Adapted from

Oral presentation: ADR in Cyberspace Symposium, Nov 11, 1999
Discussion paper: ADR Cyberweek 2000
One Accord Publication: 15 Ohio S. J. on Disp. Resol. 643 (2000)

Ernest M. Thiessen, PEng, PhD, President
SmartSettle, a division of ICAN Systems Inc.

Joseph P. McMahon, Jr., PE, MS, JD
Mediator & Settlement Counsel, LLP, Denver, CO

1. Introduction

   Cyberspace and sophisticated computer technology are presenting new opportunities for overcoming the challenges of conventional negotiation with mutual gains bargaining. This paper briefly introduces the theory used in solving negotiation problems with web-based negotiation support systems and then describes a new product and negotiation process called SmartSettle that uses optimization to take negotiators "beyond win-win". The new concepts introduced in this paper are illustrated in the context of a hypothetical environmental negotiation.

2. Challenges and Opportunities

   Following are some current challenges of conventional negotiation that often prevent participants from achieving better outcomes.

   1. Challenges to effective and efficient negotiations
      
      
      1. Adversarial Tactics

         Our society places a high value on an individual or entity’s power and rights, encouraging them to work hard and compete in order to succeed. Although competition can encourage high individual goals, competition in negotiation often results in harmful adversarial tactics, and inefficient results.

      2. Piecemeal Thinking

         Without sophisticated tools to deal with the inherent complexity of most real-world negotiations, decision-makers are forced to deal with them piecemeal, i.e., one issue at a time. Moreover, a piecemeal approach to negotiation encourages positional rather than mutual gains bargaining.

      3. Tedium

         Giving adequate attention to important decisions tends to be very time consuming, especially in multiparty negotiations. Because of the demands of other important management and business tasks, negotiators often are unprepared and have too little time or resources for the decisions required.

      4. High Cost

         Drawn-out negotiations involving professionals can be very expensive, both in time and negotiation energy.

      5. Irrational Decisions

         Reasonable outcomes are compromised when decision-makers make logic errors, take short-cuts, or permit emotions to get the upper hand when under the stress of intensive negotiations. Without properly assessing the risks, parties are often unrealistically confident of a favorable outcome, should the matter be taken to court.

      6. Complexity

         Multiple issues and numerous possible outcomes can overwhelm negotiators, causing them to make decisions based on heuristics and emotion rather than reason. These problems are even greater in cases involving many parties.

      7. Win-Lose Outcomes
         
         

      Negotiation based on low disclosure levels results in vague ideas about reservation and aspiration levels of other parties and often misinterpretation of the parties’ real interests and needs (footnote 1). Tackling issues individually, as in a one-dimensional win-lose tug-of-war, inevitably leaves value on the table—unclaimed by any party. Parties either agree to a sub-optimal (footnote 2) conclusion or miss finding a mutually satisfactory outcome, even when it was actually achievable.

   2. Opportunities with Negotiation Support Systems

New negotiation support systems built with powerful optimization algorithms and enhanced by a maturing cyberspace, are now providing a real alternative to conventional negotiation, in business arrangements as well as the settlement of litigation. These new systems reduce negotiating time and cost for decision-makers in simple or complex cases, by putting them in control of a process that quickly clarifies tradeoffs, recognizes party satisfaction on all types of negotiation issues, and generates optimal solutions.

Internet connectivity now makes communication possible at a distance, facilitating the exchange of offers and counteroffers while simultaneously managing confidential information at a neutral site. When effectively used with a trained neutral, this process eliminates or minimizes some of the challenges of conventional negotiation, and can provide much improved negotiated outcomes. As such, the parties’ decisions are based on more complete and more thoroughly evaluated information. The parties not only achieve the win-win goals of cooperation but are also able to go beyond win-win with optimization.

