The sessions with the objective outsiders give the analyst a preliminary conflict model that serves as a starting point for the constituency sessions.

During both the proxy interviews and the spokespeople group sessions, the analyst:

  1. Discusses the problem in general terms and notes examples of goals, issues, and levels of resolution. The analyst asks questions to help define the problem and ensure that its key elements are understood.
  2. Asks for descriptions of the parties: Who are the proponents? Who are the opponents? How do they view the conflict? What would each side like to see in terms of a resolution? Why?
  3. Asks for a list of the goals that drive the parties, and gives examples of goals so participants know the meaning of each. Brainstorming is a good way to draw up a comprehensive list.
  4. Asks for an exhaustive list of the issues that are dividing the sides, to make sure they are as independent of each other as possible (i.e., there is no overlap). Possible questions include:
    • What is the underlying conflict?
    • On which issues do the opponents agree or disagree?
    • Which issues must be resolved for the sides to reach agreement?
  5. Asks for a list of all levels of resolution for each issue. Some levels will be preferred by one side, and other levels by the opponents. Some levels will be compromises that are not preferred by any side but may be acceptable to all. The analyst tries to identify as many levels of resolution as possible on each issue even though only a few may be considered in the final analysis. The reason for seeking so many levels is to promote the development of creative levels. Ask questions like:
    • What is this side’s position on the issue?
    • What solutions might it accept?
    • What solutions are totally unacceptable?
    • Why are the levels considered in this way?

Goals, issues, and levels of resolution can be identified in several ways, including a questionnaire, a survey, Delphi technique, and/or individual interviews. A private meeting of each constituency is held to obviate the risky and possibly dysfunctional step of assembling the opponents before the analyst knows the depth and nature of the conflict. Later, after the conflict analysis model has been formed and some trade-off resolutions have been mapped out, the constituencies may be assembled to search for solutions, or treaties.

Step 5:  Identify issues & level of resolutions

Issues are the basic building blocks of conflict; they are fundamental factors that must be understood and addressed in order to reach resolution. Once the issues are identified, we can classify them on a continuum from agreement to disagreement. This will allow us to concentrate on finding acceptable compromises or trade-offs on intensely disputed issues. The classification may reveal, for example, that the constituencies agree on several issues, that they are mildly opposed on others, and that the conflict really concerns just a couple of issues. This makes the conflict appear more manageable and focused and thus simplifies its resolution.

If disputed issues are not ripe for resolution because their political time has not arrived, or if they are too thorny to be resolved, the analyst can try to develop a partial solution by concentrating on more tractable issues.  Either way, separating the conflict into component issues allows the analyst to localize the problem to specific areas and use resources to solve problems more effectively.

Finally, the levels of resolution—the specific actions, laws, services, etc., that can resolve an issue—must be identified. Typically, constituencies identify levels of resolution that range from optimal to unacceptable. This step tells the analyst what each constituency is considering or fearing in terms of suitable and unsuitable solutions, and it serves as a foundation for generating new solutions.

Let’s return to our example of the issue of parental notification. Suppose that after performing these interviews with two objective outsiders and with meetings of the spokespeople, the list of goals were to:

  • reduce unwanted pregnancies,
  • teach children to be responsible,
  • preserve the family, and
  • reduce the number of abortions.

Furthermore, suppose the interviews revealed that the conflict can be distilled into two issues: which components of family planning should be available to minors and under what conditions. Suppose also that both constituencies agree that any family planning program must have at least three components:

  • Education includes topics such as values, morals, biological processes, birth control, decision making, goal setting, sex roles, pregnancy, and parenting skills.
  • Counseling focuses on some of the same issues as education but involves more interaction between provider and client. The focuses of counseling might include crises, pregnancy, abortion, elective non-parenthood, and preparation for childbirth.
  • Services might include birth control devices, adoption services, abortions, prenatal care, sexually transmitted disease testing and treatment, and financial assistance.

