To paraphrase Jane Austen, it is a truth universally acknowledged that mediation is confidential. Go on any training course, listen to any mediator’s opening speech, and the secrecy/privacy of the process will be affirmed and reaffirmed. In the commercial mediation arena, and these days most other practice areas, you will also sign a contractual undertaking about confidentiality. Article 7 of the 2008 EU Directive on Cross Border Mediation states that “mediation is intended to take place in a manner which respects confidentiality”. It is simply assumed that this is desirable, even essential, for mediation to take place. When mediation confidentiality is challenged, as it has been recently in both the US and English courts, mediators are quick to deplore the state of affairs.

In this blog I question both the possibility and the desirability of mediation confidentiality. I must first acknowledge my debt to Bill Wood, QC, for his excellent article summarising the position in England and Wales (see William Wood, 2009, ‘When Girls Go Wild: The Debate Over Mediation Privilege’ This is particularly relevant to my own jurisdiction of Scotland because our courts have not yet had to deal with challenges to mediation confidentiality and so English decisions are likely to be “persuasive” (a lovely term that means something like “I will do what these judges did, but not because I have to.”)

First some terminology. It does not help clarity that at least five different terms are used sometimes interchangeably, each with a distinct meaning (in English – is it the same in other languages?). They are:
Confidentiality – this usually refers to an undertaking made by all parties to a mediation not to share or disclose any information about what occurred in the mediation.
Privilege – legal privilege applies to certain conversations that the law recognises as deserving special protection, such as communications between solicitor and client. It seems unlikely that the courts will offer the same protection to mediation.
Without prejudice – meaning that offers can be made in negotiation as part of the bargaining process without their being relied on in subsequent court proceedings. This has underpinned negotiations for centuries and it seems sensible that mediation should also be a ‘without prejudice’ conversation. However, it is not automatic, and so must be agreed in advance.
Non-compellability – refers to a more limited protection for mediators, meaning that they cannot be compelled to give evidence about what occurred in mediation in subsequent court proceedings. The 2008 EU Directive on Cross Border Mediation placed a duty on States to ensure the non-compellability of mediators on cross border matters. Crucially, this does not protect parties from each other.
Admissibility – governs whether or not the courts will allow certain evidence to be used, including both verbal testimony and documents. So, even if a mediator is “compellable” (and so has to turn up), her evidence may not be “admissible” (and so has to be ignored).

In a nutshell, it seems that mediators wish their work to be confidential, which means attempting to render it legally privileged by means of a non-compellability clause combined with a without prejudice undertaking in the hope that the courts will treat any evidence from the process as inadmissible. Clear? Actually it’s not that simple. Wood lists at least five situations where the courts will overrule even the most cunningly drafted Agreement to Mediate:
1) Impropriety – where there is an allegation of some improper conduct (such as threats or intimidation) by one of the parties to the mediation, the courts will almost certainly wish to hear evidence.
2) The fact or terms of settlement – if there is a dispute over the terms of the settlement (or whether it occurred at all) the courts will probably override any contractual confidentiality clause in the interests of justice.
3) Fraud and duress – if a settlement is achieved through deceit or pressure, it would be surprising if parties lost their legal remedy because the best evidence is protected by mediation confidentiality. As Bill Wood puts it: “If a party is said to have made a fraudulent representation (a fortiori where he has done so through the mediator) should the mediator not be available and even keen to give his or her evidence?”
4) The consent of both parties – most confidentiality clauses are written in such a way as to protect the parties from each other. If both of them waive confidentiality, the courts may not think it is up to the mediator to withhold information.
5) Where the settlement is being challenged by a third party – for example, when a contractor believes that a sub-contractor “caved in” and struck an onerous deal with an insurer that then binds the contractor. The logic applied by the English courts is that “without prejudice” exists to protect negotiations in that matter and so the policy justification no longer exists when faced with a different dispute.

Enough on the possibility of confidentiality. Privacy still matters, but we should be a little cautious in giving blanket assurances. I turn to my other question: is mediation confidentiality desirable?

A number of critics (e.g.Laura Nader, Owen Fiss and Richard Abel) have accused mediation of sweeping injustices under the carpet by dealing with each individual dispute privately, so that bad practice attracts no negative publicity and no precedents are set for the future. This is contrasted with cases fought out in the justice system, where the glare of publicity ensures that “wrongdoers” are made an example of, thus setting public norms and expectations for future conduct.