The sophistication of systems such as those described above, is best employed by a qualified neutral facilitator, whose role is to provide orientation and guidance through the process. The combination of neutral facilitator and Internet site serves several purposes:

• all negotiating parties and the facilitator are automatically networked with one another, no matter where they are located,
• parties can easily exchange offers and counteroffers via the Internet at their own convenience;
• all the power that is needed to number-crunch a huge problem doesn’t need to reside on each party’s desktop, and,
• private and confidential data is kept safe and secure.

3. Efficiency Frontier

   The basic theory underlying the optimization algorithms of negotiation support systems can be effectively illustrated using the efficiency frontier concept. As illustrated for a hypothetical case in Figure 1, the efficiency frontier represents the best possible outcomes for all parties in a negotiation case.

   

   Figure 1: Efficiency Frontier

   Every potential resolution in a negotiation problem is associated with an expected satisfaction level for each negotiating party. This figure illustrates a plot of the expected relative satisfaction of one party named Riverside against the expected relative satisfaction of another party named DEC. These parties are engaged in a hypothetical negotiation called DEC vs Riverside, which is used for illustration later in this paper. Very rarely can two negotiating parties both obtain 100% of what they want (top right corner as (100,100)). However, parties can almost always achieve more than 50/50 (center of graph). The efficiency frontier is a line defining the greatest level of joint satisfaction that parties can get in a particular negotiation (footnote 3).

4. The SmartSettle Process

   SmartSettle (footnote 4) is a new negotiation support system that networks multiple parties located anywhere in the world and manages their confidential information with a neutral Internet site. This section describes a multi-phased process in which SmartSettle uses optimization to quickly transform conflicting objectives into fair and efficient solutions. Of course, it is expected that during the process, parties will continually compare options for a negotiated agreement with trial risks or other alternatives to settlement. The parties will continue with the process as long as they can rationally see the advantages of settlement. With the aid of a facilitator, the parties proceed through the various phases in the following sequence.

   1. Qualify Interests

The focus in this phase is on quality rather than quantity. The parties engage in a creative and cooperative process that describes the problem and identifies all issues that must be resolved. Recognizing that a well-defined problem is already half solved, this task requires adequate attention to allow success in subsequent phases. This initial phase is best conducted in joint sessions with all parties and the facilitator present.

The main task in this phase is to build a Single Negotiating Form (SNF), an outline of the desired agreement, that contains blanks for the unresolved issues. The emphasis of the SNF is on those unquantified blanks. When parties have built the SNF, they should be able to refer to the blanks to identify a comprehensive and non-redundant list of all the unresolved issues of the conflict (whether qualitative or quantitative).

While building the SNF, the facilitator should also help the parties identify preliminary bargaining ranges for the issues represented by the SNF. The facilitator should emphasize that this is not an opportunity to claim value but rather an exercise in collaboration. Establishing some common ground early in the process will help parties establish confidence that possible resolutions exist within their collective view of reasonability.

2. Quantify Satisfaction

After identifying the issues and possible ranges of resolution, the facilitator assists parties individually to analyze their own priorities among their interests and possible outcomes. Each party’s task now is to teach the SmartSettle software how that party can become satisfied on each of the issues, and thereby with any potential final outcome. An intuitive preference elicitation method called "Even Swaps" simplifies complex tradeoffs by having parties deal with comparable issues in pairs. The private preferences quantified in this phase are not shared with other parties. Rather, a patented neutral site on the Internet keeps that information confidential and secure.

3. Establish Equity

Having worked individually to assess their own preferences, parties now move to searching for an equitable resolution, i.e., one that fairly divides the benefits of resolution. Understanding that SmartSettle will use the "equitable resolution" as a tentative reference point from which to search for improvements, parties do not need to try to find an efficient resolution in this phase. That work is best left for SmartSettle in the next phase.

4. Maximize Benefits

The tentative resolution found in the previous phase becomes a reference point from which improvements can be generated. When requested, SmartSettle generates a single package that proportions benefits according to previously agreed influence (footnote 5).

