Mediation Research: What Really Works?


Mediation Research: What Really Works?

By Gary Weiner

Attorneys have been taking their clients to mediation for years now. Many cases settle and a lot don’t. Lawyers and mediators have all kinds of ideas about what works and some can even give you a theory about why. Lawyers have a whole host of ideas about why to choose one mediator or another. Mediators have ideas about what they do and why it works when it does.

[Page 24]

But, do we really know? Surprisingly, we don’t really have much good empirical information on what really works. Some new research seems to say that even mediators don’t really know what it is they actually do when they mediate.

Here’s the bottom line: What you and your clients do matters more than what the mediator does or who the mediator is. Now, to be truthful, there’s a real paucity of good, solid social science on what mediators do in mediation and what happens when they do it. There’s a lot of information out there about what lawyers think they like about mediators, what parties think they like and what trainers think we ought to do when we mediate. But what do we actually know?

What We Know: People Like Mediators and Mediation Even in Cases That Don’t Settle

We have solid evidence, based on surveys of mediation participants which shows that:

• People like mediation;

• People like mediators;

• People settle cases when they are made to go to mediation almost as often as when they voluntarily choose to go;

• People like mediators even when the case didn’t settle;

• People like mediation even when the case didn’t settle; and

• There is no correlation between training or subject matter expertise and settlement.

There is some positive correlation between the number of cases the mediator has done and the settlement rate in that mediator’s cases. For example, during my tenure as Mediation Program Administrator at the California Court of Appeal, First Appellate District, I had an analysis performed of the data collected by the program over the course of 11 years. We looked at evaluations of the mediation process, the mediator, and the court program itself. We found that there is a statistically significant difference in how participants rate the mediator and the mediation process that is correlated with whether the case settled. Not surprisingly, if a case settled, it’s more likely that the participants will rate the mediator and the mediation process more highly. What is surprising is how small the difference really is. Whereas 75% of respondents whose cases settled reported high satisfaction with the mediator, "only" 70% reported high satisfaction in cases that didn’t.

[Page 25]

Can Mediators Actually Get People to Settle Their Cases?

Even more intriguing is what mediators can do that gets people to settle their cases, repair their relations, or improve their communications. Historically, and until very recently, there has been a dearth of empirical evidence addressing what mediators do and can do to facilitate settlement. (Wall & Chan-Serafin, Processes in Civil Case Mediations (2009) 26 Conflict Res. Q. 261-291 (hereafter "Processes") ["[T]here are no empirical studies that report mediators’, plaintiffs’ and defendants’ behaviors in mediation or more importantly that indicate how the various participants’ behaviors may influence the process or the outcome."].)

In 2010, I wrote a paper calling for the development of evidence-based standards in mediation quality. (Weiner, A Call for Evidence Based Standards for Mediator Quality (2012) [as of May 4, 2014].) Now, as co-chair of the ABA Dispute Resolution Section’s Task Force on Mediation Research, I have endeavored to answer that call. Over the past few years, I’ve helped convene the first mini-conference on mediation research and have gathered input from scholars in a variety of disciplines who’ve researched or reviewed the research that has been done.

One of those scholars is Professor Craig McEwen of the Department of Sociology and Anthropology at Bowdoin College in Maine. He prepared a summary of the findings he’s reviewed for that mini-conference in 2012. He found, not surprisingly, that most cases settle with or without mediation, but wanted to know what mediation actually contributed to the settlement of cases.

Professor McEwen says that the evidence suggests that earlier settlement may be likely if the context is right. Litigated general civil cases are more likely to settle if mediators are more evaluative and recommend a specific settlement. But, in divorce cases, "problem-solving approaches" (in which the parties are primarily guided by the mediator to look for solutions through collaborative efforts) are more effective than settlement-oriented techniques (in which mediator primarily "pushes the parties toward compromise" by asserting the weaknesses in each party’s case, the likelihood of loss and the costs of going forward without resolution). Importantly, he reports that "settlement is more likely if lawyers prepare clients and act cooperatively." (McEwen and Wissler, Finding Out If It Is True: Comparing Mediation and Negotiation Through Research (2002) J. Dispute Res. 131.)

[Page 26]

He says that the research shows that, in mediation of general civil matters, only limited changes can come from better mediator techniques and strategies. That is because mediators face serious constraints as a result of when and how they receive cases and the expectations of attorneys about what mediation should be for (e.g., seeing it as a tool in advancing litigation).

The grandfather of mediation research is Dean Pruitt of the School for Conflict Analysis and Resolution at George Mason University in Arlington, Virginia. He, along with Ken Kressel, Department of Psychology at Rutgers University in Newark, New Jersey, published the only book on mediation research in 1989. (Kressel & Pruitt, Mediation Research: The Process and Effectiveness of Third-Party Intervention (1989).)

