10 Sep 2019
Limits of mediator responsibility
by Tony Allen, Solicitor and CEDR Chambers Mediator.
The hour is late at a long-running mediation and there is a broad agreement between the parties, but much needs to be done to turn the agreed terms into a practical binding agreement, and complex conveyancing documents will be needed now or later. If those heads of terms are written up and signed by each party, will they be binding, or are they going to be prey to an argument that there was no intention to create legal relations, or that they represent a mere “agreement to agree”, or were insufficiently certain or comprehensive?
This is a dilemma which quite often confronts those at a mediation, and it certainly arose following a mediation which occurred in 2011 between Mr and Mrs Abberley (AA) and their two sons David (D) and Mark (M), leading to a court finding in Abberley v Abberley  EWHC1564 (Ch) that the terms were indeed binding on them all, despite the need for many further formalities to be concluded before the deal was strictly effectuated in law.
This will be a relief for mediators. But the case has wider practical interest for mediators interested in how the law affects what they do and how they do it.
It is noteworthy that the mediator gave evidence at the trial without any apparent controversy, and the judgment throws up some interesting questions about the role of the mediator when settlement terms emerge.
The broad facts were that AA and D farmed in partnership in North Wales until 2002, when D went his way and M subsequently joined AA in the business. As the finally adjusted partition of assets between D and AA could not be agreed despite many years of trying, a mediation was arranged in December 2011, with an experienced mediator. A term of the mediation agreement signed on the morning of the mediation read:
Any settlement reached in the mediation will not be legally binding until it has been reduced to writing and signed by, or on behalf of, each of the parties.
Towards the end of the mediation, agreement on a complex partition of land with a cash adjustment appears to have emerged, described in the following way in paras 19 and 20 of the judgment of HHJ Jarman QC:
19. By about 8.30 [the mediator] went into the parties’ respective rooms one after the other and announced that a deal had been arrived at. As he often did, he typed out an agreement. As by now the office staff had left, he did so on [the claimants’ solicitor’s] desktop computer. This took about an hour. However, before it could be printed, the draft disappeared from the screen and could not be retrieved. Subsequently, IT staff at the office have failed to retrieve the draft, which suggests that it may have been written on an external website, possibly involving a template.
20. [The mediator] then wrote out the heads of terms in his own hand. All those present were invited into the room which he had been using, and when they were all there, he read out the heads of terms. He and the two solicitors then signed this written document. Everyone then shook hands, except David and Mark. By now it was after 10pm and everyone then left.
Desultory e-mail correspondence followed which added certain points, and a fresh draft agreement was drafted by AA/M’s solicitors as a suggested substitute for the hastily hand-written document signed at the mediation. Eventually D changed solicitors and denied being bound by the mediated terms. Litigation followed, heard by HHJ Jarman QC in the Cardiff CC in June 2019 – 17 years after the start of the dispute and 8 years after the mediation – who defined the issues to be decided as:
Whether the heads of terms constituted a binding contract between the parties, as the claimants say, or was intended merely to set out some matters agreed in principle as part of a process of arriving at a full and effective compromise, as the defendant says. Alternatively, the defendant says that if the heads of terms did constitute a contract, then such is unenforceable for want of compliance with the requirements of section 2 of the Law of Property (Miscellaneous Provisions) Act 1989.
His decision firmly found in favour of the claimants, and he rejected that anything materially agreed at the mediation was not embodied in the handwritten terms, so that the terms met the requirements of the 1989 Act for a written signed comprehensive agreement for the transfer of interests in land. He even forgave a drafting inconsistency between clauses, as “a slip which cannot of itself render the heads of terms uncertain.”
