The Case for Calderbank: A Procedural Reform Too Far?

Family law is not a ‘desert island’

Many family practitioners will recall the prominent role played by Calderbank offers pre-2006.[[1]] Although some practitioners will remember their ubiquity, others may be less familiar with their historical role and it is important to revisit their origins.

The Calderbank principle originated in Calderbank v Calderbank [1976] Fam 93, in which the court offered guidance – with reference to the practice adopted in Admiralty and Lands Tribunals – that an offer can be in the form that it is without prejudice as to the issues, but reserving the right to refer to it on costs; what is now termed ‘without prejudice, save as to costs’.

Although the observations in Calderbank were, strictly speaking, obiter,[[2]] the principle was quickly established in financial remedy (then ‘ancillary relief’) proceedings and ultimately found favour in wider application across civil litigation as a whole, per the judgments of Oliver and Fox LJJ in the Court of Appeal case of Cutts v Head [1984] Ch 290. It is important to pause here for a moment and consider the public policy rationale underpinning the Calderbank principle. Per Oliver LJ:

‘… parties should be encouraged so far as possible to settle their disputes without resort to litigation and should not be discouraged by the knowledge that anything that is said in the course of such negotiations (and that includes, of course, as much the failure to reply to an offer as an actual reply) may be used to their prejudice in the course of the proceedings,’

‘… a consciousness of a risk as to costs if reasonable offers are refused can only encourage settlement whilst, on the other hand, it is hard to imagine anything more calculated to encourage obstinacy and unreasonableness than the comfortable knowledge that a litigant can refuse with impunity whatever may be offered to him even if it is as much as or more than everything to which he is entitled in the action.’

In layman’s terms, the public policy was to protect negotiations by permitting them to be made without prejudice, as this would encourage parties to settle disputes without the fear that something said or offered in the course of negotiations might subsequently be deployed against them. But the without prejudice veil could then be lifted after trial to determine if a party had not been negotiating reasonably, which could result in a costs order being made against them. Absent this, where is the incentive for reasonable engagement? Already, one may sense the tension with our current rules (more of which below).

The Calderbank rule has been referred to and approved in countless cases, up to the House of Lords (cf Rush & Tompkins Ltd v Greater London Council [1989] AC 1280 and Ofulue v Bossert [2009] UKHL 16, for instance) and was enshrined in the Rules of the Supreme Court, then in CPR Part 44 and explicitly in the precursor to the Family Procedure Rules 2010 (FPR 2010), the Family Proceedings Rules 1991 (FPR 1991) at r 2.69, which provided, inter alia:

‘(1) Either party to the application may at any time make a written offer to the other party which is expressed to be “without prejudice except as to costs” and which relates to any issue in the proceedings relating to the application. …

2.69B Judgment or order more advantageous than an offer made by the other party

(1) This rule applies where the judgment or order in favour of the applicant or respondent is more advantageous to him than an offer made under rule 2.69(1) by the other party.

(2) The court must, unless it considers it unjust to do so, order that other party to pay any costs incurred after the date beginning 28 days after the offer was made.’

There it surely ends: the careful development within our common law, over many decades, of an important mechanism by which the court could impose cost consequences on a failure by a litigant to engage reasonably in the settlement of a dispute, whilst preserving the confidentiality which makes meaningful settlement negotiations possible.

2006 reforms

Actually, no. Quite the opposite, in fact. To the ongoing frustration of many family practitioners, the Calderbank rule was canned in 2006 with the insertion of r 2.71 into the FPR 1999 and ‘the general rule in ancillary relief proceedings … that the court will not make an order requiring one party to pay the costs of the other party’. And so, the (damaging) presumption of ‘no order as to costs’ came to be. Incidentally, when the FPR 2010 came into effect on 6 April 2011 (to consolidate the rules operative in family law), FPR 1999, r 2.71 was mirrored in FPR 2010, r 28.3 (with which we are all very familiar), subject to one very important future addition (see below).

Paragraph 7.1 of the Explanatory Memorandum to the Family Proceedings (Amendment) Rules 2006, which introduced this volte face, is illuminating:

‘The purpose of these rule amendments is to make changes to the courts’ power to make cost orders in ancillary relief proceedings. The new rules will only give the court power to make cost orders in ancillary relief proceedings where it is justified by the litigation conduct of a party. The policy objective is to discourage the making of costs orders in ancillary relief cases other than because of the conduct of one of the parties. Cost orders have a destabilising effect on financial settlements that have been carefully constructed by the court.