Professor Pruitt finds that mediation is likely to be successful when the conflict is moderate rather than severe and the parties are highly committed to reaching agreement. It becomes more difficult when issues of principle drive the dispute. He told me in an "oral history" that his conclusion is that the state of mediation research leads to this recommendation: "mediators should keep quiet if the parties are engaged in problem solving but should become active if the parties are stuck or are simply exchanging accusations." But Professor Pruitt cautioned that there are few studies examining the effectiveness of specific mediation strategies. There has been insufficient use of control groups and very few observational studies. Researchers have relied too heavily on surveys and focused too little on mediators’ thinking.

Indeed, it is not even clear that mediators can accurately describe effective stragtegies. Professor Kressel, working with the ombuds office at the National Institutes of Health, has found that mediators have an "explicit schema of practice," what they say they do in mediation, and an "implicit schema of practice," what they actually do when observed mediating. (Kressel, How Do Mediators Decide What to Do? Implicit Schemas of Practice and Mediator Decisionmaking (2013) 28 Ohio State J. on Dispute Res. 709-735.) What mediators say they do may not be what they actually do.

So, What Can Mediators Actually Do to Facilitate Settlement?

[Page 27]

Mediators are taught that they should try to facilitate communication and understanding, help identify issues, and challenge disputants to create solutions in order to get cases settled. They are often cautioned against using strong and assertive techniques that smack of bullying. Many of those who have reviewed the extant research have concluded that, at least as far as we have been able to demonstrate empirically, what the mediator does may not be as important as we may believe.

One of the best and most recent pieces of good research suggests that where the parties want settlement to be a priority, assertive tactics may be just what are called for. This is the result of research reported by Jim Wall, University of Missouri Curators’ Distinguished Teaching Professor of Management at Trulaske College of Business, and his colleague. (Processes, supra, at pp. 261-291.) The researchers closely observed participants in 62 civil case mediations. These included the full range of general civil cases such as personal injury, employment, and contract cases. They found that when mediators applied assertive techniques to plaintiffs and defendants, they expected to get greater concessions from plaintiffs, and that plaintiffs, in fact, made greater concessions than defendants. Therefore, mediators generally applied more assertive techniques to plaintiffs. However, in cases where the mediator thought the plaintiff’s demands were unusually high, even though the mediators used assertive techniques to get the demand lowered, fewer of those mediations ended with agreements. (Id. at p. 277.)

[Page 28]

But, even here, Professor Wall concludes that, although in many cases, mediators’ assertive behaviors helped get the case settled, mediators’ influence was actually, in many of the cases, limited. In about one-half of the cases the mediator’s behavior had no effect on the agreements and, even when there was an effect, the plaintiff’s or defendant’s behavior frequently determined the mediator’s behavior. "Our evidence, as well as that from other studies indicates mediators do not have substantial control over the process…. Rather, it seems that the case type and the plaintiffs’ behavior are the more influential factors." (Processes, supra, at p. 291.) Thus, the parties drive the mediator’s behavior and the results.

Finally, and perhaps most importantly to the advocate, a very new study out of the Netherlands of 540 employment mediations finds that when the mediators actually met in person in pre-mediation caucuses with the parties, the cases were much more likely to settle. This effect is most highly pronounced when the pre-mediation caucuses focused on saving face and building trust, but not when the caucus focused on substantive issues. (Swaab & Brett, Caucus with Care: The Impact of Pre-Mediation Caucuses on Conflict Resolution (2007) Internat. Assn. for Conflict Management Electronic Meetings Paper [as of May 4, 2014].)

What Counsel Can Do to Facilitate Settlement

So, what to make of this if you’re thinking of taking your client to mediation? Based on the research: prepare to mediate — not to try the case, not to "win" the mediation, not to show the client or the mediator or the other side how tough you are. Prepare your client and yourself for one task only: to negotiate a resolution that works. I suggest you ask your mediator to meet with your client — as I now do in every mediation I conduct — in person, in the week before the session, to talk about everything having to do with resolution and nothing to do with the merits.

In these sessions, ask the mediator to work on getting the client to like them. Be sure the mediator establishes his or her authority as a mediator, talks about how many cases similar to yours he or she has done, and his or her personal experience with disputes and conflict in general, and this kind in particular. Have the mediator tell your client about cases just like theirs that settled in mediation and about how the mediation is a once in a lifetime chance to save a whole lot of trouble and expense. Ask the mediator to help you to get your client to commit to negotiation and to start letting go of believing that "only you and your client see the facts as they really are." Work with your client and the mediator to really internalize the truth about all disputes: Both sides think the other side is wrong and is not facing "reality." The mediator’s job with your client at the pre-mediation meeting is to persuade your client to think differently about the dispute. The tools I describe come directly from the psychology of persuasion. Read the book by Robert Cialdini if you don’t believe me. (Cialdini, Influence: The Psychology of Persuasion (2006).)

At the session, don’t turn the case over to the mediator. Know the other side’s views as well if not better than your own. Expect the mediator to press you if the case isn’t settling and the mediator thinks you aren’t giving enough. Find something to give up in exchange for the promise to work at settlement and not at victory.

It’s what you do and what your client does that count.

Gary Weiner is an attorney and mediator in Emeryville. 510-597-0492 or

[Page 29]