Although this is a County Court first instance decision, albeit by an experienced specialist Chancery judge, it is comforting for mediators that courts will readily support the enforceability of a framework agreement negotiated with sufficient certainly and clarity at a mediation, even when much documentation remains to be done to formalise what has been agreed. On the question of an agreement which contemplates the need for further formal documentation he cites Parker J in Von Hatzfield-Wildenburg v Alexander  1 Ch 284,
It appears to be well settled by the authorities that if the documents or letters relied on as constituting a contract, contemplate the execution of a further contract between the parties, it is a question of construction whether the execution of the further contract is a condition or term of the bargain or whether it is a mere expression of the desire of the parties as to the manner in which the transaction already agreed to, will in fact go through. In the former case, there is no enforceable contract either because the condition is unfulfilled or because the law does not recognise a contract entering into a contract. In the latter case, there is a binding contract and the reference to the more formal document may be ignored.
On whether there was an intention to create legal relations, the judge very sensibly commented:
Given that the whole point of the mediation was to resolve the long running family dispute, and given that a written agreement was signed on behalf of the parties, as required by the mediation agreement for there to be a binding resolution, then if the heads of terms are sufficiently certain it is difficult to see how it can be said that the parties did not intend to enter into a legal relationship.
On certainty of terms, he cites Lord Wright in G Scammel & Nephew Ltd v HC and JG Ouston  AC 251, 268, as follows:
The object of the court is to do justice between the parties, and the court will do its best, if satisfied that there was an ascertainable and determinate intention to contract, to give effect to that intention, looking at substance and not mere form. It will not be deterred by mere difficulties of interpretation. Difficulty is not synonymous with ambiguity so long as any definite meaning can be extracted. But the test of intention is to be found in the words used. If these words, considered however broadly and untechnically and with due regard to all the just implications, fail to evince any definite meaning on which the court can safely act, the court has no choice but to say that there is no contract. Such a position is not often found.
Applying that approach to each lack of certainty alleged by D in Abberley, HHJ Jarman QC found that the agreement was clear enough. It is obviously necessary to achieve an acceptably high level of clarity and comprehensiveness for heads of terms to be regarded as binding, but if achieved, the courts will enforce such an agreement despite the need for further formal documentation.
There are other interesting aspects for mediators and mediation practice to this case, however.
Firstly, a good deal of evidence was given about what happened during the mediation. No attempt was made to rely upon either without prejudice privilege or confidentiality to exclude it. In effect, both parties waived their rights to such protections in the hope that what emerged would support their case. This has certainly happened before in such cases as Malmesbury v Strutt & Parker and Chantry Vellacott v Convergence Group, when one party eventually discovers that they got it wrong. 
Secondly, as noted above, the mediator himself gave both written and oral sworn evidence about what happened at the mediation, without any apparent demur.
This is only the second recorded time that this has happened (the other case being AB v CD). Whether this evidence was given voluntarily by the mediator or under pressure from either or both of the parties is unreported. It seems not to have been particularly helpful because of the eight-year time lapse between mediation and trial. The judge noted that the mediator “had little recall” of what happened at the mediation.
His judgment also reports some of the further interventions of the mediator in post-mediation contacts, though none of these seem to have taken matters further, being limited to enquiring whether finality had been achieved, and later being told of the breakdown between the parties. BY contract, in AB v CD, the mediator actuallyfurthered the later negotiations and gave evidence of those communications, but this was not the case in Abberley. I have always argued that mediators should not readily consent to giving evidence.
Very often, as in this case, they cannot add to such other evidence as is admissible, whether privilege does or does not apply, or is waived. In essence, Abberley was decided simply upon the clarity or otherwise of the content of the document signed at the end of the mediation. It happened to have been in the handwriting of the mediator, but this was of no evidential significance, and its potential effect would have been the same if it had been written out by one of the lawyers.
In many mediation agreements the parties undertake not to call the mediator as a witness. Whether the Abberley agreement provided for this we do not know, and the enforceability of such a clause has never been tested to my knowledge. All we know is that the judge was not called upon to rule on whether the Abberley mediator should give evidence, however useful or not.
Each mediator must make up their mind as to where the balance lies between helping the parties and preserving the safe framework of mediation practice.