The rules also provide that courts will no longer be able to consider any “without prejudice” or “without prejudice save as to costs” offers (also known as Calderbank offers) when making decisions on cost orders to avoid the procedural gamesmanship and uncertainty for the parties which these can produce.’

The case for this reform and the abolition of the Calderbank rule is also neatly summarised in GW v RW [2003] EWHC 611 (Fam) by Mostyn J, and in Norris v Norris; Haskins v Haskins [2003] EWCA Civ 1084 by Butler-Sloss and Thorpe LJJ. It can be summarised as follows:

(1) FPR 1999, r 2.69 was in parts incomprehensible.

(2) The nature of financial remedy proceedings had changed dramatically following White v White [2000] UKHL 54, [2001] 1 AC 596. As the costs committee noted at the time:

‘Family proceedings arise out of the breakdown of a marriage, which may be seen as a misfortune falling on both parties. The fact that the court has to assist the parties to re-adjust their finances should not of itself imply blame on the part of either party. … As Mr Mostyn QC points out at para 86 of his judgment [in GW v RW], it may often be that “each party has had to come to the court to obtain an order which fairly disposes of the issues between them”.’

This gave traction to the appeal of a presumption of no order as to costs.

(3) The Calderbank rule imposed on the parties the obligation to engage in a form of spread betting: guess the outcome of the case and take a position, leading to (harsh) binary outcomes, such that a party who beats their offer by even a ‘tiny amount’ would result in them being awarded their costs post-offer.

(4) The ‘procedural gamesmanship’ referred to the creation of two-tier negotiations. If an open offer carries little or no immediate costs risk, there is considerably less incentive to make it realistic. An ambitious open position can be maintained without necessarily exposing the party to the same costs consequences which accompanied the rejection of a reasonable Calderbank proposal. The consequence can be a curious two-tier negotiation. The open position becomes a piece of litigation positioning. The real negotiation takes place elsewhere. This could also upset a carefully crafted judgment, if a judge then felt duty-bound to make a costs order once the parties’ without prejudice save as to costs positions were unveiled.

(5) The Calderbank rule resulted in keenly fought satellite costs assessment litigation.

Financial remedy litigation is unusually discretionary (and this is not always to its credit, especially with increasing numbers of litigants in person – we expand on this below). The outcome is rarely susceptible to the sort of binary prediction which is possible in much conventional civil litigation.

The reforms which followed were therefore not without foundation. But the solution went further than removing an undesirable degree of costs gambling and removed the costs protection which accompanied private settlement negotiations altogether. The new regime sought instead to encourage parties to negotiate openly and reasonably. Open offers could be made, and conduct could be taken into account when determining costs (and which could include consideration of any open offers made).

The intention was laudable. If parties were no longer able to deploy Calderbank offers, they would instead be encouraged to put sensible proposals openly before one another and thereby facilitate settlement.

Unintended consequences

With the removal of costs consequences for failing to negotiate reasonably, parties’ open positions were often hugely ambitious and/or barely arguable, but, at the same time, possibly coupled with a slightly more reasonable without prejudice position that had no teeth. In effect, gone was the incentive for parties to engage reasonably, leading to entrenched positions,[[3]] which could only really be stymied, if at all, by attendance at an FDR, where parties may expose themselves to criticism from the FDR ‘judge’; though (importantly) still with no costs consequences.

Did the changes to the rules have a further unintended consequence of leading some lawyers to being less focused on advising clients on the legal merits of their position and, instead, pitching proposals as high as could be argued? Intellectual rigour amongst practitioners would quite often fall away and so, too, discipline in the round: there were precious little consequences for the clients and their lawyers.

As aforementioned, while there was reference to ‘any open offer to settle made by a party’ under the umbrella of conduct under FPR 2010, r 28.3, this was of little practical effect. The received wisdom was that it was too risky to advance sensible (i.e. reasonable and responsible) open proposals for fear that any judge would assume that there was a more sensible without prejudice proposal lurking beneath.

Inevitably, this led to concern, not just amongst practitioners, but also the judiciary. In a bid to encourage parties to openly negotiate sensibly, Moor J said in MAP v MFP [2015] EWHC 627 (Fam)[[4]] at [87]:

‘Now that we no longer have Calderbank offers, litigants must be encouraged to make open proposals as early as possible that are designed to encourage settlement. If the other party spurns such an offer, the court is entitled to ignore it completely and decide the case entirely on the merits. I will have no hesitation in a suitable case in awarding an applicant more than an open offer he or she has made if that is justified.’