5. Secure Commitment
   
   

The goal of all negotiations is to secure a commitment from parties to implement the decisions agreed upon. The parties’ ability to secure commitment, usually represented by a well-written final agreement, determines the degree to which they will realize the benefits of the entire process.

5. DEC vs. Riverside Pulp and Paper

   In the remainder of this paper, the SmartSettle process is illustrated in the context of a hypothetical negotiation problem between the State Department of Environmental Conservation (DEC) and Riverside Pulp and Paper (footnote 6). The conflict is briefly described as follows.

   Riverside is located beside the Deep River, which happens to be a convenient place to discharge the effluent from their operation. The Deep River water quality has deteriorated over the years and an administrative order has been issued by DEC requiring Riverside to improve the situation. Because a demonstration project would be of great value, DEC is quite keen on having Riverside install a new technology called the Technoclean Scrubber. DEC is so motivated toward the Technoclean Scrubber that it seems willing to include a guarantee, subsidy, insurance, and other incentives. However, DEC also claims the authority to require Riverside to shut down for as long as it takes to properly comply and go through a testing period. Not surprisingly, Riverside is unwilling to cooperate with DEC’s ideas for resolution of the conflict.

   In the dynamic of real world negotiations, one phase is seldom completed before the next one begins, and an iterative process may bring a party back to an earlier phase any number of times as the problem evolves and is more thoroughly understood. However, for the sake of brevity, this illustration assumes static party preferences and follows a simple sequential progression through each phase.

6. Qualify Interests

Suppose that parties to this hypothetical begin negotiating without SmartSettle. DEC says that they should require Riverside to shut down for several months while a scrubber is installed. However, if they agree to install the new Technoclean, that requirement could probably be waived. In this vein, DEC declares that the following proposal is generous in order to encourage Riverside’s use of the Technoclean.

DEC and Riverside agree to the following conditions for disposal of Riverside’s effluent:

• Riverside will install a Technoclean scrubber.
• DEC will guarantee the scrubber and subsidize $10,000.
• DEC will provide $100,000 liability insurance.
• Riverside will not be required to shut down the paper and pulp operations.
• DEC will spend $50,000 on a positive public relations campaign.
• DEC will give Riverside a compliance incentive of $100,000.

--- to be signed and dated ---

Riverside does not have confidence in the Technoclean Scrubber technology and responds with a proposal based on the tried and proven Rotoblu Scrubber. Because DEC wants Riverside to close down for several months while a scrubber is installed, Riverside offers a compromise of a two-month closure to appease DEC’s constituency. Riverside bluffs a lot about the great burden of that concession, but in private in-house discussions Riverside has decided that they could actually make good use of the down time.

DEC and Riverside agree to the following conditions for disposal of Riverside’s effluent:

• Riverside will install a Rotoblu scrubber.
• Because the Rotoblu is proven technology, DEC need not guarantee the scrubber or subsidize it.
• DEC will provide $200,000 liability insurance.
• Riverside will close down for two months.
• DEC will spend $100,000 on a positive public relations campaign.
• DEC will give Riverside a compliance incentive of $150,000.

--- to be signed and dated ---

1. Building a Single Negotiating Form

Compare the two proposals above and especially notice the differences in bold print. DEC and Riverside have been concentrating on the quantitative aspects of their negotiation. With SmartSettle, the parties are encouraged by the facilitator to first rather build an SNF such as the following.

DEC and Riverside agree to the following conditions for disposal of Riverside’s effluent:

• Riverside will install ___________(a Technoclean, a Rotoblu, no) scrubber
• DEC ___________ (will/will not) guarantee the scrubber and subsidize $______________.
• DEC will provide $_________ liability insurance.
• Riverside will close down for _______ months.
• DEC will spend $_________ on public relations.
• DEC will give Riverside a compliance incentive of $___________.

--- to be signed and dated ---

The difference between this SNF and the proposals illustrated previously is that there are blanks representing the unresolved issues. It is relatively easy to agree to the use of this form. Each party can readily say, "That’s fine with me. Just fill in the blanks with what I want."

