Trying not to say “I told you so”: A Halsey Chronology

by Tony Allen


A particularly irresistible temptation generated by the November 2023 judgment in Churchill v Merthyr Tydfil CBC[1] has been to review what I have written on the subject of courts making mediation orders over the twenty-three years that I and others have been writing on the topic and to trace the journey up to and around Halsey towards Churchill.  Having written thousands of words, mostly, but not all critical, about the Halsey judgment, there is an especial pleasure in finding that the criticisms were not misconceived, and that at last a frustrating barrier to the natural development of mediation has been acknowledged as such and removed.  So here is a broad chronology of how thinking has developed on this topic during the 21st century.

Post CPR and pre-Halsey: 1999 to 2004

The 2nd edition of The ADR Practice Guide by Mackie, Marsh, Miles and Allen was published in August 2000, fifteen months after the CPR came into force, and two months before the Human Rights Act 1998 and thus the European Convention on Human Rights (ECHR), became part of English law. This was written 18 months before Dunnett v Railtrack.

“There are bound to be new developments in the ways in which courts think it appropriate to generate pressure on parties to use ADR.  The introduction of art 6 of the ECHR, with its guarantee of a ‘fair and public hearing within a reasonable time by an independent and impartial tribunal established by law’ might put some fetter on mandatory referral, for instance. But it must be remembered that ADR does not exclude the court’s jurisdiction. Parties are free not to settle within ADR and to move back to the litigation process.  As we have seen, ADR and litigation are in effect alternative to each other.  It would be surprising indeed if the human rights legislation was used to make settlement of litigation more difficult, in effect compelling litigants to incur the expense of trial.”

This eerily prophetic passage anticipates precisely what the Court of Appeal in Halsey did four years later.  Their unanimous judgment almost seems to have assumed that referral to mediation meant compulsion to settle.  Perhaps in those early days of the CPR the Court of Appeal was a little more willing to give general guidance when cases gave rise to novel issues.   Neither the judges nor any other counsel challenged counsel for the Law Society who raised the art 6 point as to whether it was properly before the Court.  It seems from submissions made in Churchill that even the Law Society did not want their counsel to raise it!

As an aside, it is interesting to see how an article written by supporters of ADR in 2000 was so comfortable with the concept of its being regarded as “alternative” to litigation.  How things are changing – the new concept of “integrated dispute resolution” currently under discussion is very different.

Halsey and its aftermath: 2004 to 2006

The reaction to Halsey was palpable.  Courts that had been ordering parties to use ADR stopped doing so.  The one or two decisions where judges had made peremptory orders[2] were quietly ignored, and no one queried the rightness or authority of the Court’s pronouncement on the effect of ECHR art 6 on ADR Orders.  At first we commentators simply accepted this proposition without question, and it went virtually unchallenged for a number of years.  Attention turned to the six factors which might excuse a refusal to mediate, and much has been aired over these ever since.  One of them – that “mediation had no reasonable prospect of success” – was worryingly attractive to judges, especially those with no personal experience of how mediation could close apparently intractable positions.  In a slightly tetchy article entitled Halsey; myths about impossible mediations written in November 2005, I commented as follows:

“Sadly, there is little that can be done to feed back to procedural and trial judges the fact that what looked like a hopelessly intractable dispute did settle through mediation, and the constraints of confidentiality properly make case studies somewhat anaemic.  Few judges have had the opportunity to experience such successes at first hand: indeed, very few of the current judiciary have ever had a chance to attend a mediation at all.  The truths about what mediation can achieve out of the most unpromising of starts will have to percolate through slowly as experience broadens.  Mediators are developing the confidence to tackle the toughest of disputes in the knowledge that many will crack once the parties are given a true alternative to adversarialism without risking loss of face for themselves or their clients.  Not every case will settle, but is there any real reason why the potential value to be derived from the process should not be expected of the professional gatekeepers to the dispute resolution system as a norm in all but exceptional circumstances by both parties and the judiciary.” 