Strictly speaking, the courts were permitted to consider parties’ open offers under the umbrella of conduct when making decisions about costs orders. However: (1) there was no further encouragement within the rules; (2) it conflicted with the overarching presumption; and (3) there was considerable resistance amongst the profession to advise clients to advance sensible proposals. The rule change had become problematic and (belatedly) the Costs Working Group of the Family Procedure Rule Committee (FPRC) had this to say preceding the May 2019 amendment to PD 28A:

‘FPRC is concerned that insufficient emphasis is given to encouraging parties to engage reasonably and responsibly in negotiations. In particular, there is concern that little positive guidance was given in PD28A to assist the parties to understand the likely costs consequences of failing to litigate sensibly and failing to engage in sensible negotiations and/or of making an open proposal which is significantly higher or lower than the award ultimately made by the court’[[5]]

What is frustrating about this is quite how obvious this issue was. We refer to Oliver LJ’s comments above, as well as Fox LJ’s comments in Cutts v Head, all made some 25 years prior to the FPRC’s observations:

‘If a party is exposed to a risk as to costs if a reasonable offer is refused, he is more rather than less likely to accept the terms and put an end to the litigation. On the other hand, if he can refuse reasonable offers with no additional risk as to costs, it is more rather than less likely to encourage mere stubborn resistance.’

The fudge

Therefore, and in the light of calls in many corners for a reintroduction of the Calderbank principle (albeit with some procedural reform) from Resolution[[6]] and many other practitioners, in 2019, in order to right the wrongs of the abolition of Calderbank, the FPRC introduced the following addition to FPR 2010, PD 28A, para 4.4:

‘The court will take a broad view of conduct for the purposes of this rule and will generally conclude that to refuse openly to negotiate reasonably and responsibly will amount to conduct in respect of which the court will consider making an order for costs. This includes in a “needs” case where the applicant litigates unreasonably resulting in the costs incurred by each party becoming disproportionate to the award made by the court. Where an order for costs is made at an interim stage the court will not usually allow any resulting liability to be reckoned as a debt in the computation of the assets.’

This addition to PD 28A was later sharpened into an oft-relied-upon tool for seeking costs orders following open proposals, and perhaps most memorably by Mostyn J in OG v AG [2020] EWFC 52 at [31]:

‘It is important that I enunciate this principle loud and clear: if, once the financial landscape is clear, you do not openly negotiate reasonably, then you will likely suffer a penalty in costs. This applies whether the case is big or small, or whether it is being decided by reference to needs or sharing.’

The force of OG v AG has extended well beyond the facts of that case. The passage is now routinely deployed by practitioners in correspondence as a warning that a failure to engage reasonably with an open proposal may carry costs consequences. In practical terms, it has become a familiar substitute for the negotiating pressure once supplied by a Calderbank offer: parties are told, in effect, that they may not be able to shelter behind their own litigation strategy if they fail to negotiate reasonably (often read, make significant concessions) once the financial landscape is sufficiently clear.

The addition to PD 28A, para 4.4 and the mounting case law following OG v AG which supported judicial treatment of open offers being referable to costs, has consequentially compelled the ‘guessing’ to the open arena to protect parties on costs by early admissions, and a paradox was created. If the abolition of Calderbank was justified in part because its costs consequences produced undesirable gamesmanship and uncertainty, why is the modern regime increasingly seeking to recreate costs incentives through open negotiation and conduct? The answer is that settlement still requires incentives.

The difficulty is that litigation conduct and open offers are now being asked to do the work once performed by Calderbank, but without the protection of allowing parties to negotiate on a without prejudice basis and without fear that those concessions will later be deployed against them. Moor J can say in MAP v MFP that judges are entitled to ignore the open offer and decide a case on its merits – and this may be right in principle – but, in practice, can this be assured in all cases? Judges are humans. It would be impossible to rule out open concessions colouring their judgments, and yet this very risk is positively imposed on litigants, else it will sound in costs.

That refusal to negotiate reasonably and responsibly on an open basis may result in a costs order against a party strikes at the very core of the without prejudice rule:

‘The public policy justification [of the without prejudice rule], in truth, essentially rests on the desirability of preventing statements or offers made in the course of negotiations for settlement being brought before the court of trial as admissions on the question of liability.’[[7]]

The Calderbank benefit of a sealed offer, only opened at the point of costs (i.e. after judgment), was that the primary judgment could not be coloured by such without prejudice concessions.

This is not merely an arid theoretical concern; it affects us all in practice. Take the following examples:

(1) A party may be willing to be over-generous to ‘kill’ the litigation to avoid the cost or uncertainty of a hearing.