The process of qualifying interests is far from trivial, and the complexity of the process depends on many factors, including the following:

• Type of case (e.g., commercial, environment, personal injury)
• Number of issues
• Interrelationships among issues
• Number of parties
• Number of party representatives
• Types of parties (e.g., corporations, individuals, institutions, governmental, etc.)
• Whether third-party or constituency ratification is required
• Customs, bargaining styles, and experience of the parties
• Degree of cooperation among the parties
• Value of outcome to parties
• Risk tolerance of the parties
• Degree of litigation uncertainty
• Resource limitations (e.g., time, money, deadlines, quality of technical and legal support)

  Fortunately, there is much good literature on the above listed subjects. Not surprisingly, every author presents their process steps somewhat differently, depending on their assumptions about the character of the negotiation. The following list of steps, for a typical complex case, is compiled and adapted from several sources; Raiffa (footnote 7), Hansen (footnote 8), and Brin (footnote 9), as well as the experience of the authors.

1. Identify Interest Groups
2. Structure Participation (Contracting)

1. Appoint facilitator
2. Discuss process guidelines and goals
3. Arrange financing for the process
4. Select party representatives
5. Reach agreement on the process to be used
   

3. Conduct Training

1. Establish proficiency in support tools
2. Convey physical and legal reality of the context

4. Share Information, Interests, and Visions
5. Create Value without Claiming in a Search for all Possible Resolutions
1. Paraphrase the issues
2. Brainstorm resolutions
3. Find new associations and arrangements
4. Look at the issues and resolutions from the broadest possible perspective
6. Develop Single Negotiating Form

Among the references listed, only Raiffa includes the final step of developing the SNF (footnote 10). However, developing this form is a critical component required for compatibility with the rest of the SmartSettle process. Referring back to the SNF developed for the DEC vs Riverside case, the following issues must be resolved:

• Type of Scrubber
• Guarantee provided by DEC
• Subsidy provided by DEC
• Liability insurance provided by DEC
• Plant Closure
• DEC funded Publicity
• Compliance Incentive provided by DEC

The facilitator will play a key role in assisting the parties in the process of building the SNF, ensuring that parties both assert their settlement needs and listen to the needs of the other parties.

2. Entering Shared Information into SmartSettle

Having completed an initial draft of the SNF, the parties or facilitator can enter the information from the SNF into SmartSettle’s Shared Information window, as illustrated in Figure 2 (footnote 11). The Shared Information window organizes three aspects of a negotiation case; parties, issues and versions. The most important thing, however, is missing, i.e., our goal; the final resolution. A final resolution is achieved by all parties (in this case, two) agreeing on a decision to be implemented on each of the issues. Each combination of the possible outcomes is called a package. Any party may form a package that can become an offer or counteroffer from that party.

Figure 2: Shared Information

SmartSettle’s Shared Information window has three panes; one for listing Parties, one for Issues or case variables, and one for Versions. There is no theoretical limit to the number of parties or issues that can be added to a negotiation case.

1. How the SmartSettle Process uses ‘Packages’

In cases such as the one illustrated in this paper, with many issues and many possible options for each issue, thousands or even millions of significantly different packages are possible. In fact, research shows that there are often so many combinations that parties rarely find an optimal one. The ultimate goal of SmartSettle is not to merely find a good package, but rather to find an optimal one, i.e. fair, efficient and acceptable to the parties. Because the "package" concept is key to understanding how SmartSettle works, the use of packages in the SmartSettle process is summarized below, highlighting (with bold print) each reference to a package.

7. Quantify Satisfaction

   Returning to the DEC vs Riverside case, we assume now that parties have established a mutually acceptable shared description of their negotiation problem (footnote 12), and commence entering private information independently, each at their own location. The goal of this phase is for facilitators to help parties describe their preferences in a way that enables them to easily evaluate negotiation options.