With this information, the analyst is now in a position to explore the issues at the heart of the conflict. In this example, the crucial issues are:

  • Should values and morals be taught when delivering family planning services?
  • Should counseling of adolescents start from the position that premarital sex among adolescents is bad?
  • Which is more important, allowing easy access to services or having services controlled by organizations with what are considered high morals?
  • What are the optimum technical qualifications of the personnel?
  • Who, if anybody, should regulate the provision of family planning services to adolescents?

Once the issues have been identified, the objective outsiders and spokespeople are asked, separately, to define a set of feasible resolutions to each issue. A set of issues and their possible levels of resolution are listed in Table 3.

Once the issues and levels of resolution have been formulated by the full group of spokespeople, they must be checked, detailed, and rephrased in separate sessions with spokespeople from each constituency. It is critical to obtain a consensus on the phrasing and substance of all issues and all levels of resolution from all parties before proceeding. Note that at this time the goal is to have all the constituencies agree on the components of the conflict, not on a solution to it.   In many conflicts, merely listing the issues and the possible resolution levels reduces the conflict.

  Issues   Levels of Resolutions
A To what extent should family planning programs try to convince clients that adolescent sex is bad? A1 Should not do so
A2 Should be available to clients
A3 Should be required of all clients
A4 Should be a fundamental part of every service of a family planning program
B To what extent should family planning programming be oriented toward strengthening the family? B1 Not at all
B2 Depends on the client
B3 Always, with all clients
C What limitations to access to family planning programs should exist for adolescents? C1 No parental notification
C2 Parental notification before counseling or services
C3 Parental permission before counseling or services
D What type of supervision should be required for people who provide family planning services to -adolescents? D1 Social work supervision
D2 Physician supervision
D3 Theologist supervision
D4 Experience as parent of adolescent
E What organization (with what moral qualifications) should be allowed to deliver family planning services? E1 Nonprofit
E2 Educational
E3 Governmental
E4 Health care (doctor’s office or hospital)
E5 Religious
F Who should regulate the provision of family planning services for adolescents? F1 Peer review
F2 Local government
f3 State government
F4 Federal government
F5 Community

Table 3:  Issues and Levels of Resolution

 

Step 6:  Elicit weights and preferences

In this step, the analyst helps each constituency estimate the importance of various issues and the values assigned to different resolution of issues.  Methods to elicit weights and preferences are found in a previous section and in Huber (1980) and Keeney and Raiffa (1976).  Two sets of parameters need to be estimated.  First, we must identify the value of different resolutions to each constituencies.   This is done by listing the various levels from least to most preferred and assigning the most preferred option the value of 100 and the least preferred option the value of 0.   All resolutions in between the best and worst levels are assigned a proportional value.  Alternatively, if more precision is needed, the double anchored estimation method can be used.  The most preferred level is assigned 100, the least preferred level is assigned 0, and the spoke person is asked to rate the remaining levels.  

The relative priorities between the issues are assigned using the method of ratio estimates.  The spoke person is asked to list the issues in their order of importance.  The least important issue is assigned a value of 10.  The spokesperson is asked to rate how many more times the next issue is more important.  The process is continued until all issues are judged relative to one lesser important issue.  The ratio judgments are used to assign a score to each issue.   The scores are divided by the total of scores to produce the weight for each issue.  This manner of developing issue weights makes sure that weight for issues are constrained to add to one, thus if the spokesperson weights one issue heavily he/she is forced to weight other issues less important. 

When there are multiple spoke persons involved for one constituency, and if they differ in their estimates or weights for issues or values of different resolution levels,  they are asked to discuss their differences and come to consensus.  If their differences is small, responses from different spokesperson for the same constituency are averaged.  Note that during the discussion among the spokespersons, the analyst might find certain factions of a constituency to be more amenable to compromise. This is signaled by the faction having priorities similar to the opposing parties. In the extreme cases, the analyst might find that the spokespeople on one side actually represent several positions, not just one. If so, the analyst may need to develop a distinct model for each position.