Jumping ahead for a moment, the high point of mediator frustration with judicial opinion on this topic came with the reaction of mediators to the Court of Appeal decision in Swain v Mills & Reeve[3], in which Davis LJ went on record as saying (of a £750,000 “gap”):

At all stages the parties were in reality a hundred miles apart,” 

and chose to reverse a sanction for not mediating imposed at trial.   This was a case where a judge had earlier recommended mediation (which was ignored} and the trial judge had penalised the defendants for their refusal to mediate.  Davis LJ’s assertion led me to circulate a large number of experienced mediators for their views and to put together an article entitled Mediations where parties are “a hundred miles apart”: thoughts on Swain v Mills & Reeve.  The article expressed shared surprise and criticism at such an approach, and included seven case studies of cases where the parties appeared t be “hundreds of miles apart” nevertheless were settled at mediation.  Without Halsey, one of the earlier judges might well have contemplated making a formal ADR Order. 

The beginnings of doubt : 2007 to 2009 

Returning to Halsey and ECHR Art 6, it was not for three years – in 2007 – that doubts began to be expressed as to the rightness of Halsey on the Art 6 point.  The change in approach was gradual. It was in 2007 that CEDR modified its formal definition of mediation to exclude the word “voluntary”, having distinguished the essential voluntariness of continued participation in a mediation once started from advance application of pressure to participate in the process, reflecting both international and direct experience.  With judicial and rule-based encouragement (though short of compulsion) to mediate, coupled with possible costs sanctions for failure to do so, and no evidence that external pressure to mediate undermines the prospects of settlement, this seemed a proper adjustment.

The third edition of The ADR Practice Guide was published in 2007, Chapter 5 of which reviewed the effect of Halsey in relation to ordering mediation.  We pointed out that: 

the actual facts of the case themselves concern whether costs sanctions should be imposed for ignoring an inter-party request to mediate…..but the court took the opportunity to express views about ADR Orders too which, while strictly obiter, were intended to give guidance.  

We even queried the status of the submission that an ADR Order violated ECHR Art 6:

It is not clear whether this was fully argued.  It was certainly a last-minute submission made in the name of the Law Society, and the only case cited ­- Deweer v Belgium – dealt with waiver of court access through an agreement to arbitrate…..ordering a party to try a settlement process which is emphatically not an adjudicative process, from which he can withdraw without penalty once started, is a strange breach of Art 6.

These two short paragraphs were also prophetic in their own way, as they effectively paralleled the submissions made by CEDR, CMC and Ciarb to the court in Churchill.  In the closing paragraph of Chapter 5 of the 2007 ADR Practice Guide, we noted that courts remain reluctant to compel parties into mediation, but added;

As we write, this approach has been challenged in a lecture by a senior Chancery judge”.

This was of course the late and greatly lamented Sir Gavin Lightman.  I was present one evening in July 2007 on the roof garden of S.J Berwin’s London offices to hear him assert that the Court of Appeal’s views on this subject in Halsey were “clearly wrong and unreasonable”.  He went on:

“Turning to the [first] proposition regarding the European Convention, my reasons for saying this are twofold: (1) the court appears to have been uninformed about the mediation process and the distinction between an order for mediation and an order for arbitration or some other order which places a permanent stay on proceedings. An order for mediation does not interfere with the right to a trial: at most it merely imposes a short delay to afford an opportunity for settlement and indeed the order for mediation may not even do that, for the order for mediation may require or allow the parties to proceed with preparation for trial; and (2) the Court of Appeal appears to have been left in the dark as to the practice of ordering parties to proceed to mediation regardless of their wishes is prevalent elsewhere throughout the Commonwealth, the United States and the world at large, and indeed at home in matrimonial property disputes in the Family Division.

He remained a lonely voice among the judiciary in arguing both that ADR Orders were proper, and that the onus should be on a refusing party to justify their refusal to mediate, rather than on the losing party who had proposed mediation.  He had canvassed the latter view as early as 2001 in Hurst v Leeming. 