   Figure 3 is a screen shot of SmartSettle’s Package Editor illustrating the bargaining ranges defined by Riverside. Displayed are two hypothetical packages; Worst on the left and Best on the right. This is how the negotiation problem would look like from Riverside’s point of view (footnote 13). The Worst package possible within defined bargaining ranges is usually below a negotiating party’s minimum acceptable level (i.e., lower than the value of what that party expects to achieve if no agreement is reached).

   

   Figure 3: Bargaining Ranges defined by Riverside

   SmartSettle’s Package Editor is used for displaying and editing packages within defined bargaining ranges. SmartSettle orients the ranges so that preferred values for each issue appear on the right-hand side (footnote15).

   Riverside expects to litigate with DEC if no satisfactory agreement can be found in this current negotiation with SmartSettle. Figure 4 shows a package with decision values that Riverside might expect in court. This assessment is Riverside’s "Best Alternative to a Negotiated Agreement" (BATNA) (footnote 14). Consequently, Riverside believes that it must achieve a better result than its BATNA in negotiations with DEC.

   

   Figure 4: Walkaway to Court

   Displayed here is a private package with the resolutions that Riverside expects from a resolution of the conflict in court. This package represents Riverside’s BATNA and is a satisfaction level below which Riverside would walk away from negotiations.

The values displayed in Figure 4 do not represent a fixed bottom line for each issue. Rather, there are many packages that could represent a minimum acceptable level for Riverside. Riverside’s objective is not to get better resolutions on each individual issue, but rather to agree upon a comprehensive settlement package that results in greater satisfaction and value than some defined minimum acceptable level (such as Riverside’s BATNA). Therein lies a key strength of the SmartSettle software – to permit a party to look at packages analytically and comprehensively.

Riverside will determine the acceptability of any particular package by assessing the tradeoffs between issues and between options for particular issues. The SmartSettle satisfaction graphs model a party’s satisfaction on individual issues (footnote 16). The SmartSettle depiction for a particular issue depends on whether the issue is qualitative or quantitative. Potential outcomes on qualitative issues are represented by different options, whereas quantitative issues are measured within quantitative ranges. Figure 5 illustrates Riverside’s satisfaction graph for the Scrubber issue -- an example of a qualitative issue with three options.

Figure 5: A Qualitative Issue called ‘Scrubber’

The Scrubber issue is a qualitative issue with three defined options; Technoclean, Rotoblu, or No Scrubber. Issues with discrete options are represented in SmartSettle with bar graphs. The height of each bar represents its relative value, or importance, to a party as compared to the other options. In this case, the Riverside’s engineers have informed the Riverside negotiators that the difference between a Technoclean and No Scrubber would be $300,000. Deciding to measure Issue Importance in terms of thousands of dollars, this issue is assigned an importance of 300.

Figure 6 illustrates the satisfaction graph for the quantitative Subsidy issue, i.e., the issue that resolves whether and how much of a technology subsidy DEC would provide to Riverside in a settlement. The satisfaction graph shown in Figure 6 is a simple, straight line. However, when using SmartSettle, a party has the option to define points along a curve to represent any complex, non-linear shape.

Figure 6: A Quantitative Issue called ‘Subsidy’

The Subsidy issue is a quantitative issue with a bargaining range defined between $0 and $450,000. With Issue Importance measured in thousands of dollars, and the satisfaction graph being a simple straight line, the importance for this issue is 450.

The importance assigned to each issue is a relative measure of the tradeoffs among issues. Direct specification of relative importance for the issues, as described above, may seem quite straightforward. However, many decision-makers have great difficulties accurately specifying the relative importance among issues by merely referring to a list of issues (footnote 17).

Fortunately, there exists another solution that is rational and powerful, yet so simple and elegant that it has withstood the test of time. More than two centuries ago, Benjamin Franklin outlined his approach to the challenge in a letter to his friend and noted scientist Joseph Priestly, who was trying to choose between two alternatives. This charming letter (footnote 18) is a "must read" for every decision-maker.