Table 4 shows the resulting weights and value scores for different constituencies: 

Issues Levels of Resolutions Constituency I Constituency II
Weight Value Weight Value
A To what extent should family planning programs try to convince clients that adolescent sex is bad?  A1 Should not do so 0.14 0 0.07 100
A2 Should be available to clients 30 90
A3 Should be required of all clients 80 20
A4 Should be a fundamental part of every service 100 0
B To what extent should family planning programming be oriented toward strengthening the family?  B1 Not at all  0.20 0 0.04 100
B2 Programs available  30 90
B3 Programs required 90 20
B4 Built into all components 100 0
C What limitations to access to family planning programs should exist for adolescents?  C1 No parental notification  0.25 0 0.48 100
C2 Parental notification before counseling or services  80 10
C3 Parental permission before counseling or services  100 0
D What type of supervision should be required for people who provide family planning services to -adolescents?  D1 Social work supervision  0.11 0 0.14 100
D2 Physician supervision  60 50
D3 Theologist supervision  100 30
D4 Experience as parent of adolescent  50 0
E What organization (with what moral qualifications) should be allowed to deliver family planning services?  E1 Nonprofit  0.23 0 0.16 100
E2 Educational  60 30
E3 Governmental  30 15
E4 Health care (doctor’s office or hospital)  80 10
E5 Religious 100 0
F Who should regulate the provision of family planning services for adolescents?  F1 Peer review 0.07 0 0.11 100
F2 Local government 90 20
F3 State government 40 30
F4 Federal government 10 40
F5 Community 100 0

Table 4:  Issues and Levels of Resolution

Once levels of resolution for all issues have been identified and the weights and preferences assessed, it is useful to do an initial review. For instance, we know that preserving the family is important to constituency I and of relatively little importance to constituency II. This suggests that resolutions that enhance the family might be included in the model as long as they do not impair other goals. For example, family planning organizations could teach parenting skills to parents of adolescents or develop educational and counseling programs to teach adolescents to get along with their families. 

Phase 3. Explore Solutions

 

Step 7: Form and score treaties

A treaty is a set of resolutions for each issue in the conflict: one level of resolution per issue. Each resolution has a particular value to the constituency and each issue has a given priority by each constituency. The goal is to find a treaty—a combination of resolutions—that has a high value score for all constituencies.

Calculate the value of a treaty by multiplying the value of one level of resolution by its issue importance weight, doing this for all issues, and summing the results. Suppose family planning legislation passed with these levels of resolution (taken from Table 4):

  • A2 Programs stress the negative aspects of adolescent sex.
  • B2 Programs to strengthen the family would be required in any adolescent family planning service.
  • C2 Parents must be notified when an adolescent uses a family planning program.
  • D2 All providers of family planning services must have medical qualifications.
  • E2 Educational institutions will carry out family planning programs.
  • F2 State governments must regulate family planning programs.

The value of this treaty is calculated by multiplying the relative importance weight of an issue by the value assigned to the level of resolution, then adding across all issues. That is, the score of treaty k for constituency c is equal to:

kc = ∑i=1, ..., nWci Vcij 

Where:

 kc  is the score for treaty "k" (a set of levels of resolution for all n issues) evaluated for constituency "c"
Wci is the importance weight of issue "i" for constituency "c"
Vcij is the value of level of resolution "j" to issue "i" for constituency "c"
n is the number of issues underlying the conflict
i is the issue number
j is the resolution level within the issue

For constituency I, the value of the treaty described above is (values taken from Table 4):

(.14 X 30) + (.20 x 30) + (.25 x 80) + (.11 x 60) + (.23 x 60) + (.07 x 40) = 53.4

For constituency II, the value of the treaty is:

(.07 x 90) + (.04 x 90) + (.48 x 10) + (.14 x 50) + (.16 x 30) + (.11 x 30) = 29.8

How good is this treaty? Can we satisfy both constituencies better? These questions are addressed in the next section.

The point of analyzing treaties is to identify those treaties acceptable to all parties. A large number of treaties is typically available—in our example on parental notification, 4,800 treaties could be formed. In general, the number of possible treaties N is equal to:

N = ∏i=1, ..., n  xi

where xi is the number of levels of resolution attached to issue i, and n is the number of issues. One way to reduce the number of possible treaties is to look at only two values, best and worst resolution for each issue.  In these circumstances, the number of possible treaties becomes two to the power of number of issues involved.  With our model on family planning, the possible number of treaties with best and worst levels is 2 to the power of 6 or 64 possible treaties.  This reduces the number of possible treaties but how could we choose among these treaties?   