However, in 2008 a speech by Sir Anthony Clarke MR added judicial weight to doubts about Halsey, although not going so far as to suggest that this part of Halsey was obiterHe said:

Taken together, what could be described as the European and US approach to ADR appears to demonstrate that compulsory ADR does not in and of itself give rise to a violation of Article 6 or of the equivalent US constitutional right of due process. This suggests, admittedly without hearing argument, that the Halsey approach may have been overly cautious. This was not a point that was investigated in detail in Halsey and (who knows) may be open to review – either by judicial decision or in any event by rule change. 

Strengthening criticism and admitted error: 2009 to 2012

It was in 2009 (as far as I can trace) that I first unequivocally criticised the Halsey decision on Art 6, in an article entitled Ordering mediation: does it infringe human rights?  This started by a reminder about the positive message that emerged from the European case notoriously cited in Halsey by Lord Lester, Deweer v Belgium[4]: 

Deweer showed that the European Court clearly encourages settlement extra-judicially, and the courts are thus prepared to concede jurisdiction to a properly constituted arbitration process, subject only to checking for any undue constraint.  How then could mediation be regarded as in potential breach?  We looked at the only obvious argument to the contrary, that a stay of litigation for mediation might be regarded as hindering an Article 6 trial “within a reasonable time”.  But no stay is necessary with court-ordered mediation, as the process can be set up so quickly, and could never be likely to be regarded as lasting sufficiently long to amount to a breach of Article 6.  So far as we know, this passage was not mentioned to the Court by Lord Lester in Halsey.  Dare one respectfully suggest that this is one of the least persuasive sections of the Court’s very useful judgement in Halsey?  Perhaps more notice of the point from counsel who advanced it might have led to fuller argument on all sides, and a more considered approach. 

Despite the article’s politeness about the “very useful” Halsey judgment ­– this looks surprising characterisation today! – the rest of the article tries to summarise what would have been CEDR’s submissions if only Lord Lester and the Law Society had given advance notice of the Article 6 point, the thrust being that there was plenty of compulsion embodied in the CPR and the Pre-action Protocols which looked far more of an obstruction to a swift public trial than  a brief diversion into mediation.  It does however concede that to strike out a case for ignoring a judicial order to mediate might be going too far, and that a costs sanction should be a sufficient penalty.  What is interesting since the 2021 Civil Justice Council report Compulsory ADR and the 2023 decision in Churchill is whether courts might yet assert that this extreme form of case management is appropriate after all.

It was in 2010, as is now well known, that Lord Dyson himself, who had delivered the unanimous judgment in Halsey, recanted the principle stated in Halsey on Art 6 in a speech, though still expressing a personal preference for not ordering mediation.  In doing so, he puzzlingly said:

It doesn’t seem to me that it is the role of the court of law to force compromise upon people who do not want to compromise”. 

But no one is ever “forced to compromise” at a mediation.  It was sad that even as late as 2010, a senior judge could still publicly express a view that to order mediation amounted to the court forcing compromise on unwilling parties, rather than pressing on parties an occasion when they could safely consider compromise without being compelled to settle, such a choice being entirely free from fear of any adverse consequence if returning to litigation.  

Because mediation had become the subject of so much judicial decision-making over the years, the section on mediation law in The ADR Practice Guide was by agreement with my fellow authors superseded by my own book Mediation Law and Civil Practice, first published in November 2012.  The new book again argued that the Art 6 assertion in Halsey was obiter.  But by now the EU Directive on Mediation had raised its status internationally, even though it only required rather marginal amendments to be made to the CPR for cross-border disputes.  However, its preamble certainly recognised that mediation could be mandatory in Member States, and that voluntariness applied to continued participation and not to requiring its use, an approach which resonated in England & Wales.