Franklin’s method compared pros and cons of an issue, striking out combinations that were approximately equal in importance. He continued to make such cancellations until one side (either pro or con) prevailed. The obvious limitation of this method was the need for approximations when pros and cons were not exactly balanced. Although many situations can accommodate this lack of precision and still produce good results, SmartSettle’s preference elicitation methodology (footnote 19) overcomes that limitation. SmartSettle provides a mechanism called Even Swaps (footnote 20) that allows a decision-maker to define the relative importance of issues by creating a hypothetical set of equivalent alternative packages. These equivalent packages are then used for comparison purposes and to adjust the relative importance assigned to each negotiation issue.

Parties’ negotiation preferences are dynamic rather than static. As such, they may change radically over the course of negotiations as information is exchanged, and the problem description evolves. Therefore, adjustments to preferences made during the SmartSettle process are as important as the initial estimations of preferences. In practice, therefore, a party may be satisfied with rough approximations of their preferences in the early stages of negotiation. That party may later fine-tune their preferences as the negotiation progresses.

8. Establish Equity

   With jointly identified issues and ranges of possible outcome, and individually determined preferences, the next step is to establish equity, i.e., agree on how the benefits should be divided among the parties. An experienced facilitator will be helpful to parties as they choose among various available routines to exchange packages via SmartSettle’s Internet connections.

   1. Opening Proposals

The SmartSettle process encourages parties to make optimistic opening proposals. Research has shown that this strategy is by far the most successful (footnote 21). In our case, the optimistic packages proposed by each party are illustrated by Figure 7. Both parties are prepared to be flexible with respect to these proposals.

2. Divisions

If at any stage, parties agree that they are equally distant from a fair solution, they can generate a Division or Compromise. Because SmartSettle accounts for party preferences, this procedure actually produces a resolution that is perceived to be fairer than simply splitting the difference on every issue. A SmartSettle function called Multiple Division will generate a series of packages to fill in the holes between proposals least acceptable to each party. Division functionality is not illustrated in this example.

3. Equivalent Packages

At any time during the SmartSettle process, either party can request SmartSettle to "Generate Equivalent Packages." In doing this, SmartSettle will identify the least acceptable package for each party, and generate for each of them a different but equivalent package in terms of party satisfaction. Early in the negotiation process, when parties are still relatively far from resolution, this routine assists parties in fine-tuning their preferences.

Figure 7: Initial Optimistic Proposals

Listed in the Proposals group in the right pane are DEC 1 and Riverside 1. These packages are the initial optimistic (footnote 22) proposals from each party. Their issue values (footnote 23) are displayed in the left pane. DEC 1 has a rating of -758. Because Riverside is measuring satisfaction in terms of thousands of dollars, this rating implies a cost of $758,000. This sum is more than DEC expects to pay if the outcome were decided in court. On the other hand, the proposal identified as "Riverside 1" would cost Riverside only $362,000 (footnote 24).

4. Dealing with an Apparent Impasse
   
   

The hypothetical at this point in the SmartSettle process assumes that the negotiation dance has reached the point of impasse. Figure 8 illustrates the impasse as seen from Riverside’s private point of view. Figure 9 plots this impasse in satisfaction space.



Figure 8: Impasse

DEC 2 and Riverside 2 are supposedly final offers made by each party. The negotiators for each party think that they can go no farther. The issue ‘values’ of both packages are displayed in the left pane. Riverside’s proposal is displayed on top. As shown in the left pane, the two packages are different only on the guarantee issue. Riverside wants a guarantee but DEC is unwilling to give one.

 



Figure 9: Impasse in Satisfaction Space

The same two packages that are displayed in Figure 8 are plotted here in satisfaction space. The ratings shown on the vertical (Y) axis are the same as those seen by Riverside in its private view in Figure 8. The composite view, however, is known only at the neutral SmartSettle Internet site, and is not communicated to either party.