Step 8:  Analyze treaties

A treaty is Pareto optimal if it allows one party to get as much as possible without damaging the opponent’s position, and vice versa. This concept of mutual improvement is called the Pareto optimality criterion (Raiffa 1982). Pareto optimality is reached when one side cannot improve its position without degrading the position of its opponents. In Figure 2, the circled treaties are Pareto Optimal and all other treaties are not.  All the treaties outside the circle can be improved for at least one party (without depreciating the position of the other party) by moving to a treaty in the circle. 

Figure 2:  Value Scores for Two Constituents and Pareto Optimal Treaties

In a graph of value of treaties to two constituents, you can find the Pareto optimal treaties by connecting a line from the treaty having the maximum possible value to one constituent to the treaty having the maximum possible value to the other constituent.   Typically these are treaties having the value of 100 and zero to the two constituent and the treaty having the reverse value.  If there is conflict between the two constituents, all treaties should fall below this line.  The treaties closest to the line are Pareto optimal. 

The purpose of analyzing treaties is to devise a few that are acceptable to all constituencies and that can recast the conflict as a win-win situation. The analyst and constituencies should explore all trade-offs and compromises in a cooperative manner.

The analysis of treaties may be a time-consuming process, requiring calculations, figures, and computer programs, so it should be performed by the team of analysts. After the analysis is done, convene meetings with spokespeople from each constituency separately to review the promising treaties and do more analysis if necessary. Again, consult a few objective outsiders on the preliminary results before conducting these meetings.

The examination of alternative treaties is an iterative process that can be simplified with computer programs or examined intuitively by using Table 4 to ponder improvements in the negotiations. The important point is that understanding the relative importance of issues and levels of resolution often allows us to find treaties that trade off and improve the outcome for everybody.

It is common to approach conflict resolution by seeking a compromise on each issue in turn. Let’s see how compromising on issues one by one can lead to inferior solutions. Suppose that both sides negotiate and settle on something in between the two extremes in all issues (i.e. something close to value of 50, halfway between 0 and 100).  Table 5 shows the resulting treaty and the total value of the treaty to each constituent:

Issues Levels of Resolutions Constituency I Constituency II
Weight Value Weight Value
A To what extent should family planning programs try to convince clients that adolescent sex is bad? 

A2

Should be available to clients 0.14 30 0.07 90
B To what extent should family planning programming be oriented toward strengthening the family?  B2 Programs available 0.20 30 0.04 90
C What limitations to access to family planning programs should exist for adolescents?  C2 Parental notification before counseling or services  0.25 80 0.48 10
D What type of supervision should be required for people who provide family planning services to -adolescents?  D2 Physician supervision  0.11 60 0.14 50
E What organization (with what moral qualifications) should be allowed to deliver family planning services?  E2 Educational  0.23 60 0.16 30
F Who should regulate the provision of family planning services for adolescents?  F3 State government 0.07 40 0.11 30
 

Overall value of this treaty:

53.4 29.8

Table 5:  Value of Satisfying Both Constituencies on All Issues

The value of this treaty for constituent I is 53.4 and the for constituent II is 29.8. 

However, we suggest an alternative approach called logrolling. In log-rolling gains in some issues are traded off against losses in other issues so each party can win the ones it considers most important. Researchers have found that conflict resolution processes that allow logrolling lead to more Pareto optimal settlements and higher value to each participant than processes based on issue-by-issue compromise (Froman and Cohen 1970).  Logrolling’s advantages over compromising within each issue are that it

  • Defuses ideological disputes
  • Increases Pareto optimality
  • Satisfies both parties better
  • Encourages looking at the big picture
  • Does not divide the original conflict into several conflicts

On the other hand, the one-by-one compromise approach tends to exacerbate the total conflict. If one side wins issue "x," the losers may try all the harder to win issue "y." The very nature of the procedure encourages the parties not to search for the best overall solution.

