Find Mediators Near You:

“Private Caucusing” In Civil Pretrial Mediations

This June we had the pleasure of attending a week-long program on “Mediating Disputes” offered by the Harvard Law School Program on Negotiation.  The program included an ongoing dialogue between mediation and negotiation guru Robert Mnookin (co-author of “Beyond Winning”) and Boston family law specialist David Hoffman (co-author of “Bringing Peace Into the Room”) on the efficacy of caucus-style mediation vs. joint session mediation.  

Having participated as lawyers and mediators in both commercial and family cases, our experience is that mediators now predominantly employ the caucus method in both business and marital property disputes, often from start to finish.  In caucus-based mediations, after a brief joint session (or none at all), opposite parties and their lawyers are sequestered in separate rooms.  Once the sides are separated, the mediator acts as a go-between who tries to broker a deal through “shuttle diplomacy” using information the mediator receives during the private discussions.  In contrast, joint mediation is a settlement meeting in which all parties, their lawyers and the mediator remain in the same room, although the parties are free to leave the room and consult privately with their advisors as needed.

Before the Harvard program, we had not questioned the dominance of caucus-style mediation, much less appreciated its weaknesses.  We did notice that caucusing often wastes a lot of time and is far from a sure shot to resolve a case.  We took away a newfound awareness of the caucus method’s potential pitfalls:

Mediator Empowerment vs. Party Empowerment:  By its nature, caucusing transfers power to the mediator by giving the mediator control of the information flow and much of the negotiating responsibility.  Parties provide the mediator with information, negotiating strategies, arguments and positions, and essentially hope that the mediator will advocate for them with the other side.  The mediator alone hears what both sides are saying and witnesses the body language.  Much can be lost in translation – or manipulation by a mediator who regards achieving a settlement as more important than a frank process.  In contrast, joint sessions leave the power to request information, interpret it, and negotiate over it, directly in the hands of the parties and their advisors.  The filter is gone.  It’s all put out into the room – the honest statements, the lies, good information and bad information, tears, admissions, poker faces, frustration – all of it.  Rejection of the mediator as the information controller and negotiating agent is a fundamental structural difference between caucus and joint mediation.   

Does it really matter whether control of the exchange rests with the mediator or the parties?  In our own experience, it does not matter in some cases if the mediator does the back and forth – in fact, it may be quite helpful.  These are the cases in which the parties are strongly disposed to settle in the mediation.  In such cases, the parties will use the mediator’s “good offices”, by the end of the day, to horse-trade to their inevitable deal.  For highly-motivated parties, the mediator can even function as a buffer against the risk that one party will over-exploit another’s negotiating concessions (or overreact to the lack of concession), risking a collapse or delay in reaching the deal.  To us, caucusing works pretty well in these highly-motivated situations.

Yet the most elusive goal of mediation, critically important for our burdened system, is early resolution – settlements that preempt the stresses and costs of litigation.  Early settlements – as well as settlements by parties who are long-entrenched in positions, have the resources to fight, and who do not enter mediation with a desire to settle – will happen only if parties who do not start-out ready to settle can become ready to settle.  To achieve a difficult settlement, mediation must transform parties – both financially and emotionally – to value resolution more highly than their dispute positions.  They must grow into a settlement.  The mediator-empowered approach often fails in these situations because, among other things, it simultaneously disempowers the parties and constricts the flow of hard information and body language that is essential to building mutual understanding and a willingness to settle.

When early mediations fail, we often hear that it was simply “too early” – that the warring parties must “litigate a while” before they are ready to settle.  This explanation often comes out of the mouths of a mediator, and is rarely contradicted by the professional litigators in the room.  When long-entrenched parties fail to settle, we hear the failure attributed to the personal characteristics of the other side, obstructive or greedy lawyers or a refusal on the part of a party or lawyer to accept the law.

We believe the failure of early mediations results more often from a failure of the mediation process than the parties’ “need to litigate” or personal deficiencies of the parties or their legal counsel.  The caucus method of mediation, with its party sequestration and disempowerment, and its inefficient information exchange, makes early resolution more difficult.  At precisely the point when the parties need to exchange information most efficiently, engage intensively on the issues, size each other up, build or rebuild communication and understanding, and when the parties will obtain the maximum benefit in terms of saving time and money if they make a deal, physical sequestering impedes the parties from an active dialogue with the other side.  While caucus mediation correctly presumes that an exchange of information is good, shuttling is an awkward, inefficient method to exchange information – especially complex or intangible information or body language.