In this hypothetical, the parties have tentatively agreed on everything except the guarantee issue. Riverside contends that there is no way that it can agree to a Technoclean without a guarantee, particularly in light of Riverside’s prior concessions in this negotiation. Similarly, DEC contends that that it can not give the requested guarantee because DEC has already made other substantial concessions.

It is not possible to split the difference on the guarantee issue because the only options are yes or no. Therefore, this impasse looks like it could lead to a win-lose or even a lose-lose situation. The difference is really not that great to either party but either one giving in would suffer loss of face. Can SmartSettle be used to solve the impasse? The answer is shown in Figure 10.



Figure 10: Equivalent in Satisfaction Space

The impasse positions of DEC and Riverside, as plotted here and in Figure 9, are close enough for SmartSettle to be able to generate an equivalent package, i.e., one that would provide each party with the same value as their current proposal.

By this time, Riverside’s preferences are quite well represented and the equivalent package is easily accepted. After all, it would provide the same satisfaction as Riverside’s last proposal. Unbeknownst to Riverside, DEC has received the identical package and for similar reasons also accepts it. Their win-win deal is illustrated in Figure 11.



Figure 11: Equivalent Solves Impasse

From Riverside’s point of view, the generated Equivalent and Riverside’s last proposal each have a rating of -545. The Equivalent is in the Tentative group, indicating that both parties have already accepted it.

9. Maximize Benefits

   Seeking to find an even better resolution, the parties now request improvements from SmartSettle. Another package is generated as illustrated in Figure 12 and Figure 13.

   

   Figure 12: Improvement

   SmartSettle has generated an improvement to the Tentative solution. The Improvement is being compared to Riverside’s final proposal. According to the ratings, the improvement is better by 90 points. If both parties accept it, it will become the new Tentative solution.

   The improvement generated by SmartSettle is worth nearly $100,000 more to Riverside than the current Tentative agreement. Riverside readily accepts it. Because SmartSettle attempts to divide benefits fairly to each party, DEC sees similar improvements and also accepts the improved package. In this hypothetical, the improved package generated by SmartSettle becomes the framework for a final, written settlement agreement.

   

   Figure 13: Improvement in Satisfaction Space

   The generated improvement, plotted on the efficiency frontier, is better by 90 points for each party. In this hypothetical research case, for ease of comparison, the same scale was used to measure satisfaction for each party. In application of SmartSettle to real cases, satisfaction scales are chosen independently by each party. Resulting ratings remain confidential and are never compared by anybody.

10. Conclusions

The authors have presented a new negotiating process made possible with the SmartSettle negotiation support system. Preliminary research shows significant potential for negotiating parties to gain from using such systems. In a Cornell study (footnote 25) with a similarly complex case, the average gains were 16% for each party. In the authors’ opinions, this process will reduce time, cost, and stress if parties pay attention to avoiding the pitfalls that often exhaust conventional negotiators (footnote 26).

The facilitated negotiation described in this paper is intended to both (1) encourage mutual gains bargaining and (2) to work toward the goal of Pareto efficiency in negotiation. The SmartSettle process, when used with a skilled neutral, can encourage both goals. As with any technology, it cannot be mechanically applied to complex situations. The effective use of this process involves many of the principles underlying mutual gains bargaining, and effective mediation/facilitation. The software is no substitute for proper preparation and thoughtful participation in negotiation. However, the structure proposed and SmartSettle software work in synergy with, rather than supplanting, those basic negotiation principles. Multi-party and multi-issue negotiations are frequent and usually complex. These tools and processes can take parties to a place of effectively viewing their options, and making the best possible decisions in negotiation.