I myself would deploy any reasonable argument to resist being called, as I am almost certainly going to be regarded as less than impartial or neutral by one or other party, depending on which of them my evidence supports.
Abberley is striking for a different reason, because the mediator disclosed that he “often” typed out terms of settlement. Not only did he type it out and then unfortunately lose the draft, he then re-wrote out the terms in manuscript, apparently on his own in his private room, and then invited the lawyers for each party (not the parties themselves) to sign this handwritten document, also signing it himself.
I have been training mediators for CEDR worldwide for the last 20 years, and have asserted that it is not the responsibility of the mediator to draft or take responsibility for the settlement agreement: this is the territory for the lawyers for each party.
Of course, mediators have an important job to do in helping parties to shape up the detail of a settlement, but that settlement must be the parties’, and not the product of the mediator’s advice or opinion. The low level of professional indemnity insurance premiums paid by mediators, especially as contrasted with those levied on advisory professionals, surely underlines the risks of giving advice and of getting it wrong.
The salutary tale of the mediator who drafted and advised on the settlement agreement in the Australian case of Tapoohi v Louwenberg should never be forgotten, nor the reported increase in mediators’ insurance professional policy premiums said to have followed!
In Abberley, it is true, the mediator’s intention seems to have been to help the parties to summarise what they had agreed, in an effort to generate unanimity. But responsibility for advising whether the terms are sensible and acceptable must rest unequivocally with each party’s lawyers.
There are risks for mediators if they take too much responsibility either for encouraging a settlement or for the terms that are accepted. Neutrality is so easily perceived to have been compromised. The defendant, as in Abberley, who later regrets a deal and tries to break it, might well feel negatively about a mediator perceived to have striven to make that deal possible. The mediator’s advice might be based on an only partial knowledge of the relevant facts or a misunderstanding or misreading of the law.
I have also taught CEDR trainee mediators that they should not themselves add their signature to a settlement agreement as a witness to the signatures of others, let alone as if a party to the agreement (which in no circumstances is appropriate, in my view).
While the latter did not appear to be the case in Abberley, the former certainly was. This inevitably increases the risk of the mediator becoming a potentially compellable witness, which seems to me self-evidently undesirable, if not completely avoidable.
I would suggest that, rather than typing or penning a settlement agreement, the mediator’s paramount job is to bring an independent eye to its content as drafted by the lawyers and to check it for comprehensiveness, consistency, clarity and workability.
If the Abberley mediator had adopted that approach, he might well have pointed out the inconsistency between clauses that (fortunately for the claimants and the mediator) the judge was prepared to forgive. What if the judge had declined to forgive that inconsistency between two different clauses, as drafted by the mediator, and this had of itself prevented the deal from taking effect? It would then have been the claimants who would have viewed the mediator’s performance negatively.
Oddly, paragraph 33 of the judgment suggests that the lawyers for each party later agreed “not to pay [the mediator] until the final document had been resolved”. As no final document ever emerged, perhaps the mediator was never paid. Maybe both parties were expressing a degree of dissatisfaction with the mediator by this means, and maybe the mediator felt incentivised to get the parties to settle, albeit without success. Neither possibility is clear. But this is just another illustration of how important it is that the mediator has no stake in whether settlement a mediation leads to settlement.
There is I know a range of views about what mediators should and should not do, harking back to the old evaluative/facilitative debate.
Some colleagues may not agree with what I suggest in this article. But I believe that mediators do need to retain a sense of responsibility not just to those who hire them but also to the framework of the process which they deliver, so as to keep it safe for the benefit of all.
The mediator in Farm Assist v DEFR (No.2) bravely sought to resist being compelled to give evidence of what occurred during a mediation when she felt it might undermine the general security of the mediation process.
We should I believe always give careful consideration to where the limits of our responsibility lie.
 There is a full discussion of these and all other cases cited in the rest of this article in the second edition of Tony Allen’s book Mediation Law and Civil Practice, published by Bloomsbury Professional in December 2018.