(2) A party may be emotionally unable to contemplate the stress of a final hearing (including the stress of cross-examination, etc) and so makes an offer against their interests (a particular risk in a case involving abuse/coercive control).

(3) A party may accept a lower spousal maintenance figure or term than they are entitled to for similar reasons. Doing so on an open basis is hugely risky, as maintenance is usually confined to ‘need’.

(4) A party may agree to a particular division of capital because they want to retain an asset which has personal significance.

(5) Conversely, a party may offer more than they consider strictly justified because the financial and emotional cost of continued litigation has become disproportionate and/or the other side’s position is completely unreasonable.

None of those decisions necessarily represents a party’s true assessment of their legal entitlement; they are compromises, often significant. But, in order to apply costs pressure to the other side, the current regime effectively pressures parties to make such compromises openly and, yet, once the genie is out of the bottle, retreat becomes considerably harder.

This goes beyond tactical inconvenience and fundamentally strikes at the rationale for the without prejudice and public policy principles. Furthermore, if a party fails to engage in without prejudice discussions, the consequence is that a party has incurred costs in unsuccessful without prejudice or non-court dispute resolution attempts, and the spurning party cannot be held to account. Can that be regarded as fair?

Not so simple

There is a further practical difficulty.

The costs regime is not even uniform across closely connected financial proceedings. FPR 2010, r 28.3 only applies to ‘financial remedy proceedings’. Proceedings which are merely ‘in connection with’ financial remedy proceedings fall outside that regime. The distinction is not merely semantic. At [51] in Judge v Judge [2008] EWCA Civ 1458, Wilson LJ drew a deliberate distinction between proceedings ‘for’ and ‘in connection with’ (then) ancillary relief, with the former attracting the ‘no-order for costs’ rule; the latter did not (thereby retaining the Calderbank jurisdiction):

‘Although the proceedings before the judge were in connection with ancillary relief, they were not for ancillary relief…. The general rule in [the modern FPR 2010, r 28.3(6)] is only a concomitant of the modern approach in applications for ancillary relief that the sum owed by each party in respect of his own costs will be treated as his liability for the purposes of calculating the substantive award.’

The result is that closely connected financial litigation may be governed by materially different costs regimes. An application to vary a maintenance order falls within the financial remedy ‘no order for costs’ regime, whereas enforcement proceedings do not (even though they are commonly heard together). Thus, the same parties could attend the same hearing, concerning the same financial circumstances, with the applications heard together and often arising from the same underlying dispute, yet find themselves subject to two entirely different costs regimes.

The position is even more difficult for litigants in person. The Law Commission expressly observed that the interaction of the rules was difficult to follow and recommended that the costs rules applicable to enforcement of family financial orders be consolidated into a stand-alone regime.[[8]] However, the answer to the apparently straightforward question ‘Can I make a private settlement offer which may have costs consequences?’ continues to depend upon whether the proceedings are characterised as variation, enforcement, maintenance pending suit, an application for payment of legal services or some other form of financial application.

Indeed, if the concern was that private offers created gamesmanship and uncertainty, it is difficult to see why the answer should be to prohibit them in one category of financial litigation whilst leaving them potentially available in another closely connected category.

The modern costs position requires a litigant in person (and practitioner) to understand, amongst other things:

(1) the difference between proceedings for a financial order and those ‘in connection with’ a financial order;

(2) the general rule that there will be no order as to costs in proceedings for a financial order, and the rule that costs follow the event in proceedings in connection with a financial order;

(3) the exception based upon conduct (and then what qualifies as relevant litigation conduct);

(4) the importance of open offers;

(5) the limits of without prejudice offers; and

(6) the potential significance of departing from a reasonable proposal.

That is a considerable amount of procedure and procedural strategy to expect a lay party to absorb, with now ever more serious cost consequences for error.

It is particularly acute in financial remedy proceedings because the jurisdiction is not a conventional damages jurisdiction. The ultimate outcome is discretionary. A litigant may therefore be required to formulate an offer which reflects their best assessment of a highly uncertain judicial outcome, whilst not understanding or knowing that the offer may ultimately be relevant to the costs consequences of their conduct. That difficulty is not unique to open offers: it is inherent in any costs regime which attaches consequences to the gap between an offer and the eventual outcome. The important distinction is that a Calderbank offer permits the parties to test those positions within the protection of without-prejudice negotiations and explicitly referable to costs, whereas an open offer exposes the negotiating position itself to scrutiny and whether it was ‘reasonable’ (a highly uncertain concept, except perhaps in retrospect).