Lets examine logrolling with our data.  In Table 6, we provide a treaty in which the constituent who cares most about an issue receives the best possible resolution on that issue. 

Issues Levels of Resolutions Constituency I Constituency II
Weight Value Weight Value
A To what extent should family planning programs try to convince clients that adolescent sex is bad?  A4 Should be a fundamental part of every service 0.14 100 0.07 0
B To what extent should family planning programming be oriented toward strengthening the family?  B4 Built into all components 0.20 100 0.04 0
C What limitations to access to family planning programs should exist for adolescents?  C1 No parental notification  0.25 0 0.48 100
D What type of supervision should be required for people who provide family planning services to -adolescents?  D1 Social work supervision  0.11 0 0.14 100
E What organization (with what moral qualifications) should be allowed to deliver family planning services?  E5 Religious 0.23 100 0.16 0
F Who should regulate the provision of family planning services for adolescents?  F1 Peer review 0.07 0 0.11 100
  Overall treaty score 57 73

Table 6:   Letting Each Constituent Win on Issues They Care Most About

The values for this treaty would be 57 and 73 improving the satisfaction of both constituents over the treaty in Table 6. Thus, we believe this type of logrolling should be encouraged at the expense of compromising on issues one by one.  Logrolling produces Pareto Optimal treaties, in which no one can improve their situation further without hurting other participants. 

It is crucial that before the next step—in which the analyst and all constituencies jointly generate the resolution treaty—each constituency understands the other side’s view and values. This understanding can be increased and clarified through the analysis of treaties presented here.

Step 9:  Develop a strategy of negotiation

Now we have a model of conflict acceptable to all parties, a set of weights representing the issue importance, and preferences for the various levels of resolution. Each set of weights represents the value structure of one constituency. With our model of conflict and value structures, we have created, scored, and analyzed treaties and discovered, after discussions with spokespeople for each constituency, a few treaties that might be acceptable to all parties.

Coming up with an "agreed-upon" conflict model is a big achievement in itself. The opposing constituencies have been delineated, the conflict has been specified, the structure of the model (goals, issues, and levels of resolution) has been defined, and the value structures of the constituencies have been articulated.

Now we assemble the constituencies for the first time. The delay emphasizes the need for each constituency to understand its own position before discussing it with its opponents. To minimize conflict caused by lack of understanding of one’s own position, the CAM spends a great deal of time clarifying what those positions are and what feasible treaties might look like.

Now it is time to choose a course of action. Note that at this point the parties have agreed on the issues and feasible levels of resolution but not on the relative importance of the issues or on their preferences for levels of resolution. It is unlikely that agreement will be reached on all these matters, because some issues are purely ideological, but it is realistic to aim for agreement on certain issues and levels of resolution and to hope to compromise on the rest. The following discussion offers suggestions and guidelines to increase the probabilities of success at this stage.

Because the issues and level of resolution have been agreed upon, the emphasis of the meeting is to reduce differences of perception and judgment on the relative importance of disputed issues and levels. But which issue should be addressed first? The answer arises from an examination of the source of the disagreements, which the analyst performs before the plenary meeting. In our view, the source of disagreements can be classified along a continuum ranging from purely ideological to purely technical. 

The analyst must determine an order in which to present the issues to the constituencies. Here it might help to classify the issues as follows:

  1. Issues about which parties agree about the relative importance and the level of resolution to implement
  2. Issues about which parties disagree about relative importance but agree about the level of resolution to implement
  3. Issues about which parties agree about relative importance but disagree about the level of resolution
  4. Issues about which parties disagree about both relative importance and level of resolution -

The order in which the cluster of issues should be presented to all spokespeople is shown in Figure 1. Figure 1 shows that the process should start with those issues on which the constituencies agree (number 1). It is important to quickly establish a mood of goodwill and cooperation. By showing the areas of agreement, the analyst can enhance the mood for the later negotiations.   The analyst should then discuss issues where both constituencies have agreed about a preferred level of’ resolution. Again, this is to prove ground for agreement exists on some issues.