Footnotes

1. H. RAIFFA, THE ART AND SCIENCE OF NEGOTIATION, 142 (1981)
2. Economists sometimes use the term "Pareto Optimal" to describe an outcome that cannot be made better for any party without making some other party worse off.
3. In dynamic real world cases, the location of the efficiency frontier is actually a function of an evolving problem description and constantly changing party preferences. This makes it a moving target in contrast to the static representation in the example of this paper.
4. For more information about SmartSettle, visit the website.
5. By default, SmartSettle treats all parties equally. However, parties may agree ahead of time to change the default by setting levels of relative influence for each party.
6. This case is used here with permission from the Program on Negotiation at Harvard Law School (www.pon.harvard.edu). It is an adaptation of, "DEC v. Riverside Lumber" written by David Lax, James Sebenius, Lawrence Susskind, and Thomas Weeks. For the purposes of this paper, and to conserve the uniqueness of the PON case, details of the case have been substantially changed. Copyright is held by the Program on Negotiation from which "DEC v. Riverside Lumber" and other teaching materials may be purchased for a nominal fee.
7. H. RAIFFA, Lectures on Negotiation Analysis (1996)("Lectures").
8. J. HANSEN, Round Table Manners for Facilitators (1998).
9. D. Brin, Avoiding Fatal Errors, The Future of Civilization and the Internet, __ OHIO ST. J. DISP. RES. (1999)
10. In "Lectures", Raiffa (cited above), on p. 50, uses the term "Template" where "Single Negotiating Form" is used in this paper.
11. Normally, only one party or the facilitator will enter the Shared Information and then give the access codes to other parties. Thereafter, linked by the SmartSettle neutral site on the Internet, the Shared Information is available to all parties.
12. In an actual case, the parties may have further joint sessions or teleconferences from time to time to better define the problem and modify/improve the SNF.
13. In the remainder of the illustration, the process is described from Riverside's perspective. All screen shots show how SmartSettle appears on Riverside's computer monitor.
14. R. FISHER AND W. URY, GETTING TO YES: NEGOTIATING AGREEMENT WITHOUT GIVING IN. 10 (1981)
15. The reader might look at the values on the right-hand side and recognize that it would be impossible for DEC to guarantee a scrubber if the no scrubber option was chosen. Parties can teach SmartSettle about such infeasible conditions with constraints. Detailed discussion of constraints is beyond the scope of this paper.
16. How a party becomes satisfied on a particular issue often depends on outcomes for other issues. How SmartSettle deals such interdependencies is not dealt with in this paper.
17. Hammond, Keeney, and Raiffa, Even Swaps: A Rational Method for Making Trade-offs, HARV.BUS. REV. 137-149 (March-April 1998) (Hammond, Keeney, and Raiffa state that making wise trade-offs is one of the most important challenges in decision-making.)
18. quoted in the SmartSettle User Guide (1999).
19. Thiessen and Loucks, Computer-Assisted Negotiation of Multi-objective Water Resources Conflicts, WATER RES. BULL., Am. Water Res. A., 28(1), 163-177, Feb.(1992).
20. Hammond, Keeney and Raiffa (1998) (cited above). (These authors use the term "even swaps" to describe a similar methodology for simplifying tradeoffs that can be employed without computer assistance.)
21. R. SHELL, BARGAINING FOR ADVANTAGE: NEGOTIATION STRATEGIES FOR REASONABLE PEOPLE. 165 (1999).
22. The term "optimistic" as used here means that the party making the proposal doesn't really expect it to be accepted, and is prepared to make further concessions.
23. By definition, a complete package is composed of a set of issue values. An issue value is a possible decision for the subject issue.
24. The white dot is a private marker, indicating acceptance. It may be put on any package, whether private or published. By definition, a proposal will always have a white dot.
25. Thiessen , Loucks & Stedinger, Computer-Assisted Negotiation of Water Resources Conflicts, GROUP DECISION AND NEGOTIATION. 109-129 (March 1998).
26. Id. Shell reports from his own research that when most negotiators had finished "a hotly contested, complex deal, they were tired and wanted to stop" rather than participate in a post-settlement settlement exercise.