The supposed simplicity of ‘make a reasonable open offer and negotiate reasonably’ is therefore rather less simple in operation.

The way forward: restore, do not recreate

We are not saying that the answer is to return uncritically to the pre-2006 regime. The concerns which prompted reform were real and there were legitimate objections to a costs system in which parties could face substantial consequences for failing to predict with sufficient accuracy the outcome of a highly discretionary financial remedy exercise. The court must retain a broad discretion. Nor should settlement become a mathematical game in which the litigant who guesses the eventual award most accurately automatically wins the costs argument.

However, it does not follow that Calderbank itself was the problem, but rather how its application and interpretation evolved over time. It was never the case that it was designed to provide binary outcomes and nor should it (though that does, in part, appear to be the thrust of Mostyn J’s concerns in GW), especially in the fraught and highly discretionary area of financial remedy. Indeed, we understand that in civil litigation, the Calderbank jurisdiction is relied upon less than the Part 36 procedure under the CPR, precisely because it results in less binary outcomes in respect of costs.

The better answer may be to restore a properly defined Calderbank jurisdiction and allow the FPRC to finesse its operation. Factors such as the timing of the offer, the reasonableness of the proposal, the information available to the parties when it was made and level of disclosure available, the extent of the compromise sought, the complexity of the case, the parties’ respective conduct and the eventual outcome could all remain relevant.

Crucially, with this reinstatement, the distinction between a settlement concession (made without prejudice) and an admission would also once again be respected.

Conclusion

The abolition of Calderbank was driven by legitimate concerns. Financial remedy litigation is unusually discretionary, and the 2006 reforms sought to reduce the tactical gamesmanship associated with private costs offers.

But remove the costs consequence of rejecting a reasonable private offer and there is less incentive to make one. The subsequent reliance on open offers and conduct has, instead, created a potentially more serious collision with the public policy underlying without prejudice protection.

Settlement requires candour, but candour requires protection. The answer cannot be to continue blurring that distinction through open offers and conduct.

Calderbank should be restored, with its operation refined by the FPRC. The court should retain discretion as to whether an offer was reasonable, whether its rejection was unreasonable and what costs consequence, if any, should follow. But that is very different from near abolishing its jurisdiction altogether.

[[1]]: Family law is not a ‘desert island’ is Lord Sumption’s observation in Prest v Petrodel Resources Ltd [2013] UKSC 34. Family law is not a self-contained jurisdiction in which established legal principles acquire a different meaning simply because the dispute happens to arise between spouses. As Lord Sumption observed in Prest, where a legal right exists, it exists across the different jurisdictions of the High Court and county courts.

[[2]]: Indeed, the one thing strikingly absent from Calderbank was a Calderbank offer! Also curiously, it doesn’t actually have much to say about the without prejudice save as to costs principle; the far better read is Cutts v Head [1984] Ch 290. However, it is worth your time, if only to read Scarman LJ take counsel for the wife (she being the wealthier party) to task for his somewhat antiquated submissions: ‘Mr. Hordern is really saying that we must read section 25 [Matrimonial Causes Act 1973] as stating impliedly, because it certainly does not say so expressly, that financial provision may be made for a husband only in exceptional circumstances. The section says nothing of the sort’. He doesn’t leave it there: ‘It is very difficult to fault the judge’s conclusion except upon the theoretical or conceptual basis advanced by Mr. Hordern that really this statute does not provide that a wife should make financial provision for her husband save in exceptional circumstances. Such a provision cannot be found in the relevant sections’.

[[3]]: The point being that, with the passage of time, clients often become very attached to proposals they advance, notwithstanding advice at the time that those proposals may be unrealistic or, worse still, fanciful.

[[4]]: This was oft-quoted by practitioners with open proposals in a bid to give them substance (or teeth).

[[5]]: Family Procedure Rules 2010, ‘Consultation in relation to the treatment of Calderbank offers when determining issues relating to costs’, para 5.

[[6]]: ‘Consultation in relation to the treatment of Calderbank offers when determining issues relating to costs: Resolution’s response to the Family Procedure Rule Committee’. In answer to ‘Do you consider that offers made “without prejudice save as to costs” should be admissible in considering the “conduct” of a party for the purposes of FPR r28.3?’, 118 were in support and 17 in opposition.

[[7]]: Oliver LJ (again).

[[8]]: Enforcement of Family Financial Orders, Law Com No 370, 2016, para 16.10.