By its nature, party sequestration seems to presume a high level of irreparable breakdown (the parties cannot even be in the room together, much less engage in direct communications, to settle their dispute).  In caucus mediation, the message from the start is that direct discussion is impossible to achieve – therefore, a direct, efficient information exchange must be sacrificed in whole to shuttling – with responsibility placed in the hands of a mediator who often knows the least.  Why is the most inefficient method of sharing information and negotiating now the default in litigation dispute resolution, rather than the last resort?  For people involved in mediation for the last three decades, the answer is simple and unsatisfactory – that’s what we were taught.

The Perception of Manipulation vs. Transparency:  Parties and their lawyers often anticipate that caucus-style mediation will be a manipulative process – which, too often, becomes a self-fulfilling prophecy.  Lawyers advise their clients that the mediator’s primary approach is to resolve the dispute by undermining each side’s confidence in its case – the goal being that both sides will give ground to a mutually-agreeable compromise.  Predictably, rather than approach the mediation as an exercise in compromise, resolution and healing, the parties gird for battle not just with the other side but with the mediator as well.  When mediator manipulation yields a mutually-acceptable settlement, all’s well that ends well – until a party who was manipulated thinks about what happened.  When mediation is manipulative and fails, the parties often are disgusted at how much time and money they wasted.

There is a far sounder basis for the mediation process than manipulation.  In a well-conducted joint session, the parties and mediators work toward a realistic understanding of each other’s interests, perceptions and positions, and explore the law and possible outcomes in a process that is not limited to ominously-delivered “parade of horribles” scenarios.  To achieve early settlements, when both parties are often most optimistic about how their positions will play-out in court, mediation cannot depend on the parties’ willingness to embrace an overly-negative assessment of their case.  A better goal and practice of mediation is to jointly explore what real-world benefits litigation can and cannot provide in relation to settlement, and at what cost.

A standard objection to a joint session is that rather than exchange information, the parties will posture and irritate one another – which will waste time, make settlement more difficult and cause parties to feel the session was exploited to advance positions in the case rather than to settle.  This is where the mediator’s skill, far more than his or her power, can make a big difference – by helping the parties work through the legal and financial positions, emotions and personal styles that drive disputes and settlements.  The mediator adds value not with her ability to carry positions between rooms, adding a personal gloss to make the parties want to settle.  The mediator adds value by helping the parties explore the forces and interests underlying their positions so each party may assess not only the positions but the rationales and emotions behind them.  

Does Material Shared With A Mediator “In Confidence” Really Settle Cases? Caucusing is often justified on the ground that parties might be comfortable sharing information with the mediator – but not with the other side.  It is said that caucusing may be used to overcome the “prisoner’s dilemma” caused by the parties’ mutual fear of exploitation and “adverse selection” from the parties’ failure to disclose material information.  [see D. Hoffman, “Mediation and the Art of Shuttle Diplomacy,” at 272-275].  Proponents suggest that caucusing can be effective in assisting the parties with negotiation coaching (where the parties’ respective negotiation skills and mediation experience create an uneven playing field); safely managing parties’ unrealistic expectations; facilitating the sharing of material information (provided the mediator can convince the parties to allow disclosure); separately floating a mediator proposal to both sides; allowing the parties to save face; and managing extraordinarily difficult parties and/or attorneys [Id. at 263-298].

While no doubt there is plenty of information that might be exploited by the other side, the lingering question is: how much can a piece of information advance the settlement process if it is available only to the mediator?  Even if the undisclosed information helps the mediator shape a notion of where the deal should be done, a restriction on disclosure pushes the mediator into a “trust me” mode with the other side – the mediator now is holding back and manipulating.  It is naïve to assume parties cannot detect when a mediator is less than forthright.  When the parties’ trust in the process starts to break down, which happens when the mediator is perceived as advocating a settlement that is unwarranted by the known facts, the mediator’s ability to assist the parties to reach a settlement begins to dissipate.  More often than not, when the mediator’s apparent position on a settlement diverges from what the disclosed facts warrant, the process gets stuck in a rut.  That is where we end so often in caucus–style mediation.   

