The ADR field has a tendency to make large claims in many things, an inevitable result of mixing a great deal of social commitment, a dash of professional insecurity, and lots of lawyers whose ebullience would be worthy of Teddy Roosevelt.

One of those claims is to keep what parties say in confidence. Does this particular claim measure up, in logic, practice or the reasonable expectations of the parties and the public? This short article will discuss some practical facets of the problem. Call it a reality check, if you like.

Keeping it from the other side
Confidentiality involves two quite different sets of concerns, depending on whom the information is to be kept confidential from. The first is when the confidential information is the property of one party that wishes it kept from the other. When we speak of keeping things in confidence, this is probably what most negotiators think of most of the time.

I have no methodologically rigorous way to be sure, but I get the impression that programs and individual mediators may often use a broad brush in describing this feature of mediation – some phraseology like “I will keep everything you tell me in confidence, unless you tell me it can be disclosed.” In practice, I believe, we have no way actually to follow through on this and still do our jobs. While mediators may be able to keep from disclosing many specifics, there are inherent hints in anything the mediator says than make the hermetic concept of confidentiality untenable.

Mediators cannot avoid giving off certain verbal and nonverbal cues every time we change caucuses. The other party, especially if competently represented, doesn’t ignore these. For example, as soon as we start to ask about specifics, we betray the probability if not the details of the other party’s interest in that particular line of questioning.

Also, any competent negotiator will draw inferences from “the dog that did not bark,” such as proposals, arguments or questions that were anticipated from the opponent, but that don’t seem to be arriving through the mediator. The inadvertent signals become stronger as the mediator takes on more of a role in the formulation of proposals.

For example, the innocent question “If they did X, could you do Y?” is the soul and core of much mediation deal-making. But while ostensibly it reveals nothing about the opponent’s confidential position, in fact it is fraught with implications – starting with the reasonable presumption that the mediator is not there to waste time, and therefore not only that Y is seen as important by the other side and that X may now be on the table for the first time as a real possibility, but that by implication, Z may be less firmly desired by the opposing party than had been thought.

At some level, parties already know this, and in my experience, use this feature of mediation deliberately to explore ideas without committing to them. This kind of half-disclosure is, arguably, one of the key features of assisted negotiation, and one of the reasons parties who distrust each other may be willing to work together through a mediator. Yet to claim that “everything you have told me is kept confidential” under these circumstances is to claim too much. Perhaps we should coin a word that describes what we can actually offer, vis-a-vis the other party, rather than confidentiality; something like “nonattributability.”

A similar problem arises with documents. Often, a term of a written mediation agreement, or even a statute or court rule, runs something like this: “Any statements made or documents produced for the mediation are not admissible at trial, unless the information can be discovered through some means other than the mediation itself.”

This sounds, on its face, like a fairly strong protection. But at a minimum, if the case does not settle, an opposing attorney is likely to have newly noticed either the existence, or a possible different interpretation, of certain documentation. Add lawyerly creativity into the mix, and the exception can easily come to overwhelm the rule, as the attorney thinks up some quite unrelated reason why the document simply has to be admissible.

Keeping it in the room
The second set of concerns arises when someone wants information kept confidential from anyone outside the negotiations.

I don’t propose to discuss the “usual and customary” exceptions: threats of violence, of serious harm to third parties, and other well-rehearsed limitations on promises of confidentiality will get attention by other authors here. But well short of these obviously important problems, we may not, in fact, be able to deliver on claims of keeping things secret. And – surprise – sometimes, it may not matter!

There are at least three types of circumstances under which it seems unlikely that confidences will be kept to the degree the parties are routinely led to expect. In two the result is presumptively benign, or even in parties’ long-term interest. In the third, the public interest is – rightly or wrongly – cited as the reason behind the disclosure.

The first is advice-seeking. In mediator work groups, it is common for a “mediator with a problem” to turn to another for a bit of advice. The parties may not be told this occurs. Yet reasonable expectations of the parties might, in fact, dictate this as a form of professionalism. It has a logical analogy to customary practice in older professions, such as medicine. Just about everybody who has ever stood in a hospital corridor knows that the standard claim of confidentiality in the doctor-patient relationship is subject, de facto, to such advice-seeking. Its value to the patient is obvious, and nobody objects as long as it doesn’t go further.