Figure 1:  Order of Presentation of Analysis

Discuss third the issues on which the constituencies agree about relative importance but disagree about the level of resolution to implement. Start by discussing issues having a low importance rating.  Finally, negotiate the differences on issues about which the constituencies disagree on both levels of resolution and importance. These are the crux of the matter and will take the most negotiating skill and likely the most time as well.

One way to decide on the order of issues to discuss is to examine the difference in weights assigned by the constituents.  Issues with large ranges would contain more conflict.  For example, in Table 4, the biggest conflict is in "To what extent should family planning programming be oriented toward strengthening the family?".  The difference in weights in this issue is 0.16 points, larger than any other issue.  It is likely that discussion of this issue would contain most of the conflict. The more disagreement and polarization there is about an issue, the larger the range (or variance) of the weights on it and the later in the discussion it should be presented. As before, this ordering typically implies dealing with purely technical issues first, and then moving gradually to the more ideological issues.

Why spend so much time determining the order to present and discuss issues? Primarily because when the parties to a hot conflict assemble, they can create such an explosive atmosphere that it’s good to get some agreement on some issues as quickly as possible. If a feeling of accomplishment and understanding can be instilled, the constituencies may suspend or water down their negative preconceptions of the other side and adopt a constructive, problem-solving attitude. Also, if a controversial issue cannot be resolved, the fact that some issues were settled diminishes the frustration of deadlocked negotiations and assures the constituencies that their efforts were at least partially successful.

Throughout the negotiations of differences, the analyst must keep in mind the Pareto optimal treaties that were developed earlier with the objective outsiders. At the beginning, when relatively non-controversial issues are being discussed, it is better not to bring up these “fair” treaties, so the constituencies cannot focus on the final outcome and shortcut the early stages, which help establish a positive attitude. Later, if negotiations are deadlocked, use the fair treaties to generate a breakthrough. These Pareto optimal treaties can prove that one constituency can improve its satisfaction without damaging the satisfaction of its opponents.

Finally, it is important to end the meeting with a concrete plan for action. If the analyst succeeded in bringing about an overall resolution and drafting a final treaty, then actions to implement this treaty should be agreed upon. On the other hand, if no treaty or a partial treaty was obtained, the analyst should list actions that will continue the conflict analysis and resolution. Remember that conflict analysis is being done with spokespeople who must sell the resulting agreement to their constituencies.

Step 10: Present results to all constituents

We have presented the conflict analysis methodology as a strategy to increase the understanding of the underlying sources of conflict. In this section, we show how CAM can deescalate a conflict and increase the probability of a negotiated settlement to conflicts that seem intractable.

CAM checks this escalation by:

  • Preventing the parties from negotiating on the overall treaty until some agreements have been reached
  • Discussing issues in a sequence that minimizes frustration
  • Dividing the conflict into component issues and increasing the probability of finding a few areas of agreement

Parties to escalating conflicts frequently forget their initial concerns and turn to trying to beat each other. The parties may think about “saving face,” “getting even,” “teaching the others a lesson,” or “showing them they can’t get away with this.” At this point, they more closely resemble a battered prizefighter than a reasonable participant in a public policy debate. When this stage has been reached, the conflict is likely to expand to other areas where it should not logically exist. These ancillary conflicts may be created by a party to make sure it has some ground to retaliate if it loses the main battle. Examples can be found when one party blocks Pareto optimal improvements for the other party.

Escalation is likely to increase the number and size of the issues under dispute and the hostile and competitive relations among the parties. The opponents may pursue increasingly extreme demands or objectives, using more and more coercive tactics. At the same time, whatever trust existed between the parties is likely to corrode. The ultimate stage in escalation is reached when the parties think differences exist across many issues (even some that were created solely for the sake of bargaining). A feeling of frustration settles in, along with the impression that the parties are incompatible, that compromise is impossible, and a fight for total victory is the wisest course.