Caucusing Can Waste A Lot Of Time:  It is a sad axiom of caucus mediation that, if your position is reasonable, you will spend most of the mediation day checking email with your client in a breakout room – while the mediator “works on” the other side.  If a settlement is not reached, parties may leave thousands of dollars lighter in fees, with no greater insight into the dispute.

If we want parties to mediate more, sooner and more earnestly, parties will need to get much more out of mediation than a day in a small room.  The parties should gain insight and clarity into the case – the real case – not just wild scenarios offered by a mediator as a reason to “settle now.”  The parties should be able to evaluate the other side, both as adversaries and potential compromisers.  The parties should see their counsel advocate openly, and be tested, as they will be in a courtroom.  The parties should get the satisfaction of knowing they spent an active day attempting to resolve their dispute at a “settlement meeting” and, if it did not resolve, a good sense of what made the deal fail.

We have found that parties who have never participated in mediation are utterly surprised to learn they may have no contact with the other side.  Certainly a joint session mediation is demanding for all involved.  It is not a dull, wasted, day in a breakout room.  At the end of a joint session, deal or not, the parties feel they have engaged with each other and the mediator and their efforts to settle are often rewarded in unexpected ways.  While parties often come into mediation as adversaries, and may leave as adversaries, having sat through a joint session they leave with a better and deeper understanding of the other side.  Many cases that do not settle during mediation settle, in whole or in part, a short time after the mediation – sometimes by the parties themselves.  In these cases, the parties’ ability to settle directly is often the result of communication bridges that were rebuilt during the mediation.  While such bridges sometimes are built during caucus mediation, the opportunities to build direct bridges are enhanced in a joint session.

Caucusing Raises a Variety of Ethical Issues:  Caucusing creates a mother lode of ethical issues for the mediator, who will be tempted (even requested) to shade the truth or manipulate the facts to bring about a settlement.  While joint sessions are not free from ethical dilemmas, the vast array of ethical issues that stem from the mediator withholding undisclosed, possibly material, information is reduced.  One megatrend in our society, which the next generation of disputants already embraces, is transparency.  Deep suspicion of secrecy (especially by persons in power roles) and contempt for “spin” highlight our new sensitivities.  Given these trends, the Wizard of Oz mediator who uses the illusion of omniscience, enhanced by exclusive access to “what is being said in the other room”, to control the discussion is outdated, if not obsolete.

We also found merit to the counterpoint – that caucusing often can be helpful:

Early Disclosure of Sensitive Information:  There is little doubt that caucusing creates a “safe” place where a party may disclose his or her true values, interests, emotions and perspectives only because they will be kept confidential. 

A joint session may be uncomfortable and power dynamics in the parties’ relationship may factor-in and discourage one party from being frank.  Caucusing can yield material information that a joint session will not yield early in the process, and may never yield.  A party may be reluctant to be open in a joint session, especially early on, about emotional or legal vulnerabilities.  Even if the information shared in caucus might not be useful until it is appreciated by the other side, it may help the mediator evaluate settlement options or avoid hurting feelings.  Of course, as described above, this “safety” has its costs: the facts and feelings that motivate one side likely will have less impact if not shared.

Avoidance of Time Consumed by Interpersonal Problems Between the Parties:  Joint sessions can consume a great deal of time as difficulties in the parties’ interactions are revealed and managed.  Family law in particular involves a high level of emotion and the parties’ resources may not allow for several sessions to hash out hostility and resistance before they are open to discussing a resolution.  In some instances the parties’ relationship fracture may be too new or too bitter to achieve an early joint discussion.  Timing, of course, plays an important role.  However, we do not believe that the inability or unwillingness of some parties to work together on a joint resolution warrants a default to caucusing.  If parties can work together, there is no reason to separate them.

Abuse:  In some cases, the parties’ interpersonal dynamic does not allow for a physically or emotionally safe joint mediation session (e.g. cases involving abuse, intimidation).  There are cases in which it is better for the mediation to be conducted by telephone, or where the mediator travels to separate locations to conduct the sessions.  This is a minority of cases. 