The pressure for reform increases: 2013 to 2021

Thereafter, judges like Sir Alan Ward in Wright v Michael Wright Supplies[5] and Norris J in Bradley v Heslin[6] used judgments to articulate increasing frustration about the impact of Halsey.  The Civil Justice Council then turned its attention to the place of mediation in civil justice.  It first produced a working party report on ADR and Civil Justice in 2018, which on review today seems a startlingly cautious production, with no frontal assault on Halsey and Art 6; rather a critique of the Halsey six factors, with a plea for them to be both narrowed and applied consistently.  However, the CJC’s report Compulsory ADR, published in 2021, very frankly asserted that Halsey was wrong on Art 6, and that it was both lawful in principle and in certain circumstances desirable that courts should be able to order parties to engage in ADR.  This was justified as much as anything by the Court of Appeal’s decision in Lomax v Lomax[7].   Although that case was about judge-ordered neutral evaluation[8] even when one party objects, the court was referred to the Halsey judgment, but declined to find that to order neutral evaluation offended Article 6 principles.  Moylan LJ said:

In any event, ENE does not prevent the parties from having their disputes determined by the court if they do not settle their case at or following an ENE hearing.  It does not, in any material way, obstruct a party’s access to the court.  Insofar as it includes an additional step in the process, this is not in any sense an “unacceptable constraint”, to use the expression from Halsey.  In my view, it is a step in the process which can assist with the fair and sensible resolution of cases. 

The point can immediately be made (and was by me!) that when you substitute “mediation” for ENE in the above paragraph, the same holds just as true for mediation as for ENE.

The final lap: 2021 to 2023

With the appointment of Sir Geoffrey Vos as Master of the Rolls in 2021, matters have moved on apace, with his determination to evolve a civil justice system that meets the needs of real people.  As has typified the development of civil justice since the 1990s, the vanguard of reform has been led not by government or the legal professions themselves but by leading judges.  That honourable line stretches from Lord Woolf through Sir Brian Neill, Lord Slynn, Sir Gavin Lightman, Sir Henry Brooke, Sir Alan Ward, Sir Rupert Jackson, and many others including the last two MRs. Sir Terry Etherton and Sir Geoffrey Vos.  We now have a Lady Chief Justice (who presided in Churchill) whose biography mentions that she is a trained arbitrator and mediator.  One of my predictions long ago was that the climate for mediation development would change as soon as the senior judiciary was populated by lawyers who had themselves experience of mediation both as advisers and mediators.  That time seems very happily to have arrived.

The CJC took a further step in advocating change in relation to the place of mediation in civil justice in Part 1 of its final report on the Pre-action Protocols, published in 2023.  This appended a new draft General Pre-action Protocol which after required information exchange has occurred, effective requires adoption of a dispute resolution process.  Unless this is a process involving a third party neutral (such as a mediator or evaluator) the court would be able to order such a neutrally managed process later in litigation.  The report is a long way from having been implemented, but it provides evidence of a major evolution in thinking at a high judicial level.

So we reach the November 2023 judgement of the Court of Appeal in Churchill v Merthyr Tydfil CBC[9] and the review which it was able to undertake of Halsey.  As always with the common law, it was largely a jurisprudential accident that a set of facts emerged which made this possible.  Who would have thought that a dispute over an incursion of Japanese knotweed would be the trigger for a review of mediation law?  And rather like Mrs Donoghue, who purchased a ginger beer bottle containing a snail (but did it?) and reformed the law of negligence, neither Mr Churchill nor indeed Mrs Halsey played much of a personal part in the appeals brought in their name.

It gives particular pleasure to read the concluding paragraphs of Vos MR’s judgment in Churchill v Merthyr Tydfil, where he, sitting with Carr LCJ and Birss LJ found unanimously that the Art 6 point as expressed in Halsey was not part of the essential reasoning in that case – i.e. it was obiter and not binding in precedental terms, and stating in terms that every mediator would applaud:

The court can lawfully stay proceedings for, or order, the parties to engage in a non-court-based dispute resolution process provided that the order made does not impair the very essence of the claimant’s right to proceed to a judicial hearing, and is proportionate to achieving the legitimate aim of settling the dispute fairly, quickly and at reasonable cost. 