The second set of circumstances in which promises of confidentiality can be overblown deals with study, evaluation and research. I was present once when a prominent mediator gave a blow-by-blow account of a complex case to a roomful of practitioner and academic colleagues, as grounding for discussion of some problems and principles. The mediator, requesting that those in the room keep the specifics to themselves, admitted that the usual promises of confidentiality had been given during the case, and that permission had not been sought from the numerous parties for this specific use of the case data.

We’ve all heard the adage that a secret can be kept by three people, if one of them is dead. Here, 30 people were let in on the story (to be fair, not the most sensitive details). Yet the purpose was benign, and the result probably innocuous for the parties. I would go further, and submit that disclosure in that instance led to a rich discussion of a kind that offers great benefit for parties generally, in the long run.

But would the parties have agreed to the disclosure? While we badly need more research into many aspects of our field, and more collaboration between researchers and practitioners – exactly the result of the mediator’s disclosure in this instance – it is clear that we run a risk with the parties by such data-sharing. The risk can be reduced by perversely refusing to be involved in research and evaluation – hardly, one would think, in the parties’ interest – or it can be reduced by appropriate rewording of what we promise in the first place.

Drawing the Line
Even those mediators and programs inclined toward the most generous offers of confidentiality draw the line somewhere. Often the line is drawn by statute or court rule. But the third set of circumstances of concern here has to do with the role of “external influences” in leading mediators to disclose information. No one, to my knowledge, has attempted a comprehensive study of influence-based or “force majeure” disclosures. A worrying note, however, was struck in a simulation which Charles Pou and I recently ran.

The setting was a day-long continuing legal education course on ethics in dispute resolution, held at the University of Texas Law School last summer. In one role play, a publicly-employed “collateral duty” mediator – i.e. one who mediates as an add-on to another job – achieved a settlement in a case in which a former employee alleged that he had been fired by a truck driver training school for being too zealous about standards, and that the school’s approach to training drivers was lax. Everybody congratulated the mediator, who returned to her “office.”

Then from the center of the audience, a “ringer” actor literally rang her on a cell phone. “This is Billy Bob Boudreaux,” he said in stentorian tones, reminding her and the audience that as chair of the Texas State Legislature’s Joint Committee on Highway Safety, truck driving training firms came under his jurisdiction, and there was a public interest in finding out anything that threatened the safety on Our State’s Great Highways. He said he would like to know what had come out in that mediation, and would like that information right quick.

Freeze action. Query audience. A private mediator spoke first: No way. Confidentiality of process; not one of the standard exceptions. Besides, there were lots of other ways for the legislator to satisfy an entirely proper curiosity without impinging on Our Field’s credibility.

But several other mediators, who mostly seemed to be employed in state agencies and other public programs, said they would turn the information right over. Some said there was clearly a public safety issue involved, which trumped the confidentiality principle in the same way as incipient violence. One was even more forthright. “I don’t know where you work,” she said, “but when a legislator calls someone in my agency, everything . . . just . . . stops.”

Lest this be interpreted as a slam on publicly-employed mediators, let me be quick to postulate a private-practice equivalent: The Big Client. Is it reasonable for One-Time Co. to worry that when Repeat Player Corp. – the backbone of Settlements-R-Us’ mediation practice – wants to know a little something about what happened in One-Time’s case, at least a hint or two might be forthcoming? “Of course not,” says Settlements, “we have Chinese walls for that problem and we maintain them inviolate. Anyway, we’d soon have no clients if we did that sort of thing.” Well, I can’t prove otherwise. But the consistency of application of such internal controls has been a source of continuing concern in some other fields, such as finance.

What’s to be done?
Disclosure of arguably confidential material is something that few programs or mediators are likely to advertise. The extent of the problem is therefore murky. But on a general level, better-elaborated principles of ethics for the field are a matter of strong current interest, in the CPR/Georgetown Ethics Commission and other groups, and may provide improved guidance to mediators caught between powerful forces.

On a less complex level, we should recognize that we are in the business fundamentally of improving communication, not bottling it up. I believe sophisticated parties know “in their bones” most or all of the problems discussed here. The steady increase of “repeat use” of mediation over the past 20 years demonstrates better than any argument that the overselling of confidentiality should not be seen as necessary to attract customers. We should have enough self-confidence to describe what we do accurately, and the parties should recognize that as care and candor.