Having joined the caucus vs. joint debate for a week, and comparing it with our own experience, we believe that joint sessions should be utilized far more in mediations, both commercial and in family asset division cases.   We recognize that an effective joint session takes work, probably more work than caucusing – especially when emotions are high and the parties are not ready to be constructive.  The mediator will need to hone certain skills that are required for a successful joint session.   A joint session mediator must be able to listen to both sides with evident understanding of all positions, even seemingly absurd ones, without alienating the side who is not speaking at the moment.  A joint session mediator needs to help the parties rebuild constructive communications and develop mutual understanding of interests and solutions, while resisting the temptation to impose rules and otherwise seize power.  A joint session mediator needs to avoid taking personal ownership of the dispute, the process or the resolution, and guide the parties instead – “more like a friend,” as our instructors described. 

We observe that many business partners and divorcing couples have a long history of effective planning and joint decision-making.  They often have built substantial enterprises and made monumental decisions together (even raised kids) – often quite well.  The emotional components of  business separations, marital dissolutions and other disputed do not vitiate all of the parties’ history and skills.  For many parties, it is more awkward to work through lawyers and a mediator than to have direct discussions – even during the dispute.  It is very costly. 

Certainly there is an overlay of psychology and often a communication breakdown in any major dispute or separation – but we question whether placing the parties in different rooms does anything more than confirm that they are supposed to be at war.  The “default” should be to signal to the parties that, although they are in conflict, mediation offers them a chance to settle their disputes jointly and avoid costly consequences – most notably, diverting a chunk of their wealth to litigation, inevitably reducing each party’s share of the pie and prolonging the impact of active litigation on employees, investors, children and others.

We add that many marital and commercial dispute settlements involve long-term obligations.  Divorcing parties often continue to share custody of their children and often the futures of parties to a commercial case are intertwined.  Many cases have opportunities to create value rather than merely divide assets or distribute settlement funds.  In a mediation structure that constricts the free flow of ideas with the mediator as a sluice-gate, such options often remain unexplored.  We fail to see how a structure that has most of the procedural elements of a time-out (physical sequestering, dominance of the process by authority figures and punitive costs) helps rebuild the parties’ relationship to the point that they are comfortable making new long-term commitments to each other.

Our ultimate conclusion is that mediators should more aggressively explore the benefits of joint sessions with the parties at the outset of mediation. The default to caucusing should be replaced with a discussion of process. We recognize there are circumstances in which caucusing can be helpful to overcome specific obstacles.  The use of caucusing – at a point – does not require the entire mediation process to be shifted to indirect “shuttle diplomacy.”  As our profession and system continue to evolve, our thinking about the mediation process must evolve.  A new generation of transparency-driven parties demands a dispute resolution process that is less obscure and manipulative.  The process presumption in mediation should shift from caucusing toward joint sessions – open, interactive, often emotional, and we anticipate more often successful, settlement meetings

We take no credit for the extraordinary insights of the Harvard team, commented upon in this piece.  We take full responsibility for anything incorrect, poorly stated or based solely on our own experiences.


Jeffrey Makoff

Jeffrey Makoff is a business litigation and mediation partner in Valle Makoff LLP, San Francisco.  He is a member of the California and District of Columbia Bars.  He is a graduate of U.C.L.A. and Hastings College of the Law.  MORE >


Jessica Grynberg

Jessica Grynberg is the principal of Grynberg Family Law in San Diego.  She is a member of the California Bar and a graduate of Boston University and the Benjamin N. Cardozo School of Law.  MORE >

Featured Members

View all

Read these next


Diligence, Consistency, and Balance: Lessons for Conflict Management Systems Design Practitioners in Fostering Sustainable Change

This article first appeared in Fall 2002 newsletter of The Canadian International Institute of Applied Negotiation. The design and implementation of integrated conflict management systems within organizations is an innovative...

By Ellen Kabcenell Wayne, Leah Borsa

The Other Side

During mediations, I often attempt to put myself into the shoes of the people I try to bring to common ground and imagine what I would have done in their...

By Alex Dukhovny

Success Leaves Clues: An Interview With Jason Dykstra

One of the best parts of blogging and maintaining an online presence is the opportunity to meet, connect and engage with smart, interesting, good people whose paths might never have...

By Tammy Lenski