Model-building has a cathartic effect, alleviating hostility

Researchers have noted that where substantial differences exist, parties must ventilate their feelings toward each other and talk about the issues dividing them before they can seek a solution that integrates the positions of all important parties (Thomas 1975). That is, a ventilation phase generally needs to precede an integration phase. CAM encourages, and actually demands, that parties state their positions, their perception of the sources of conflict, and their assumptions. The model-building phase usually has a cathartic effect, alleviating hostility and creating an atmosphere conducive to such an integration phase.

Thomas (1975), in a comprehensive review of the literature, identified several causes of escalation, among them lack of reevaluation, self-fulfilling prophecies and biases of perception, distortions in communication, increase in distrust and hostility, losing sight of the original issue and expanding the conflict to other issues, feeling of general incompatibility, and lack of a ventilating phase.

How can CAM address these causes of escalation? The reevaluation process may happen when a party, hearing the other side’s arguments, reevaluates its definition of the issue or its choice of preferred alternative. CAM allows for reevaluation at several points in the process, by facilitating communication, explaining positions, and emphasizing problem-solving behavior. The methodology insists throughout that parties develop an understanding not only of their position but of their opponent’s positions as well.

Self-fulfilling prophecies and biases in others’ perception directly affect the ability to make trade-offs. An interpretation of another’s behavior is largely a response to the image each party has of the other. A constituency that sees the other as cold and calculating will be suspicious of compromises its opponent suggests. If it sees the opponent as compassionate and respectful, the resulting goodwill can simplify the negotiations. CAM’s main thrust is to elicit rational, analytical behavior and to clarify, through an iterative process of questioning, the perceptions of all parties.

In conflicts, parties tend to use information to manipulate and coerce their opponents and/or the public. Communication becomes distorted, trust is diminished, and messages between parties lose credibility or are ignored. At this stage, a party can further distort its view of the opponents without contradiction. CAM tries to dispel misunderstandings by allowing an explicit presentation of the parties’ positions, and tries, by developing treaties, to serve as catalyst for discussion of issues and potential resolutions.

Summary

This chapter describes a methodology to reduce conflict and build consensus called the Conflict Analysis Methodology (CAM). It consists of identifying the constituencies, their assumptions, their goals, and the appropriate spokespeople to represent their position; identifying the issues underlying the conflict and possible levels of resolution; developing and analyzing treaties; and following a structured process of negotiation and consensus building to agree on a final treaty. A summary of the methodology is shown in Table 7.

  Action Purpose
1 Hold informal interviews with people informed about the conflict. To get an overview of the conflict and write a preliminary list of goals and their importance.
2 Identify constituencies and their spokespeople. To define the coalitions and identify individuals who will later develop the conflict model and choose treaties.
3 Analyze assumptions. To obtain a general understanding of the problem, identify ideological and technical sources of conflict.
4 Perform an in-depth interview with one or more objective outsiders. To refine the goals identified in step 1, check understanding of issues and different set of values involved, and begin exploring grounds for resolution. 
5 In separate sessions with each constituency, identify issues, and levels of resolution.. To develop the conflict model and identify key levels of resolution that lead to an overall compromise.
6 Assess importance of issues & value of resolution levels To quantify the importance and preference of the component issues and the levels of resolution,
7 Form and score treaties To generate a set of feasible solutions and explore Pareto optimal solutions.
8 Analyze the treaties To generate treaties that are likely to resolve the conflict.
9 Develop a strategy of negotiation To increase the likelihood of a positive outcome.
10 Present results to all constituents. To generate an acceptable treaty and develop guidelines to implement the treaty or agree on further actions in the resolution process.

Table 7: Summary of the Conflict Analysis Methodology

The main sources of conflict escalation are a lack of understanding among the parties, a lack of information about the opponent’s position, and an emphasis on bargaining behavior. CAM addresses these problems by using a joint definition of the problem, emphasizing win-win solutions, and helping to minimize the impact of ideological differences by identifying areas of agreement.

CAM promotes consensus building by underscoring the importance of a clear and structured resolution process, by eliciting an understanding of the positions held by the constituencies, and by finding potential trade-offs among them.