As the appeal was based on the Council’s argument that the deputy district judge did have power to order ADR and was not prevented in law from making such an order by the obiter part of the Hasley judgment, we can safely operate on the basis that the Court of Appeal’s view in Churchill that this is right in law cannot be regarded as itself obiter.  But what the Churchill court declined to do was either to formulate guidance as to when orders to mediate might be made, or indeed to make a ruling as to what should happen in the unresolved dispute between Mr. Churchill and Merthyr Tydfil Council.  Rather touchingly, the last sentence of the judgment recommends that they might well consider a mediation for that purpose. Vos MR places trust in the discretion of each judge to decide what is appropriate in each case when either ordering a stay or ordering the use of what he elects to call (rather clunkily) a form of “non-court-based dispute resolution process”[10].

Nor does the Churchill judgment give guidance as to what is the proper extent of sanctioning for ignoring an order to mediate.  Will they extend as far as to strike out a claim or defence in default, as hinted at by the CJC report Compulsory ADR?  Or would that “impair the very essence of the claimant’s[11] right to proceed to a judicial hearing?”  We do not yet know the answer to that.  But in any case, while we now know that judges may properly order mediation, it may turn out that they are only rarely called upon to do so.  If one party wants to mediate, it will be a rash opponent who opposes this now.  An application to the court will probably follow, with the judge empowered to make an order and quite probably to make an adverse costs order summarily against the opposer.  It will be safer for parties simply to agree to mediate mutually.  Maybe judges will be more tempted to make a mediation order at a CCMC or other directions hearing where both or all parties decline to mediate, but where the judge thinks that a confidential settlement discussion should at least be tried.  As I said in Mediation Law and Civil Practice (a paragraph quoted at para 85 of the CJC’s Compulsory ADR report):

A civil justice system is surely able to protect its users from themselves and to try to make sure that whatever is litigated in front of the courts justifies that level of judicial input. Moreover it should only do so [i.e. convene a trial] if all parties unshakeably resolve to litigate despite examining every alternative. 

In other words, used often in relation to the intention of the CPR reforms from Lord Woolf onwards, litigation and litigated trial should be “the last resort”. However, the best outcome of all for this will be that parties will simply elect to mediate at a sensible stage of any claim, and come to recognise the value of doing so.  “Success” in mediation will not merely be measured by whether settlement is achieved, though statistical evidence suggests that a large majority of mediations will continue to lead to consensual settlements at or after the mediation event.  Where no settlement is achieved, parties will have discerned the level of determination of their opponent, seen where there is common ground and where the dispute really lies, revisited their risk assessments, and refined their approach to continued litigation.  A judge will then decide any continuing good faith disagreement over parties’ rights and obligations at a trial well worth convening.  Meanwhile mediation will become more and more “voluntary” because its worth as a process is better appreciated as experience grows. 

There is more to be discovered about the implications of this dramatic reversal of an “authority” which has dominated mediation law and civil justice for 18 years, apparently wrongly all that time.  But it is good to have removed this unfortunate feature of the litigation landscape at last, and to discover that all the criticism increasingly heaped upon it was justifiable.

[1]  [2023] EWCA Civ 1416

[2]  Like Guinle v Kirreh: Kinstreet v Balmargo unrep 3 Aug 1999and Shiryama v Danovo [2003] EWHC 3006 (Ch), both of which predated Halsey.  Both were recognised again in the Churchill judgment as having been in essence right after all.

[3][3][3]  [2012] EWCA Civ 498

[4]   [1980] EHRR 439

[5]   [2013] EWCA Civ 234

[6]   [2014] EWHC 3267 {Ch}

[7]   [2019] EWCA Civ 1467

[9]  [2023] EWCA Civ 1416

[10]  Have we really only lost the acronym “ADR” to move to “NCBDRP”?

[11]  or indeed the defendant’s right

Tailored Insights for You

Sign-up to receive regular insights on topics ranging from effective Conflict Management and Negotiation to Commercial Mediation and ADR Thought Leadership.