While mediators have staged a staunch defence (the justice system is no great shakes for the less privileged; mediation is empowering; many mediators do ensure that legal norms are followed) I wonder whether this secrecy needs to be a given. I can think of at least two reasons. First, quite a number of clients, when confidentiality is first explained, declare, “I don’t care who knows. I would say this in front of anyone.” It may be that confidentiality is more important to mediators than their clients, in some cases. Second, and this is particularly relevant in employment disputes, those outside the mediation often have a legitimate interest in the outcome. An employer may well have to take action or make changes as a result, and indeed may also be paying the bill. Other colleagues may be deeply affected. Is it not right that they should know what has been agreed?

This highlights a muddle created by the blanket term “confidentiality”. We are in danger of confusing process with outcome. A “without prejudice” process can still produce an agreement for public consumption. I am not suggesting that we reveal every detail of the conversation, but should the cloak of confidentiality prevent important information from reaching those in a position to do something about it? Why not share the outcome with others, with the parties’ permission? My current practice is to raise this possibility at the outset, thus directing clients’ attention to the wider, systemic issues that may be thrown up by the conflict. At the end of the mediation we can ask: “Who else needs to know?” “What can the organisation learn from the way this dispute has been handled?” “What needs to change?”

Another possibility, elegantly described by Susan Sturm and Howard Gadlin, is for a group of mediators to engage in long-term, anonymised reporting. In this way organisations can learn lessons from disputes and modify practice.

My radical proposal is that mediators around the world, or throughout a jurisdiction, engage in a similar exercise. Why not have an international digest of mediation decisions? Kluwer has itself led the way in shedding light on arbitration decisions (see Perhaps the time has come for mediators to accept rather than attack our critics. If the privacy of mediation causes some problem while solving others, let us address those problems head on. I would argue that mediation can make a claim to be developing societal norms as effectively as litigation: in some respects better, in that decisions are arrived at on the basis of more than purely legal norms. Parties themselves shape the outcomes. What a fascinating resource it would be if we could reflect the full range of our work. And if it turns out that people are somehow settling for “less” in mediation, let’s face up to this, and adapt our practice accordingly.

I’m sure this will not be the last word on confidentiality, but next time you are telling clients that the process is confidential I predict there will be a little voice in your head saying: “Is it?” and “Should it be?”


To make sure you do not miss out on regular updates from the Kluwer Mediation Blog, please subscribe here.

Kluwer Arbitration
This page as PDF


  1. Great post Charlie.

    I’m curious. Do you think there’s something wrong, unfair, unjust about someone “settling for less” in mediation? Less that what? Less than what they would have received at trial? But, of course, who knows for sure what that would have been.

    I see parties settle for “less” in mediation all the time. But if we’re satisfied that the settlement represents a good decision for that party, all things considered, where’s the problem. Mediation is based on party self-determination. If a particular party puts a certain price tag on finality, certainty, closure etc., does this mean there’s a flaw in the process.

    Thank you for a thought provoking post.


  2. A very interesting article and I agree that it will not be the last word on the subject. As mediation becomesmore widley used, it follows that there will be more people who may want to test the confidentiality of a mediation, notwithstanding the agreement.
    Boulle and Nesic in “Mediator Skills and Techniques: Triangle of Influence” 2010 have given a succinct summary of the current state of play.


  3. I work at a community mediation center in a state, where based on state statutory law, mediators are basically granted a privilege if they meed certain criteria.

    We deal with confidentiality as, mediation content is confidential – as to mediators. We incorporate party expectations of confidentiality – for one another – into the mediation discussion. It’s not always realistic to expect that a mediation party won’t leave and tell a spouse, colleague, or family member about what happened in the mediation. Perhaps more people would be interested in mediation if they heard first-hand accounts of how it had helped people that they know.

    In our center’s pre-mediation meetings with clients, we discuss that sanitized information about their cases may be used for grant reporting, fundraising, or publicity (people are able to opt out if there is a concern). It’s part of the model standards that we improve the profession of mediation by research, mentoring, and other methods which may require a certain level of “sharing” of information. The most important things is for parties to have accurate expectations for each other and their mediators regarding what will happen with their information.

  4. Developing societal norms? I think it is sufficient that mediation is an alternative to litigation for the settlement of disputes. In fact it is still an important job to get this across to the public before even considering educating about what went on at mediation. As an employment mediator I agree wholeheartedly about the individualisation of problems and the social isolation of the process for individuals who come to mediation for settlement without support and workplace solidarity.
    That is why I have a real concern about mediation being used in abuse of power and repeated breaches of law, especially workplace bullying.

Leave a Reply

Your email address will not be published. Required fields are marked *