References

Balke, W. M., K. R. Hammond, and G. D. Meyer. 1973. “An Alternative Approach to Labor Management Relations.” Administrative Science Quarterly 18: 311-27.

Froman, L. A., and M. D. Cohen. 1970. “Compromise and Logroll; Comparing the Efficiency of Two Bargaining Processes.” Behavioral Science 15: 180 - 86.

Huber, G, P. 1980. Managerial Decision Making. Glenview, IL: Scott, Foresman.

Janis, I. L., and L. Mann. 1977. Decision Making: A Psychological Analysis of Conflict, Choice, arid Commitment. New York: Free Press.

Keeney, R. L., and H. Raiffa. 1976. Decisions with Multiple Objectives.’ Preferences and Value Trade-Offt, New York: John Wiley.

Raiffa, H. 1982. The Art and Science of Negotiation. Cambridge, MA: Harvard University Press.

Rowe, A. J., R. 0. Mason, and K. Dickel. 1982. Strategic Management and Business Policy: A Methodological Approach. Reading, MA: Addison-Wesley.

Thomas, K. W. 1975. “Conflict and Conflict Management.” In The Handbook of Industrial and Organizational Psychology, Vol. II, edited by M. D. Dunnette. Chicago: Rand McNally.

Von Neumann, J., and 0. Morgenstern. 1944. Theory of Games and Economic Behavior. New York: John Wiley.

More

Analyze Data

  1. Using the data in Table 4, generate 20 possible treaties and calculate the value score for each constituency.  Plot your data and show the treaties that are Pareto Optimal in bold (here are instructions on how to identify Pareto Optimal treaties Video► IE► 

  2. Produce a list of at least 6 issues and the possible resolution for each issue for either resolving end of life conflict among family members or negotiating a compensation package.  Assign what you think are the appropriate weights and values that the parties would have assigned to these issues.   Use the following format to report the issues, the issue resolutions, the value of each resolution and the importance of each issue to both constituents.  
     

    Issues Levels of Resolutions Constituency I Constituency II
    Weight Value Weight Value
    A Description of
    Issue A
    A1 Description of worst resolution of B 0.xx 0 0.xx xx
    A2 Intermediate resolution of A xx xx
    A3 Intermediate resolution of A xx xx
    A4 Best resolution of A 100 xx
    B Description of
    Issue B
    B1 Description of worst resolution of B 0.xx 0 0.xx xx
    B2 Intermediate resolution of B xx xx
    B3 Intermediate resolution of B xx xx
    B4 Best resolution of B 100 xx

    Using log rolling and your assigned weights, identify a Pareto optimal solution to the conflict.  Here are instructions on how to identify Pareto Optimal treaties Video► IE►
     

  3. The following data were obtain from 3 people opposed to family planning.  Estimate what are the values associated with the attribute levels.  Make sure your estimates for values on each attribute are normalized to range from 0 to 100.  (Please review the convention on constructing single attribute value functions Read►

Issues Levels of Resolutions Constituency II
First Second Third
A To what extent should family planning programs try to convince clients that adolescent sex is bad?  A1 Should not do so 10 20 20
A2 Should be available to clients 20 30 40
A3 Should be required of all clients 50 80 100
A4 Should be a fundamental part of every service 80 80 90

Data From Three Respondent Opposed to Family Planning On One Issue

Presentations

 

To assist you in reviewing the material in this lecture, please see the following resources:

  1. Lecture on conflict analysis Listen► IE► Slides►

  2. See how to identify Pareto Optimal treaties Video► IE►

  3. Standardizing single attribute values Video► IE►

Narrated lectures require use of Flash.


This page is copyright protected by Farrokh Alemi, Ph.D..  This page is part of the course on Decision Analysis, lecture on Conflict Analysis.  This page was first made on Wednesday, November 06, 2002 and most recent revision was on 10/19/2014.  This page is based on a chapter with the same name published by Gustafson DH, Cats-Baril WL, Alemi F. in the book Systems to Support Health Policy Analysis:  Theory, Model and Uses, Health Administration Press: Ann Arbor, Michigan, 1992.